![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F8193876%2F25ddf78aa93b4adfa7d853f14e5e85f6%2F11111.png?generation=1668608504095158&alt=media\")
I just found a post on XiaoHongShu(Chinese Instagram): http://xhslink.com/S6Ttfl
\nWhich means, if you paid, you will be guaranteed to get at least a silver medal, or even a gold medal.
\nThis is an extension of a comment I made in another post. Some participants said his LB dropped from 60 to 110 within a single day. This new post is for everyone to better collect and organize evidence.
\nAnother post for Feedback competition: http://xhslink.com/cFdufl
------------------------------------Translate for the first link-------------------------------------------
\nTitle: Silver medal guaranteed! The latest edition of the Kaggle competition is hot for registration!
Full text：
\n-----------------------------------------Suspecious patterns---------------------------------------------
\nMany of the team's accounts have very strangely similar behavior.
\nFor example, I think most teams ranking 20~37 currently are suspecious, except team chromosom and Shuji Suzuki.
The only records that almost all of the accounts in these suspicious teams had on kaggle before were posting a meaningless comment in a random post. (e.g. Thank you! Wonderful idea!) I guess this will turn them from novice to contributor, so they won't be detected as easily!
\nThere should be more suspicious teams on the whole leaderboard. But the contributors teams ranking 20~37 now are too wired.
I just found a post on XiaoHongShu(Chinese Instagram): http://xhslink.com/S6Ttfl
\nWhich means, if you paid, you will be guaranteed to get at least a silver medal, or even a gold medal.
\nThis is an extension of a comment I made in another post. Some participants said his LB dropped from 60 to 110 within a single day. This new post is for everyone to better collect and organize evidence.
\nAnother post for Feedback competition: http://xhslink.com/cFdufl
------------------------------------Translate for the first link-------------------------------------------
\nTitle: Silver medal guaranteed! The latest edition of the Kaggle competition is hot for registration!
Full text：
\n-----------------------------------------Suspecious patterns---------------------------------------------
\nMany of the team's accounts have very strangely similar behavior.
\nFor example, I think most teams ranking 20~37 currently are suspecious, except team chromosom and Shuji Suzuki.
The only records that almost all of the accounts in these suspicious teams had on kaggle before were posting a meaningless comment in a random post. (e.g. Thank you! Wonderful idea!) I guess this will turn them from novice to contributor, so they won't be detected as easily!
\nThere should be more suspicious teams on the whole leaderboard. But the contributors teams ranking 20~37 now are too wired.
Since this comment is mainly intended to be seen by more Chinese Kagglers, please allow me to use Chinese first and I will translate and explain in English below.
\n······························································································
\n我昨天在得知这件事和写那条长回复时，也觉得我们没什么能做的，只能交给官方去审核作弊，而且从本质上无法避免这种行为，他们完全可以进一步伪装，最终无法分辨。
\n但 @kaito510 的回复提醒了我，我们作为能看懂中文的群体，有些事或许是我们可以做的，而且只有我们可以做。
1。我们可以向淘宝官方的审核团队和小红书的审核团队进行沟通，或者让Kaggle官方出面和他们沟通，要求他们对这个灰色产业的广告进行审核。
\n这是有先例可循的，在淘宝创立初期，我们几乎可以在上面购买到任何不直接违反国家法律的服务，大量灰色产业在淘宝肆无忌惮，但是随着越来越多的审核机制，淘宝已经禁止了相当一部分灰色产业。比如你可以尝试搜索英语6级代考，你不会得到任何结果。对于Kaggle的奖牌我想也是可以并且应该被禁止公开售卖的。但这不会自动发生，需要有人主动与淘宝的审核团队沟通，Kaggle的官方不会想到这个方法，所以我在此提出。如果你们愿意作为官方出面，我想我和其他的中文Kaggler会愿意尽可能提供帮助。对于小红书和其他电子商务和社交媒体平台也同样适用，每个平台都有内容审核团队。 @ryanholbrook @wcukierski
\n2。我们可以对发现的售卖奖牌的广告进行举报。如果上述的审核内容成果被纳入各个平台愿意打击的内容，我想些信息会很快被清除掉。
\n当然这也无法从本质上完全避免这类交易，但是可以将他们从主要的电商和社交媒体平台删除，这将会有极大的帮助。我在准备留学的时候听说过太多离谱的这类交易了，甚至传说可以付高价直接买到托福、雅思、GRE等考试的成绩，我不知道是通过枪手代考还是内部修改成绩，但是因为这几个考试的主办方与国内平台沟通，这项交易被各个平台严厉打击，普通人很难获取到相关广告和报价，所以只有非常少量的交易进行，尽可能的保护了考试的公平性。
\n对于Kaggle奖牌交易，我想完全可以遵循以上先例。目前即使各平台不加以审核，公开宣传，也不是一个很大的产业，我猜提供这项服务的只有不到100人。如果能将他们的广告从平台删除，私下进行的交易量将会很小，甚至无法支撑这个产业形成能够持续运作的规模，从而可能解决大规模作弊的现象。
以上是我个人对这件事的一些看法和建议，从这个交易的源头下手，而不是在竞赛之后判断谁是作弊者然后删除成绩。我与 @forcewithme 私下沟通了这个想法，我们认为这可能是目前最好的办法了。
\n作为Kaggle社区中的中文群体，其他的国人做了这样的事，影响到了整个国际社区，我为此感到羞耻。而当想到了这个或许是只有作为中国人才能帮助到整个社区的方法，我认为我没有理由不提出来并承担这份责任。如果你作为一个中国人与我有同样的感觉并且赞同或愿意帮助实施这项方案，欢迎在下面评论，如果你有其他的建设性建议也请发出来。这件事关系到每一位在Kaggle平台认真竞赛的中文Kaggler和所有Kaggler，你一定不希望之后参加的比赛中越来越多的遇到这个比赛类似的事情。
\n如果你的账号涉及到个人信息隐私，请注意保护自己，我们面对的是一个大概几十人的灰色产业。但是相比认真竞赛的中文Kaggler和整个Kaggle社区，我们没什么好害怕的。
\n······························································································
\nWhen I was told about this and writing that long reply yesterday, I too thought there was not much we could do but leave it to the officials to check for cheating, and that there was no way to inherently avoid this kind of deal, and that they could have disguised it further and ultimately not been able to distinguish it.
But @kaito510 's reply reminded me that there are things we, as a group of people who can understand Chinese, might be able to do, and only we can do.
\n1。We can communicate with Taobao's official audit team and Xiaohongshu's audit team, or engage Kaggle officials to communicate with them and ask them to audit the ads in this grey industry.
\nThere are precedents for this. In the early years of Taobao, we could buy almost any service on it that did not directly violate national laws, and a large number of grey industries ran rampant on Taobao, but with more and more auditing mechanisms, Taobao has banned quite a number of them. For example, you can try searching for an English level 6 exams agency and you won't get any results. For Kaggle medals I think they can and should be banned from being sold publicly too. But this wouldn't happen automatically, someone would need to take the initiative to communicate with the Taobao audit team, and the Kaggle officials wouldn't think of this method, so I'm proposing it here. If you are willing to step in as an official, I think I and the other Chinese Kagglers would be willing to help as much as possible. The same applies to Xiaohongshu and other e-commerce and social media platforms, every one of which has a content audit team. @ryanholbrook @wcukierski
\n2。We can report advertisements we find that are selling medals. I think some information will be removed quickly if the above-mentioned audit content results are included in the content that each platform is willing to combat.
\nOf course, this wouldn't prevent these types of deals inherently and completely, but it would help immensely to remove them from the main e-commerce and social media platforms. I've heard of too many ridiculous deals of this kind when I was preparing to study abroad, and it's even rumoured that you can pay high prices to buy test scores directly for TOEFL, IELTS, GRE, etc. I don't know if it's through a ghostwriter or an internal revision of the scores, but because the organisers of these exams communicate with the domestic platforms, this deal is cracked down on by each platform and it's very difficult for ordinary people to access the relevant ads and quotes, so only a very small number of deals are conducted and the fairness of the exams is protected as much as possible.
\nFor Kaggle medal trading, I think it is perfectly possible to follow this precedent. At the moment, even if the platforms are not audited and are publicly advertised, it is not a very big industry, and I would guess that there are fewer than 100 people offering this service. If their ads could be removed from the platforms, the amount of trading that takes place privately would be too small to even support an industry on a scale that could sustainably operate and thus potentially address the massive cheating.
\nThe above are some of my personal thoughts and suggestions on this matter, start at the origin of this deal rather than judging who is the cheater after the competition and then removing the results. I communicated this idea privately with @forcewithme and we think this is probably the best way to proceed for now.
\nAs a group of Chinese in the Kaggle community, other Chinese have done something like this which has affected the international community as a whole, and I am ashamed of it. And when thinking of this as perhaps only being Chinese can help the whole community, I see no reason not to raise it and take this responsibility. If you, as a Chinese, feel the same way as I do and agree or would like to help implement this idea, please feel free to comment below, and if you have other constructive suggestions, please post them too. This matter concerns every Chinese Kaggler and all Kagglers who compete seriously on the Kaggle platform, and you certainly don't want to encounter more and more things like this competition in the competitions you participate in afterwards.
\nIf your account involves personal information privacy, please take care to protect yourself, we are facing a grey industry of about a few dozen people. But compared to the Chinese Kaggler who are serious about competing and the whole Kaggle community, we have nothing to fear.
", "rawMarkdown": "Since this comment is mainly intended to be seen by more Chinese Kagglers, please allow me to use Chinese first and I will translate and explain in English below.\n······························································································\n我昨天在得知这件事和写那条长回复时，也觉得我们没什么能做的，只能交给官方去审核作弊，而且从本质上无法避免这种行为，他们完全可以进一步伪装，最终无法分辨。\n但 @kaito510 的回复提醒了我，我们作为能看懂中文的群体，有些事或许是我们可以做的，而且只有我们可以做。\n\n1。我们可以向淘宝官方的审核团队和小红书的审核团队进行沟通，或者让Kaggle官方出面和他们沟通，要求他们对这个灰色产业的广告进行审核。\n\n这是有先例可循的，在淘宝创立初期，我们几乎可以在上面购买到任何不直接违反国家法律的服务，大量灰色产业在淘宝肆无忌惮，但是随着越来越多的审核机制，淘宝已经禁止了相当一部分灰色产业。比如你可以尝试搜索英语6级代考，你不会得到任何结果。对于Kaggle的奖牌我想也是可以并且应该被禁止公开售卖的。但这不会自动发生，需要有人主动与淘宝的审核团队沟通，Kaggle的官方不会想到这个方法，所以我在此提出。如果你们愿意作为官方出面，我想我和其他的中文Kaggler会愿意尽可能提供帮助。对于小红书和其他电子商务和社交媒体平台也同样适用，每个平台都有内容审核团队。 @ryanholbrook @wcukierski\n\n2。我们可以对发现的售卖奖牌的广告进行举报。如果上述的审核内容成果被纳入各个平台愿意打击的内容，我想些信息会很快被清除掉。\n\n当然这也无法从本质上完全避免这类交易，但是可以将他们从主要的电商和社交媒体平台删除，这将会有极大的帮助。我在准备留学的时候听说过太多离谱的这类交易了，甚至传说可以付高价直接买到托福、雅思、GRE等考试的成绩，我不知道是通过枪手代考还是内部修改成绩，但是因为这几个考试的主办方与国内平台沟通，这项交易被各个平台严厉打击，普通人很难获取到相关广告和报价，所以只有非常少量的交易进行，尽可能的保护了考试的公平性。\n对于Kaggle奖牌交易，我想完全可以遵循以上先例。目前即使各平台不加以审核，公开宣传，也不是一个很大的产业，我猜提供这项服务的只有不到100人。如果能将他们的广告从平台删除，私下进行的交易量将会很小，甚至无法支撑这个产业形成能够持续运作的规模，从而可能解决大规模作弊的现象。\n\n以上是我个人对这件事的一些看法和建议，从这个交易的源头下手，而不是在竞赛之后判断谁是作弊者然后删除成绩。我与 @forcewithme 私下沟通了这个想法，我们认为这可能是目前最好的办法了。\n\n作为Kaggle社区中的中文群体，其他的国人做了这样的事，影响到了整个国际社区，我为此感到羞耻。而当想到了这个或许是只有作为中国人才能帮助到整个社区的方法，我认为我没有理由不提出来并承担这份责任。如果你作为一个中国人与我有同样的感觉并且赞同或愿意帮助实施这项方案，欢迎在下面评论，如果你有其他的建设性建议也请发出来。这件事关系到每一位在Kaggle平台认真竞赛的中文Kaggler和所有Kaggler，你一定不希望之后参加的比赛中越来越多的遇到这个比赛类似的事情。\n\n如果你的账号涉及到个人信息隐私，请注意保护自己，我们面对的是一个大概几十人的灰色产业。但是相比认真竞赛的中文Kaggler和整个Kaggle社区，我们没什么好害怕的。\n······························································································\nWhen I was told about this and writing that long reply yesterday, I too thought there was not much we could do but leave it to the officials to check for cheating, and that there was no way to inherently avoid this kind of deal, and that they could have disguised it further and ultimately not been able to distinguish it.\n\nBut @kaito510 's reply reminded me that there are things we, as a group of people who can understand Chinese, might be able to do, and only we can do.\n\n1。We can communicate with Taobao's official audit team and Xiaohongshu's audit team, or engage Kaggle officials to communicate with them and ask them to audit the ads in this grey industry.\n\nThere are precedents for this. In the early years of Taobao, we could buy almost any service on it that did not directly violate national laws, and a large number of grey industries ran rampant on Taobao, but with more and more auditing mechanisms, Taobao has banned quite a number of them. For example, you can try searching for an English level 6 exams agency and you won't get any results. For Kaggle medals I think they can and should be banned from being sold publicly too. But this wouldn't happen automatically, someone would need to take the initiative to communicate with the Taobao audit team, and the Kaggle officials wouldn't think of this method, so I'm proposing it here. If you are willing to step in as an official, I think I and the other Chinese Kagglers would be willing to help as much as possible. The same applies to Xiaohongshu and other e-commerce and social media platforms, every one of which has a content audit team. @ryanholbrook @wcukierski\n\n2。We can report advertisements we find that are selling medals. I think some information will be removed quickly if the above-mentioned audit content results are included in the content that each platform is willing to combat.\n\nOf course, this wouldn't prevent these types of deals inherently and completely, but it would help immensely to remove them from the main e-commerce and social media platforms. I've heard of too many ridiculous deals of this kind when I was preparing to study abroad, and it's even rumoured that you can pay high prices to buy test scores directly for TOEFL, IELTS, GRE, etc. I don't know if it's through a ghostwriter or an internal revision of the scores, but because the organisers of these exams communicate with the domestic platforms, this deal is cracked down on by each platform and it's very difficult for ordinary people to access the relevant ads and quotes, so only a very small number of deals are conducted and the fairness of the exams is protected as much as possible.\n\nFor Kaggle medal trading, I think it is perfectly possible to follow this precedent. At the moment, even if the platforms are not audited and are publicly advertised, it is not a very big industry, and I would guess that there are fewer than 100 people offering this service. If their ads could be removed from the platforms, the amount of trading that takes place privately would be too small to even support an industry on a scale that could sustainably operate and thus potentially address the massive cheating.\n\nThe above are some of my personal thoughts and suggestions on this matter, start at the origin of this deal rather than judging who is the cheater after the competition and then removing the results. I communicated this idea privately with @forcewithme and we think this is probably the best way to proceed for now.\n\nAs a group of Chinese in the Kaggle community, other Chinese have done something like this which has affected the international community as a whole, and I am ashamed of it. And when thinking of this as perhaps only being Chinese can help the whole community, I see no reason not to raise it and take this responsibility. If you, as a Chinese, feel the same way as I do and agree or would like to help implement this idea, please feel free to comment below, and if you have other constructive suggestions, please post them too. This matter concerns every Chinese Kaggler and all Kagglers who compete seriously on the Kaggle platform, and you certainly don't want to encounter more and more things like this competition in the competitions you participate in afterwards.\n\nIf your account involves personal information privacy, please take care to protect yourself, we are facing a grey industry of about a few dozen people. But compared to the Chinese Kaggler who are serious about competing and the whole Kaggle community, we have nothing to fear.", "votes": 32, "replies": [ { "id": 2033994, "postDate": "2022-11-17T17:52:39.837Z", "content": "I think this approach is more effective in limiting such behavior than detecting with technical approaches. @ryanholbrook @wcukierski Please think about this suggestion. 🙏
", "rawMarkdown": "I think this approach is more effective in limiting such behavior than detecting with technical approaches. @ryanholbrook @wcukierski Please think about this suggestion. 🙏", "votes": 10 } ] }, { "id": 2032915, "postDate": "2022-11-16T21:49:42.910Z", "content": "Although I did not participate in this competition, I was shocked and disgusted to hear about this today. I went to the Xiaohongshu platform to look up the content and found that these illegal ads are still there with impunity and there are dozens, hundreds of people replying below to seek further information on prices.
\nThese dirty deals are mainly targeted at Chinese undergraduates who are desperately looking for ways to improve their academic background and thereby increase their chances of getting an offer from a prestigious university when applying to overseas master's programs. The Kaggle medal obviously is a perfect choice.
\nTwo years ago, I experienced the same thing as these undergraduates, applying to master's programs and enhancing my academic background through extracurricular research and certificates. I applied to the MSc AI program as an interdisciplinary student, and I had learned or heard from friends around me who were preparing to study abroad together, they participated in various programs to enhance their academic background in AI or data science, but at that time I had never heard of this dirty deal of paying for Kaggle's silver medal.
\nToday, this deal is taking shape. If you look closely at some of the very dubious accounts on the leaderboard, some of them have had more than just this one dubious entry. If you look at their teammates from previous competitions, they even fit @forcewithme 's description above, which is very suspicious. So these dirty deals didn't just start at this competition, but also at previous ones, and some people would buy medals multiple times, they just weren't big enough to be discovered before.
\nJudging from the number of relevant ads and replies on just one platform, Xiaohongshu, at least a few hundred people have already inquired about the quotes. Judging from the suspicious accounts of this competition and the teammates of these accounts from previous competitions, at least a few dozen people have already paid for medals.
\nThe point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term. This is only the beginning, I had never heard of this when I applied two years ago; Now, I estimate that less than 10 people are offering these deals, and with dozens of people paying to participate in this one competition, it is being discovered and exposed by everyone for the first time; If it is not stopped, in a few years time, when this disgusting deal becomes popular and is heard of by more and more undergraduates, believe me, just the annual number of students in related majors in China, even if only a small percentage of them choose to pay, It is also enough to ruin the value of the Kaggle competition medals and the Kaggle community.
\nI would like to keep the Chinese version of this thread and write some words to the cheaters of this competition and to future students who want to pay to cheat, in the hope that you will see it and stop doing so. I don't want to just say something about the philosophy of life with honesty.
\nA Kaggle medal may help you get one more offer, but it will never completely change the trajectory of your life. Perhaps you now think that getting to apply to a certain top school is the most important thing and you will do whatever it takes to achieve it, even if it's this dirty trick. But once you've actually completed the degree, you'll realise that the ranking of the school you were so focused on is simply not that important. The different offers that a Kaggle medal can help you get, most likely won't make any difference to your life. If you've heard about Kaggle and have already paid to cheat, you've chosen a data science-related field as your career path. The Kaggle platform is a unique treasure in the field of data science, and you won't find a similar platform in most other fields where you can learn cutting-edge knowledge and techniques as well as compete to prove your ability and practice what you've learned. Your actions are destroying the community, but also a part of your own career field. Your career journey hasn't even started yet, it's a long journey, take a longer view than just thinking about the immediate offer, the Kaggle community can help you for decades afterwards in your career, it's an asset to you and everyone in this field, please keep it healthy and don't ruin this for the sake of immediate petty gains.
