Although I always have a tendency to moan about where I think Kaggle gets it wrong - on the whole the community is second to none.
\n\nJerome is a great example of this - integrity and honesty in showing the leak that he could have sat on until the last day.
\n\nIf this competition is reset I would like to think some Kaggle Swag will be on its way to Jerome
", "messages": [ { "id": 402078, "postDate": "2018-10-11T06:51:32.143Z", "content": "Although I always have a tendency to moan about where I think Kaggle gets it wrong - on the whole the community is second to none.
\n\nJerome is a great example of this - integrity and honesty in showing the leak that he could have sat on until the last day.
\n\nIf this competition is reset I would like to think some Kaggle Swag will be on its way to Jerome
", "rawMarkdown": "Although I always have a tendency to moan about where I think Kaggle gets it wrong - on the whole the community is second to none.\n\nJerome is a great example of this - integrity and honesty in showing the leak that he could have sat on until the last day.\n\nIf this competition is reset I would like to think some Kaggle Swag will be on its way to Jerome", "votes": 72 }, { "id": 402110, "postDate": "2018-10-11T07:32:49.187Z", "content": "Thanks @Scirpus & @mezoganet I appreciate.
\n\n\n\n\n...that he could have sat on until the last day... <
\n
Omg, don't make me regret :)))
", "rawMarkdown": "Thanks @Scirpus & @mezoganet I appreciate.\n\n> ...that he could have sat on until the last day... <\n\n\nOmg, don't make me regret :)))", "votes": 37, "replies": [ { "id": 402118, "postDate": "2018-10-11T07:42:30.633Z", "rawMarkdown": "", "isDeleted": true }, { "id": 402119, "postDate": "2018-10-11T07:43:36.927Z", "content": "I am sure of that about you : \"you could not have sat on until the last day.\"
", "rawMarkdown": "I am sure of that about you : \"*you could **not** have sat on until the last day.*\"", "votes": 1 }, { "id": 402190, "postDate": "2018-10-11T09:55:09.243Z", "content": "Everyone should honor your honesty :D
", "rawMarkdown": "Everyone should honor your honesty :D" }, { "id": 402242, "postDate": "2018-10-11T12:06:56.407Z", "rawMarkdown": "", "isDeleted": true }, { "id": 402291, "postDate": "2018-10-11T13:16:51.730Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 402197, "postDate": "2018-10-11T10:01:49.813Z", "content": "Be a man like Jerome
", "rawMarkdown": "Be a man like Jerome", "votes": 13 }, { "id": 402112, "postDate": "2018-10-11T07:36:42.667Z", "content": "A big thank you to Jérôme who investigated, released this issue and accepted to submit a leaky submission to prove data was freely available on Google Analytics !
\n\nOptions are (1) the available data is published as a Public Dataset, (2) the competition is reset or (3) the competition is cancelled
\n\nI suspect option 1 will be chosen.
", "rawMarkdown": "A big thank you to Jérôme who investigated, released this issue and accepted to submit a leaky submission to prove data was freely available on Google Analytics !\n\nOptions are (1) the available data is published as a Public Dataset, (2) the competition is reset or (3) the competition is cancelled\n\nI suspect option 1 will be chosen.", "votes": 10, "replies": [ { "id": 402117, "postDate": "2018-10-11T07:41:53.697Z", "content": "I'm not even sure they will bother with 1 - they might just point people to a link, if past exp is anything to go by.
", "rawMarkdown": "I'm not even sure they will bother with 1 - they might just point people to a link, if past exp is anything to go by.", "votes": 3 }, { "id": 402194, "postDate": "2018-10-11T10:01:08.350Z", "content": "Dear Kaggle,
\n\nSo many leak-full competitions recently. Santander, Airship prediction and now this.
\n\nI agree that framing problems and testing out for a leak is a really challenging task. But, the recent Airbus problem just had fragments of images and this competition had opened data freely available to look at.
\n\nI believe that causation of such obvious leak might be the communication gap between competition setters and the hosts. It's at an edge and the time that we need a leak-free and smooth going competitions (a few of them too). Too many competitions spoil the broth!
\n\nPeace.
", "rawMarkdown": "Dear Kaggle, \n\nSo many leak-full competitions recently. Santander, Airship prediction and now this.\n\nI agree that framing problems and testing out for a leak is a really challenging task. But, the recent Airbus problem just had fragments of images and this competition had opened data freely available to look at. \n\nI believe that causation of such obvious leak might be the communication gap between competition setters and the hosts. It's at an edge and the time that we need a leak-free and smooth going competitions (a few of them too). ***Too many competitions spoil the broth***!\n\nPeace.", "votes": -1 }, { "id": 402246, "postDate": "2018-10-11T12:13:25.750Z", "content": "The problem with option 1 is I suspect the data will take us to 0 error...
