{ "id": 335986, "title": "Two notebook medals for the same code", "url": "/competitions/amex-default-prediction/discussion/335986", "author_name": "HinePo", "post_date": "2022-07-08T19:42:23.745000", "votes": 34, "comment_count": 27, "views": 0, "content": "

Unfortunately the increasingly common practice in kaggle also happened to me.

\n

I published a notebook that 11 days later was used by another kaggler who simply removed some lines (first markdown cell, and last code cell (feature importances)) and republished it. It didn't add anything or any line to my work. Didn't even change hyperparameters. It achieves exactly the same CV (0.78542) on the very same epoch (15).

\n

I reported the notebook to discourage such practices on kaggle.

\n

notebook I published

\n

copied notebook

", "messages": [ { "id": 1848645, "postDate": "2022-07-08T19:42:23.747Z", "content": "

Unfortunately the increasingly common practice in kaggle also happened to me.

\n

I published a notebook that 11 days later was used by another kaggler who simply removed some lines (first markdown cell, and last code cell (feature importances)) and republished it. It didn't add anything or any line to my work. Didn't even change hyperparameters. It achieves exactly the same CV (0.78542) on the very same epoch (15).

\n

I reported the notebook to discourage such practices on kaggle.

\n

notebook I published

\n

copied notebook

", "rawMarkdown": "Unfortunately the increasingly common practice in kaggle also happened to me.\n\nI published a notebook that 11 days later was used by another kaggler who simply removed some lines (first markdown cell, and last code cell (feature importances)) and republished it. It didn't add anything or any line to my work. Didn't even change hyperparameters. It achieves exactly the same CV (0.78542) on the very same epoch (15).\n\nI reported the notebook to discourage such practices on kaggle.\n\n[notebook I published](https://www.kaggle.com/code/hinepo/amex-tabnet-training)\n\n[copied notebook](https://www.kaggle.com/code/dungdao706/amex-tabnet/notebook)", "votes": 32 }, { "id": 1851107, "postDate": "2022-07-11T03:14:21.503Z", "content": "

Hello to everyone, I am the one who copied the author's notebook here. I am truly sorry for my lack of knowledge and awareness.
\nI thought copying the notebook and sharing through clicking on Share and then selecting Public is the most convenient way of letting my group of friends access the notebook for further discussion and editing - and I did think that Kaggle has an auto-detection system when I copied your notebook so I will not receive any credit.
\nI feel very terrible about my lack of knowledge of how this platform work and the way to credit other people's work has affected the author. I have made the notebook private for additional modification and acknowledgment. I have also sent a personal email to apologize to the author @hinepo
\nThis is a big lesson for me.

", "rawMarkdown": "Hello to everyone, I am the one who copied the author's notebook here. I am truly sorry for my lack of knowledge and awareness.\nI thought copying the notebook and sharing through clicking on Share and then selecting Public is the most convenient way of letting my group of friends access the notebook for further discussion and editing - and I did think that Kaggle has an auto-detection system when I copied your notebook so I will not receive any credit.\nI feel very terrible about my lack of knowledge of how this platform work and the way to credit other people's work has affected the author. I have made the notebook private for additional modification and acknowledgment. I have also sent a personal email to apologize to the author @hinepo \nThis is a big lesson for me.", "votes": 17, "replies": [ { "id": 1851650, "postDate": "2022-07-11T12:16:13.070Z", "content": "

Hi @dungdao706

\n

I really appreciate your words and attitudes, and I'm glad we were able to resolve this issue with communication.

\n

I also learned from this episode. Don't feel bad, it was just a mistake and it was also a lesson for me. Thanks again for your post and your email, they were very polite and made me understand your side.

\n

I hope the notebook will be useful to you and your friends. I wish good luck to you all for the rest of the competition

\n

Regards,
\nHinePo

", "rawMarkdown": "Hi @dungdao706\n\nI really appreciate your words and attitudes, and I'm glad we were able to resolve this issue with communication.\n\nI also learned from this episode. Don't feel bad, it was just a mistake and it was also a lesson for me. Thanks again for your post and your email, they were very polite and made me understand your side.\n\nI hope the notebook will be useful to you and your friends. I wish good luck to you all for the rest of the competition\n\nRegards,\nHinePo", "votes": 12 }, { "id": 1851930, "postDate": "2022-07-11T16:43:21.787Z", "content": "

Just wanted to appreciate @dungdao706 's attitude of coming out in open and apologising.