············································································································································
\nSince this happened in the Chinese community, please allow me to repeat the above in Chinese for others to see
\n············································································································································
虽然没有参加这个比赛，但是今天得知此事后感到震惊和恶心。我去小红书上查找相关内容后发现这些违规的广告还肆无忌惮的在那里，而且有几十，几百个人在下面回复，来进一步了解价格。
\n这些肮脏的交易主要是针对中国的大学生，他们急于寻找各种手段提升自己的学术背景，以此在申请海外硕士项目时增加拿到名校offer的机会。Kaggle的奖牌自然是很好的选择。
我在两三年前经历了和这些大学生同样的事，申请硕士项目，通过课外研究和证书提升学生背景，我是跨专业申请的MSc AI项目，我了解过或者听说身边一起准备出国的朋友参加了各种不同的项目来提升AI或者数据科学的学术背景，但是那时我还从来没有听说过这种肮脏的交易，付费购买Kaggle的银牌。
\n如今，这项交易已经初具规模。如果仔细查看排行榜上一些非常可疑的账号，有些不止这一次可疑的参赛经历。如果去查看他们以前比赛的队友，他们甚至也符合 @forcewithme 在上面的描述，非常可疑，所以这些肮脏交易不止是从这次比赛开始的，之前的比赛也有，有些人会多次购买奖牌，只不过之前规模不大，没有被发现。
\n仅从小红书一个平台上的相关广告数量和回复的数量来看，至少已经有几百个人咨询了解过相关的报价了。从这次比赛可疑的账号和这些账号以前比赛的队友来看，至少有几十人已经付费购买奖牌了。
我想强调的是，仅仅删除掉这次比赛的作弊者是不够的，一定要想办法建立一个审查机制，能长期有效的阻止这样的交易。这才只是刚刚开始，两年前我在申请的时候从来没有听说过这事；现在，我估计有不到10个人在提供这项交易，这一次比赛有几十人付费参与，第一次被大家发现和曝光；如果不加以制止，几年后，这项恶心的交易流行起来，被越来越多的大学生听说后，相信我，仅仅中国每年相关专业的学生数量，哪怕只有很小比例的人选择付费，也足以毁掉Kaggle竞赛奖牌的含金量和Kaggle这个社区。
\n我想要保留这条帖子的中文版本，写一些话给这次比赛的作弊者和未来想要付费作弊的学生，希望你们能看到后停止这样的行为。
\nKaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹，或许你现在认为能申请到某个名校是最重要的事情，你会不惜一切手段实现它，即使是这种肮脏的手段。但当你真正完成这个学位之后你会发现，你当初看重的学校排名根本没那么重要，一个Kaggle奖牌能帮你得到的offer差距大概率不会有实质影响。如果你了解到了Kaggle并且已经选择了付费作弊，说明你已经选择了数据科学相关的领域作为自己的职业道路。Kaggle平台是数据科学这个领域独有的财富，在其他大部分领域你无法找到一个类似的平台，既能学习到最前沿的知识和技术，又能通过参赛证明自己的能力和实践自己的所学。你的行为在毁掉这个社区，同时也在毁掉你自己的职业领域的一部分。你的职业道路还没有开始，还很长，把眼光放长远一些，不只想着眼前的offer，Kaggle社区能在你之后几十年的职业生涯中都帮助到你，它是你和这个领域里所有人的财富，请保持它的健康发展，不要为了眼前的蝇头小利毁了这些。
@ryanholbrook @wcukierski Please take this seriously, this is just the beginning, there is still a huge potential market in China alone and many people are still unknown for the moment, once it becomes widespread, all the competitions will be greatly affected!
", "rawMarkdown": "@ryanholbrook @wcukierski Please take this seriously, this is just the beginning, there is still a huge potential market in China alone and many people are still unknown for the moment, once it becomes widespread, all the competitions will be greatly affected!", "votes": 7 }, { "id": 2032974, "postDate": "2022-11-17T00:08:09.233Z", "content": "Well, I saw two posts from @group16 two years ago, this kind of deal was on taobao.com two years ago, and now it's on Xiaohongshu, and peoples keep making such deals😔😔
\nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156
\nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635
How about Kaggle creates a cheating dataset and competition for everyone to contribute cheat detection models?
\nSome cheaters can be easily identified simply by looking at a few features of their profile.
@chenlin1999 Firstly, thank you for the information you have collected, most of us (myself included) struggle to find evidence for this issue because of the language barrier.
\n\n\nThe point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term.
\n
I agree with you because it's too easy to make a new Kaggle account. All you need is one email and phone number. For example, to make a gambling account, you need to provide some form of ID to prove your age. So if you do something illegal on it, the same person will not be able to make a new account because that ID will be on the blacklist database. However on Kaggle, a cheater can just comeback to Kaggle with one email and phone number.
\nThe organizers have agreed to my idea so all I can hope is to be prioritized high by the Kaggle team so this kind of issue occurs less in future competitions.
", "rawMarkdown": "@chenlin1999 Firstly, thank you for the information you have collected, most of us (myself included) struggle to find evidence for this issue because of the language barrier.\n\n> The point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term.\n\nI agree with you because it's too easy to make a new Kaggle account. All you need is one email and phone number. For example, to make a gambling account, you need to provide some form of ID to prove your age. So if you do something illegal on it, the same person will not be able to make a new account because that ID will be on the blacklist database. However on Kaggle, a cheater can just comeback to Kaggle with one email and phone number.\n\nThe organizers have agreed to my idea so all I can hope is to be prioritized high by the Kaggle team so this kind of issue occurs less in future competitions.", "votes": 1 }, { "id": 2033214, "postDate": "2022-11-17T05:06:25.040Z", "content": "@chenlin1999 - you have made a good and thoughtful post with much to consider for all sides of this. And as you have discovered, it is not the first or most recent example. It has been noted by many recently the tactics by any means to get medals to get rankings then announce to get congratulated and whether these medals have any value except perhaps in competitions. So one would like/hope competitions medals could at least have meaning for those involved.
\nHow much value they have for offers of study/work may be hard to qualify. In part, if universities or organisations do not have rigorous applications and interviews to access if an applicant has done their own work, has the necessary skills for their programs or jobs, but just look at a medal that is inappropriate and reflects on them. There are certainly full fee paying university programs for overseas students that take students with or without medals purely for the money. So the problems are not just about cheating on Kaggle. Students are probably exploited in many ways, by many people. But as you have posted, even with degrees or medals, that is no guarantee of future jobs, money, success.
\nKaggle has always been pretty open and accessible and the quality of datasets and competitions are most valuable to those that come to the site. It would be unfortunate if hosts are turned off by this. But most people that know how the real world works know all kinds of things happen. Would not like to see it become so onerous to join or participate because of the actions of some.
\nAs for the medals, an important person for me used to say about them - \"that's nice, now with your medal and [insert price here] you can get a cup of coffee\"
", "rawMarkdown": "@chenlin1999 - you have made a good and thoughtful post with much to consider for all sides of this. And as you have discovered, it is not the first or most recent example. It has been noted by many recently the tactics by any means to get medals to get rankings then announce to get congratulated and whether these medals have any value except perhaps in competitions. So one would like/hope competitions medals could at least have meaning for those involved. \n\nHow much value they have for offers of study/work may be hard to qualify. In part, if universities or organisations do not have rigorous applications and interviews to access if an applicant has done their own work, has the necessary skills for their programs or jobs, but just look at a medal that is inappropriate and reflects on them. There are certainly full fee paying university programs for overseas students that take students with or without medals purely for the money. So the problems are not just about cheating on Kaggle. Students are probably exploited in many ways, by many people. But as you have posted, even with degrees or medals, that is no guarantee of future jobs, money, success. \n\nKaggle has always been pretty open and accessible and the quality of datasets and competitions are most valuable to those that come to the site. It would be unfortunate if hosts are turned off by this. But most people that know how the real world works know all kinds of things happen. Would not like to see it become so onerous to join or participate because of the actions of some. \n\nAs for the medals, an important person for me used to say about them - \"that's nice, now with your medal and [insert price here] you can get a cup of coffee\" " }, { "id": 2033786, "postDate": "2022-11-17T14:31:53.093Z", "content": "@something4kag I understand what you're trying to say about how we should consider medals, that we shouldn't be satisfied with the medals we receive and slow down or stop our learning because of them. But that's not a reason for us not to defend the value of Kaggle medals.
", "rawMarkdown": "@something4kag I understand what you're trying to say about how we should consider medals, that we shouldn't be satisfied with the medals we receive and slow down or stop our learning because of them. But that's not a reason for us not to defend the value of Kaggle medals.", "votes": 1 }, { "id": 2033790, "postDate": "2022-11-17T14:35:09.130Z", "content": "@kaito510 Thanks for your reply, it gave me some inspiration, maybe as a Chinese Kaggler there are some things that only we can do, I will post a separate one below.
", "rawMarkdown": "@kaito510 Thanks for your reply, it gave me some inspiration, maybe as a Chinese Kaggler there are some things that only we can do, I will post a separate one below.", "votes": 1 }, { "id": 2034142, "postDate": "2022-11-17T21:07:18.123Z", "content": "Thanks @chenlin1999 I am hoping your comprehensive investigation will have a positive impact for Kaggle's future. As a side note, a Kaggle account's creation timestamp can be found by hovering your cursor over the \"Joined x years ago\" part. This could be useful to gain one more small piece of evidence.
", "rawMarkdown": "Thanks @chenlin1999 I am hoping your comprehensive investigation will have a positive impact for Kaggle's future. As a side note, a Kaggle account's creation timestamp can be found by hovering your cursor over the \"Joined x years ago\" part. This could be useful to gain **one more small piece of evidence**.", "votes": 1 }, { "id": 2036605, "postDate": "2022-11-20T01:00:00.353Z", "content": "Agree!zsbd
", "rawMarkdown": "Agree!zsbd" }, { "id": 2046296, "postDate": "2022-11-28T04:55:13.367Z", "content": "\"Kaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹\"
\n就我知道的，作弊的人，最后混得还不差（不少”赛圈大佬“ 狗头.jpg）…只能说作弊也算是一种实力了
", "rawMarkdown": "\"Kaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹\" \n\n就我知道的，作弊的人，最后混得还不差（不少”赛圈大佬“ 狗头.jpg）...只能说作弊也算是一种实力了", "votes": 1 } ] }, { "id": 2030829, "postDate": "2022-11-15T16:39:48.547Z", "content": "Hi everyone,
\nPlease report any suspicions of cheating here: https://www.kaggle.com/contact#/competitions/cheating
\nWe'll conduct an investigation after the competition ends and remove any teams that have violated the rules of the competition.
", "rawMarkdown": "Hi everyone,\n\nPlease report any suspicions of cheating here: https://www.kaggle.com/contact#/competitions/cheating\n\nWe'll conduct an investigation after the competition ends and remove any teams that have violated the rules of the competition.", "votes": 17, "replies": [ { "id": 2031241, "postDate": "2022-11-16T00:42:36.833Z", "content": "But also, please keep reporting this publicly in order to at least put some pressure at Kaggle to improve their detection mechanisms!
", "rawMarkdown": "But also, please keep reporting this publicly in order to at least put some pressure at Kaggle to improve their detection mechanisms! ", "votes": 10 }, { "id": 2032356, "postDate": "2022-11-16T15:19:05.790Z", "content": "@group16 I agree the detection mechanism may need to improve but on the flipside, they need to make it harder to join competitions (or harder to make new accounts).
\nFinding the cheaters manually for this competition is the best thing we can do at this point but it is only putting a band-aid as a solution to the problem. The exact same cheaters will comeback in future competitions with a new account because they can easily make a new one with just one email and phone number.
", "rawMarkdown": "@group16 I agree the detection mechanism may need to improve but on the flipside, they need to make it harder to join competitions (or harder to make new accounts). \n\nFinding the cheaters manually for this competition is the best thing we can do at this point but it is **only putting a band-aid as a solution to the problem**. The exact same cheaters will comeback in future competitions with a new account because they can easily make a new one with just one email and phone number.", "votes": 2, "replies": [ { "id": 2032444, "postDate": "2022-11-16T16:01:52.300Z", "content": "@group16 @kaito510 I wanted to post to let you know we have heard you, agree, and are working to better detect and prevent this kind of activity. What we do may not always be publicly visible--and the nature of the issue requires that we're often required to act with discretion--but we hear you.
\nAs our tooling improves, we also will not hesitate to ban accounts which are discovered to have cheated in previous competitions. While we make every effort to catch the first infraction, it's equally important to send the message that competing unfairly jeopardizes your Kaggle achievements and professional reputation indefinitely.
", "rawMarkdown": "@group16 @kaito510 I wanted to post to let you know we have heard you, agree, and are working to better detect and prevent this kind of activity. What we do may not always be publicly visible--and the nature of the issue requires that we're often required to act with discretion--but we hear you.\n\nAs our tooling improves, we also will not hesitate to ban accounts which are discovered to have cheated in previous competitions. While we make every effort to catch the first infraction, it's equally important to send the message that competing unfairly jeopardizes your Kaggle achievements and professional reputation indefinitely.", "votes": 11 } ] }, { "id": 2032654, "postDate": "2022-11-16T18:21:18.623Z", "content": "@wcukierski nice to read this but what currently concerns me most is the fact that (according to the screenshot provided by ForcewithMe) there are already active businesses selling solutions for all the other 4 active Kaggle competitions (Feedback, Novozymes, G2net and OTTO). As only Feedback Prize is a coding competition we can expect very similar development at the end of all of them.
\nYou can hardly prevent businesses from offering some solutions but it is fully in your hands to limit the buyers' demand if you really consistently terminate the accounts of those who play false and limit their chance to come back.
", "rawMarkdown": "@wcukierski nice to read this but what currently concerns me most is the fact that (according to the screenshot provided by ForcewithMe) there are already active businesses selling solutions for all the other 4 active Kaggle competitions (Feedback, Novozymes, G2net and OTTO). As only Feedback Prize is a coding competition we can expect very similar development at the end of all of them.\n\nYou can hardly prevent businesses from offering some solutions but it is fully in your hands to limit the buyers' demand if you really consistently terminate the accounts of those who play false and limit their chance to come back.", "votes": 3 }, { "id": 2034137, "postDate": "2022-11-17T21:01:10.863Z", "content": "Thanks for your reply to confirm that my idea has been respected. But unfortunately I disagree that their “professional reputation is jeopardized” because we can currently make a Kaggle account with just one email and phone number (i.e. no personal ID).
\nSo the sad reality is that cheating on Kaggle is currently a free shot, if you get caught, one email and one phone number attached to the account is caught cheating and banned. The cheating “person” cannot be identified, so the same person can easily make a new account and cheat again. For example, the same \"person\" could be trying to buy medals on the other competitions mentioned by @blankaf. That is why I am pushing the idea of making it harder to create Kaggle accounts.
", "rawMarkdown": "Thanks for your reply to confirm that my idea has been respected. But unfortunately I disagree that their “professional reputation is jeopardized” because we can currently make a Kaggle account with just one email and phone number (i.e. no personal ID).\n\nSo the sad reality is that **cheating on Kaggle is currently a free shot**, if you get caught, one email and one phone number attached to the account is caught cheating and banned. The cheating “person” cannot be identified, so the same person can easily make a new account and cheat again. For example, the same \"person\" could be trying to buy medals on the other competitions mentioned by @blankaf. That is why I am pushing the idea of making it harder to create Kaggle accounts.", "votes": 1 }, { "id": 2039476, "postDate": "2022-11-22T08:39:05.783Z", "content": "Though I don't know who they are but I believe there are someone who have bought kaggle's medals more than once, I hope kaggle organizers can also find them. They are usually organized by a group og novices in the form of a team and usually stable in the silver medal area. They may be more than one organization sell there models so there may be more than one solution. Every teams which submitted similar solutions should be taken into consideration.
", "rawMarkdown": "Though I don't know who they are but I believe there are someone who have bought kaggle's medals more than once, I hope kaggle organizers can also find them. They are usually organized by a group og novices in the form of a team and usually stable in the silver medal area. They may be more than one organization sell there models so there may be more than one solution. Every teams which submitted similar solutions should be taken into consideration.", "votes": 1 } ] }, { "id": 2030929, "postDate": "2022-11-15T18:15:03.293Z", "content": "There are more than 10 groups are selling kaggle medals in xiaohongshu, WeChat, 1688.com and so on.
", "rawMarkdown": "There are more than 10 groups are selling kaggle medals in xiaohongshu, WeChat, 1688.com and so on.", "votes": 16, "replies": [ { "id": 2030939, "postDate": "2022-11-15T18:21:36.797Z", "content": "How much are they selling? I am curious why people are selling/buying a sliver medal. I think it is not that super helpful in job/school seeking? At least not worth the money.
", "rawMarkdown": "How much are they selling? I am curious why people are selling/buying a sliver medal. I think it is not that super helpful in job/school seeking? At least not worth the money.", "votes": 5 }, { "id": 2030940, "postDate": "2022-11-15T18:22:36.697Z", "content": "Please collect evidence my Chinese friends! Most of us don't speak Chinese.
", "rawMarkdown": "Please collect evidence my Chinese friends! Most of us don't speak Chinese.", "votes": 5 }, { "id": 2030948, "postDate": "2022-11-15T18:28:33.670Z", "content": "About thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard.
", "rawMarkdown": "About thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard.", "votes": 11 }, { "id": 2030956, "postDate": "2022-11-15T18:33:31.660Z", "content": "Hi, gzguevara. It's still difficult to find out cheaters even we know all sellers.
\nBut rank 20-40 are tooooooo crowded.
", "rawMarkdown": "Hi, gzguevara. It's still difficult to find out cheaters even we know all sellers.\n\nBut rank 20-40 are tooooooo crowded.", "votes": 7 }, { "id": 2030965, "postDate": "2022-11-15T18:39:39.240Z", "content": "There are several methods to check cheaters. Check correlation of your submission with bought results. Proportionally Subtract bought results from you submission and see by how much the LB scores drops. That's two I can think of. But for that we need the sold predictions.
", "rawMarkdown": "There are several methods to check cheaters. Check correlation of your submission with bought results. Proportionally Subtract bought results from you submission and see by how much the LB scores drops. That's two I can think of. But for that we need the sold predictions.", "votes": 5 }, { "id": 2031006, "postDate": "2022-11-15T19:13:37.477Z", "content": "The most difficult thing is that we can’t find the relationship between the sellers on social media and suspicious teams on Kaggle. I can only imagine we do it by tracking the IP. But there are privacy issues, and I don't think any of us have the authority to track and disclose those IP relationships. Kaggle can track the IP of those accounts. But still they can’t track the IP of those sellers on social media. So I think it’s really hard to get the evidence.
", "rawMarkdown": "The most difficult thing is that we can’t find the relationship between the sellers on social media and suspicious teams on Kaggle. I can only imagine we do it by tracking the IP. But there are privacy issues, and I don't think any of us have the authority to track and disclose those IP relationships. Kaggle can track the IP of those accounts. But still they can’t track the IP of those sellers on social media. So I think it’s really hard to get the evidence.", "votes": 3 }, { "id": 2031035, "postDate": "2022-11-15T19:31:09.923Z", "content": "not only silver part, bronze part 60-140 has the same thing happened ..
", "rawMarkdown": "not only silver part, bronze part 60-140 has the same thing happened ..", "votes": 3 }, { "id": 2032326, "postDate": "2022-11-16T15:00:19.273Z", "content": "@forcewithme Thanks for this important post. I think @gzguevara 's ideas are pretty good, if we find one cheater (e.g. the deleted gold-winning account or ones identified by @liji11 are probably good chances) the rest will become increasingly easier to find. But I have two more ideas.
\nMy 1st idea is too late this time round but, @ryanholbrook can you please make it harder to create several Kaggle accounts (or harder to join competitions)? Like how gambling accounts require you to send a form of ID to prove your age. These cheaters will comeback with a new account because currently, you can make an account to join a competition with just one phone number and one e-mail address, that is too easy for a website that hosts competitions which people take very seriously.
\"sudden rises\" up the public leaderboard near the deadline. Is there a database on Kaggle where we can find the daily changes in the public leaderboard? I know \"sudden rises\" can legally happen with a clever idea but this can narrow down the potential cheating accounts.
\n\nAbout thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard
\n
This is about one month salary in China!?
", "rawMarkdown": "@forcewithme Thanks for this important post. I think @gzguevara 's ideas are pretty good, if we find one cheater (e.g. the deleted gold-winning account or ones identified by @liji11 are probably good chances) the rest will become increasingly easier to find. But I have two more ideas.\n\n1. My 1st idea is too late this time round but, @ryanholbrook can you please **make it harder to create several Kaggle accounts (or harder to join competitions)?** Like how gambling accounts require you to send a form of ID to prove your age. These cheaters will comeback with a new account because currently, you can make an account to join a competition with just one phone number and one e-mail address, that is too easy for a website that hosts competitions which people take very seriously. \n\n2. \"sudden rises\" up the public leaderboard near the deadline. Is there a database on Kaggle where we can find the daily changes in the public leaderboard? I know \"sudden rises\" can legally happen with a clever idea but this can narrow down the potential cheating accounts.\n\n> About thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard\n\nThis is about one month salary in China!?", "votes": 2 }, { "id": 2034270, "postDate": "2022-11-18T01:58:52.740Z", "content": "This is about one month salary in China!?
