\n
![\"\"](\"https://i.imgur.com/VlVO46i.png\")
\nAnd to quote Kaggle's terms of use:
There are four differences between the notebooks Baseline - Starter and Genetic Algorithm for Naive Baseline:
\nSEED, the first has 42 and in the second is 2021Many of us are here to learn, while doing this, we migh take borrowed pieces of code from another's work but, present a work as yours just because you change the SEED value it's not fair for the author who spend hours doing a really good work.
", "messages": [ { "id": "1238354", "postDate": "03/14/2021 23:25:57", "content": "There are four differences between the notebooks Baseline - Starter and Genetic Algorithm for Naive Baseline:
\nSEED, the first has 42 and in the second is 2021Many of us are here to learn, while doing this, we migh take borrowed pieces of code from another's work but, present a work as yours just because you change the SEED value it's not fair for the author who spend hours doing a really good work.
", "rawMarkdown": "There are four differences between the notebooks [Baseline - Starter](https://www.kaggle.com/ghaiyur/baseline-starter) and [Genetic Algorithm for Naive Baseline](https://www.kaggle.com/andypenrose/genetic-algorithm-for-naive-baseline):\n\n- The name\n- The `SEED`, the first has 42 and in the second is 2021\n- In the section \"Comments\" of the first you can read \"The author has turned off further commenting on this notebook.\"\n- The first is not a original work, is a copy paste of the second\n\nMany of us are here to learn, while doing this, we migh take borrowed pieces of code from another's work but, present a work as yours just because you change the SEED value it's not fair for the author who spend hours doing a really good work.", "votes": null }, { "id": "1238373", "postDate": "03/14/2021 23:58:37", "content": "I don't think this is plagiarism. Kaggle's platform is built for sharing. In the notebook, there is actually an attribution:
\n
\nAnd to quote Kaggle's terms of use:
\n", "rawMarkdown": "I don't think this is plagiarism. Kaggle's platform is built for sharing. In the notebook, there is actually an attribution:\n\nAnd to quote Kaggle's terms of use:\n> you grant all other users of the Services a license to access that Public User Submission, and to use and exercise all rights in it, as permitted by the functionality of the Services.", "votes": null }, { "id": "1238376", "postDate": "03/15/2021 00:08:16", "content": "you grant all other users of the Services a license to access that Public User Submission, and to use and exercise all rights in it, as permitted by the functionality of the Services.
\n
You have a point, it should be nice to read others opinions, to enrich the discussion.
", "rawMarkdown": "You have a point, it should be nice to read others opinions, to enrich the discussion.", "votes": null }, { "id": "1238690", "postDate": "03/15/2021 08:04:48", "content": "@matthewmasters Kaggle allows public sharing and forking in order to \"build upon\" and improve a kernel, but not to take away upvotes from original authors. Also, the right-hand \"copied from ..\" is auto-generated by Kaggle, whereas a real attribution would be to add text at the top of the forked kernel such as:
\n@hiramcho has valid points that:
\n@sirishks I respectfully disagree. By your definition, any Github fork without major changes to the code would be plagiarism, which just isn't true. Also, what makes you say the auto-generated attribution isn't \"real\" but a hand-typed one is? Finally, implying that the author had bad intentions and simply did this to steal upvotes is based on conjecture. And from what I can tell, they haven't actually taken away any upvotes from the original notebook.
", "rawMarkdown": "sirishks I respectfully disagree. By your definition, any Github fork without major changes to the code would be plagiarism, which just isn't true. Also, what makes you say the auto-generated attribution isn't \"real\" but a hand-typed one is? Finally, implying that the author had bad intentions and simply did this to steal upvotes is based on conjecture. And from what I can tell, they haven't actually taken away any upvotes from the original notebook.", "votes": null }, { "id": "1239284", "postDate": "03/15/2021 15:31:37", "content": "Thanks for share your opinions @matthewmasters and @sirishks. This question is for the first of you, do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it? My questions have no bad intentions, i just want to have a broader perspective of this topic.