", "rawMarkdown": "Just wanted to appreciate @dungdao706 's attitude of coming out in open and apologising. ", "votes": 4 } ] }, { "id": 1851538, "postDate": "2022-07-11T10:46:29.893Z", "content": "

I hope kaggle sort of improving the rules.

", "rawMarkdown": "I hope kaggle sort of improving the rules.", "votes": 3 }, { "id": 1850760, "postDate": "2022-07-10T17:57:51.923Z", "content": "

Maybe kaggle should start a competition to detect plagiarism so you can help them sort this nonsense. :) Jokes aside, I have no idea why people find it hard to credit work properly.

", "rawMarkdown": "Maybe kaggle should start a competition to detect plagiarism so you can help them sort this nonsense. :) Jokes aside, I have no idea why people find it hard to credit work properly.", "votes": 3 }, { "id": 1849695, "postDate": "2022-07-09T18:01:58.770Z", "content": "

well.. sorry to hear that.. but no one of us can say.. they always do all the coding from scratch.. we will depend on someone's work and learn from it.. it's always great to give credits though.. I think we are all hear to learn from ourselves.. good luck !! ๐Ÿ™‚

", "rawMarkdown": "well.. sorry to hear that.. but no one of us can say.. they always do all the coding from scratch.. we will depend on someone's work and learn from it.. it's always great to give credits though.. I think we are all hear to learn from ourselves.. good luck !! ๐Ÿ™‚", "votes": 3 }, { "id": 1849569, "postDate": "2022-07-09T16:28:57.687Z", "content": "

If there are similarities like you describe above they should take down the notebook from participating in any competition. Otherwise, if he/she made a fork: changing the variables' names, modifying hyperparameters, changing comments, using another model, and adding or deleting some implementations, they should be legit. Most of the users are in the second category. Cheers and good luck! ๐Ÿ™‹

", "rawMarkdown": "If there are similarities like you describe above they should take down the notebook from participating in any competition. Otherwise, if he/she made a fork: changing the variables' names, modifying hyperparameters, changing comments, using another model, and adding or deleting some implementations, they should be legit. Most of the users are in the second category. Cheers and good luck! ๐Ÿ™‹", "votes": 4 }, { "id": 1849535, "postDate": "2022-07-09T15:56:08.237Z", "content": "

Yes Kaggle should take this into accountโ€ฆ. Sometimes you never got to know know your real rank in competition because there are many people who just copied someone's notebook and submit it.

", "rawMarkdown": "Yes Kaggle should take this into account.... Sometimes you never got to know know your real rank in competition because there are many people who just copied someone's notebook and submit it.", "votes": 4 }, { "id": 1848666, "postDate": "2022-07-08T20:13:26.320Z", "content": "

Of course, sharing has its pleasures, advantages and learnings as well.

\n

I take this oppotunity to promote @medali1992 notebook .

\n

He/she took my notebook and added great functionalities, improved score AND explainability. And gave credits to all the works he used. This was awesome and made me happy!

", "rawMarkdown": "Of course, sharing has its pleasures, advantages and learnings as well.\n\nI take this oppotunity to promote @medali1992 [notebook](https://www.kaggle.com/code/medali1992/amex-tabnetclassifier-feature-eng-0-791) .\n\nHe/she took my notebook and added great functionalities, improved score AND explainability. And gave credits to all the works he used. This was awesome and made me happy!", "votes": 3, "replies": [ { "id": 1848715, "postDate": "2022-07-08T20:56:14.103Z", "content": "

Thank you for promoting my notebook. Your notebook was very helpful and clean so please don't lose motivation and I hope that the kaggle stuff will someday find ร  solution for plagiarisim.

\n

Good luck.