", "rawMarkdown": "The problem with option 1 is I suspect the data will take us to 0 error...", "votes": 1 }, { "id": 402257, "postDate": "2018-10-11T12:28:08.643Z", "content": "Even if they reset the competition, first they would have to plug up the leak, right?
", "rawMarkdown": "Even if they reset the competition, first they would have to plug up the leak, right? " }, { "id": 402390, "postDate": "2018-10-11T16:17:21.113Z", "content": "Really??? I always suspected this, but did not want to be the person messing up this competition! See screenshot in my EDA that shows 1.8M sessions in Google Analytics while we have 1.7 in the train+test. I thought/hoped they took a random sample somehow.....
", "rawMarkdown": "Really??? I always suspected this, but did not want to be the person messing up this competition! See screenshot in my EDA that shows 1.8M sessions in Google Analytics while we have 1.7 in the train+test. I thought/hoped they took a random sample somehow....." } ] }, { "id": 402099, "postDate": "2018-10-11T07:21:10.130Z", "content": "LOL I am glad that I chose doodle recognition :)
", "rawMarkdown": "LOL I am glad that I chose doodle recognition :)", "votes": 6 }, { "id": 402103, "postDate": "2018-10-11T07:23:55.533Z", "content": "Fully agree about your opinion concerning Jerome : integrity and honesty.
\n\nThe question is now : what Kaggle will do within the next hours/days ?
", "rawMarkdown": "Fully agree about your opinion concerning Jerome : integrity and honesty.\n\nThe question is now : what Kaggle will do within the next hours/days ? ", "votes": 1, "replies": [ { "id": 402121, "postDate": "2018-10-11T07:44:50.637Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 402128, "postDate": "2018-10-11T07:53:13.913Z", "content": "I agree with you. He saved my time!
", "rawMarkdown": "I agree with you. He saved my time!", "votes": 2 }, { "id": 403659, "postDate": "2018-10-14T09:05:10.743Z", "content": "Congratulations! :-)
", "rawMarkdown": "Congratulations! :-)" }, { "id": 402911, "postDate": "2018-10-12T14:54:11.850Z", "content": "I was getting my skills up so as to try this as my first featured competition. It's sad that people out there just want to cop paste/ find ways of cheating. Thanks Jerome and I hope Kaggle find a way to solve all this.
", "rawMarkdown": "I was getting my skills up so as to try this as my first featured competition. It's sad that people out there just want to cop paste/ find ways of cheating. Thanks Jerome and I hope Kaggle find a way to solve all this. ", "replies": [ { "id": 402915, "postDate": "2018-10-12T15:08:39.757Z", "content": "I'm not sure people want to cheat in a malicious way.. This is part of Data Science, finding all the data they can gather, wherever it is (in a legal way, of course...), to achieve a goal (in Kaggle, to have the highest rank for some, for others like you and me, to learn).
", "rawMarkdown": "I'm not sure people want to cheat in a malicious way.. This is part of Data Science, finding all the data they can gather, wherever it is (in a legal way, of course...), to achieve a goal (in Kaggle, to have the highest rank for some, for others like you and me, to learn). " } ] }, { "id": 402384, "postDate": "2018-10-11T16:09:18.027Z", "content": "I see a rally at the top LB now for several other participants. Is this just the beginning?)
", "rawMarkdown": "I see a rally at the top LB now for several other participants. Is this just the beginning?)", "replies": [ { "id": 402391, "postDate": "2018-10-11T16:19:47.723Z", "content": "quite disappointed now. I spent hours on feature engineering to reach a top position but now others can easily achieve an exceptionally high score just by exploiting the leak. I hope the kaggle admins can give an official response asap and handle the situation in a fair and just manner.
", "rawMarkdown": "quite disappointed now. I spent hours on feature engineering to reach a top position but now others can easily achieve an exceptionally high score just by exploiting the leak. I hope the kaggle admins can give an official response asap and handle the situation in a fair and just manner.\n ", "votes": 4 }, { "id": 402396, "postDate": "2018-10-11T16:29:38.907Z", "content": "Don't worry. I am sure that the situation will be corrected and everyone will continue to fight honestly.