", "rawMarkdown": "Thanks for share your opinions @matthewmasters and @sirishks. This question is for the first of you, do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it? My questions have no bad intentions, i just want to have a broader perspective of this topic.", "votes": null }, { "id": "1239665", "postDate": "03/15/2021 22:32:18", "content": "To all those who find this unacceptable behavior, I have proposed to enable downvotes on kernels so that kernels that add no value get downvoted. This would remove incentives to claim other people's work.
\nIf you want to support this votes are still possible: https://www.kaggle.com/product-feedback/76394
", "rawMarkdown": "To all those who find this unacceptable behavior, I have proposed to enable downvotes on kernels so that kernels that add no value get downvoted. This would remove incentives to claim other people's work.\n\nIf you want to support this votes are still possible: https://www.kaggle.com/product-feedback/76394", "votes": null }, { "id": "1240124", "postDate": "03/16/2021 08:38:40", "content": "Upvoted (122 votes now)! but Kaggle is still not responding?
\nPS: Downvotes should be enabled on Datasets too …
@hiramcho Thanks for raising this. I was most disappointed that my header comment was removed, and the explanation that the notebook was a naive baseline, not a serious solution to the problem at hand, was gone. I was going to leave a comment on the notebook, but of course I could not.
\nThankfully, given the very low number of upvotes for the \"Baseline - Starter\" notebook - it seems most people have seen it for exactly what it is.
", "rawMarkdown": "hiramcho Thanks for raising this. I was most disappointed that my header comment was removed, and the explanation that the notebook was a naive baseline, not a serious solution to the problem at hand, was gone. I was going to leave a comment on the notebook, but of course I could not.\n\nThankfully, given the very low number of upvotes for the \"Baseline - Starter\" notebook - it seems most people have seen it for exactly what it is.", "votes": null }, { "id": "1244092", "postDate": "03/18/2021 17:34:43", "content": "\"Plagiarism is the representation of another author's language, thoughts, ideas, or expressions as one's own original work\" As @matthewmasters has pointed out, this definition does not apply above. It is definitely NOT plagiarism by definition. The author as a good-will gesture could and should have credited the original authors explicitly in the kernel, but that is a different story. Kaggle rules dont mandate that. We use lot of open source libraries but dont necessarily credit all the authors explicitly.
\nJust sharing a contrarian point of view. But I do understand the pain…Maybe each time someone forks a kernel, 5 votes could be given to original author or something. Or maybe every subsequent vote on the forked kernel contributes (in some % way) to original author's vote bank. I am sure there are good alternatives available..
", "rawMarkdown": "\"Plagiarism is the representation of another author's language, thoughts, ideas, or expressions as one's own original work\" As @matthewmasters has pointed out, this definition does not apply above. It is definitely NOT plagiarism by definition. The author as a good-will gesture could and should have credited the original authors explicitly in the kernel, but that is a different story. Kaggle rules dont mandate that. We use lot of open source libraries but dont necessarily credit all the authors explicitly. \n\nJust sharing a contrarian point of view. But I do understand the pain...Maybe each time someone forks a kernel, 5 votes could be given to original author or something. Or maybe every subsequent vote on the forked kernel contributes (in some % way) to original author's vote bank. I am sure there are good alternatives available..", "votes": null }, { "id": "1244173", "postDate": "03/18/2021 18:29:42", "content": "Wanted to comment on the notebook and saw:
\n\n", "rawMarkdown": "Wanted to comment on the notebook and saw:\n> The author has turned off further commenting on this notebook.", "votes": null }, { "id": "1244179", "postDate": "03/18/2021 18:36:25", "content": "The author has turned off further commenting on this notebook.
\n
The auto-generated attribution is clearly not clear enough, as I discussed here:
\n\n\nI wouldn't be bothered by this if it was an honest mistake (e.g. accidentally made public after forking), but it is clear that the name change and erasing reference to the original notebook were not accidental or a beginner's mistake. What disappointed me is that many of the people that upvoted the plagiarized notebook didn't even upvote the original notebook, which indicates that even with the two indicators (ref. to the screenshots), it's still hard for people to find the original notebook and determine if there's any worthy change.