\nIt depends. For IT workers living in Chinese big cities like Beijing, Shanghai etc, 300-500 dollars is in average about 10% of one month salary.
Rank 15-40 are to crowded.
", "rawMarkdown": "Rank 15-40 are to crowded.", "votes": 1 } ] }, { "id": 2036123, "postDate": "2022-11-19T13:41:41.233Z", "content": "I hope the cheaters will be also removed from the Kaggle platform.
", "rawMarkdown": "I hope the cheaters will be also removed from the Kaggle platform.", "votes": 11, "replies": [ { "id": 2036242, "postDate": "2022-11-19T15:55:18.293Z", "content": "I can't agree more but unfortunately, because we can make Kaggle accounts with just one phone number and email, these cheaters cannot be \"personally\" identified. Thus, cheating on Kaggle is currently a free shot, they will comeback again with a new account using a different email and phone number. Kaggle organizers have agreed to my point but I reeeally (x10) want to push this point and hope action will be taken before other competitions get effected like this.
", "rawMarkdown": "I can't agree more but unfortunately, because we can make Kaggle accounts with just one phone number and email, these cheaters cannot be \"personally\" identified. Thus, cheating on Kaggle is currently a free shot, **they will comeback again** with a new account using a different email and phone number. Kaggle organizers have agreed to my point but I reeeally (x10) want to push this point and hope action will be taken before other competitions get effected like this.", "votes": 3 }, { "id": 2036251, "postDate": "2022-11-19T16:04:26.667Z", "content": "ID verified accounts would have gone a long way in fighting cheating like this. I raised why kaggle should start ID verifying a while ago.
\nhttps://www.kaggle.com/discussions/product-feedback/300028
I agree, it lowers the quality of the platform
", "rawMarkdown": "I agree, it lowers the quality of the platform", "votes": 3 }, { "id": 2041070, "postDate": "2022-11-23T16:49:59.313Z", "content": "Upvoted your post @desalegngeb Thanks for sharing, I 100% agree, if Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!
", "rawMarkdown": "Upvoted your post @desalegngeb Thanks for sharing, I 100% agree, if Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!", "votes": 1 }, { "id": 2041279, "postDate": "2022-11-23T19:44:47.263Z", "content": "\n\nif Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!
\n
@kaito510 noted! We are currently and will continue to investigate what options make sense to keep the platform fair.
", "rawMarkdown": "> if Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!\n\n@kaito510 noted! We are currently and will continue to investigate what options make sense to keep the platform fair.", "votes": 2, "replies": [ { "id": 2113115, "postDate": "2023-01-24T04:55:35.997Z", "content": "@wcukierski, any updates on personal verification for Kaggle accounts?
", "rawMarkdown": "@wcukierski, any updates on personal verification for Kaggle accounts?" }, { "id": 2113879, "postDate": "2023-01-24T15:05:52.033Z", "content": "@kaito510 We've kicked off a project to investigate options and had some vendor conversations, but in the event we proceed with some form of verification, it's going to take time to implement (we have to get privacy, legal, engineering reviews, etc.)
", "rawMarkdown": "@kaito510 We've kicked off a project to investigate options and had some vendor conversations, but in the event we proceed with some form of verification, it's going to take time to implement (we have to get privacy, legal, engineering reviews, etc.)" }, { "id": 2115175, "postDate": "2023-01-25T14:39:11.077Z", "content": "Thank you a lot for your reply. All the best and hope everything goes well for Kaggle's future!
", "rawMarkdown": "Thank you a lot for your reply. All the best and hope everything goes well for Kaggle's future!" }, { "id": 2115413, "postDate": "2023-01-25T18:41:06.060Z", "content": "@wcukierski you have known about the severity of the problem for one year.
\nNow you finally (or allegedly, who knows?) as you say \"kicked off a project to investigate options\".
\nAt this speed Kaggle competitions will have lost the last remains of the former reputation before you start really implementing anything.
BTW, please, don't try to persuade anybody how you value users' privacy. Your compliance with GDPR is more than disputable, e.g. you automatically store non-essential cookies before asking users for any consent.
", "rawMarkdown": "@wcukierski you have known about the severity of the problem for one year.\n\nNow you finally (or allegedly, who knows?) as you say \"kicked off a project to investigate options\".\nAt this speed Kaggle competitions will have lost the last remains of the former reputation before you start really implementing anything.\n\nBTW, please, don't try to persuade anybody how you value users' privacy. Your compliance with GDPR is more than disputable, e.g. you automatically store non-essential cookies before asking users for any consent. " } ] }, { "id": 2041341, "postDate": "2022-11-23T21:00:24.257Z", "content": "@wcukierski better late than never.
\nBut the same problem appeared in the same monster size already 7 months ago in HappyWhale competition and despite all suggestions and promises nothing systemic has been done yet.
It feels strange to have the same ranking promotion @forcewithme
", "rawMarkdown": "\n\nIt feels strange to have the same ranking promotion @forcewithme ", "votes": 12, "replies": [ { "id": 2032277, "postDate": "2022-11-16T14:23:54.017Z", "content": "And three of the same result scores
", "rawMarkdown": "And three of the same result scores", "votes": 7 }, { "id": 2032307, "postDate": "2022-11-16T14:45:58.667Z", "content": "Totally same LB, Pb and shake👀
", "rawMarkdown": "Totally same LB, Pb and shake👀", "votes": 5 }, { "id": 2032349, "postDate": "2022-11-16T15:16:52.003Z", "content": "\"One_team\"… that's for sure.
", "rawMarkdown": "\"One_team\"... that's for sure.", "votes": 11 } ] }, { "id": 2031238, "postDate": "2022-11-16T00:40:48.107Z", "content": "Nicely spotted @forcewithme! Also on taobao:
\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=640669004957&toSite=main
\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=613430072293&toSite=main
And looks like they are not even trying to hide it…
", "rawMarkdown": "Nicely spotted @forcewithme! Also on taobao:\n\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=640669004957&toSite=main\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=613430072293&toSite=main\n\nAnd looks like they are not even trying to hide it...", "votes": 8 }, { "id": 2031001, "postDate": "2022-11-15T19:10:58.660Z", "content": "Becoming increasingly common for these competitions where you do not have to submit a notebook
", "rawMarkdown": "Becoming increasingly common for these competitions where you do not have to submit a notebook", "votes": 8, "replies": [ { "id": 2032345, "postDate": "2022-11-16T15:12:15.893Z", "content": "Yeah because it's too easy to make a Kaggle account. If they get caught, they won't care much because they can just use a different e-mail address and phone number to make a new Kaggle account.
", "rawMarkdown": "Yeah because it's too easy to make a Kaggle account. If they get caught, they won't care much because they can just use a different e-mail address and phone number to make a new Kaggle account." } ] }, { "id": 2030892, "postDate": "2022-11-15T17:27:57.360Z", "content": "I was noticing it!!! PLEAES @ryanholbrook put all effort into this investigation. Most of us spend the last 3 months for this competition :(
", "rawMarkdown": "I was noticing it!!! PLEAES @ryanholbrook put all effort into this investigation. Most of us spend the last 3 months for this competition :(", "votes": 8 }, { "id": 2039375, "postDate": "2022-11-22T06:02:09.407Z", "content": "Thaks for your accusation! I divided the data by batch just like Silogram, but I experienced a huge unlucky shake. Though I do have some things I didn't do well, now I think I found some reasons which didn't caused by myself. If I am incompetent, I have nothing to say, but if they make progress by cheating, I think it's unfair to everyone who spent so much time in this competition. I strongly support the organizers to investigate seriously and ensure the fairness of the game and the Kaggle community.
", "rawMarkdown": "Thaks for your accusation! I divided the data by batch just like Silogram, but I experienced a huge unlucky shake. Though I do have some things I didn't do well, now I think I found some reasons which didn't caused by myself. If I am incompetent, I have nothing to say, but if they make progress by cheating, I think it's unfair to everyone who spent so much time in this competition. I strongly support the organizers to investigate seriously and ensure the fairness of the game and the Kaggle community.", "votes": 5 }, { "id": 2032465, "postDate": "2022-11-16T16:15:53.817Z", "content": "Not feeling good to see the platform of kaggle being infiltrated with unfair means
", "rawMarkdown": "Not feeling good to see the platform of kaggle being infiltrated with unfair means ", "votes": 5 }, { "id": 2031243, "postDate": "2022-11-16T00:43:41.120Z", "content": "If the deleted account is related to this, they sold a gold solution this time.
", "rawMarkdown": "If the deleted account is related to this, they sold a gold solution this time.", "votes": 5 }, { "id": 2047608, "postDate": "2022-11-28T22:33:47.873Z", "content": "I am so glad that you point out this problem, and I strongly suggest Kaggle to implement anti-cheat mechanism as soon as possible.
\nBeyond that, there are also organizations that selling the actual Kaggle accounts which contains competition medals on social media as well. Anyone buy an account can edit their profile page to make it looks legit for job interviews or college applications. Currently, anyone can register a Kaggle account with an email address and a phone number to get access. There is basically no cost for cheating. Kaggle should consider adding additional verification process to make it difficult for cheating.
\nApart from that, I am grateful that Kaggle provides the free opportunities for me to practice the skills. I don't really understand why they weight medals higher than the actual skills. I wish Kaggle could soon fix those problems.
", "rawMarkdown": "I am so glad that you point out this problem, and I strongly suggest Kaggle to implement anti-cheat mechanism as soon as possible. \n\nBeyond that, there are also organizations that selling the actual Kaggle accounts which contains competition medals on social media as well. Anyone buy an account can edit their profile page to make it looks legit for job interviews or college applications. Currently, anyone can register a Kaggle account with an email address and a phone number to get access. There is basically no cost for cheating. Kaggle should consider adding additional verification process to make it difficult for cheating. \n\nApart from that, I am grateful that Kaggle provides the free opportunities for me to practice the skills. I don't really understand why they weight medals higher than the actual skills. I wish Kaggle could soon fix those problems.", "votes": 3, "replies": [ { "id": 2048712, "postDate": "2022-11-29T16:15:26.127Z", "rawMarkdown": "", "votes": -1, "isDeleted": true } ] }, { "id": 2041575, "postDate": "2022-11-24T04:08:49.793Z", "content": "My friend's account was blocked by mistake. Are there any approaches we can use to help him? I do agree that we need to work against to cheating cases, but we should not hurt innocent people.
", "rawMarkdown": "My friend's account was blocked by mistake. Are there any approaches we can use to help him? I do agree that we need to work against to cheating cases, but we should not hurt innocent people. ", "votes": 3, "replies": [ { "id": 2043264, "postDate": "2022-11-25T13:57:17.767Z", "content": "Yes, my friend is suffering from the same problem
", "rawMarkdown": "Yes, my friend is suffering from the same problem", "votes": 1 } ] }, { "id": 2039383, "postDate": "2022-11-22T06:24:44.117Z", "content": "I think organizers can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.@Daniel Burkhardt@Ryan Holbrook.
", "rawMarkdown": "I think organizers can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.@Daniel Burkhardt@Ryan Holbrook.", "votes": 3 }, { "id": 2032121, "postDate": "2022-11-16T12:50:35.833Z", "content": "As a top-167 team member I want to believe in deletion of 40+ teams, but it won't be so. Even if bought competitors added just a little random noise, the precisous similarity will be incredibly hard to measure. BTW, I believe that Kaggle GMs can solve this problem :0
", "rawMarkdown": "As a top-167 team member I want to believe in deletion of 40+ teams, but it won't be so. Even if bought competitors added just a little random noise, the precisous similarity will be incredibly hard to measure. BTW, I believe that Kaggle GMs can solve this problem :0 ", "votes": 3, "replies": [ { "id": 2051149, "postDate": "2022-12-01T08:49:34.353Z", "content": "Funnily, I've got only 5 places between me and medal
", "rawMarkdown": "Funnily, I've got only 5 places between me and medal", "votes": 1 }, { "id": 2052087, "postDate": "2022-12-01T22:43:37.747Z", "content": "Slightly ahead of you….. I'd rather there not be so many cheaters and to be further away from a medal than this…
", "rawMarkdown": "Slightly ahead of you..... I'd rather there not be so many cheaters and to be further away from a medal than this... ", "votes": 1 } ] }, { "id": 2030895, "postDate": "2022-11-15T17:28:52.247Z", "content": "@forcewithme Thank you soo much for posting this!!!
", "rawMarkdown": "@forcewithme Thank you soo much for posting this!!!", "votes": 3 }, { "id": 2038635, "postDate": "2022-11-21T14:32:36.660Z", "content": "6 days after competition end and the results are still preliminary. Is it the usual case? Or is it because the investigation is going on?
", "rawMarkdown": "6 days after competition end and the results are still preliminary. Is it the usual case? Or is it because the investigation is going on?", "votes": 4, "replies": [ { "id": 2038851, "postDate": "2022-11-21T17:21:09.583Z", "content": "@danielburkhardt can we please get an update on this?
", "rawMarkdown": "@danielburkhardt can we please get an update on this? ", "votes": 1 }, { "id": 2038888, "postDate": "2022-11-21T17:57:24.237Z", "content": "Hi @gzguevara and @artemfedorov, I wish I had an update for you. @ryanholbrook and his team are handling the investigation.
\nAs the competition host, I see the same information as you, but I think Kaggle has more sophisticated ways of diagnosing cheating. I'm sure there will be a report back at some point in the coming days.
\nIf there's anything else I can do, please let me know.
\nOtherwise I hope the competition was a good experience for you. We want to learn what we can do better, so if you have 10 minutes to fill out our survey (link), we would greatly appreciate it!
\nBest,
\nDaniel
Hi @artemfedorov,
\nWe are continuing the investigation and will finalize the leaderboard once that is complete. I can confirm at this time that there will be a number of teams removed from the competition as well as a number of accounts that will be banned.
\nWe want to make sure everyone's efforts are fairly rewarded, so we'll take as long as necessary to feel confident about the results. (I imagine it won't be too much longer, however.)
", "rawMarkdown": "Hi @artemfedorov,\n\nWe are continuing the investigation and will finalize the leaderboard once that is complete. I can confirm at this time that there will be a number of teams removed from the competition as well as a number of accounts that will be banned.\n\nWe want to make sure everyone's efforts are fairly rewarded, so we'll take as long as necessary to feel confident about the results. (I imagine it won't be too much longer, however.)", "votes": 15 }, { "id": 2038928, "postDate": "2022-11-21T18:52:10.503Z", "content": "Hi @danielburkhardt ,
\nI've filled out the survey already.
\nJust wanted to know if this is the usual case for all the kaggle competitions or something special is going on with this contest.
\nDaniel, I understand the question is for kaggle staff, not for you. Everything is perfectly ok, thank you again for hosting this competition.
Hi @ryanholbrook ,
\nDidn't see your comment when I was typing a reply to Daniel. Thank you for the update!
Hi @ryanholbrook @wcukierski ,
\nThey are already advertising the OTTO competition and from the number of replies below the ad, there have been at least a couple of dozen people asking for quotes. So at the end of that competition, you probably have to do this again.
Thank you for reporting. We will keep our eyes on it.
", "rawMarkdown": "Thank you for reporting. We will keep our eyes on it.", "votes": 2 } ] }, { "id": 2039225, "postDate": "2022-11-22T02:29:19.107Z", "content": "\n\nyou probably have to do this again.
\n
@chenlin1999 I agree to what you are saying. But instead of \"doing it again\" like what myself, @uesugierii, @desalegngeb and many more are saying, it will save time, effort and anger for everyone in the future if we can solve the underlying problem of this \"cheating loop\":
\nEasy registration -> Buy high scoring solution -> Win a gold/silver medal if not caught cheating, just one email and phone number banned if caught cheating -> Repeat the loop with new email and phone number
\n@ryanholbrook please provide us some updates on my point, one of the staff have told me I have been \"heard\" but I am still yet to receive an answer to why Kaggle does not have a stricter registration/competition joining system.
", "rawMarkdown": "> you probably have to do this again.\n\n@chenlin1999 I agree to what you are saying. But instead of \"doing it again\" like what myself, @uesugierii, @desalegngeb and many more are saying, it will **save time, effort and anger for everyone in the future** if we can solve the underlying problem of this \"cheating loop\":\n\nEasy registration -> Buy high scoring solution -> Win a gold/silver medal if not caught cheating, just one email and phone number banned if caught cheating -> Repeat the loop with new email and phone number\n\n@ryanholbrook please provide us some updates on my point, one of the staff have told me **I have been \"heard\" but I am still yet to receive an answer** to why Kaggle does not have a stricter registration/competition joining system.", "votes": 1 }, { "id": 2039346, "postDate": "2022-11-22T05:21:10.273Z", "content": "Many thanks to Team kaggle for making great contributions to the fairness of the competition!! 👍👍👍
", "rawMarkdown": "Many thanks to Team kaggle for making great contributions to the fairness of the competition!! 👍👍👍" }, { "id": 2039350, "postDate": "2022-11-22T05:23:52.790Z", "content": "Hi
\nThey are already advertising the \"Feedback Award\" competition, hopefully the authorities will dig deeper!!🙏🙏
I think you can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.
", "rawMarkdown": "I think you can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them. ", "votes": 2 } ] }, { "id": 2031275, "postDate": "2022-11-16T01:26:33.277Z", "content": "The last several days ,the competition is very intense,many kaggle invoice(new account) come in,and Leaderboard is Rapid change these days，my team is about 30th+ -> 90th+,it's amazing and confused😟
", "rawMarkdown": "The last several days ,the competition is very intense,many kaggle invoice(new account) come in,and Leaderboard is Rapid change these days，my team is about 30th+ -> 90th+,it's amazing and confused😟", "votes": 4 }, { "id": 2047699, "postDate": "2022-11-29T01:04:40.383Z", "content": "![\"123\"](\"https://i.imgur.com/Ea0QKo6.jpg\")
@yuanzhezhou
\nindustry will damage the activity of Kaggle community.
\n深度之眼 only provides 199 yuan (about 27 dollar) Kaggle guided learning course, which involves the introduction of basic knowledge (for example, what is CNN. XGBoost. Transformer, etc.), and how they apply it to the competition. It does not provide any promise of any brand.
\nI don't know how you got the conclusion \". You may find some teams cooking (even top1 winner) from reading it.\".
\nDo you think 199 RMB (27 dollar) will buy any medal？, or do you have a conflict of interest with them？
\nAs far as I know, any student who gets a medal in 深度之眼 through his own efforts.
\nI have participated in many courses of 深度之眼 which have never provided any brand guarantee service.
\nIf you think this basic explanation course is cheating, please list your reasons. If it is reasonable, I will also help you maintain the Kaggle community. If not, please apologize.
\nTherefore, please read and understand the advertisement carefully before you guess. Do not use the greatest malice to guess innocent people, so as to avoid harm to others.
灰色产业会损害Kaggle社区活跃度。
\n深度之眼只提供199 RMB(27 dollar)的Kaggle带学课程，涉及的内容包括：基础知识的讲解（例如：什么是CNN. XGBoost. Transformer等），以及他们如何应用到这个比赛，并不会提供承诺任何牌子的事情。
\n我不知道你是如何得到“. You may find some teams cheating (even top1 winner) from reading it.”结论的，你觉得199 RMB(27 dollar)会买到任何奖牌的吗？，或者你是否跟他们存在利益冲突吗？
\n据我所知，任何在深度之眼拿到奖牌的同学都是通过自己努力得到牌子的。
\n我参加过很多期深度之眼的课程，从来没有保牌服务。
\n如果你觉得这种基础讲解课程是欺骗，请列举你的理由，如果是合理的，我也会帮你维护kaggle社区。如果不是，请道歉。
\n所以，请仔细阅读并理解广告再去猜测，不要用最大的恶意揣测无辜的人，避免有人受到伤害。
@yuanzhezhou
\nWhy did you delete the previous comment.？
\nCan you explain the other two teammates in Google AI4Code – Understand Code in Python Notebooks 13th
你为什么要删除之前的评论。
\n你能否解释Google AI4Code – Understand Code in Python Notebooks 13th中的另外两个队友的异常。
I delete that comment because it is up to Kaggle to decide if these participants have \"cheated\" or not, although they have indeed paid to \"get\" a medal.
\nI witnessed many obvious cheatings not punished by the host, so I decide to just leave the evidence here.
\nAbout my teammate in Google AI4Code – Understand Code in Python Notebooks 13th, they are more proficient than me in NLP and AI4code. I'm just a newbie who had never actively participated in any NLP competitions before.