", "rawMarkdown": "Thank you for promoting my notebook. Your notebook was very helpful and clean so please don't lose motivation and I hope that the kaggle stuff will someday find ร  solution for plagiarisim.\n\nGood luck.", "votes": 3 } ] }, { "id": 1854655, "postDate": "2022-07-13T21:56:53.677Z", "content": "

Hello guysโ€ฆ!
\n @hinepo how did you find it is plagiarized i am new to kaggle tooโ€ฆ
\nI'm just curious to know how did you find itโ€ฆ
\nWould it be okay to share??

", "rawMarkdown": "Hello guys...!\n @hinepo how did you find it is plagiarized i am new to kaggle too...\nI'm just curious to know how did you find it... \nWould it be okay to share??", "votes": 1, "replies": [ { "id": 1854675, "postDate": "2022-07-13T22:39:27.930Z", "content": "

Hi @shankar1t and welcome to kaggle :)

\n

I opened this discssion post because I've seen it happen a few times before on other competitions and I wanted to see what people think about this case. When it happens, people usually agree on a common-sense but I don't actually know the details that kaggle takes into account when judging these things. Since they didn't take down the notebook this time, they are probably ok with it in this case.

\n

I am just another kaggler and I found about it at random, browsing the forum. I just compared the notebooks.

\n

In my opinion, sharing should be encouraged, and when sharing it is nice to reference previous works and to add some change or functionallity. But apparently not everyone agrees with that and there are people with a more flexible concept of sharing. And it is ok. We should respect every opinion on every matter.

", "rawMarkdown": "Hi @shankar1t and welcome to kaggle :)\n\nI opened this discssion post because I've seen it happen a few times before on other competitions and I wanted to see what people think about this case. When it happens, people usually agree on a common-sense but I don't actually know the details that kaggle takes into account when judging these things. Since they didn't take down the notebook this time, they are probably ok with it in this case.\n\nI am just another kaggler and I found about it at random, browsing the forum. I just compared the notebooks.\n\nIn my opinion, sharing should be encouraged, and when sharing it is nice to reference previous works and to add some change or functionallity. But apparently not everyone agrees with that and there are people with a more flexible concept of sharing. And it is ok. We should respect every opinion on every matter.", "votes": 1 }, { "id": 1854719, "postDate": "2022-07-14T00:25:39.910Z", "content": "

Okay Thank you

", "rawMarkdown": "Okay Thank you", "votes": 1 } ] }, { "id": 1851821, "postDate": "2022-07-11T15:00:27.750Z", "content": "

Hello HinePo ๐Ÿ‘‹
\nFor later times, I suggest you consider distributing your code by applying a *GPL license or other equivalent free software license. This way you will discourage anyone from removing credits without feeling worried to be blamed at least a little bit afterwards ;-)

\n

My best regards :-)

", "rawMarkdown": "Hello HinePo ๐Ÿ‘‹\nFor later times, I suggest you consider distributing your code by applying a *GPL license or other equivalent free software license. This way you will discourage anyone from removing credits without feeling worried to be blamed at least a little bit afterwards ;-)\n\nMy best regards :-)", "votes": 2 }, { "id": 1848825, "postDate": "2022-07-08T23:55:33.440Z", "content": "

Very sad @hinepo

", "rawMarkdown": "Very sad @hinepo ", "votes": 1 }, { "id": 1848837, "postDate": "2022-07-09T00:37:33.840Z", "rawMarkdown": "", "votes": -9, "isDeleted": true, "replies": [ { "id": 1848840, "postDate": "2022-07-09T00:42:49.580Z", "content": "

Acknowledgment doesn't hurt even if the code is a baseline.

", "rawMarkdown": "Acknowledgment doesn't hurt even if the code is a baseline.", "votes": 3 }, { "id": 1848842, "postDate": "2022-07-09T00:46:49.543Z", "rawMarkdown": "", "votes": -9, "isDeleted": true }, { "id": 1848845, "postDate": "2022-07-09T00:54:13.337Z", "content": "

I invite you to see for example all the notebooks of @cdeotte who is one of the best data scientist in the world and he gives credits to the people who helped him build his baseline.
\nDude chill out ok, this is a friendly environment where everyone enjoys learning new things in data science.