\n
I've proposed UI changes to make it clearer when you are forking a notebook; unfortunately no one from the @kaggleteam acknowledged it so I'm guessing it's not gonna be implemented any time soon.
\nConcerning the discussion about Github, it's a lot easier to see when a repository is forked, see:
\n![\"\"](\"https://i.imgur.com/JIgmJ3H.jpg\")
I've made suggestions to improve the UI here. Unfortunately no feedback from the Kaggle team :/
", "rawMarkdown": "I've made [suggestions to improve the UI here](https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/207431). Unfortunately no feedback from the Kaggle team :/", "votes": null }, { "id": "1244536", "postDate": "03/19/2021 03:43:24", "content": "This is nice and practical. Upvoted.
\nKaggle definitely needs to consider this
yea, I just dont understand why this feature is present. People should not be allowed to turn off comments. If comments are toxic, they can always be reported.
", "rawMarkdown": "yea, I just dont understand why this feature is present. People should not be allowed to turn off comments. If comments are toxic, they can always be reported.", "votes": null }, { "id": "1245751", "postDate": "03/20/2021 05:38:47", "content": "Great suggestion @cpmpml . But sometimes people just fork a top public kernel and submit it. I'm afraid that some of them could want to downvote the source kernel after losing on private LB due to overfitting or shake up.
", "rawMarkdown": "Great suggestion @cpmpml . But sometimes people just fork a top public kernel and submit it. I'm afraid that some of them could want to downvote the source kernel after losing on private LB due to overfitting or shake up.", "votes": null }, { "id": "1246413", "postDate": "03/20/2021 18:02:05", "content": "Great idea. I've commented on that post.
", "rawMarkdown": "Great idea. I've commented on that post.", "votes": null }, { "id": "1248414", "postDate": "03/22/2021 15:29:42", "content": "It's not necessarily plagiarism in the legal sense, but it's definitely not \"good form\" to share a notebook that you have forked and only made cosmetic changes to. You are obviously not adding any value to the community. The practice of posting simply for the sake of gathering potential votes, without any thought for adding even the slightest bit of real value or information should be discouraged.
", "rawMarkdown": "It's not necessarily plagiarism in the legal sense, but it's definitely not \"good form\" to share a notebook that you have forked and only made cosmetic changes to. You are obviously not adding any value to the community. The practice of posting simply for the sake of gathering potential votes, without any thought for adding even the slightest bit of real value or information should be discouraged.", "votes": null }, { "id": "1248442", "postDate": "03/22/2021 15:50:41", "content": "Maybe i was too far calling it \"plagiarism\", thanks for share your opinion.
", "rawMarkdown": "Maybe i was too far calling it \"plagiarism\", thanks for share your opinion.", "votes": null }, { "id": "1248458", "postDate": "03/22/2021 16:01:48", "content": "@hiramcho Regardless of the strict semantic and legal definitions of that word, I fully support the spirit of your post about not posting what is essentially someone else's work simply to gain votes :)
\nGamification (in the form of medals, badges etc) can be a good way to get people more engaged with a site, but sometimes it causes people to get obsessed with short-term rewards…
…\"do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it?\"…
\nI was trying to provide a perspective to Hiram since he requested for it. However I do agree that the spirit in which this post is done is a good one. The first thing I did when I saw this post was to quickly go thru' the original kernel, upvot it and book-mark it for a later leisure read.
\nI believe there are some simple ways to tackle this issue like (a) not allowing the forker to turn off commenting (b) automatic upvote each time someone forks (c) 50% upvote credit for original kernel for each upvote downstream (though this will be difficult to implement) d) the nice UI changes suggested by Xhulu above etc..