\nchinese：
\n你不需要跟我争辩，你们做了什么，会导致什么后果，都是由kaggle来决定的。
I just observed that no one in the front area of the medal will take the novice without any record to participate in the competition, which has to be doubted.
\nBy the way, I'm just refuting your conclusion that there is no evidence, because it will really hurt many people.
\nAs you said that this is decided by Kaggle, you also said that some institutions are cheating.
我只是观察到，奖牌前区没有任何人会带没有任何记录的新手参加比赛，这不得不让人怀疑。
\n顺便说一句，我只是在反驳你没有任何证据的说辞，因为这真的会伤害很多人。
\n正如你自己一边说这是由kaggle决定的，一边又说某些机构拿牌的都是作弊。
For your concerns, there are many specialists who had never participated in a Kaggle competition before.
\nKaggle only bans a participant with \"more than sufficient\" evidence. So you do not need to worry about people getting falsely hurt.
\nForcewithMe raised a case where his places changed. I believe it's because some team's score is calculated late and appeared late.
\n我只见过明显作弊的人没有被取消成绩，还没见过几个没作弊的人被误封的……如果你担心买了课会被”误封“，那是不可能的，你只要没作弊就不需要担心。
", "rawMarkdown": "For your concerns, there are many specialists who had never participated in a Kaggle competition before. \n\nKaggle only bans a participant with \"more than sufficient\" evidence. So you do not need to worry about people getting falsely hurt. \n\nForcewithMe raised a case where his places changed. I believe it's because some team's score is calculated late and appeared late. \n\n我只见过明显作弊的人没有被取消成绩，还没见过几个没作弊的人被误封的......如果你担心买了课会被”误封“，那是不可能的，你只要没作弊就不需要担心。" }, { "id": 2048845, "postDate": "2022-11-29T18:11:41.633Z", "content": "I won't worry about being banned, because I will never cheat.
\nIn addition, I firmly believe that practical skills are the most important.
\nI'm just arguing about what you said or comments below that \"the \"深度之眼\" students who have won medals are bought\".
\nBecause this is really a groundless accusation.
我不会担心被封，因为我绝不作弊。
\n另外，我坚信实际的技能是最重要的。
\n我只是在争论你之前的言论“深度之眼”学员拿到奖牌的都是买的。
\n因为这真是无端的指责。
I think competitions should be fair and equally open to all kinds of people. If cheating will be stopped then we will have true winners and also all kinds of people will get motivated. So, this unethical practices prevailing inside it, needs to be banned throughout this competition. I hope everyone will agree with this point.
", "rawMarkdown": "**I think competitions should be fair and equally open to all kinds of people. If cheating will be stopped then we will have true winners and also all kinds of people will get motivated. So, this unethical practices prevailing inside it, needs to be banned throughout this competition. I hope everyone will agree with this point**.", "votes": 1 }, { "id": 2038891, "postDate": "2022-11-21T18:04:21.047Z", "content": "I hope these cheaters are removed from the competition. Maybe someone could make a competition about trying to detect cheating accounts.
", "rawMarkdown": "I hope these cheaters are removed from the competition. Maybe someone could make a competition about trying to detect cheating accounts.", "votes": 1 }, { "id": 2034309, "postDate": "2022-11-18T03:12:48.527Z", "content": "We saw a lot of good submissions in the last day or two on lb, but they didn't all have the same score. I don't think they just sell a csv with some noise (or no noise at all) to these people. Most likely, a notebook was sold, and the teams made their own improvements to the better-performing notebook. In this way, it will be more difficult for anti-cheating.
", "rawMarkdown": "We saw a lot of good submissions in the last day or two on lb, but they didn't all have the same score. I don't think they just sell a csv with some noise (or no noise at all) to these people. Most likely, a notebook was sold, and the teams made their own improvements to the better-performing notebook. In this way, it will be more difficult for anti-cheating.", "votes": 1 }, { "id": 2031879, "postDate": "2022-11-16T09:33:17.283Z", "content": "Nice, that means we all will get a free shakeup :-).
", "rawMarkdown": "Nice, that means we all will get a free shakeup :-).", "votes": 1 }, { "id": 2031231, "postDate": "2022-11-16T00:29:49.273Z", "content": "Saluate to every geek!
", "rawMarkdown": "Saluate to every geek!", "votes": 1 }, { "id": 2046293, "postDate": "2022-11-28T04:51:46.957Z", "content": "Check here for kaggle 1st place \"sold\"?
\nhttps://zhuanlan.zhihu.com/p/555516229?utm_source=wechat_session&utm_medium=social&utm_oi=1400297217975062528
There are many gold medal sold by these groups.
", "rawMarkdown": "Check here for kaggle 1st place \"sold\"?\nhttps://zhuanlan.zhihu.com/p/555516229?utm_source=wechat_session&utm_medium=social&utm_oi=1400297217975062528\n\nThere are many gold medal sold by these groups.\n", "votes": 2, "replies": [ { "id": 2046308, "postDate": "2022-11-28T05:19:15.770Z", "content": "It was deleted
", "rawMarkdown": "It was deleted" }, { "id": 2046311, "postDate": "2022-11-28T05:22:31.980Z", "content": "”我们可以看一下右图，kaggle第一名，有12,000美金。 下面是一个比赛的介绍，这是最近结束的一场比赛，由深度之眼组队我们讲解的kaggle肠胃分割比赛，第一名其实就是来自我们深度之眼的同学。“
\nsome people told them to delete it.
", "rawMarkdown": "”我们可以看一下右图，kaggle第一名，有12,000美金。 下面是一个比赛的介绍，这是最近结束的一场比赛，由深度之眼组队我们讲解的kaggle肠胃分割比赛，第一名其实就是来自我们深度之眼的同学。“\n\nsome people told them to delete it." }, { "id": 2047692, "postDate": "2022-11-29T00:52:22.580Z", "content": "@forcewithme check here for the backup: https://imgur.com/a/U73IF2w . They \"bought\" 1st place of a recently completed kaggle competition. BTW, I noticed that many \"student\" of X佬 got gold medals in kaggle competitions.
", "rawMarkdown": "@forcewithme check here for the backup: https://imgur.com/a/U73IF2w . They \"bought\" 1st place of a recently completed kaggle competition. BTW, I noticed that many \"student\" of X佬 got gold medals in kaggle competitions.\n\n", "votes": 1 }, { "id": 2048837, "postDate": "2022-11-29T17:52:47.370Z", "content": "How do you judge \"bought\"？
\nAfter reading the whole advertisement, I just found that they were emphasizing that the first one was the student of their basic courses, and they had won medals after three years of efforts.
\nI carefully observing the official website and official account of 深度之眼, They have never provided any brand guarantee service. so how did you come to this conclusion?
\nWhy are you trying to discredit 深度之眼? Do you have a certain purpose?
\nI can't find the information of BaoDianBang. You can provide any website or WeChat official account. If there is brand guarantee service, I can help you support.
你是如何判断出\"bought\"的。
\n阅读完整篇广告，我只是发现他们在强调第一名是他们基础课程的学员，并且经过3年的努力积累得到了奖牌，没有存在任何买奖牌的问题。
\n并且我仔细观察了深度之眼的官网和公众号，没有任何保牌服务，因此你是如何得到这个结论的？
\n为什么你在努力的抹黑深度之眼，你是有一定的目的性吗？
\n我没有找到 BaoDianBang的信息，你可以提供任何网址或者微信公众号。如果真是有保牌服务，我可以帮你声援。
\"如果真是有保牌服务，我可以帮你声援。\"
\nAs you have required, I tried and did find out some evidence about their(SHENDUZHIYAN's) behaviour against Kaggle's rule. If Kaggle requires it, I'm glad to provide it with Kaggle.
1) Please show it here and let people in the community judge.
\n2) Please provide more information about \"BaoDianBang\". not just SHENDUZHIYAN.
\n3) I still don't understand why you only said a period in a paragraph？
\nThis paragraph has two periods, but you only intercepted the first sentence and intentionally blackened it in the English expression. Do you have a certain purpose?
\nI am very serious to ask you to answer this question.
\nIt doesn't mean anything about \"bought\" , Are you deliberately smearing？
\nBecause your behavior makes Chinese lose face in this community. Please don't hide anything.
I just pasted the original words of the post you posted above, without any modification.
\nIf you are Chinese, please read it directly. I believe that a normal person will not get the same conclusion as @yuanzhezhou .
\nIf you are a foreigner, please use translation software to translate.
\n\"我们可以看一下右图，kaggle第一名，有12,000美金。 下面是一个比赛的介绍，这是最近结束的一场比赛，由深度之眼组队我们讲解的kaggle肠胃分割比赛，第一名其实就是来自我们深度之眼的同学。 得到第一名的同学其实在深度之眼已经参加了非常多的比赛了，从前年开始到今年不断学习积累打到了一第一名。
It seems that you are an endeavour (stayed up late) employee from SHENDUZHIYAN. I'm sorry but I'm not going to mess up with your SIBI（撕逼）。
\nI will answer in short with your own question:
\nWhy did you delete the ZHIHU POST within 10 minutes as soon as I found it？
\n你们为什么在我发现你们的知乎文章后的十分钟内就把它删掉？
What are you afraid of?
\n你们是在因为什么而感到害怕？
Q1：It seems that you are an endeavour (stayed up late) employee from SHENDUZHIYAN ？I'm sorry but I'm not going to mess up with your SIBI（撕逼）。
\nlol. no. It's really rumor making at zero cost（造谣零成本）. Why do you like to draw some unfounded conclusions so much？I just want your reply to the above questions. You've been avoiding talking about it. Why? and just said **Students who take the SHENDUZHIYAN‘s course cheat ** with no evidence . You must apologize, or you will pay a price. What do you think if I said that all the competitions you participated in were bought? For me, it's just one sentence. For you, you need to say dozens of sentences to explain.
Q2: Why did you delete the ZHIHU POST within 10 minutes as soon as I found it？ What are you afraid of?
\nYou can find a lot of the same ads in Baidu including ZHIHU or BillBill. Nobody deleted him. I suspect you acted on your own.（You've been avoiding questions., which makes me wonder. ） You first reported Zhihu, deleted the post, and then you saw that BillBill was not deleted, so you took a screenshot and created it. said “Students who take the SHENDUZHIYAN‘s course cheat“ . A normal Chinese with reading comprehension ability sees that your emphasis is two sentences, and you deliberately only say one sentence and spread rumors. And you avoid talking about other institutions, such as BaoDianBang. why??????
I agree to maintain Kaggle community and resist gray industry, because it is really a great platform.
\nHowever, if you feel that an organization has an illegal problem(e.g. SHENDUZHIYAN, BAODIANBANG and so on), please go to Chinese platforms(e.g. ZHIHU) to post or file a lawsuit. the heat will be higher. Instead of hiding here and distorting rumors to show foreigners.
Well, you avoid the fact that you delete the ZHIHU POST within 10 minutes as soon as I found it by denying it 😂.
\nYour are trying to distract people from the fact that the students did pay for a competition/competitions. No matter how much they paid, it is called \"paid/sold/bought\".
\nWhat do you think if I said that all the competitions you participated in were bought? For me, it's just one sentence. For you, you need to say dozens of sentences to explain.
\nPlease say it to anyone you know 😂. It will be great if you can even buy an advertisement for it.
However, if you feel that an organization has an illegal problem(e.g. SHENDUZHIYAN, BAODIANBANG and so on), please go to Chinese platforms(e.g. ZHIHU) to post or file a lawsuit. the heat will be higher. Instead of hiding here and distorting rumors to show foreigners.
\nSadly, SHENDUZHIYAN is \"legally\" violating Kaggle's rules. If you are trying to defend your company, it is pointless SIBI here. Everyone except you knows what kind of organization it is.
Q1: Well, you avoid the fact that you delete the ZHIHU POST within 10 minutes as soon as I found it by denying it 😂.
\nI don't know about this situation, because I see that the same ads on other platforms of SHENDUZHIYAN still exist.
\nIf there is a problem, it is impossible to just delete one.
Q2: Your are trying to distract people from the fact that the students did pay for a competition/competitions. No matter how much they paid, it is called \"paid/sold/bought\".
\nThey \"bought\" 1st place of a recently completed kaggle competition. means \"obtain in exchange for payment\". No advertising text is related to \"bought\" for brand guarantee service.
Q3: Please say it to anyone you know 😂. It will be great if you can even buy an advertisement for it.
\nYou are such a genius for playing rascals👎
Moreover, there are still some teams which definitely cheat in this competition. Although we reported them, it deos not work. Are there any approaches we can keep on reporting these cases? Thanks.
", "rawMarkdown": "Moreover, there are still some teams which definitely cheat in this competition. Although we reported them, it deos not work. Are there any approaches we can keep on reporting these cases? Thanks.", "votes": 2, "replies": [ { "id": 2041623, "postDate": "2022-11-24T04:59:44.387Z", "content": "Your best chance is to go to a Kaggle staff's profile page and directly send an e-mail with proper details about why a cheater has been undetected or why an innocent has been detected as a cheater.
", "rawMarkdown": "Your best chance is to go to a Kaggle staff's profile page and directly send an e-mail with proper details about why a cheater has been undetected or why an innocent has been detected as a cheater. ", "votes": 1 }, { "id": 2043239, "postDate": "2022-11-25T13:29:45.013Z", "content": "@kaito510 is right. Wish you and your friend good luck.
", "rawMarkdown": "@kaito510 is right. Wish you and your friend good luck. " } ] }, { "id": 2038616, "postDate": "2022-11-21T14:17:55.260Z", "content": "Yes, this kind of behavior is a lot in China. You can find them on 淘宝(Taobao), 闲鱼, 小红书 and other Chinese platforms. I think Kaggle needs to find ways to combat this industry chain, such as stricter registration process and stricter team merge conditions.
\nAs far as I know, a piece of silver is worth about 4000RMB (558.25USD).
Looking at those fake accounts I get the feeling that it is not undergraduates who buy results. It is very high-scoring teams which create accounts, get a few medal and then sell the accounts.
\nWhat is the most sad thing about that is that those high scores must have been sold from masters or grandmaster. They must come from teams which are in the top 10 :( Especially from those people we as a community expect the most integrity as we look up to them …
I don't know the business logic behind this, in fact with the profits they are making, they could have hired someone with the ability to get a silver medal themselves. If the quote given by @hydantess is for a guy to provide a solution for just a group of 5 people, he could earn about $2000 for that one competition, which is quite substantial as a part-time job outside of work. If he could provide a solution for about 5 teams of 25 people, that would be $10,000, and this capable person could even make this a full-time job, which would be better paid and more relaxed than working for a big Chinese tech company. So probably not from the top10 of this competition and for this grey industry, they won't be able to go and contact a different person every competition and expect to sell them a solution, I think they can handle the silver medal solution themselves.
", "rawMarkdown": "I don't know the business logic behind this, in fact with the profits they are making, they could have hired someone with the ability to get a silver medal themselves. If the quote given by @hydantess is for a guy to provide a solution for just a group of 5 people, he could earn about $2000 for that one competition, which is quite substantial as a part-time job outside of work. If he could provide a solution for about 5 teams of 25 people, that would be $10,000, and this capable person could even make this a full-time job, which would be better paid and more relaxed than working for a big Chinese tech company. So probably not from the top10 of this competition and for this grey industry, they won't be able to go and contact a different person every competition and expect to sell them a solution, I think they can handle the silver medal solution themselves.", "votes": 3 }, { "id": 2033714, "postDate": "2022-11-17T13:19:32.127Z", "content": "Yes, probably you are right. We/I shouldn’t speculate who is responsible for this. But those accounts are definitely for sell. They have fake pictures and look all the same. This does not look like somebody trying to profile himself. More like somebody creating a blanco check which is easy to personalise when sold.
", "rawMarkdown": "Yes, probably you are right. We/I shouldn’t speculate who is responsible for this. But those accounts are definitely for sell. They have fake pictures and look all the same. This does not look like somebody trying to profile himself. More like somebody creating a blanco check which is easy to personalise when sold.", "votes": 3 }, { "id": 2033719, "postDate": "2022-11-17T13:25:21.023Z", "content": "\n\nMore like somebody creating a blanco check which is easy to personalise when sold.
\n
Haha, it is like those fake \"AI\" players in video games nowadays. You think you have teammates? No, they are all fake.
", "rawMarkdown": "> More like somebody creating a blanco check which is easy to personalise when sold.\n\nHaha, it is like those fake \"AI\" players in video games nowadays. You think you have teammates? No, they are all fake.", "votes": 1 } ] }, { "id": 2032267, "postDate": "2022-11-16T14:19:01.193Z", "content": "It is very suspicious that the score of many teams has suddenly improved significantly in the last few days
\nI fell back 60 in last three days
Dear all! If you speak Chinese and are able to get those bought/fake predictions, which were sold, then please get them and send them to Kaggle! It will be easier for Kaggle to check who used them by checking correlations of submissions, i belive.
", "rawMarkdown": "Dear all! If you speak Chinese and are able to get those bought/fake predictions, which were sold, then please get them and send them to Kaggle! It will be easier for Kaggle to check who used them by checking correlations of submissions, i belive.", "votes": 2 }, { "id": 2030862, "postDate": "2022-11-15T16:59:07.107Z", "content": "17 lb account deleted. Don’t you think such a good scorer would also improve in the meanwhile?
\nI think top 17 subs may be compromised to. Mine are just speculation.
", "rawMarkdown": "17 lb account deleted. Don’t you think such a good scorer would also improve in the meanwhile?\n\nI think top 17 subs may be compromised to. Mine are just speculation.", "votes": 2, "replies": [ { "id": 2030865, "postDate": "2022-11-15T17:04:11.477Z", "content": "Or some teams close to the deleted accounts.
", "rawMarkdown": "Or some teams close to the deleted accounts.", "votes": 2 }, { "id": 2031038, "postDate": "2022-11-15T19:38:04.050Z", "rawMarkdown": "", "isDeleted": true }, { "id": 2031055, "postDate": "2022-11-15T19:57:00.110Z", "content": "but I still don't understand why that account is deleted, what is the meaning. And we all can see the most new teams above 0.814 are very close the deleted one, strange thing
", "rawMarkdown": "but I still don't understand why that account is deleted, what is the meaning. And we all can see the most new teams above 0.814 are very close the deleted one, strange thing ", "votes": 1 } ] }, { "id": 2030852, "postDate": "2022-11-15T16:49:33.830Z", "content": "The post in XiaoHongshu has been deleted. Here is the screenshot. 👋
\n![\"1.jpg\"](\"https://s2.loli.net/2022/11/16/ZMtuVbiFGkDJUe9.jpg\")
There were just 790 teams at the time of posting (proven by the screen), this is really funny. A lot of time to improve the model and reach gold spot 😅
", "rawMarkdown": "There were just 790 teams at the time of posting (proven by the screen), this is really funny. A lot of time to improve the model and reach gold spot 😅", "votes": 2 } ] }, { "id": 2038329, "postDate": "2022-11-21T10:31:14.500Z", "content": "oh, this is unbelievable,9-9
", "rawMarkdown": "oh, this is unbelievable,9-9" }, { "id": 2035688, "postDate": "2022-11-19T05:53:16.883Z", "content": "it's good to see someone speaking against cheatings
", "rawMarkdown": "it's good to see someone speaking against cheatings", "votes": 1 }, { "id": 2041893, "postDate": "2022-11-24T09:24:12.430Z", "content": "Damn, i thought this only happens in chinese's academic papers but on kaggle fr?
", "rawMarkdown": "Damn, i thought this only happens in chinese's academic papers but on kaggle fr? " }, { "id": 2044209, "postDate": "2022-11-26T11:29:29.163Z", "content": "Good one! This was helpful
", "rawMarkdown": "Good one! This was helpful", "votes": -2 }, { "id": 2043837, "postDate": "2022-11-26T01:54:11.030Z", "content": "wow, thank you for sharing. very helpful
", "rawMarkdown": "wow, thank you for sharing. very helpful", "votes": -1 }, { "id": 2043001, "postDate": "2022-11-25T07:21:27.907Z", "content": "This was great.
", "rawMarkdown": "This was great.", "votes": -1 }, { "id": 2038430, "postDate": "2022-11-21T11:27:17.260Z", "content": "in my opinion, we cannot blame Kaggle for this situation. Kaggle has to maintain a delicate balance between ease of access and security of competition. There are fraud companies all over the world, that cheat in such competitions having no professional ethics. their only goal is to make money. Such fraudsters should be identified and prosecuted.