", "rawMarkdown": "I invite you to see for example all the notebooks of @cdeotte who is one of the best data scientist in the world and he gives credits to the people who helped him build his baseline.\nDude chill out ok, this is a friendly environment where everyone enjoys learning new things in data science.", "votes": 4 }, { "id": 1848846, "postDate": "2022-07-09T00:55:45.927Z", "rawMarkdown": "", "votes": -6, "isDeleted": true }, { "id": 1849027, "postDate": "2022-07-09T06:09:50.310Z", "content": "

It is ok if you've seen something similar applied to another problem.

\n

But I care because I think people should put in at least some work to get medals and upvotes. Otherwise we devalue kaggle medals.

\n

I repeat: he added nothing. Not a single character. That is really zero work.

", "rawMarkdown": "It is ok if you've seen something similar applied to another problem.\n\nBut I care because I think people should put in at least some work to get medals and upvotes. Otherwise we devalue kaggle medals.\n\nI repeat: he added nothing. Not a single character. That is really zero work.", "votes": 2 }, { "id": 1849039, "postDate": "2022-07-09T06:22:42.983Z", "rawMarkdown": "", "votes": -4, "isDeleted": true }, { "id": 1849045, "postDate": "2022-07-09T06:32:28.423Z", "content": "

It may be this way for you and for most people. But I think you can understand my point as I understand yours.

", "rawMarkdown": "It may be this way for you and for most people. But I think you can understand my point as I understand yours.", "votes": 2 }, { "id": 1849254, "postDate": "2022-07-09T10:47:40.053Z", "rawMarkdown": "" }, { "id": 1849290, "postDate": "2022-07-09T11:20:48.947Z", "rawMarkdown": "", "votes": -2, "isDeleted": true }, { "id": 1849321, "postDate": "2022-07-09T12:22:51.023Z", "content": "

Dude seriously you're not normal, the guy you're talking to is a recognized ML engineer in France and his a master of discussion and notebook. His notebooks are well known in the community and this is the whole purpose of kaggle so please don't tell me kaggle is all about medal competition because you joined Kaggle 7 years ago and you only have 4 medals and I think you read countless public notebooks and shared solution of humble kagglers to become who you are today.

\n

People who share articles and GitHub repositories put a Reference section that you use for competing on Kaggle.
\nYou're not better than anyone and you're talking as if you are a grandmaster competition. ๐Ÿ‘Ž
\nStop disrespecting people on kaggle.

", "rawMarkdown": "Dude seriously you're not normal, the guy you're talking to is a recognized ML engineer in France and his a master of discussion and notebook. His notebooks are well known in the community and this is the whole purpose of kaggle so please don't tell me kaggle is all about medal competition because you joined Kaggle 7 years ago and you only have 4 medals and I think you read countless public notebooks and shared solution of humble kagglers to become who you are today.\n\nPeople who share articles and GitHub repositories put a Reference section that you use for competing on Kaggle.\nYou're not better than anyone and you're talking as if you are a grandmaster competition. ๐Ÿ‘Ž\nStop disrespecting people on kaggle.", "votes": 1 }, { "id": 1849324, "postDate": "2022-07-09T12:25:35.400Z", "content": "

You guys seem a bit heated over me saying basic Cross Validation code shouldnโ€™t count as plagiarism ๐Ÿ˜‚, making all these other assumptions and resorting to name calling.. ๐Ÿ˜‰

\n

I think you should all go outside its very sunny today.. Ending this conversation here now ๐Ÿ‘‹

", "rawMarkdown": "You guys seem a bit heated over me saying basic Cross Validation code shouldnโ€™t count as plagiarism ๐Ÿ˜‚, making all these other assumptions and resorting to name calling.. ๐Ÿ˜‰\n\nI think you should all go outside its very sunny today.. Ending this conversation here now ๐Ÿ‘‹ ", "votes": -4 } ] } ], "comments": [ { "id": 1851107, "author_name": "Dung Minh Dao", "author_url": "", "post_date": "2022-07-11T03:14:21.503000", "content": "

Hello to everyone, I am the one who copied the author's notebook here. I am truly sorry for my lack of knowledge and awareness.
\nI thought copying the notebook and sharing through clicking on Share and then selecting Public is the most convenient way of letting my group of friends access the notebook for further discussion and editing - and I did think that Kaggle has an auto-detection system when I copied your notebook so I will not receive any credit.
\nI feel very terrible about my lack of knowledge of how this platform work and the way to credit other people's work has affected the author. I have made the notebook private for additional modification and acknowledgment. I have also sent a personal email to apologize to the author @hinepo
\nThis is a big lesson for me.