", "rawMarkdown": "...\"*do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it?*\"...\n\nI was trying to provide a perspective to Hiram since he requested for it. However I do agree that the spirit in which this post is done is a good one. The first thing I did when I saw this post was to quickly go thru' the original kernel, upvot it and book-mark it for a later leisure read. \n\nI believe there are some simple ways to tackle this issue like (a) not allowing the forker to turn off commenting (b) automatic upvote each time someone forks (c) 50% upvote credit for original kernel for each upvote downstream (though this will be difficult to implement) d) the nice UI changes suggested by Xhulu above etc..", "votes": null }, { "id": "1248672", "postDate": "03/22/2021 18:37:09", "content": "Update: I've decided to share some tips for beginners to avoid plagiarism and ensure that the sources are known to the users. Might be a small step but I hope it helps some of the new users here.
", "rawMarkdown": "Update: I've decided to share [some tips for beginners](https://www.kaggle.com/getting-started/227919) to avoid plagiarism and ensure that the sources are known to the users. Might be a small step but I hope it helps some of the new users here.", "votes": null } ], "comments": [ { "id": 1238373, "author_name": "matthewmasters", "author_url": "", "post_date": "03/14/2021 23:58:37", "content": "I don't think this is plagiarism. Kaggle's platform is built for sharing. In the notebook, there is actually an attribution:
\n
\nAnd to quote Kaggle's terms of use:
\n", "votes": null, "replies": [ { "id": 1238376, "author_name": "hiramcho", "author_url": "", "post_date": "03/15/2021 00:08:16", "content": "you grant all other users of the Services a license to access that Public User Submission, and to use and exercise all rights in it, as permitted by the functionality of the Services.
\n
You have a point, it should be nice to read others opinions, to enrich the discussion.
", "votes": null, "replies": [] }, { "id": 1238690, "author_name": "sirishks", "author_url": "", "post_date": "03/15/2021 08:04:48", "content": "@matthewmasters Kaggle allows public sharing and forking in order to \"build upon\" and improve a kernel, but not to take away upvotes from original authors. Also, the right-hand \"copied from ..\" is auto-generated by Kaggle, whereas a real attribution would be to add text at the top of the forked kernel such as:
\n@hiramcho has valid points that:
\n@sirishks I respectfully disagree. By your definition, any Github fork without major changes to the code would be plagiarism, which just isn't true. Also, what makes you say the auto-generated attribution isn't \"real\" but a hand-typed one is? Finally, implying that the author had bad intentions and simply did this to steal upvotes is based on conjecture. And from what I can tell, they haven't actually taken away any upvotes from the original notebook.
", "votes": null, "replies": [] }, { "id": 1239284, "author_name": "hiramcho", "author_url": "", "post_date": "03/15/2021 15:31:37", "content": "Thanks for share your opinions @matthewmasters and @sirishks. This question is for the first of you, do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it? My questions have no bad intentions, i just want to have a broader perspective of this topic.
", "votes": null, "replies": [] }, { "id": 1244092, "author_name": "allohvk", "author_url": "", "post_date": "03/18/2021 17:34:43", "content": "\"Plagiarism is the representation of another author's language, thoughts, ideas, or expressions as one's own original work\" As @matthewmasters has pointed out, this definition does not apply above. It is definitely NOT plagiarism by definition. The author as a good-will gesture could and should have credited the original authors explicitly in the kernel, but that is a different story. Kaggle rules dont mandate that. We use lot of open source libraries but dont necessarily credit all the authors explicitly.
\nJust sharing a contrarian point of view. But I do understand the pain…Maybe each time someone forks a kernel, 5 votes could be given to original author or something. Or maybe every subsequent vote on the forked kernel contributes (in some % way) to original author's vote bank. I am sure there are good alternatives available..
", "votes": null, "replies": [] }, { "id": 1244179, "author_name": "xhlulu", "author_url": "", "post_date": "03/18/2021 18:36:25", "content": "The auto-generated attribution is clearly not clear enough, as I discussed here:
\n\n\nI wouldn't be bothered by this if it was an honest mistake (e.g. accidentally made public after forking), but it is clear that the name change and erasing reference to the original notebook were not accidental or a beginner's mistake. What disappointed me is that many of the people that upvoted the plagiarized notebook didn't even upvote the original notebook, which indicates that even with the two indicators (ref. to the screenshots), it's still hard for people to find the original notebook and determine if there's any worthy change.