", "rawMarkdown": "in my opinion, we cannot blame Kaggle for this situation. Kaggle has to maintain a delicate balance between ease of access and security of competition. There are fraud companies all over the world, that cheat in such competitions having no professional ethics. their only goal is to make money. Such fraudsters should be identified and prosecuted. ", "votes": -3 }, { "id": 2036914, "postDate": "2022-11-20T09:26:04.517Z", "content": "You are the best, I love you.
", "rawMarkdown": "You are the best, I love you.", "votes": -5 }, { "id": 2047149, "postDate": "2022-11-28T16:39:58.940Z", "content": "If it is true, this needs to be stopped.
", "rawMarkdown": "If it is true, this needs to be stopped. " }, { "id": 2039337, "postDate": "2022-11-22T05:19:09.267Z", "content": "Indeed, these people are really too hateful, if you don't have the strength please don't disturb the kaggle competition environment; kaggle should deal with these people strictly.🙏
", "rawMarkdown": "Indeed, these people are really too hateful, if you don't have the strength please don't disturb the kaggle competition environment; kaggle should deal with these people strictly.🙏" }, { "id": 2039302, "postDate": "2022-11-22T04:22:05.710Z", "content": "Thanks for speaking out. If you see something, say something.
", "rawMarkdown": "Thanks for speaking out. If you see something, say something." }, { "id": 2047463, "postDate": "2022-11-28T19:33:54.737Z", "rawMarkdown": "", "votes": 2, "isDeleted": true }, { "id": 2039772, "postDate": "2022-11-22T14:03:39.130Z", "rawMarkdown": "", "votes": -6, "isDeleted": true }, { "id": 2043835, "postDate": "2022-11-26T01:51:53.317Z", "content": "Thank you for your sharing
", "rawMarkdown": "Thank you for your sharing" } ], "comments": [ { "id": 2033961, "author_name": "Chen Lin", "author_url": "", "post_date": "2022-11-17T17:34:26.300000", "content": "Since this comment is mainly intended to be seen by more Chinese Kagglers, please allow me to use Chinese first and I will translate and explain in English below.
\n······························································································
\n我昨天在得知这件事和写那条长回复时，也觉得我们没什么能做的，只能交给官方去审核作弊，而且从本质上无法避免这种行为，他们完全可以进一步伪装，最终无法分辨。
\n但 @kaito510 的回复提醒了我，我们作为能看懂中文的群体，有些事或许是我们可以做的，而且只有我们可以做。
1。我们可以向淘宝官方的审核团队和小红书的审核团队进行沟通，或者让Kaggle官方出面和他们沟通，要求他们对这个灰色产业的广告进行审核。
\n这是有先例可循的，在淘宝创立初期，我们几乎可以在上面购买到任何不直接违反国家法律的服务，大量灰色产业在淘宝肆无忌惮，但是随着越来越多的审核机制，淘宝已经禁止了相当一部分灰色产业。比如你可以尝试搜索英语6级代考，你不会得到任何结果。对于Kaggle的奖牌我想也是可以并且应该被禁止公开售卖的。但这不会自动发生，需要有人主动与淘宝的审核团队沟通，Kaggle的官方不会想到这个方法，所以我在此提出。如果你们愿意作为官方出面，我想我和其他的中文Kaggler会愿意尽可能提供帮助。对于小红书和其他电子商务和社交媒体平台也同样适用，每个平台都有内容审核团队。 @ryanholbrook @wcukierski
\n2。我们可以对发现的售卖奖牌的广告进行举报。如果上述的审核内容成果被纳入各个平台愿意打击的内容，我想些信息会很快被清除掉。
\n当然这也无法从本质上完全避免这类交易，但是可以将他们从主要的电商和社交媒体平台删除，这将会有极大的帮助。我在准备留学的时候听说过太多离谱的这类交易了，甚至传说可以付高价直接买到托福、雅思、GRE等考试的成绩，我不知道是通过枪手代考还是内部修改成绩，但是因为这几个考试的主办方与国内平台沟通，这项交易被各个平台严厉打击，普通人很难获取到相关广告和报价，所以只有非常少量的交易进行，尽可能的保护了考试的公平性。
\n对于Kaggle奖牌交易，我想完全可以遵循以上先例。目前即使各平台不加以审核，公开宣传，也不是一个很大的产业，我猜提供这项服务的只有不到100人。如果能将他们的广告从平台删除，私下进行的交易量将会很小，甚至无法支撑这个产业形成能够持续运作的规模，从而可能解决大规模作弊的现象。
以上是我个人对这件事的一些看法和建议，从这个交易的源头下手，而不是在竞赛之后判断谁是作弊者然后删除成绩。我与 @forcewithme 私下沟通了这个想法，我们认为这可能是目前最好的办法了。
\n作为Kaggle社区中的中文群体，其他的国人做了这样的事，影响到了整个国际社区，我为此感到羞耻。而当想到了这个或许是只有作为中国人才能帮助到整个社区的方法，我认为我没有理由不提出来并承担这份责任。如果你作为一个中国人与我有同样的感觉并且赞同或愿意帮助实施这项方案，欢迎在下面评论，如果你有其他的建设性建议也请发出来。这件事关系到每一位在Kaggle平台认真竞赛的中文Kaggler和所有Kaggler，你一定不希望之后参加的比赛中越来越多的遇到这个比赛类似的事情。
\n如果你的账号涉及到个人信息隐私，请注意保护自己，我们面对的是一个大概几十人的灰色产业。但是相比认真竞赛的中文Kaggler和整个Kaggle社区，我们没什么好害怕的。
\n······························································································
\nWhen I was told about this and writing that long reply yesterday, I too thought there was not much we could do but leave it to the officials to check for cheating, and that there was no way to inherently avoid this kind of deal, and that they could have disguised it further and ultimately not been able to distinguish it.
But @kaito510 's reply reminded me that there are things we, as a group of people who can understand Chinese, might be able to do, and only we can do.
\n1。We can communicate with Taobao's official audit team and Xiaohongshu's audit team, or engage Kaggle officials to communicate with them and ask them to audit the ads in this grey industry.
\nThere are precedents for this. In the early years of Taobao, we could buy almost any service on it that did not directly violate national laws, and a large number of grey industries ran rampant on Taobao, but with more and more auditing mechanisms, Taobao has banned quite a number of them. For example, you can try searching for an English level 6 exams agency and you won't get any results. For Kaggle medals I think they can and should be banned from being sold publicly too. But this wouldn't happen automatically, someone would need to take the initiative to communicate with the Taobao audit team, and the Kaggle officials wouldn't think of this method, so I'm proposing it here. If you are willing to step in as an official, I think I and the other Chinese Kagglers would be willing to help as much as possible. The same applies to Xiaohongshu and other e-commerce and social media platforms, every one of which has a content audit team. @ryanholbrook @wcukierski
\n2。We can report advertisements we find that are selling medals. I think some information will be removed quickly if the above-mentioned audit content results are included in the content that each platform is willing to combat.
\nOf course, this wouldn't prevent these types of deals inherently and completely, but it would help immensely to remove them from the main e-commerce and social media platforms. I've heard of too many ridiculous deals of this kind when I was preparing to study abroad, and it's even rumoured that you can pay high prices to buy test scores directly for TOEFL, IELTS, GRE, etc. I don't know if it's through a ghostwriter or an internal revision of the scores, but because the organisers of these exams communicate with the domestic platforms, this deal is cracked down on by each platform and it's very difficult for ordinary people to access the relevant ads and quotes, so only a very small number of deals are conducted and the fairness of the exams is protected as much as possible.
\nFor Kaggle medal trading, I think it is perfectly possible to follow this precedent. At the moment, even if the platforms are not audited and are publicly advertised, it is not a very big industry, and I would guess that there are fewer than 100 people offering this service. If their ads could be removed from the platforms, the amount of trading that takes place privately would be too small to even support an industry on a scale that could sustainably operate and thus potentially address the massive cheating.
\nThe above are some of my personal thoughts and suggestions on this matter, start at the origin of this deal rather than judging who is the cheater after the competition and then removing the results. I communicated this idea privately with @forcewithme and we think this is probably the best way to proceed for now.
\nAs a group of Chinese in the Kaggle community, other Chinese have done something like this which has affected the international community as a whole, and I am ashamed of it. And when thinking of this as perhaps only being Chinese can help the whole community, I see no reason not to raise it and take this responsibility. If you, as a Chinese, feel the same way as I do and agree or would like to help implement this idea, please feel free to comment below, and if you have other constructive suggestions, please post them too. This matter concerns every Chinese Kaggler and all Kagglers who compete seriously on the Kaggle platform, and you certainly don't want to encounter more and more things like this competition in the competitions you participate in afterwards.
\nIf your account involves personal information privacy, please take care to protect yourself, we are facing a grey industry of about a few dozen people. But compared to the Chinese Kaggler who are serious about competing and the whole Kaggle community, we have nothing to fear.
", "votes": 32, "replies": [ { "id": 2033994, "author_name": "ForcewithMe", "author_url": "", "post_date": "2022-11-17T17:52:39.837000", "content": "I think this approach is more effective in limiting such behavior than detecting with technical approaches. @ryanholbrook @wcukierski Please think about this suggestion. 🙏
", "votes": 10, "replies": [] } ] }, { "id": 2032915, "author_name": "Chen Lin", "author_url": "", "post_date": "2022-11-16T21:49:42.910000", "content": "Although I did not participate in this competition, I was shocked and disgusted to hear about this today. I went to the Xiaohongshu platform to look up the content and found that these illegal ads are still there with impunity and there are dozens, hundreds of people replying below to seek further information on prices.
\nThese dirty deals are mainly targeted at Chinese undergraduates who are desperately looking for ways to improve their academic background and thereby increase their chances of getting an offer from a prestigious university when applying to overseas master's programs. The Kaggle medal obviously is a perfect choice.
\nTwo years ago, I experienced the same thing as these undergraduates, applying to master's programs and enhancing my academic background through extracurricular research and certificates. I applied to the MSc AI program as an interdisciplinary student, and I had learned or heard from friends around me who were preparing to study abroad together, they participated in various programs to enhance their academic background in AI or data science, but at that time I had never heard of this dirty deal of paying for Kaggle's silver medal.
\nToday, this deal is taking shape. If you look closely at some of the very dubious accounts on the leaderboard, some of them have had more than just this one dubious entry. If you look at their teammates from previous competitions, they even fit @forcewithme 's description above, which is very suspicious. So these dirty deals didn't just start at this competition, but also at previous ones, and some people would buy medals multiple times, they just weren't big enough to be discovered before.
\nJudging from the number of relevant ads and replies on just one platform, Xiaohongshu, at least a few hundred people have already inquired about the quotes. Judging from the suspicious accounts of this competition and the teammates of these accounts from previous competitions, at least a few dozen people have already paid for medals.
\nThe point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term. This is only the beginning, I had never heard of this when I applied two years ago; Now, I estimate that less than 10 people are offering these deals, and with dozens of people paying to participate in this one competition, it is being discovered and exposed by everyone for the first time; If it is not stopped, in a few years time, when this disgusting deal becomes popular and is heard of by more and more undergraduates, believe me, just the annual number of students in related majors in China, even if only a small percentage of them choose to pay, It is also enough to ruin the value of the Kaggle competition medals and the Kaggle community.
\nI would like to keep the Chinese version of this thread and write some words to the cheaters of this competition and to future students who want to pay to cheat, in the hope that you will see it and stop doing so. I don't want to just say something about the philosophy of life with honesty.
\nA Kaggle medal may help you get one more offer, but it will never completely change the trajectory of your life. Perhaps you now think that getting to apply to a certain top school is the most important thing and you will do whatever it takes to achieve it, even if it's this dirty trick. But once you've actually completed the degree, you'll realise that the ranking of the school you were so focused on is simply not that important. The different offers that a Kaggle medal can help you get, most likely won't make any difference to your life. If you've heard about Kaggle and have already paid to cheat, you've chosen a data science-related field as your career path. The Kaggle platform is a unique treasure in the field of data science, and you won't find a similar platform in most other fields where you can learn cutting-edge knowledge and techniques as well as compete to prove your ability and practice what you've learned. Your actions are destroying the community, but also a part of your own career field. Your career journey hasn't even started yet, it's a long journey, take a longer view than just thinking about the immediate offer, the Kaggle community can help you for decades afterwards in your career, it's an asset to you and everyone in this field, please keep it healthy and don't ruin this for the sake of immediate petty gains.
············································································································································
\nSince this happened in the Chinese community, please allow me to repeat the above in Chinese for others to see
\n············································································································································
虽然没有参加这个比赛，但是今天得知此事后感到震惊和恶心。我去小红书上查找相关内容后发现这些违规的广告还肆无忌惮的在那里，而且有几十，几百个人在下面回复，来进一步了解价格。
\n这些肮脏的交易主要是针对中国的大学生，他们急于寻找各种手段提升自己的学术背景，以此在申请海外硕士项目时增加拿到名校offer的机会。Kaggle的奖牌自然是很好的选择。
我在两三年前经历了和这些大学生同样的事，申请硕士项目，通过课外研究和证书提升学生背景，我是跨专业申请的MSc AI项目，我了解过或者听说身边一起准备出国的朋友参加了各种不同的项目来提升AI或者数据科学的学术背景，但是那时我还从来没有听说过这种肮脏的交易，付费购买Kaggle的银牌。
\n如今，这项交易已经初具规模。如果仔细查看排行榜上一些非常可疑的账号，有些不止这一次可疑的参赛经历。如果去查看他们以前比赛的队友，他们甚至也符合 @forcewithme 在上面的描述，非常可疑，所以这些肮脏交易不止是从这次比赛开始的，之前的比赛也有，有些人会多次购买奖牌，只不过之前规模不大，没有被发现。
\n仅从小红书一个平台上的相关广告数量和回复的数量来看，至少已经有几百个人咨询了解过相关的报价了。从这次比赛可疑的账号和这些账号以前比赛的队友来看，至少有几十人已经付费购买奖牌了。
我想强调的是，仅仅删除掉这次比赛的作弊者是不够的，一定要想办法建立一个审查机制，能长期有效的阻止这样的交易。这才只是刚刚开始，两年前我在申请的时候从来没有听说过这事；现在，我估计有不到10个人在提供这项交易，这一次比赛有几十人付费参与，第一次被大家发现和曝光；如果不加以制止，几年后，这项恶心的交易流行起来，被越来越多的大学生听说后，相信我，仅仅中国每年相关专业的学生数量，哪怕只有很小比例的人选择付费，也足以毁掉Kaggle竞赛奖牌的含金量和Kaggle这个社区。
\n我想要保留这条帖子的中文版本，写一些话给这次比赛的作弊者和未来想要付费作弊的学生，希望你们能看到后停止这样的行为。
\nKaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹，或许你现在认为能申请到某个名校是最重要的事情，你会不惜一切手段实现它，即使是这种肮脏的手段。但当你真正完成这个学位之后你会发现，你当初看重的学校排名根本没那么重要，一个Kaggle奖牌能帮你得到的offer差距大概率不会有实质影响。如果你了解到了Kaggle并且已经选择了付费作弊，说明你已经选择了数据科学相关的领域作为自己的职业道路。Kaggle平台是数据科学这个领域独有的财富，在其他大部分领域你无法找到一个类似的平台，既能学习到最前沿的知识和技术，又能通过参赛证明自己的能力和实践自己的所学。你的行为在毁掉这个社区，同时也在毁掉你自己的职业领域的一部分。你的职业道路还没有开始，还很长，把眼光放长远一些，不只想着眼前的offer，Kaggle社区能在你之后几十年的职业生涯中都帮助到你，它是你和这个领域里所有人的财富，请保持它的健康发展，不要为了眼前的蝇头小利毁了这些。
@ryanholbrook @wcukierski Please take this seriously, this is just the beginning, there is still a huge potential market in China alone and many people are still unknown for the moment, once it becomes widespread, all the competitions will be greatly affected!
", "votes": 7, "replies": [] }, { "id": 2032974, "author_name": "Chen Lin", "author_url": "", "post_date": "2022-11-17T00:08:09.233000", "content": "Well, I saw two posts from @group16 two years ago, this kind of deal was on taobao.com two years ago, and now it's on Xiaohongshu, and peoples keep making such deals😔😔
\nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/175156
\nhttps://www.kaggle.com/competitions/siim-isic-melanoma-classification/discussion/174635
How about Kaggle creates a cheating dataset and competition for everyone to contribute cheat detection models?
\nSome cheaters can be easily identified simply by looking at a few features of their profile.
@chenlin1999 Firstly, thank you for the information you have collected, most of us (myself included) struggle to find evidence for this issue because of the language barrier.
\n\n\nThe point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term.
\n
I agree with you because it's too easy to make a new Kaggle account. All you need is one email and phone number. For example, to make a gambling account, you need to provide some form of ID to prove your age. So if you do something illegal on it, the same person will not be able to make a new account because that ID will be on the blacklist database. However on Kaggle, a cheater can just comeback to Kaggle with one email and phone number.
\nThe organizers have agreed to my idea so all I can hope is to be prioritized high by the Kaggle team so this kind of issue occurs less in future competitions.
", "votes": 1, "replies": [] }, { "id": 2033214, "author_name": "something4kag", "author_url": "", "post_date": "2022-11-17T05:06:25.040000", "content": "@chenlin1999 - you have made a good and thoughtful post with much to consider for all sides of this. And as you have discovered, it is not the first or most recent example. It has been noted by many recently the tactics by any means to get medals to get rankings then announce to get congratulated and whether these medals have any value except perhaps in competitions. So one would like/hope competitions medals could at least have meaning for those involved.
\nHow much value they have for offers of study/work may be hard to qualify. In part, if universities or organisations do not have rigorous applications and interviews to access if an applicant has done their own work, has the necessary skills for their programs or jobs, but just look at a medal that is inappropriate and reflects on them. There are certainly full fee paying university programs for overseas students that take students with or without medals purely for the money. So the problems are not just about cheating on Kaggle. Students are probably exploited in many ways, by many people. But as you have posted, even with degrees or medals, that is no guarantee of future jobs, money, success.
\nKaggle has always been pretty open and accessible and the quality of datasets and competitions are most valuable to those that come to the site. It would be unfortunate if hosts are turned off by this. But most people that know how the real world works know all kinds of things happen. Would not like to see it become so onerous to join or participate because of the actions of some.
\nAs for the medals, an important person for me used to say about them - \"that's nice, now with your medal and [insert price here] you can get a cup of coffee\"
", "votes": 0, "replies": [] }, { "id": 2033786, "author_name": "Chen Lin", "author_url": "", "post_date": "2022-11-17T14:31:53.093000", "content": "@something4kag I understand what you're trying to say about how we should consider medals, that we shouldn't be satisfied with the medals we receive and slow down or stop our learning because of them. But that's not a reason for us not to defend the value of Kaggle medals.
", "votes": 1, "replies": [] }, { "id": 2033790, "author_name": "Chen Lin", "author_url": "", "post_date": "2022-11-17T14:35:09.130000", "content": "@kaito510 Thanks for your reply, it gave me some inspiration, maybe as a Chinese Kaggler there are some things that only we can do, I will post a separate one below.
", "votes": 1, "replies": [] }, { "id": 2034142, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-17T21:07:18.123000", "content": "Thanks @chenlin1999 I am hoping your comprehensive investigation will have a positive impact for Kaggle's future. As a side note, a Kaggle account's creation timestamp can be found by hovering your cursor over the \"Joined x years ago\" part. This could be useful to gain one more small piece of evidence.
", "votes": 1, "replies": [] }, { "id": 2036605, "author_name": " Zhang Qizhi", "author_url": "", "post_date": "2022-11-20T01:00:00.353000", "content": "Agree!zsbd
", "votes": 0, "replies": [] }, { "id": 2046296, "author_name": "yuanzhe zhou", "author_url": "", "post_date": "2022-11-28T04:55:13.367000", "content": "\"Kaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹\"
\n就我知道的，作弊的人，最后混得还不差（不少”赛圈大佬“ 狗头.jpg）…只能说作弊也算是一种实力了
", "votes": 1, "replies": [] } ] }, { "id": 2030829, "author_name": "Ryan Holbrook", "author_url": "", "post_date": "2022-11-15T16:39:48.547000", "content": "Hi everyone,
\nPlease report any suspicions of cheating here: https://www.kaggle.com/contact#/competitions/cheating
\nWe'll conduct an investigation after the competition ends and remove any teams that have violated the rules of the competition.