", "votes": 17, "replies": [ { "id": 1851650, "author_name": "HinePo", "author_url": "", "post_date": "2022-07-11T12:16:13.070000", "content": "

Hi @dungdao706

\n

I really appreciate your words and attitudes, and I'm glad we were able to resolve this issue with communication.

\n

I also learned from this episode. Don't feel bad, it was just a mistake and it was also a lesson for me. Thanks again for your post and your email, they were very polite and made me understand your side.

\n

I hope the notebook will be useful to you and your friends. I wish good luck to you all for the rest of the competition

\n

Regards,
\nHinePo

", "votes": 12, "replies": [] }, { "id": 1851930, "author_name": "monte carlo", "author_url": "", "post_date": "2022-07-11T16:43:21.787000", "content": "

Just wanted to appreciate @dungdao706 's attitude of coming out in open and apologising.

", "votes": 4, "replies": [] } ] }, { "id": 1851538, "author_name": "zhehao liang", "author_url": "", "post_date": "2022-07-11T10:46:29.893000", "content": "

I hope kaggle sort of improving the rules.

", "votes": 3, "replies": [] }, { "id": 1850760, "author_name": "vivekjan14", "author_url": "", "post_date": "2022-07-10T17:57:51.923000", "content": "

Maybe kaggle should start a competition to detect plagiarism so you can help them sort this nonsense. :) Jokes aside, I have no idea why people find it hard to credit work properly.

", "votes": 3, "replies": [] }, { "id": 1849695, "author_name": "Richard Sarpong", "author_url": "", "post_date": "2022-07-09T18:01:58.770000", "content": "

well.. sorry to hear that.. but no one of us can say.. they always do all the coding from scratch.. we will depend on someone's work and learn from it.. it's always great to give credits though.. I think we are all hear to learn from ourselves.. good luck !! ๐Ÿ™‚

", "votes": 3, "replies": [] }, { "id": 1849569, "author_name": "cristian.popescu", "author_url": "", "post_date": "2022-07-09T16:28:57.687000", "content": "

If there are similarities like you describe above they should take down the notebook from participating in any competition. Otherwise, if he/she made a fork: changing the variables' names, modifying hyperparameters, changing comments, using another model, and adding or deleting some implementations, they should be legit. Most of the users are in the second category. Cheers and good luck! ๐Ÿ™‹

", "votes": 4, "replies": [] }, { "id": 1849535, "author_name": "ShubhamChauhan", "author_url": "", "post_date": "2022-07-09T15:56:08.237000", "content": "

Yes Kaggle should take this into accountโ€ฆ. Sometimes you never got to know know your real rank in competition because there are many people who just copied someone's notebook and submit it.

", "votes": 4, "replies": [] }, { "id": 1848666, "author_name": "HinePo", "author_url": "", "post_date": "2022-07-08T20:13:26.320000", "content": "

Of course, sharing has its pleasures, advantages and learnings as well.

\n

I take this oppotunity to promote @medali1992 notebook .

\n

He/she took my notebook and added great functionalities, improved score AND explainability. And gave credits to all the works he used. This was awesome and made me happy!

", "votes": 3, "replies": [ { "id": 1848715, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-08T20:56:14.103000", "content": "

Thank you for promoting my notebook. Your notebook was very helpful and clean so please don't lose motivation and I hope that the kaggle stuff will someday find ร  solution for plagiarisim.

\n

Good luck.

", "votes": 3, "replies": [] } ] }, { "id": 1854655, "author_name": "SHANKAR T", "author_url": "", "post_date": "2022-07-13T21:56:53.677000", "content": "

Hello guysโ€ฆ!
\n @hinepo how did you find it is plagiarized i am new to kaggle tooโ€ฆ
\nI'm just curious to know how did you find itโ€ฆ
\nWould it be okay to share??