\n
I've proposed UI changes to make it clearer when you are forking a notebook; unfortunately no one from the @kaggleteam acknowledged it so I'm guessing it's not gonna be implemented any time soon.
\nConcerning the discussion about Github, it's a lot easier to see when a repository is forked, see:
\n
It's not necessarily plagiarism in the legal sense, but it's definitely not \"good form\" to share a notebook that you have forked and only made cosmetic changes to. You are obviously not adding any value to the community. The practice of posting simply for the sake of gathering potential votes, without any thought for adding even the slightest bit of real value or information should be discouraged.
", "votes": null, "replies": [] }, { "id": 1248442, "author_name": "hiramcho", "author_url": "", "post_date": "03/22/2021 15:50:41", "content": "Maybe i was too far calling it \"plagiarism\", thanks for share your opinion.
", "votes": null, "replies": [] }, { "id": 1248458, "author_name": "thomasbrekkunnvik", "author_url": "", "post_date": "03/22/2021 16:01:48", "content": "@hiramcho Regardless of the strict semantic and legal definitions of that word, I fully support the spirit of your post about not posting what is essentially someone else's work simply to gain votes :)
\nGamification (in the form of medals, badges etc) can be a good way to get people more engaged with a site, but sometimes it causes people to get obsessed with short-term rewards…
…\"do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it?\"…
\nI was trying to provide a perspective to Hiram since he requested for it. However I do agree that the spirit in which this post is done is a good one. The first thing I did when I saw this post was to quickly go thru' the original kernel, upvot it and book-mark it for a later leisure read.
\nI believe there are some simple ways to tackle this issue like (a) not allowing the forker to turn off commenting (b) automatic upvote each time someone forks (c) 50% upvote credit for original kernel for each upvote downstream (though this will be difficult to implement) d) the nice UI changes suggested by Xhulu above etc..
", "votes": null, "replies": [] } ] }, { "id": 1239665, "author_name": "cpmpml", "author_url": "", "post_date": "03/15/2021 22:32:18", "content": "To all those who find this unacceptable behavior, I have proposed to enable downvotes on kernels so that kernels that add no value get downvoted. This would remove incentives to claim other people's work.
\nIf you want to support this votes are still possible: https://www.kaggle.com/product-feedback/76394
", "votes": null, "replies": [ { "id": 1240124, "author_name": "sirishks", "author_url": "", "post_date": "03/16/2021 08:38:40", "content": "Upvoted (122 votes now)! but Kaggle is still not responding?
\nPS: Downvotes should be enabled on Datasets too …
Great suggestion @cpmpml . But sometimes people just fork a top public kernel and submit it. I'm afraid that some of them could want to downvote the source kernel after losing on private LB due to overfitting or shake up.
", "votes": null, "replies": [] } ] }, { "id": 1242893, "author_name": "andypenrose", "author_url": "", "post_date": "03/17/2021 22:15:38", "content": "@hiramcho Thanks for raising this. I was most disappointed that my header comment was removed, and the explanation that the notebook was a naive baseline, not a serious solution to the problem at hand, was gone. I was going to leave a comment on the notebook, but of course I could not.
\nThankfully, given the very low number of upvotes for the \"Baseline - Starter\" notebook - it seems most people have seen it for exactly what it is.
", "votes": null, "replies": [] }, { "id": 1244173, "author_name": "xhlulu", "author_url": "", "post_date": "03/18/2021 18:29:42", "content": "Wanted to comment on the notebook and saw:
\n\n", "votes": null, "replies": [ { "id": 1244540, "author_name": "allohvk", "author_url": "", "post_date": "03/19/2021 03:47:09", "content": "The author has turned off further commenting on this notebook.
\n
yea, I just dont understand why this feature is present. People should not be allowed to turn off comments. If comments are toxic, they can always be reported.