", "votes": 17, "replies": [ { "id": 2031241, "author_name": "Gilles Vandewiele", "author_url": "", "post_date": "2022-11-16T00:42:36.833000", "content": "But also, please keep reporting this publicly in order to at least put some pressure at Kaggle to improve their detection mechanisms!
", "votes": 10, "replies": [] }, { "id": 2032356, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-16T15:19:05.790000", "content": "@group16 I agree the detection mechanism may need to improve but on the flipside, they need to make it harder to join competitions (or harder to make new accounts).
\nFinding the cheaters manually for this competition is the best thing we can do at this point but it is only putting a band-aid as a solution to the problem. The exact same cheaters will comeback in future competitions with a new account because they can easily make a new one with just one email and phone number.
", "votes": 2, "replies": [ { "id": 2032444, "author_name": "Will Cukierski", "author_url": "", "post_date": "2022-11-16T16:01:52.300000", "content": "@group16 @kaito510 I wanted to post to let you know we have heard you, agree, and are working to better detect and prevent this kind of activity. What we do may not always be publicly visible--and the nature of the issue requires that we're often required to act with discretion--but we hear you.
\nAs our tooling improves, we also will not hesitate to ban accounts which are discovered to have cheated in previous competitions. While we make every effort to catch the first infraction, it's equally important to send the message that competing unfairly jeopardizes your Kaggle achievements and professional reputation indefinitely.
", "votes": 11, "replies": [] } ] }, { "id": 2032654, "author_name": "Allie K.", "author_url": "", "post_date": "2022-11-16T18:21:18.623000", "content": "@wcukierski nice to read this but what currently concerns me most is the fact that (according to the screenshot provided by ForcewithMe) there are already active businesses selling solutions for all the other 4 active Kaggle competitions (Feedback, Novozymes, G2net and OTTO). As only Feedback Prize is a coding competition we can expect very similar development at the end of all of them.
\nYou can hardly prevent businesses from offering some solutions but it is fully in your hands to limit the buyers' demand if you really consistently terminate the accounts of those who play false and limit their chance to come back.
", "votes": 3, "replies": [] }, { "id": 2034137, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-17T21:01:10.863000", "content": "Thanks for your reply to confirm that my idea has been respected. But unfortunately I disagree that their “professional reputation is jeopardized” because we can currently make a Kaggle account with just one email and phone number (i.e. no personal ID).
\nSo the sad reality is that cheating on Kaggle is currently a free shot, if you get caught, one email and one phone number attached to the account is caught cheating and banned. The cheating “person” cannot be identified, so the same person can easily make a new account and cheat again. For example, the same \"person\" could be trying to buy medals on the other competitions mentioned by @blankaf. That is why I am pushing the idea of making it harder to create Kaggle accounts.
", "votes": 1, "replies": [] }, { "id": 2039476, "author_name": "BarryZhou", "author_url": "", "post_date": "2022-11-22T08:39:05.783000", "content": "Though I don't know who they are but I believe there are someone who have bought kaggle's medals more than once, I hope kaggle organizers can also find them. They are usually organized by a group og novices in the form of a team and usually stable in the silver medal area. They may be more than one organization sell there models so there may be more than one solution. Every teams which submitted similar solutions should be taken into consideration.
", "votes": 1, "replies": [] } ] }, { "id": 2030929, "author_name": "hyd", "author_url": "", "post_date": "2022-11-15T18:15:03.293000", "content": "There are more than 10 groups are selling kaggle medals in xiaohongshu, WeChat, 1688.com and so on.
", "votes": 16, "replies": [ { "id": 2030939, "author_name": "Anthony Chiu", "author_url": "", "post_date": "2022-11-15T18:21:36.797000", "content": "How much are they selling? I am curious why people are selling/buying a sliver medal. I think it is not that super helpful in job/school seeking? At least not worth the money.
", "votes": 5, "replies": [] }, { "id": 2030940, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-15T18:22:36.697000", "content": "Please collect evidence my Chinese friends! Most of us don't speak Chinese.
", "votes": 5, "replies": [] }, { "id": 2030948, "author_name": "hyd", "author_url": "", "post_date": "2022-11-15T18:28:33.670000", "content": "About thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard.
", "votes": 11, "replies": [] }, { "id": 2030956, "author_name": "hyd", "author_url": "", "post_date": "2022-11-15T18:33:31.660000", "content": "Hi, gzguevara. It's still difficult to find out cheaters even we know all sellers.
\nBut rank 20-40 are tooooooo crowded.
", "votes": 7, "replies": [] }, { "id": 2030965, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-15T18:39:39.240000", "content": "There are several methods to check cheaters. Check correlation of your submission with bought results. Proportionally Subtract bought results from you submission and see by how much the LB scores drops. That's two I can think of. But for that we need the sold predictions.
", "votes": 5, "replies": [] }, { "id": 2031006, "author_name": "ForcewithMe", "author_url": "", "post_date": "2022-11-15T19:13:37.477000", "content": "The most difficult thing is that we can’t find the relationship between the sellers on social media and suspicious teams on Kaggle. I can only imagine we do it by tracking the IP. But there are privacy issues, and I don't think any of us have the authority to track and disclose those IP relationships. Kaggle can track the IP of those accounts. But still they can’t track the IP of those sellers on social media. So I think it’s really hard to get the evidence.
", "votes": 3, "replies": [] }, { "id": 2031035, "author_name": "no-magic", "author_url": "", "post_date": "2022-11-15T19:31:09.923000", "content": "not only silver part, bronze part 60-140 has the same thing happened ..
", "votes": 3, "replies": [] }, { "id": 2032326, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-16T15:00:19.273000", "content": "@forcewithme Thanks for this important post. I think @gzguevara 's ideas are pretty good, if we find one cheater (e.g. the deleted gold-winning account or ones identified by @liji11 are probably good chances) the rest will become increasingly easier to find. But I have two more ideas.
\nMy 1st idea is too late this time round but, @ryanholbrook can you please make it harder to create several Kaggle accounts (or harder to join competitions)? Like how gambling accounts require you to send a form of ID to prove your age. These cheaters will comeback with a new account because currently, you can make an account to join a competition with just one phone number and one e-mail address, that is too easy for a website that hosts competitions which people take very seriously.
\"sudden rises\" up the public leaderboard near the deadline. Is there a database on Kaggle where we can find the daily changes in the public leaderboard? I know \"sudden rises\" can legally happen with a clever idea but this can narrow down the potential cheating accounts.
\n\nAbout thousands of CNY(300-500 dollars). Most buyers are students who want to study aboard
\n
This is about one month salary in China!?
", "votes": 2, "replies": [] }, { "id": 2034270, "author_name": "HAO", "author_url": "", "post_date": "2022-11-18T01:58:52.740000", "content": "This is about one month salary in China!?
\nIt depends. For IT workers living in Chinese big cities like Beijing, Shanghai etc, 300-500 dollars is in average about 10% of one month salary.
Rank 15-40 are to crowded.
", "votes": 1, "replies": [] } ] }, { "id": 2036123, "author_name": "Gabriel Preda", "author_url": "", "post_date": "2022-11-19T13:41:41.233000", "content": "I hope the cheaters will be also removed from the Kaggle platform.
", "votes": 11, "replies": [ { "id": 2036242, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-19T15:55:18.293000", "content": "I can't agree more but unfortunately, because we can make Kaggle accounts with just one phone number and email, these cheaters cannot be \"personally\" identified. Thus, cheating on Kaggle is currently a free shot, they will comeback again with a new account using a different email and phone number. Kaggle organizers have agreed to my point but I reeeally (x10) want to push this point and hope action will be taken before other competitions get effected like this.
", "votes": 3, "replies": [] }, { "id": 2036251, "author_name": "A Legacy Grandmaster!", "author_url": "", "post_date": "2022-11-19T16:04:26.667000", "content": "ID verified accounts would have gone a long way in fighting cheating like this. I raised why kaggle should start ID verifying a while ago.
\nhttps://www.kaggle.com/discussions/product-feedback/300028
I agree, it lowers the quality of the platform
", "votes": 3, "replies": [] }, { "id": 2041070, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-23T16:49:59.313000", "content": "Upvoted your post @desalegngeb Thanks for sharing, I 100% agree, if Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!
", "votes": 1, "replies": [] }, { "id": 2041279, "author_name": "Will Cukierski", "author_url": "", "post_date": "2022-11-23T19:44:47.263000", "content": "\n\nif Coursera and edX can successfully implement personal verification, why can’t Kaggle do the same?!
\n
@kaito510 noted! We are currently and will continue to investigate what options make sense to keep the platform fair.
", "votes": 2, "replies": [ { "id": 2113115, "author_name": "goto_conversion", "author_url": "", "post_date": "2023-01-24T04:55:35.997000", "content": "@wcukierski, any updates on personal verification for Kaggle accounts?
", "votes": 0, "replies": [] }, { "id": 2113879, "author_name": "Will Cukierski", "author_url": "", "post_date": "2023-01-24T15:05:52.033000", "content": "@kaito510 We've kicked off a project to investigate options and had some vendor conversations, but in the event we proceed with some form of verification, it's going to take time to implement (we have to get privacy, legal, engineering reviews, etc.)
", "votes": 0, "replies": [] }, { "id": 2115175, "author_name": "goto_conversion", "author_url": "", "post_date": "2023-01-25T14:39:11.077000", "content": "Thank you a lot for your reply. All the best and hope everything goes well for Kaggle's future!
", "votes": 0, "replies": [] }, { "id": 2115413, "author_name": "Allie K.", "author_url": "", "post_date": "2023-01-25T18:41:06.060000", "content": "@wcukierski you have known about the severity of the problem for one year.
\nNow you finally (or allegedly, who knows?) as you say \"kicked off a project to investigate options\".
\nAt this speed Kaggle competitions will have lost the last remains of the former reputation before you start really implementing anything.
BTW, please, don't try to persuade anybody how you value users' privacy. Your compliance with GDPR is more than disputable, e.g. you automatically store non-essential cookies before asking users for any consent.
", "votes": 0, "replies": [] } ] }, { "id": 2041341, "author_name": "Allie K.", "author_url": "", "post_date": "2022-11-23T21:00:24.257000", "content": "@wcukierski better late than never.
\nBut the same problem appeared in the same monster size already 7 months ago in HappyWhale competition and despite all suggestions and promises nothing systemic has been done yet.
It feels strange to have the same ranking promotion @forcewithme
", "votes": 12, "replies": [ { "id": 2032277, "author_name": "蒹葭", "author_url": "", "post_date": "2022-11-16T14:23:54.017000", "content": "And three of the same result scores
", "votes": 7, "replies": [] }, { "id": 2032307, "author_name": "ForcewithMe", "author_url": "", "post_date": "2022-11-16T14:45:58.667000", "content": "Totally same LB, Pb and shake👀
", "votes": 5, "replies": [] }, { "id": 2032349, "author_name": "Alistair Martin Waddell", "author_url": "", "post_date": "2022-11-16T15:16:52.003000", "content": "\"One_team\"… that's for sure.
", "votes": 11, "replies": [] } ] }, { "id": 2031238, "author_name": "Gilles Vandewiele", "author_url": "", "post_date": "2022-11-16T00:40:48.107000", "content": "Nicely spotted @forcewithme! Also on taobao:
\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=640669004957&toSite=main
\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=613430072293&toSite=main
And looks like they are not even trying to hide it…
", "votes": 8, "replies": [] }, { "id": 2031001, "author_name": "JM", "author_url": "", "post_date": "2022-11-15T19:10:58.660000", "content": "Becoming increasingly common for these competitions where you do not have to submit a notebook
", "votes": 8, "replies": [ { "id": 2032345, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-16T15:12:15.893000", "content": "Yeah because it's too easy to make a Kaggle account. If they get caught, they won't care much because they can just use a different e-mail address and phone number to make a new Kaggle account.
", "votes": 0, "replies": [] } ] }, { "id": 2030892, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-15T17:27:57.360000", "content": "I was noticing it!!! PLEAES @ryanholbrook put all effort into this investigation. Most of us spend the last 3 months for this competition :(
", "votes": 8, "replies": [] }, { "id": 2039375, "author_name": "BarryZhou", "author_url": "", "post_date": "2022-11-22T06:02:09.407000", "content": "Thaks for your accusation! I divided the data by batch just like Silogram, but I experienced a huge unlucky shake. Though I do have some things I didn't do well, now I think I found some reasons which didn't caused by myself. If I am incompetent, I have nothing to say, but if they make progress by cheating, I think it's unfair to everyone who spent so much time in this competition. I strongly support the organizers to investigate seriously and ensure the fairness of the game and the Kaggle community.
", "votes": 5, "replies": [] }, { "id": 2032465, "author_name": "Shreya Mishra 0307", "author_url": "", "post_date": "2022-11-16T16:15:53.817000", "content": "Not feeling good to see the platform of kaggle being infiltrated with unfair means
", "votes": 5, "replies": [] }, { "id": 2031243, "author_name": "Anthony Chiu", "author_url": "", "post_date": "2022-11-16T00:43:41.120000", "content": "If the deleted account is related to this, they sold a gold solution this time.
", "votes": 5, "replies": [] }, { "id": 2047608, "author_name": "Jiahao Xu 99", "author_url": "", "post_date": "2022-11-28T22:33:47.873000", "content": "I am so glad that you point out this problem, and I strongly suggest Kaggle to implement anti-cheat mechanism as soon as possible.
\nBeyond that, there are also organizations that selling the actual Kaggle accounts which contains competition medals on social media as well. Anyone buy an account can edit their profile page to make it looks legit for job interviews or college applications. Currently, anyone can register a Kaggle account with an email address and a phone number to get access. There is basically no cost for cheating. Kaggle should consider adding additional verification process to make it difficult for cheating.
\nApart from that, I am grateful that Kaggle provides the free opportunities for me to practice the skills. I don't really understand why they weight medals higher than the actual skills. I wish Kaggle could soon fix those problems.
", "votes": 3, "replies": [ { "id": 2048712, "author_name": "", "author_url": "", "post_date": "2022-11-29T16:15:26.127000", "content": "", "votes": -1, "replies": [] } ] }, { "id": 2041575, "author_name": "TESUZI", "author_url": "", "post_date": "2022-11-24T04:08:49.793000", "content": "My friend's account was blocked by mistake. Are there any approaches we can use to help him? I do agree that we need to work against to cheating cases, but we should not hurt innocent people.
", "votes": 3, "replies": [ { "id": 2043264, "author_name": "StarxSky", "author_url": "", "post_date": "2022-11-25T13:57:17.767000", "content": "Yes, my friend is suffering from the same problem
", "votes": 1, "replies": [] } ] }, { "id": 2039383, "author_name": "BarryZhou", "author_url": "", "post_date": "2022-11-22T06:24:44.117000", "content": "I think organizers can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.@Daniel Burkhardt@Ryan Holbrook.
", "votes": 3, "replies": [] }, { "id": 2032121, "author_name": "Pizzaboi", "author_url": "", "post_date": "2022-11-16T12:50:35.833000", "content": "As a top-167 team member I want to believe in deletion of 40+ teams, but it won't be so. Even if bought competitors added just a little random noise, the precisous similarity will be incredibly hard to measure. BTW, I believe that Kaggle GMs can solve this problem :0
", "votes": 3, "replies": [ { "id": 2051149, "author_name": "Pizzaboi", "author_url": "", "post_date": "2022-12-01T08:49:34.353000", "content": "Funnily, I've got only 5 places between me and medal
", "votes": 1, "replies": [] }, { "id": 2052087, "author_name": "Markham Lee", "author_url": "", "post_date": "2022-12-01T22:43:37.747000", "content": "Slightly ahead of you….. I'd rather there not be so many cheaters and to be further away from a medal than this…
", "votes": 1, "replies": [] } ] }, { "id": 2030895, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-15T17:28:52.247000", "content": "@forcewithme Thank you soo much for posting this!!!
", "votes": 3, "replies": [] }, { "id": 2038635, "author_name": "Artem Fedorov", "author_url": "", "post_date": "2022-11-21T14:32:36.660000", "content": "6 days after competition end and the results are still preliminary. Is it the usual case? Or is it because the investigation is going on?
", "votes": 4, "replies": [ { "id": 2038851, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-21T17:21:09.583000", "content": "@danielburkhardt can we please get an update on this?
", "votes": 1, "replies": [] }, { "id": 2038888, "author_name": "Daniel Burkhardt", "author_url": "", "post_date": "2022-11-21T17:57:24.237000", "content": "Hi @gzguevara and @artemfedorov, I wish I had an update for you. @ryanholbrook and his team are handling the investigation.
\nAs the competition host, I see the same information as you, but I think Kaggle has more sophisticated ways of diagnosing cheating. I'm sure there will be a report back at some point in the coming days.
\nIf there's anything else I can do, please let me know.
\nOtherwise I hope the competition was a good experience for you. We want to learn what we can do better, so if you have 10 minutes to fill out our survey (link), we would greatly appreciate it!
\nBest,
\nDaniel
Hi @artemfedorov,
\nWe are continuing the investigation and will finalize the leaderboard once that is complete. I can confirm at this time that there will be a number of teams removed from the competition as well as a number of accounts that will be banned.
\nWe want to make sure everyone's efforts are fairly rewarded, so we'll take as long as necessary to feel confident about the results. (I imagine it won't be too much longer, however.)
", "votes": 15, "replies": [] }, { "id": 2038928, "author_name": "Artem Fedorov", "author_url": "", "post_date": "2022-11-21T18:52:10.503000", "content": "Hi @danielburkhardt ,
\nI've filled out the survey already.
\nJust wanted to know if this is the usual case for all the kaggle competitions or something special is going on with this contest.
\nDaniel, I understand the question is for kaggle staff, not for you. Everything is perfectly ok, thank you again for hosting this competition.
Hi @ryanholbrook ,
\nDidn't see your comment when I was typing a reply to Daniel. Thank you for the update!
Hi @ryanholbrook @wcukierski ,
\nThey are already advertising the OTTO competition and from the number of replies below the ad, there have been at least a couple of dozen people asking for quotes. So at the end of that competition, you probably have to do this again.
Thank you for reporting. We will keep our eyes on it.
", "votes": 2, "replies": [] } ] }, { "id": 2039225, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-22T02:29:19.107000", "content": "\n\nyou probably have to do this again.
\n
@chenlin1999 I agree to what you are saying. But instead of \"doing it again\" like what myself, @uesugierii, @desalegngeb and many more are saying, it will save time, effort and anger for everyone in the future if we can solve the underlying problem of this \"cheating loop\":
\nEasy registration -> Buy high scoring solution -> Win a gold/silver medal if not caught cheating, just one email and phone number banned if caught cheating -> Repeat the loop with new email and phone number
\n@ryanholbrook please provide us some updates on my point, one of the staff have told me I have been \"heard\" but I am still yet to receive an answer to why Kaggle does not have a stricter registration/competition joining system.
", "votes": 1, "replies": [] }, { "id": 2039346, "author_name": "StarxSky", "author_url": "", "post_date": "2022-11-22T05:21:10.273000", "content": "Many thanks to Team kaggle for making great contributions to the fairness of the competition!! 👍👍👍
", "votes": 0, "replies": [] }, { "id": 2039350, "author_name": "StarxSky", "author_url": "", "post_date": "2022-11-22T05:23:52.790000", "content": "Hi
\nThey are already advertising the \"Feedback Award\" competition, hopefully the authorities will dig deeper!!🙏🙏
I think you can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.
", "votes": 2, "replies": [] } ] }, { "id": 2031275, "author_name": "Kunming Xie", "author_url": "", "post_date": "2022-11-16T01:26:33.277000", "content": "The last several days ,the competition is very intense,many kaggle invoice(new account) come in,and Leaderboard is Rapid change these days，my team is about 30th+ -> 90th+,it's amazing and confused😟
", "votes": 4, "replies": [] }, { "id": 2047699, "author_name": "yuanzhe zhou", "author_url": "", "post_date": "2022-11-29T01:04:40.383000", "content": "
@yuanzhezhou
\nindustry will damage the activity of Kaggle community.
\n深度之眼 only provides 199 yuan (about 27 dollar) Kaggle guided learning course, which involves the introduction of basic knowledge (for example, what is CNN. XGBoost. Transformer, etc.), and how they apply it to the competition. It does not provide any promise of any brand.
\nI don't know how you got the conclusion \". You may find some teams cooking (even top1 winner) from reading it.\".
\nDo you think 199 RMB (27 dollar) will buy any medal？, or do you have a conflict of interest with them？
\nAs far as I know, any student who gets a medal in 深度之眼 through his own efforts.
\nI have participated in many courses of 深度之眼 which have never provided any brand guarantee service.