", "votes": 1, "replies": [ { "id": 1854675, "author_name": "HinePo", "author_url": "", "post_date": "2022-07-13T22:39:27.930000", "content": "

Hi @shankar1t and welcome to kaggle :)

\n

I opened this discssion post because I've seen it happen a few times before on other competitions and I wanted to see what people think about this case. When it happens, people usually agree on a common-sense but I don't actually know the details that kaggle takes into account when judging these things. Since they didn't take down the notebook this time, they are probably ok with it in this case.

\n

I am just another kaggler and I found about it at random, browsing the forum. I just compared the notebooks.

\n

In my opinion, sharing should be encouraged, and when sharing it is nice to reference previous works and to add some change or functionallity. But apparently not everyone agrees with that and there are people with a more flexible concept of sharing. And it is ok. We should respect every opinion on every matter.

", "votes": 1, "replies": [] }, { "id": 1854719, "author_name": "SHANKAR T", "author_url": "", "post_date": "2022-07-14T00:25:39.910000", "content": "

Okay Thank you

", "votes": 1, "replies": [] } ] }, { "id": 1851821, "author_name": "N. Fabio Arceri", "author_url": "", "post_date": "2022-07-11T15:00:27.750000", "content": "

Hello HinePo ๐Ÿ‘‹
\nFor later times, I suggest you consider distributing your code by applying a *GPL license or other equivalent free software license. This way you will discourage anyone from removing credits without feeling worried to be blamed at least a little bit afterwards ;-)

\n

My best regards :-)

", "votes": 2, "replies": [] }, { "id": 1848825, "author_name": "Muhammed Tausif", "author_url": "", "post_date": "2022-07-08T23:55:33.440000", "content": "

Very sad @hinepo

", "votes": 1, "replies": [] }, { "id": 1848837, "author_name": "", "author_url": "", "post_date": "2022-07-09T00:37:33.840000", "content": "", "votes": -9, "replies": [ { "id": 1848840, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-09T00:42:49.580000", "content": "

Acknowledgment doesn't hurt even if the code is a baseline.

", "votes": 3, "replies": [] }, { "id": 1848842, "author_name": "", "author_url": "", "post_date": "2022-07-09T00:46:49.543000", "content": "", "votes": -9, "replies": [] }, { "id": 1848845, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-09T00:54:13.337000", "content": "

I invite you to see for example all the notebooks of @cdeotte who is one of the best data scientist in the world and he gives credits to the people who helped him build his baseline.
\nDude chill out ok, this is a friendly environment where everyone enjoys learning new things in data science.

", "votes": 4, "replies": [] }, { "id": 1848846, "author_name": "", "author_url": "", "post_date": "2022-07-09T00:55:45.927000", "content": "", "votes": -6, "replies": [] }, { "id": 1849027, "author_name": "HinePo", "author_url": "", "post_date": "2022-07-09T06:09:50.310000", "content": "

It is ok if you've seen something similar applied to another problem.

\n

But I care because I think people should put in at least some work to get medals and upvotes. Otherwise we devalue kaggle medals.

\n

I repeat: he added nothing. Not a single character. That is really zero work.

", "votes": 2, "replies": [] }, { "id": 1849039, "author_name": "", "author_url": "", "post_date": "2022-07-09T06:22:42.983000", "content": "", "votes": -4, "replies": [] }, { "id": 1849045, "author_name": "HinePo", "author_url": "", "post_date": "2022-07-09T06:32:28.423000", "content": "

It may be this way for you and for most people. But I think you can understand my point as I understand yours.

", "votes": 2, "replies": [] }, { "id": 1849254, "author_name": "", "author_url": "", "post_date": "2022-07-09T10:47:40.053000", "content": "", "votes": 0, "replies": [] }, { "id": 1849290, "author_name": "", "author_url": "", "post_date": "2022-07-09T11:20:48.947000", "content": "", "votes": -2, "replies": [] }, { "id": 1849321, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-09T12:22:51.023000", "content": "

Dude seriously you're not normal, the guy you're talking to is a recognized ML engineer in France and his a master of discussion and notebook. His notebooks are well known in the community and this is the whole purpose of kaggle so please don't tell me kaggle is all about medal competition because you joined Kaggle 7 years ago and you only have 4 medals and I think you read countless public notebooks and shared solution of humble kagglers to become who you are today.