", "votes": null, "replies": [] } ] }, { "id": 1244180, "author_name": "xhlulu", "author_url": "", "post_date": "03/18/2021 18:37:40", "content": "I've made suggestions to improve the UI here. Unfortunately no feedback from the Kaggle team :/
", "votes": null, "replies": [ { "id": 1244536, "author_name": "allohvk", "author_url": "", "post_date": "03/19/2021 03:43:24", "content": "This is nice and practical. Upvoted.
\nKaggle definitely needs to consider this
Great idea. I've commented on that post.
", "votes": null, "replies": [] }, { "id": 1248672, "author_name": "xhlulu", "author_url": "", "post_date": "03/22/2021 18:37:09", "content": "Update: I've decided to share some tips for beginners to avoid plagiarism and ensure that the sources are known to the users. Might be a small step but I hope it helps some of the new users here.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1238354": "There are four differences between the notebooks [Baseline - Starter](https://www.kaggle.com/ghaiyur/baseline-starter) and [Genetic Algorithm for Naive Baseline](https://www.kaggle.com/andypenrose/genetic-algorithm-for-naive-baseline):\n\n- The name\n- The `SEED`, the first has 42 and in the second is 2021\n- In the section \"Comments\" of the first you can read \"The author has turned off further commenting on this notebook.\"\n- The first is not a original work, is a copy paste of the second\n\nMany of us are here to learn, while doing this, we migh take borrowed pieces of code from another's work but, present a work as yours just because you change the SEED value it's not fair for the author who spend hours doing a really good work.", "1238373": "I don't think this is plagiarism. Kaggle's platform is built for sharing. In the notebook, there is actually an attribution:\n\nAnd to quote Kaggle's terms of use:\n> you grant all other users of the Services a license to access that Public User Submission, and to use and exercise all rights in it, as permitted by the functionality of the Services.", "1238376": "You have a point, it should be nice to read others opinions, to enrich the discussion.", "1238690": "matthewmasters Kaggle allows public sharing and forking in order to \"build upon\" and improve a kernel, but not to take away upvotes from original authors. Also, the right-hand \"copied from ..\" is auto-generated by Kaggle, whereas a real attribution would be to add text at the top of the forked kernel such as:\n - \"Please remember to **upvote the original kernel** [Genetic Algorithm for Naive Baseline](https://www.kaggle.com/andypenrose/genetic-algorithm-for-naive-baseline) by Andy\"\n - and add a few lines about the contributions and changes made in the forked copy\n\n@hiramcho has valid points that:\n- simply changing a seed, or other minor changes, and making it public may cause the original author to lose upvotes\n- in this case, the person who forked did not even upvote Andy's original kernel (when I checked today)\n- another signal that Kaggler's should be aware of is to check why \"**The author has turned off further commenting on this notebook.**\"", "1239271": "sirishks I respectfully disagree. By your definition, any Github fork without major changes to the code would be plagiarism, which just isn't true. Also, what makes you say the auto-generated attribution isn't \"real\" but a hand-typed one is? Finally, implying that the author had bad intentions and simply did this to steal upvotes is based on conjecture. And from what I can tell, they haven't actually taken away any upvotes from the original notebook.", "1239284": "Thanks for share your opinions @matthewmasters and @sirishks. This question is for the first of you, do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it? My questions have no bad intentions, i just want to have a broader perspective of this topic.", "1239665": "To all those who find this unacceptable behavior, I have proposed to enable downvotes on kernels so that kernels that add no value get downvoted. This would remove incentives to claim other people's work.\n\nIf you want to support this votes are still possible: https://www.kaggle.com/product-feedback/76394", "1240124": "Upvoted (122 votes now)! but Kaggle is still not responding?\nPS: Downvotes should be enabled on Datasets too ...", "1242893": "hiramcho Thanks for raising this. I was most disappointed that my header comment was removed, and the explanation that the notebook was a naive baseline, not a serious solution to the problem at hand, was gone. I was going to leave a comment on the notebook, but of course I could not.