\nIf you think this basic explanation course is cheating, please list your reasons. If it is reasonable, I will also help you maintain the Kaggle community. If not, please apologize.
\nTherefore, please read and understand the advertisement carefully before you guess. Do not use the greatest malice to guess innocent people, so as to avoid harm to others.
灰色产业会损害Kaggle社区活跃度。
\n深度之眼只提供199 RMB(27 dollar)的Kaggle带学课程，涉及的内容包括：基础知识的讲解（例如：什么是CNN. XGBoost. Transformer等），以及他们如何应用到这个比赛，并不会提供承诺任何牌子的事情。
\n我不知道你是如何得到“. You may find some teams cheating (even top1 winner) from reading it.”结论的，你觉得199 RMB(27 dollar)会买到任何奖牌的吗？，或者你是否跟他们存在利益冲突吗？
\n据我所知，任何在深度之眼拿到奖牌的同学都是通过自己努力得到牌子的。
\n我参加过很多期深度之眼的课程，从来没有保牌服务。
\n如果你觉得这种基础讲解课程是欺骗，请列举你的理由，如果是合理的，我也会帮你维护kaggle社区。如果不是，请道歉。
\n所以，请仔细阅读并理解广告再去猜测，不要用最大的恶意揣测无辜的人，避免有人受到伤害。
@yuanzhezhou
\nWhy did you delete the previous comment.？
\nCan you explain the other two teammates in Google AI4Code – Understand Code in Python Notebooks 13th
你为什么要删除之前的评论。
\n你能否解释Google AI4Code – Understand Code in Python Notebooks 13th中的另外两个队友的异常。
I delete that comment because it is up to Kaggle to decide if these participants have \"cheated\" or not, although they have indeed paid to \"get\" a medal.
\nI witnessed many obvious cheatings not punished by the host, so I decide to just leave the evidence here.
\nAbout my teammate in Google AI4Code – Understand Code in Python Notebooks 13th, they are more proficient than me in NLP and AI4code. I'm just a newbie who had never actively participated in any NLP competitions before.
\nchinese：
\n你不需要跟我争辩，你们做了什么，会导致什么后果，都是由kaggle来决定的。
I just observed that no one in the front area of the medal will take the novice without any record to participate in the competition, which has to be doubted.
\nBy the way, I'm just refuting your conclusion that there is no evidence, because it will really hurt many people.
\nAs you said that this is decided by Kaggle, you also said that some institutions are cheating.
我只是观察到，奖牌前区没有任何人会带没有任何记录的新手参加比赛，这不得不让人怀疑。
\n顺便说一句，我只是在反驳你没有任何证据的说辞，因为这真的会伤害很多人。
\n正如你自己一边说这是由kaggle决定的，一边又说某些机构拿牌的都是作弊。
For your concerns, there are many specialists who had never participated in a Kaggle competition before.
\nKaggle only bans a participant with \"more than sufficient\" evidence. So you do not need to worry about people getting falsely hurt.
\nForcewithMe raised a case where his places changed. I believe it's because some team's score is calculated late and appeared late.
\n我只见过明显作弊的人没有被取消成绩，还没见过几个没作弊的人被误封的……如果你担心买了课会被”误封“，那是不可能的，你只要没作弊就不需要担心。
", "votes": 0, "replies": [] }, { "id": 2048845, "author_name": "rougamo", "author_url": "", "post_date": "2022-11-29T18:11:41.633000", "content": "I won't worry about being banned, because I will never cheat.
\nIn addition, I firmly believe that practical skills are the most important.
\nI'm just arguing about what you said or comments below that \"the \"深度之眼\" students who have won medals are bought\".
\nBecause this is really a groundless accusation.
我不会担心被封，因为我绝不作弊。
\n另外，我坚信实际的技能是最重要的。
\n我只是在争论你之前的言论“深度之眼”学员拿到奖牌的都是买的。
\n因为这真是无端的指责。
I think competitions should be fair and equally open to all kinds of people. If cheating will be stopped then we will have true winners and also all kinds of people will get motivated. So, this unethical practices prevailing inside it, needs to be banned throughout this competition. I hope everyone will agree with this point.
", "votes": 1, "replies": [] }, { "id": 2038891, "author_name": "pineapple", "author_url": "", "post_date": "2022-11-21T18:04:21.047000", "content": "I hope these cheaters are removed from the competition. Maybe someone could make a competition about trying to detect cheating accounts.
", "votes": 1, "replies": [] }, { "id": 2034309, "author_name": "ZhangYuan2000", "author_url": "", "post_date": "2022-11-18T03:12:48.527000", "content": "We saw a lot of good submissions in the last day or two on lb, but they didn't all have the same score. I don't think they just sell a csv with some noise (or no noise at all) to these people. Most likely, a notebook was sold, and the teams made their own improvements to the better-performing notebook. In this way, it will be more difficult for anti-cheating.
", "votes": 1, "replies": [] }, { "id": 2031879, "author_name": "Gabriel Preda", "author_url": "", "post_date": "2022-11-16T09:33:17.283000", "content": "Nice, that means we all will get a free shakeup :-).
", "votes": 1, "replies": [] }, { "id": 2031231, "author_name": "TESUZI", "author_url": "", "post_date": "2022-11-16T00:29:49.273000", "content": "Saluate to every geek!
", "votes": 1, "replies": [] }, { "id": 2046293, "author_name": "yuanzhe zhou", "author_url": "", "post_date": "2022-11-28T04:51:46.957000", "content": "Check here for kaggle 1st place \"sold\"?
\nhttps://zhuanlan.zhihu.com/p/555516229?utm_source=wechat_session&utm_medium=social&utm_oi=1400297217975062528
There are many gold medal sold by these groups.
", "votes": 2, "replies": [ { "id": 2046308, "author_name": "ForcewithMe", "author_url": "", "post_date": "2022-11-28T05:19:15.770000", "content": "It was deleted
", "votes": 0, "replies": [] }, { "id": 2046311, "author_name": "yuanzhe zhou", "author_url": "", "post_date": "2022-11-28T05:22:31.980000", "content": "”我们可以看一下右图，kaggle第一名，有12,000美金。 下面是一个比赛的介绍，这是最近结束的一场比赛，由深度之眼组队我们讲解的kaggle肠胃分割比赛，第一名其实就是来自我们深度之眼的同学。“
\nsome people told them to delete it.
", "votes": 0, "replies": [] }, { "id": 2047692, "author_name": "yuanzhe zhou", "author_url": "", "post_date": "2022-11-29T00:52:22.580000", "content": "@forcewithme check here for the backup: https://imgur.com/a/U73IF2w . They \"bought\" 1st place of a recently completed kaggle competition. BTW, I noticed that many \"student\" of X佬 got gold medals in kaggle competitions.
", "votes": 1, "replies": [] }, { "id": 2048837, "author_name": "rougamo", "author_url": "", "post_date": "2022-11-29T17:52:47.370000", "content": "How do you judge \"bought\"？
\nAfter reading the whole advertisement, I just found that they were emphasizing that the first one was the student of their basic courses, and they had won medals after three years of efforts.
\nI carefully observing the official website and official account of 深度之眼, They have never provided any brand guarantee service. so how did you come to this conclusion?
\nWhy are you trying to discredit 深度之眼? Do you have a certain purpose?
\nI can't find the information of BaoDianBang. You can provide any website or WeChat official account. If there is brand guarantee service, I can help you support.
你是如何判断出\"bought\"的。
\n阅读完整篇广告，我只是发现他们在强调第一名是他们基础课程的学员，并且经过3年的努力积累得到了奖牌，没有存在任何买奖牌的问题。
\n并且我仔细观察了深度之眼的官网和公众号，没有任何保牌服务，因此你是如何得到这个结论的？
\n为什么你在努力的抹黑深度之眼，你是有一定的目的性吗？
\n我没有找到 BaoDianBang的信息，你可以提供任何网址或者微信公众号。如果真是有保牌服务，我可以帮你声援。
\"如果真是有保牌服务，我可以帮你声援。\"
\nAs you have required, I tried and did find out some evidence about their(SHENDUZHIYAN's) behaviour against Kaggle's rule. If Kaggle requires it, I'm glad to provide it with Kaggle.
1) Please show it here and let people in the community judge.
\n2) Please provide more information about \"BaoDianBang\". not just SHENDUZHIYAN.
\n3) I still don't understand why you only said a period in a paragraph？
\nThis paragraph has two periods, but you only intercepted the first sentence and intentionally blackened it in the English expression. Do you have a certain purpose?
\nI am very serious to ask you to answer this question.
\nIt doesn't mean anything about \"bought\" , Are you deliberately smearing？
\nBecause your behavior makes Chinese lose face in this community. Please don't hide anything.
I just pasted the original words of the post you posted above, without any modification.
\nIf you are Chinese, please read it directly. I believe that a normal person will not get the same conclusion as @yuanzhezhou .
\nIf you are a foreigner, please use translation software to translate.
\n\"我们可以看一下右图，kaggle第一名，有12,000美金。 下面是一个比赛的介绍，这是最近结束的一场比赛，由深度之眼组队我们讲解的kaggle肠胃分割比赛，第一名其实就是来自我们深度之眼的同学。 得到第一名的同学其实在深度之眼已经参加了非常多的比赛了，从前年开始到今年不断学习积累打到了一第一名。
It seems that you are an endeavour (stayed up late) employee from SHENDUZHIYAN. I'm sorry but I'm not going to mess up with your SIBI（撕逼）。
\nI will answer in short with your own question:
\nWhy did you delete the ZHIHU POST within 10 minutes as soon as I found it？
\n你们为什么在我发现你们的知乎文章后的十分钟内就把它删掉？
What are you afraid of?
\n你们是在因为什么而感到害怕？
Q1：It seems that you are an endeavour (stayed up late) employee from SHENDUZHIYAN ？I'm sorry but I'm not going to mess up with your SIBI（撕逼）。
\nlol. no. It's really rumor making at zero cost（造谣零成本）. Why do you like to draw some unfounded conclusions so much？I just want your reply to the above questions. You've been avoiding talking about it. Why? and just said **Students who take the SHENDUZHIYAN‘s course cheat ** with no evidence . You must apologize, or you will pay a price. What do you think if I said that all the competitions you participated in were bought? For me, it's just one sentence. For you, you need to say dozens of sentences to explain.
Q2: Why did you delete the ZHIHU POST within 10 minutes as soon as I found it？ What are you afraid of?
\nYou can find a lot of the same ads in Baidu including ZHIHU or BillBill. Nobody deleted him. I suspect you acted on your own.（You've been avoiding questions., which makes me wonder. ） You first reported Zhihu, deleted the post, and then you saw that BillBill was not deleted, so you took a screenshot and created it. said “Students who take the SHENDUZHIYAN‘s course cheat“ . A normal Chinese with reading comprehension ability sees that your emphasis is two sentences, and you deliberately only say one sentence and spread rumors. And you avoid talking about other institutions, such as BaoDianBang. why??????
I agree to maintain Kaggle community and resist gray industry, because it is really a great platform.
\nHowever, if you feel that an organization has an illegal problem(e.g. SHENDUZHIYAN, BAODIANBANG and so on), please go to Chinese platforms(e.g. ZHIHU) to post or file a lawsuit. the heat will be higher. Instead of hiding here and distorting rumors to show foreigners.
Well, you avoid the fact that you delete the ZHIHU POST within 10 minutes as soon as I found it by denying it 😂.
\nYour are trying to distract people from the fact that the students did pay for a competition/competitions. No matter how much they paid, it is called \"paid/sold/bought\".
\nWhat do you think if I said that all the competitions you participated in were bought? For me, it's just one sentence. For you, you need to say dozens of sentences to explain.
\nPlease say it to anyone you know 😂. It will be great if you can even buy an advertisement for it.
However, if you feel that an organization has an illegal problem(e.g. SHENDUZHIYAN, BAODIANBANG and so on), please go to Chinese platforms(e.g. ZHIHU) to post or file a lawsuit. the heat will be higher. Instead of hiding here and distorting rumors to show foreigners.
\nSadly, SHENDUZHIYAN is \"legally\" violating Kaggle's rules. If you are trying to defend your company, it is pointless SIBI here. Everyone except you knows what kind of organization it is.
Q1: Well, you avoid the fact that you delete the ZHIHU POST within 10 minutes as soon as I found it by denying it 😂.
\nI don't know about this situation, because I see that the same ads on other platforms of SHENDUZHIYAN still exist.
\nIf there is a problem, it is impossible to just delete one.
Q2: Your are trying to distract people from the fact that the students did pay for a competition/competitions. No matter how much they paid, it is called \"paid/sold/bought\".
\nThey \"bought\" 1st place of a recently completed kaggle competition. means \"obtain in exchange for payment\". No advertising text is related to \"bought\" for brand guarantee service.
Q3: Please say it to anyone you know 😂. It will be great if you can even buy an advertisement for it.
\nYou are such a genius for playing rascals👎
Moreover, there are still some teams which definitely cheat in this competition. Although we reported them, it deos not work. Are there any approaches we can keep on reporting these cases? Thanks.
", "votes": 2, "replies": [ { "id": 2041623, "author_name": "goto_conversion", "author_url": "", "post_date": "2022-11-24T04:59:44.387000", "content": "Your best chance is to go to a Kaggle staff's profile page and directly send an e-mail with proper details about why a cheater has been undetected or why an innocent has been detected as a cheater.
", "votes": 1, "replies": [] }, { "id": 2043239, "author_name": "ForcewithMe", "author_url": "", "post_date": "2022-11-25T13:29:45.013000", "content": "@kaito510 is right. Wish you and your friend good luck.
", "votes": 0, "replies": [] } ] }, { "id": 2038616, "author_name": "Uesugi Erii", "author_url": "", "post_date": "2022-11-21T14:17:55.260000", "content": "Yes, this kind of behavior is a lot in China. You can find them on 淘宝(Taobao), 闲鱼, 小红书 and other Chinese platforms. I think Kaggle needs to find ways to combat this industry chain, such as stricter registration process and stricter team merge conditions.
\nAs far as I know, a piece of silver is worth about 4000RMB (558.25USD).
Looking at those fake accounts I get the feeling that it is not undergraduates who buy results. It is very high-scoring teams which create accounts, get a few medal and then sell the accounts.
\nWhat is the most sad thing about that is that those high scores must have been sold from masters or grandmaster. They must come from teams which are in the top 10 :( Especially from those people we as a community expect the most integrity as we look up to them …
I don't know the business logic behind this, in fact with the profits they are making, they could have hired someone with the ability to get a silver medal themselves. If the quote given by @hydantess is for a guy to provide a solution for just a group of 5 people, he could earn about $2000 for that one competition, which is quite substantial as a part-time job outside of work. If he could provide a solution for about 5 teams of 25 people, that would be $10,000, and this capable person could even make this a full-time job, which would be better paid and more relaxed than working for a big Chinese tech company. So probably not from the top10 of this competition and for this grey industry, they won't be able to go and contact a different person every competition and expect to sell them a solution, I think they can handle the silver medal solution themselves.
", "votes": 3, "replies": [] }, { "id": 2033714, "author_name": "gzguevara", "author_url": "", "post_date": "2022-11-17T13:19:32.127000", "content": "Yes, probably you are right. We/I shouldn’t speculate who is responsible for this. But those accounts are definitely for sell. They have fake pictures and look all the same. This does not look like somebody trying to profile himself. More like somebody creating a blanco check which is easy to personalise when sold.
", "votes": 3, "replies": [] }, { "id": 2033719, "author_name": "Anthony Chiu", "author_url": "", "post_date": "2022-11-17T13:25:21.023000", "content": "\n\nMore like somebody creating a blanco check which is easy to personalise when sold.
\n
Haha, it is like those fake \"AI\" players in video games nowadays. You think you have teammates? No, they are all fake.