\n

People who share articles and GitHub repositories put a Reference section that you use for competing on Kaggle.
\nYou're not better than anyone and you're talking as if you are a grandmaster competition. ๐Ÿ‘Ž
\nStop disrespecting people on kaggle.

", "votes": 1, "replies": [] }, { "id": 1849324, "author_name": "JM", "author_url": "", "post_date": "2022-07-09T12:25:35.400000", "content": "

You guys seem a bit heated over me saying basic Cross Validation code shouldnโ€™t count as plagiarism ๐Ÿ˜‚, making all these other assumptions and resorting to name calling.. ๐Ÿ˜‰

\n

I think you should all go outside its very sunny today.. Ending this conversation here now ๐Ÿ‘‹

", "votes": -4, "replies": [] } ] } ], "raw_markdown_by_id": { "1848645": "Unfortunately the increasingly common practice in kaggle also happened to me.\n\nI published a notebook that 11 days later was used by another kaggler who simply removed some lines (first markdown cell, and last code cell (feature importances)) and republished it. It didn't add anything or any line to my work. Didn't even change hyperparameters. It achieves exactly the same CV (0.78542) on the very same epoch (15).\n\nI reported the notebook to discourage such practices on kaggle.\n\n[notebook I published](https://www.kaggle.com/code/hinepo/amex-tabnet-training)\n\n[copied notebook](https://www.kaggle.com/code/dungdao706/amex-tabnet/notebook)", "1851107": "Hello to everyone, I am the one who copied the author's notebook here. I am truly sorry for my lack of knowledge and awareness.\nI thought copying the notebook and sharing through clicking on Share and then selecting Public is the most convenient way of letting my group of friends access the notebook for further discussion and editing - and I did think that Kaggle has an auto-detection system when I copied your notebook so I will not receive any credit.\nI feel very terrible about my lack of knowledge of how this platform work and the way to credit other people's work has affected the author. I have made the notebook private for additional modification and acknowledgment. I have also sent a personal email to apologize to the author @hinepo \nThis is a big lesson for me.", "1851538": "I hope kaggle sort of improving the rules.", "1850760": "Maybe kaggle should start a competition to detect plagiarism so you can help them sort this nonsense. :) Jokes aside, I have no idea why people find it hard to credit work properly.", "1849695": "well.. sorry to hear that.. but no one of us can say.. they always do all the coding from scratch.. we will depend on someone's work and learn from it.. it's always great to give credits though.. I think we are all hear to learn from ourselves.. good luck !! ๐Ÿ™‚", "1849569": "If there are similarities like you describe above they should take down the notebook from participating in any competition. Otherwise, if he/she made a fork: changing the variables' names, modifying hyperparameters, changing comments, using another model, and adding or deleting some implementations, they should be legit. Most of the users are in the second category. Cheers and good luck! ๐Ÿ™‹", "1849535": "Yes Kaggle should take this into account.... Sometimes you never got to know know your real rank in competition because there are many people who just copied someone's notebook and submit it.", "1848666": "Of course, sharing has its pleasures, advantages and learnings as well.\n\nI take this oppotunity to promote @medali1992 [notebook](https://www.kaggle.com/code/medali1992/amex-tabnetclassifier-feature-eng-0-791) .\n\nHe/she took my notebook and added great functionalities, improved score AND explainability. And gave credits to all the works he used. This was awesome and made me happy!", "1854655": "Hello guys...!\n @hinepo how did you find it is plagiarized i am new to kaggle too...\nI'm just curious to know how did you find it... \nWould it be okay to share??", "1851821": "Hello HinePo ๐Ÿ‘‹\nFor later times, I suggest you consider distributing your code by applying a *GPL license or other equivalent free software license. This way you will discourage anyone from removing credits without feeling worried to be blamed at least a little bit afterwards ;-)\n\nMy best regards :-)", "1848825": "Very sad @hinepo ", "1848837": "" } }