\n\nThankfully, given the very low number of upvotes for the \"Baseline - Starter\" notebook - it seems most people have seen it for exactly what it is.", "1244092": "\"Plagiarism is the representation of another author's language, thoughts, ideas, or expressions as one's own original work\" As @matthewmasters has pointed out, this definition does not apply above. It is definitely NOT plagiarism by definition. The author as a good-will gesture could and should have credited the original authors explicitly in the kernel, but that is a different story. Kaggle rules dont mandate that. We use lot of open source libraries but dont necessarily credit all the authors explicitly. \n\nJust sharing a contrarian point of view. But I do understand the pain...Maybe each time someone forks a kernel, 5 votes could be given to original author or something. Or maybe every subsequent vote on the forked kernel contributes (in some % way) to original author's vote bank. I am sure there are good alternatives available..", "1244173": "Wanted to comment on the notebook and saw:\n> The author has turned off further commenting on this notebook.", "1244179": "The auto-generated attribution is clearly not clear enough, as I discussed [here](https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/205147):\n> I wouldn't be bothered by this if it was an honest mistake (e.g. accidentally made public after forking), but it is clear that the name change and erasing reference to the original notebook were not accidental or a beginner's mistake. What disappointed me is that many of the people that upvoted the plagiarized notebook didn't even upvote the original notebook, which indicates that even with the two indicators (ref. to the screenshots), it's still hard for people to find the original notebook and determine if there's any worthy change.\n\nI've [proposed UI changes](https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/207431) to make it clearer when you are forking a notebook; unfortunately no one from the @kaggleteam acknowledged it so I'm guessing it's not gonna be implemented any time soon.\n\nConcerning the discussion about Github, it's a lot easier to see when a repository is forked, see:\n\n", "1244180": "I've made [suggestions to improve the UI here](https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/207431). Unfortunately no feedback from the Kaggle team :/", "1244536": "This is nice and practical. Upvoted. \nKaggle definitely needs to consider this", "1244540": "yea, I just dont understand why this feature is present. People should not be allowed to turn off comments. If comments are toxic, they can always be reported.", "1245751": "Great suggestion @cpmpml . But sometimes people just fork a top public kernel and submit it. I'm afraid that some of them could want to downvote the source kernel after losing on private LB due to overfitting or shake up.", "1246413": "Great idea. I've commented on that post.", "1248414": "It's not necessarily plagiarism in the legal sense, but it's definitely not \"good form\" to share a notebook that you have forked and only made cosmetic changes to. You are obviously not adding any value to the community. The practice of posting simply for the sake of gathering potential votes, without any thought for adding even the slightest bit of real value or information should be discouraged.", "1248442": "Maybe i was too far calling it \"plagiarism\", thanks for share your opinion.", "1248458": "hiramcho Regardless of the strict semantic and legal definitions of that word, I fully support the spirit of your post about not posting what is essentially someone else's work simply to gain votes :)\nGamification (in the form of medals, badges etc) can be a good way to get people more engaged with a site, but sometimes it causes people to get obsessed with short-term rewards...", "1248537": "...\"*do you think that i go too far calling what i posted here \"plagiarism\"? If the answer is true, how do you calle it?*\"...\n\nI was trying to provide a perspective to Hiram since he requested for it. However I do agree that the spirit in which this post is done is a good one. The first thing I did when I saw this post was to quickly go thru' the original kernel, upvot it and book-mark it for a later leisure read. \n\nI believe there are some simple ways to tackle this issue like (a) not allowing the forker to turn off commenting (b) automatic upvote each time someone forks (c) 50% upvote credit for original kernel for each upvote downstream (though this will be difficult to implement) d) the nice UI changes suggested by Xhulu above etc..", "1248672": "Update: I've decided to share [some tips for beginners](https://www.kaggle.com/getting-started/227919) to avoid plagiarism and ensure that the sources are known to the users. Might be a small step but I hope it helps some of the new users here." }, "source": "meta" }