", "votes": 1, "replies": [] } ] }, { "id": 2032267, "author_name": "", "author_url": "", "post_date": "2022-11-16T14:19:01.193000", "content": "", "votes": 2, "replies": [] }, { "id": 2030942, "author_name": "", "author_url": "", "post_date": "2022-11-15T18:24:10.693000", "content": "", "votes": 2, "replies": [] }, { "id": 2030862, "author_name": "", "author_url": "", "post_date": "2022-11-15T16:59:07.107000", "content": "", "votes": 2, "replies": [ { "id": 2030865, "author_name": "", "author_url": "", "post_date": "2022-11-15T17:04:11.477000", "content": "", "votes": 2, "replies": [] }, { "id": 2031038, "author_name": "", "author_url": "", "post_date": "2022-11-15T19:38:04.050000", "content": "", "votes": 0, "replies": [] }, { "id": 2031055, "author_name": "", "author_url": "", "post_date": "2022-11-15T19:57:00.110000", "content": "", "votes": 1, "replies": [] } ] }, { "id": 2030852, "author_name": "", "author_url": "", "post_date": "2022-11-15T16:49:33.830000", "content": "", "votes": 2, "replies": [ { "id": 2030864, "author_name": "", "author_url": "", "post_date": "2022-11-15T17:02:26.550000", "content": "", "votes": 2, "replies": [] } ] }, { "id": 2038329, "author_name": "", "author_url": "", "post_date": "2022-11-21T10:31:14.500000", "content": "", "votes": 0, "replies": [] }, { "id": 2035688, "author_name": "", "author_url": "", "post_date": "2022-11-19T05:53:16.883000", "content": "", "votes": 1, "replies": [] }, { "id": 2041893, "author_name": "", "author_url": "", "post_date": "2022-11-24T09:24:12.430000", "content": "", "votes": 0, "replies": [] }, { "id": 2044209, "author_name": "", "author_url": "", "post_date": "2022-11-26T11:29:29.163000", "content": "", "votes": -2, "replies": [] }, { "id": 2043837, "author_name": "", "author_url": "", "post_date": "2022-11-26T01:54:11.030000", "content": "", "votes": -1, "replies": [] }, { "id": 2043001, "author_name": "", "author_url": "", "post_date": "2022-11-25T07:21:27.907000", "content": "", "votes": -1, "replies": [] }, { "id": 2038430, "author_name": "", "author_url": "", "post_date": "2022-11-21T11:27:17.260000", "content": "", "votes": -3, "replies": [] }, { "id": 2036914, "author_name": "", "author_url": "", "post_date": "2022-11-20T09:26:04.517000", "content": "", "votes": -5, "replies": [] }, { "id": 2047149, "author_name": "", "author_url": "", "post_date": "2022-11-28T16:39:58.940000", "content": "", "votes": 0, "replies": [] }, { "id": 2039337, "author_name": "", "author_url": "", "post_date": "2022-11-22T05:19:09.267000", "content": "", "votes": 0, "replies": [] }, { "id": 2039302, "author_name": "", "author_url": "", "post_date": "2022-11-22T04:22:05.710000", "content": "", "votes": 0, "replies": [] }, { "id": 2047463, "author_name": "", "author_url": "", "post_date": "2022-11-28T19:33:54.737000", "content": "", "votes": 2, "replies": [] }, { "id": 2039772, "author_name": "", "author_url": "", "post_date": "2022-11-22T14:03:39.130000", "content": "", "votes": -6, "replies": [] }, { "id": 2043835, "author_name": "", "author_url": "", "post_date": "2022-11-26T01:51:53.317000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2030811": "I just found a post on XiaoHongShu(Chinese Instagram): [http://xhslink.com/S6Ttfl](http://xhslink.com/S6Ttfl)\nWhich means, if you paid, you will be guaranteed to get at least a silver medal, or even a gold medal.\nThis is an extension of a comment I made in [another post](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/366267). Some participants said his LB dropped from 60 to 110 within a single day. This new post is for everyone to better collect and organize evidence.\nAnother post for Feedback competition: [http://xhslink.com/cFdufl](http://xhslink.com/S6Ttfl)\n\n------------------------------------**Translate for the first link**-------------------------------------------\n**Title: Silver medal guaranteed! The latest edition of the Kaggle competition is hot for registration!**\n\n**Full text：**\n\n- Kaggle gold or silver. If we fail to take the silver medal, we will not charge you. The bronze medal is free.\n- Study Abroad, Background Enhancement, Changing careers to write code\n- Latest data mining competition: open-ended problem - multimodal single-cell integration\n- Enrollment ends: November 9\n- Result time: November 16\n\n-----------------------------------------**Suspecious patterns**---------------------------------------------\nMany of the team's accounts have very strangely similar behavior.\nFor example, I think most teams ranking 20~37 currently are suspecious, except team `chromosom` and `Shuji Suzuki `.\n\nThe only records that almost all of the accounts in these suspicious teams had on kaggle before were posting a meaningless comment in a random post. (e.g. Thank you! Wonderful idea!) I guess this will turn them from novice to contributor, so they won't be detected as easily!\n\nThere should be more suspicious teams on the whole leaderboard. But the `contributors teams` ranking 20~37 now are too wired.", "2033961": "Since this comment is mainly intended to be seen by more Chinese Kagglers, please allow me to use Chinese first and I will translate and explain in English below.\n······························································································\n我昨天在得知这件事和写那条长回复时，也觉得我们没什么能做的，只能交给官方去审核作弊，而且从本质上无法避免这种行为，他们完全可以进一步伪装，最终无法分辨。\n但 @kaito510 的回复提醒了我，我们作为能看懂中文的群体，有些事或许是我们可以做的，而且只有我们可以做。\n\n1。我们可以向淘宝官方的审核团队和小红书的审核团队进行沟通，或者让Kaggle官方出面和他们沟通，要求他们对这个灰色产业的广告进行审核。\n\n这是有先例可循的，在淘宝创立初期，我们几乎可以在上面购买到任何不直接违反国家法律的服务，大量灰色产业在淘宝肆无忌惮，但是随着越来越多的审核机制，淘宝已经禁止了相当一部分灰色产业。比如你可以尝试搜索英语6级代考，你不会得到任何结果。对于Kaggle的奖牌我想也是可以并且应该被禁止公开售卖的。但这不会自动发生，需要有人主动与淘宝的审核团队沟通，Kaggle的官方不会想到这个方法，所以我在此提出。如果你们愿意作为官方出面，我想我和其他的中文Kaggler会愿意尽可能提供帮助。对于小红书和其他电子商务和社交媒体平台也同样适用，每个平台都有内容审核团队。 @ryanholbrook @wcukierski\n\n2。我们可以对发现的售卖奖牌的广告进行举报。如果上述的审核内容成果被纳入各个平台愿意打击的内容，我想些信息会很快被清除掉。\n\n当然这也无法从本质上完全避免这类交易，但是可以将他们从主要的电商和社交媒体平台删除，这将会有极大的帮助。我在准备留学的时候听说过太多离谱的这类交易了，甚至传说可以付高价直接买到托福、雅思、GRE等考试的成绩，我不知道是通过枪手代考还是内部修改成绩，但是因为这几个考试的主办方与国内平台沟通，这项交易被各个平台严厉打击，普通人很难获取到相关广告和报价，所以只有非常少量的交易进行，尽可能的保护了考试的公平性。\n对于Kaggle奖牌交易，我想完全可以遵循以上先例。目前即使各平台不加以审核，公开宣传，也不是一个很大的产业，我猜提供这项服务的只有不到100人。如果能将他们的广告从平台删除，私下进行的交易量将会很小，甚至无法支撑这个产业形成能够持续运作的规模，从而可能解决大规模作弊的现象。\n\n以上是我个人对这件事的一些看法和建议，从这个交易的源头下手，而不是在竞赛之后判断谁是作弊者然后删除成绩。我与 @forcewithme 私下沟通了这个想法，我们认为这可能是目前最好的办法了。\n\n作为Kaggle社区中的中文群体，其他的国人做了这样的事，影响到了整个国际社区，我为此感到羞耻。而当想到了这个或许是只有作为中国人才能帮助到整个社区的方法，我认为我没有理由不提出来并承担这份责任。如果你作为一个中国人与我有同样的感觉并且赞同或愿意帮助实施这项方案，欢迎在下面评论，如果你有其他的建设性建议也请发出来。这件事关系到每一位在Kaggle平台认真竞赛的中文Kaggler和所有Kaggler，你一定不希望之后参加的比赛中越来越多的遇到这个比赛类似的事情。\n\n如果你的账号涉及到个人信息隐私，请注意保护自己，我们面对的是一个大概几十人的灰色产业。但是相比认真竞赛的中文Kaggler和整个Kaggle社区，我们没什么好害怕的。\n······························································································\nWhen I was told about this and writing that long reply yesterday, I too thought there was not much we could do but leave it to the officials to check for cheating, and that there was no way to inherently avoid this kind of deal, and that they could have disguised it further and ultimately not been able to distinguish it.\n\nBut @kaito510 's reply reminded me that there are things we, as a group of people who can understand Chinese, might be able to do, and only we can do.\n\n1。We can communicate with Taobao's official audit team and Xiaohongshu's audit team, or engage Kaggle officials to communicate with them and ask them to audit the ads in this grey industry.\n\nThere are precedents for this. In the early years of Taobao, we could buy almost any service on it that did not directly violate national laws, and a large number of grey industries ran rampant on Taobao, but with more and more auditing mechanisms, Taobao has banned quite a number of them. For example, you can try searching for an English level 6 exams agency and you won't get any results. For Kaggle medals I think they can and should be banned from being sold publicly too. But this wouldn't happen automatically, someone would need to take the initiative to communicate with the Taobao audit team, and the Kaggle officials wouldn't think of this method, so I'm proposing it here. If you are willing to step in as an official, I think I and the other Chinese Kagglers would be willing to help as much as possible. The same applies to Xiaohongshu and other e-commerce and social media platforms, every one of which has a content audit team. @ryanholbrook @wcukierski\n\n2。We can report advertisements we find that are selling medals. I think some information will be removed quickly if the above-mentioned audit content results are included in the content that each platform is willing to combat.\n\nOf course, this wouldn't prevent these types of deals inherently and completely, but it would help immensely to remove them from the main e-commerce and social media platforms. I've heard of too many ridiculous deals of this kind when I was preparing to study abroad, and it's even rumoured that you can pay high prices to buy test scores directly for TOEFL, IELTS, GRE, etc. I don't know if it's through a ghostwriter or an internal revision of the scores, but because the organisers of these exams communicate with the domestic platforms, this deal is cracked down on by each platform and it's very difficult for ordinary people to access the relevant ads and quotes, so only a very small number of deals are conducted and the fairness of the exams is protected as much as possible.\n\nFor Kaggle medal trading, I think it is perfectly possible to follow this precedent. At the moment, even if the platforms are not audited and are publicly advertised, it is not a very big industry, and I would guess that there are fewer than 100 people offering this service. If their ads could be removed from the platforms, the amount of trading that takes place privately would be too small to even support an industry on a scale that could sustainably operate and thus potentially address the massive cheating.\n\nThe above are some of my personal thoughts and suggestions on this matter, start at the origin of this deal rather than judging who is the cheater after the competition and then removing the results. I communicated this idea privately with @forcewithme and we think this is probably the best way to proceed for now.\n\nAs a group of Chinese in the Kaggle community, other Chinese have done something like this which has affected the international community as a whole, and I am ashamed of it. And when thinking of this as perhaps only being Chinese can help the whole community, I see no reason not to raise it and take this responsibility. If you, as a Chinese, feel the same way as I do and agree or would like to help implement this idea, please feel free to comment below, and if you have other constructive suggestions, please post them too. This matter concerns every Chinese Kaggler and all Kagglers who compete seriously on the Kaggle platform, and you certainly don't want to encounter more and more things like this competition in the competitions you participate in afterwards.\n\nIf your account involves personal information privacy, please take care to protect yourself, we are facing a grey industry of about a few dozen people. But compared to the Chinese Kaggler who are serious about competing and the whole Kaggle community, we have nothing to fear.", "2032915": "Although I did not participate in this competition, I was shocked and disgusted to hear about this today. I went to the Xiaohongshu platform to look up the content and found that these illegal ads are still there with impunity and there are dozens, hundreds of people replying below to seek further information on prices.\n\nThese dirty deals are mainly targeted at Chinese undergraduates who are desperately looking for ways to improve their academic background and thereby increase their chances of getting an offer from a prestigious university when applying to overseas master's programs. The Kaggle medal obviously is a perfect choice.\n\nTwo years ago, I experienced the same thing as these undergraduates, applying to master's programs and enhancing my academic background through extracurricular research and certificates. I applied to the MSc AI program as an interdisciplinary student, and I had learned or heard from friends around me who were preparing to study abroad together, they participated in various programs to enhance their academic background in AI or data science, but at that time I had never heard of this dirty deal of paying for Kaggle's silver medal.\n\nToday, this deal is taking shape. If you look closely at some of the very dubious accounts on the leaderboard, some of them have had more than just this one dubious entry. If you look at their teammates from previous competitions, they even fit @forcewithme 's description above, which is very suspicious. So these dirty deals didn't just start at this competition, but also at previous ones, and some people would buy medals multiple times, they just weren't big enough to be discovered before.\n\nJudging from the number of relevant ads and replies on just one platform, Xiaohongshu, at least a few hundred people have already inquired about the quotes. Judging from the suspicious accounts of this competition and the teammates of these accounts from previous competitions, at least a few dozen people have already paid for medals.\n\nThe point I want to emphasise is that it is not enough to just remove the cheaters from this competition, must find a way to create a censorship mechanism that can effectively stop deals like this in the long term. This is only the beginning, I had never heard of this when I applied two years ago; Now, I estimate that less than 10 people are offering these deals, and with dozens of people paying to participate in this one competition, it is being discovered and exposed by everyone for the first time; If it is not stopped, in a few years time, when this disgusting deal becomes popular and is heard of by more and more undergraduates, believe me, just the annual number of students in related majors in China, even if only a small percentage of them choose to pay, It is also enough to ruin the value of the Kaggle competition medals and the Kaggle community.\n\nI would like to keep the Chinese version of this thread and write some words to the cheaters of this competition and to future students who want to pay to cheat, in the hope that you will see it and stop doing so. I don't want to just say something about the philosophy of life with honesty.\nA Kaggle medal may help you get one more offer, but it will never completely change the trajectory of your life. Perhaps you now think that getting to apply to a certain top school is the most important thing and you will do whatever it takes to achieve it, even if it's this dirty trick. But once you've actually completed the degree, you'll realise that the ranking of the school you were so focused on is simply not that important. The different offers that a Kaggle medal can help you get, most likely won't make any difference to your life. If you've heard about Kaggle and have already paid to cheat, you've chosen a data science-related field as your career path. The Kaggle platform is a unique treasure in the field of data science, and you won't find a similar platform in most other fields where you can learn cutting-edge knowledge and techniques as well as compete to prove your ability and practice what you've learned. Your actions are destroying the community, but also a part of your own career field. Your career journey hasn't even started yet, it's a long journey, take a longer view than just thinking about the immediate offer, the Kaggle community can help you for decades afterwards in your career, it's an asset to you and everyone in this field, please keep it healthy and don't ruin this for the sake of immediate petty gains.\n\n············································································································································\nSince this happened in the Chinese community, please allow me to repeat the above in Chinese for others to see\n············································································································································\n\n虽然没有参加这个比赛，但是今天得知此事后感到震惊和恶心。我去小红书上查找相关内容后发现这些违规的广告还肆无忌惮的在那里，而且有几十，几百个人在下面回复，来进一步了解价格。\n这些肮脏的交易主要是针对中国的大学生，他们急于寻找各种手段提升自己的学术背景，以此在申请海外硕士项目时增加拿到名校offer的机会。Kaggle的奖牌自然是很好的选择。\n\n我在两三年前经历了和这些大学生同样的事，申请硕士项目，通过课外研究和证书提升学生背景，我是跨专业申请的MSc AI项目，我了解过或者听说身边一起准备出国的朋友参加了各种不同的项目来提升AI或者数据科学的学术背景，但是那时我还从来没有听说过这种肮脏的交易，付费购买Kaggle的银牌。\n如今，这项交易已经初具规模。如果仔细查看排行榜上一些非常可疑的账号，有些不止这一次可疑的参赛经历。如果去查看他们以前比赛的队友，他们甚至也符合 @forcewithme 在上面的描述，非常可疑，所以这些肮脏交易不止是从这次比赛开始的，之前的比赛也有，有些人会多次购买奖牌，只不过之前规模不大，没有被发现。\n仅从小红书一个平台上的相关广告数量和回复的数量来看，至少已经有几百个人咨询了解过相关的报价了。从这次比赛可疑的账号和这些账号以前比赛的队友来看，至少有几十人已经付费购买奖牌了。\n\n我想强调的是，仅仅删除掉这次比赛的作弊者是不够的，一定要想办法建立一个审查机制，能长期有效的阻止这样的交易。这才只是刚刚开始，两年前我在申请的时候从来没有听说过这事；现在，我估计有不到10个人在提供这项交易，这一次比赛有几十人付费参与，第一次被大家发现和曝光；如果不加以制止，几年后，这项恶心的交易流行起来，被越来越多的大学生听说后，相信我，仅仅中国每年相关专业的学生数量，哪怕只有很小比例的人选择付费，也足以毁掉Kaggle竞赛奖牌的含金量和Kaggle这个社区。\n\n我想要保留这条帖子的中文版本，写一些话给这次比赛的作弊者和未来想要付费作弊的学生，希望你们能看到后停止这样的行为。\n**Kaggle的奖牌或许可以多帮你拿到一个offer，但是它绝对无法彻底改变你的人生轨迹，或许你现在认为能申请到某个名校是最重要的事情，你会不惜一切手段实现它，即使是这种肮脏的手段。但当你真正完成这个学位之后你会发现，你当初看重的学校排名根本没那么重要，一个Kaggle奖牌能帮你得到的offer差距大概率不会有实质影响。如果你了解到了Kaggle并且已经选择了付费作弊，说明你已经选择了数据科学相关的领域作为自己的职业道路。Kaggle平台是数据科学这个领域独有的财富，在其他大部分领域你无法找到一个类似的平台，既能学习到最前沿的知识和技术，又能通过参赛证明自己的能力和实践自己的所学。你的行为在毁掉这个社区，同时也在毁掉你自己的职业领域的一部分。你的职业道路还没有开始，还很长，把眼光放长远一些，不只想着眼前的offer，Kaggle社区能在你之后几十年的职业生涯中都帮助到你，它是你和这个领域里所有人的财富，请保持它的健康发展，不要为了眼前的蝇头小利毁了这些。**", "2030829": "Hi everyone,\n\nPlease report any suspicions of cheating here: https://www.kaggle.com/contact#/competitions/cheating\n\nWe'll conduct an investigation after the competition ends and remove any teams that have violated the rules of the competition.", "2030929": "There are more than 10 groups are selling kaggle medals in xiaohongshu, WeChat, 1688.com and so on.", "2036123": "I hope the cheaters will be also removed from the Kaggle platform.", "2032274": "\n\nIt feels strange to have the same ranking promotion @forcewithme ", "2031238": "Nicely spotted @forcewithme! Also on taobao:\n\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=640669004957&toSite=main\nhttps://h5.m.taobao.com/awp/core/detail.htm?id=613430072293&toSite=main\n\nAnd looks like they are not even trying to hide it...", "2031001": "Becoming increasingly common for these competitions where you do not have to submit a notebook", "2030892": "I was noticing it!!! PLEAES @ryanholbrook put all effort into this investigation. Most of us spend the last 3 months for this competition :(", "2039375": "Thaks for your accusation! I divided the data by batch just like Silogram, but I experienced a huge unlucky shake. Though I do have some things I didn't do well, now I think I found some reasons which didn't caused by myself. If I am incompetent, I have nothing to say, but if they make progress by cheating, I think it's unfair to everyone who spent so much time in this competition. I strongly support the organizers to investigate seriously and ensure the fairness of the game and the Kaggle community.", "2032465": "Not feeling good to see the platform of kaggle being infiltrated with unfair means ", "2031243": "If the deleted account is related to this, they sold a gold solution this time.", "2047608": "I am so glad that you point out this problem, and I strongly suggest Kaggle to implement anti-cheat mechanism as soon as possible. \n\nBeyond that, there are also organizations that selling the actual Kaggle accounts which contains competition medals on social media as well. Anyone buy an account can edit their profile page to make it looks legit for job interviews or college applications. Currently, anyone can register a Kaggle account with an email address and a phone number to get access. There is basically no cost for cheating. Kaggle should consider adding additional verification process to make it difficult for cheating. \n\nApart from that, I am grateful that Kaggle provides the free opportunities for me to practice the skills. I don't really understand why they weight medals higher than the actual skills. I wish Kaggle could soon fix those problems.", "2041575": "My friend's account was blocked by mistake. Are there any approaches we can use to help him? I do agree that we need to work against to cheating cases, but we should not hurt innocent people. ", "2039383": "I think organizers can investigate from historical submission records. If thes same results are submitteed many times, then we have reason to think if they are cheating. After all, there are alse someone who use the final results with a little change or noise, we also can't ignore them.@Daniel Burkhardt@Ryan Holbrook.", "2032121": "As a top-167 team member I want to believe in deletion of 40+ teams, but it won't be so. Even if bought competitors added just a little random noise, the precisous similarity will be incredibly hard to measure. BTW, I believe that Kaggle GMs can solve this problem :0 ", "2030895": "@forcewithme Thank you soo much for posting this!!!", "2038635": "6 days after competition end and the results are still preliminary. Is it the usual case? Or is it because the investigation is going on?", "2031275": "The last several days ,the competition is very intense,many kaggle invoice(new account) come in,and Leaderboard is Rapid change these days，my team is about 30th+ -> 90th+,it's amazing and confused😟", "2047699": "", "2043053": "**I think competitions should be fair and equally open to all kinds of people. If cheating will be stopped then we will have true winners and also all kinds of people will get motivated. So, this unethical practices prevailing inside it, needs to be banned throughout this competition. I hope everyone will agree with this point**.", "2038891": "I hope these cheaters are removed from the competition. Maybe someone could make a competition about trying to detect cheating accounts.", "2034309": "We saw a lot of good submissions in the last day or two on lb, but they didn't all have the same score. I don't think they just sell a csv with some noise (or no noise at all) to these people. Most likely, a notebook was sold, and the teams made their own improvements to the better-performing notebook. In this way, it will be more difficult for anti-cheating.", "2031879": "Nice, that means we all will get a free shakeup :-).", "2031231": "Saluate to every geek!", "2046293": "Check here for kaggle 1st place \"sold\"?\nhttps://zhuanlan.zhihu.com/p/555516229?utm_source=wechat_session&utm_medium=social&utm_oi=1400297217975062528\n\nThere are many gold medal sold by these groups.\n", "2041576": "Moreover, there are still some teams which definitely cheat in this competition. Although we reported them, it deos not work. Are there any approaches we can keep on reporting these cases? Thanks.", "2038616": "Yes, this kind of behavior is a lot in China. You can find them on 淘宝(Taobao), 闲鱼, 小红书 and other Chinese platforms. I think Kaggle needs to find ways to combat this industry chain, such as stricter registration process and stricter team merge conditions.\nAs far as I know, a piece of silver is worth about 4000RMB (558.25USD).", "2033667": "Looking at those fake accounts I get the feeling that it is not undergraduates who buy results. It is very high-scoring teams which create accounts, get a few medal and then sell the accounts. \nWhat is the most sad thing about that is that those high scores must have been sold from masters or grandmaster. They must come from teams which are in the top 10 :( Especially from those people we as a community expect the most integrity as we look up to them …", "2032267": "It is very suspicious that the score of many teams has suddenly improved significantly in the last few days\nI fell back 60 in last three days", "2030942": "Dear all! If you speak Chinese and are able to get those bought/fake predictions, which were sold, then please get them and send them to Kaggle! It will be easier for Kaggle to check who used them by checking correlations of submissions, i belive.", "2030862": "17 lb account deleted. Don’t you think such a good scorer would also improve in the meanwhile?\n\nI think top 17 subs may be compromised to. Mine are just speculation.", "2030852": "The post in XiaoHongshu has been deleted. Here is the screenshot. 👋\n\n", "2038329": "oh, this is unbelievable,9-9", "2035688": "it's good to see someone speaking against cheatings", "2041893": "Damn, i thought this only happens in chinese's academic papers but on kaggle fr? ", "2044209": "Good one! This was helpful", "2043837": "wow, thank you for sharing. very helpful", "2043001": "This was great.", "2038430": "in my opinion, we cannot blame Kaggle for this situation. Kaggle has to maintain a delicate balance between ease of access and security of competition. There are fraud companies all over the world, that cheat in such competitions having no professional ethics. their only goal is to make money. Such fraudsters should be identified and prosecuted. ", "2036914": "You are the best, I love you.", "2047149": "If it is true, this needs to be stopped. ", "2039337": "Indeed, these people are really too hateful, if you don't have the strength please don't disturb the kaggle competition environment; kaggle should deal with these people strictly.🙏", "2039302": "Thanks for speaking out. If you see something, say something.", "2047463": "", "2039772": "", "2043835": "Thank you for your sharing" } }