", "rawMarkdown": "Don't worry. I am sure that the situation will be corrected and everyone will continue to fight honestly.", "votes": 2 }, { "id": 402656, "postDate": "2018-10-12T04:19:01.397Z", "content": "Kaggle hopefully will give it their best to carry on the competition! Keep it up!
", "rawMarkdown": "Kaggle hopefully will give it their best to carry on the competition! Keep it up!" } ] }, { "id": 402113, "postDate": "2018-10-11T07:38:23.373Z", "content": "I haven't fully understood your comment about the leak. \nHowever I see that Jerome's submission is a way ahead of everyone else's.
", "rawMarkdown": "I haven't fully understood your comment about the leak. \nHowever I see that Jerome's submission is a way ahead of everyone else's. ", "replies": [ { "id": 402152, "postDate": "2018-10-11T08:46:10.303Z", "content": "Jerome actually found a leak of data and made it public. The leak found, in fact, makes the whole contest under a question and the competition could be stopped/reset or conditions could be changed.
\n\nSee the discussions below:\nhttps://www.kaggle.com/c/ga-customer-revenue-prediction/discussion/68235#401880
", "rawMarkdown": "Jerome actually found a leak of data and made it public. The leak found, in fact, makes the whole contest under a question and the competition could be stopped/reset or conditions could be changed.\n\nSee the discussions below:\nhttps://www.kaggle.com/c/ga-customer-revenue-prediction/discussion/68235#401880", "votes": 2 }, { "id": 402219, "postDate": "2018-10-11T10:53:00.653Z", "content": "Thank you. I followed the thread and now I understand the context.
", "rawMarkdown": "Thank you. I followed the thread and now I understand the context." } ] }, { "id": 402727, "postDate": "2018-10-12T07:56:21.373Z", "rawMarkdown": "", "isDeleted": true }, { "id": 402105, "postDate": "2018-10-11T07:25:05.807Z", "content": "I am very sad =( But, thank you Jerome
", "rawMarkdown": "I am very sad =( But, thank you Jerome" } ], "comments": [ { "id": 402110, "author_name": "Jerome Vallet", "author_url": "", "post_date": "2018-10-11T07:32:49.187000", "content": "Thanks @Scirpus & @mezoganet I appreciate.
\n\n\n\n\n...that he could have sat on until the last day... <
\n
Omg, don't make me regret :)))
", "votes": 37, "replies": [ { "id": 402118, "author_name": "", "author_url": "", "post_date": "2018-10-11T07:42:30.633000", "content": "", "votes": 0, "replies": [] }, { "id": 402119, "author_name": "mezoganet", "author_url": "", "post_date": "2018-10-11T07:43:36.927000", "content": "I am sure of that about you : \"you could not have sat on until the last day.\"
", "votes": 1, "replies": [] }, { "id": 402190, "author_name": "Yao Zhongwei", "author_url": "", "post_date": "2018-10-11T09:55:09.243000", "content": "Everyone should honor your honesty :D
", "votes": 0, "replies": [] }, { "id": 402242, "author_name": "", "author_url": "", "post_date": "2018-10-11T12:06:56.407000", "content": "", "votes": 0, "replies": [] }, { "id": 402291, "author_name": "", "author_url": "", "post_date": "2018-10-11T13:16:51.730000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 402197, "author_name": "Alexpartisan", "author_url": "", "post_date": "2018-10-11T10:01:49.813000", "content": "Be a man like Jerome
", "votes": 13, "replies": [] }, { "id": 402112, "author_name": "olivier", "author_url": "", "post_date": "2018-10-11T07:36:42.667000", "content": "A big thank you to Jérôme who investigated, released this issue and accepted to submit a leaky submission to prove data was freely available on Google Analytics !
\n\nOptions are (1) the available data is published as a Public Dataset, (2) the competition is reset or (3) the competition is cancelled
\n\nI suspect option 1 will be chosen.
", "votes": 10, "replies": [ { "id": 402117, "author_name": "Konrad Banachewicz", "author_url": "", "post_date": "2018-10-11T07:41:53.697000", "content": "I'm not even sure they will bother with 1 - they might just point people to a link, if past exp is anything to go by.
", "votes": 3, "replies": [] }, { "id": 402194, "author_name": "Shahebaz Mohammad", "author_url": "", "post_date": "2018-10-11T10:01:08.350000", "content": "Dear Kaggle,
\n\nSo many leak-full competitions recently. Santander, Airship prediction and now this.
\n\nI agree that framing problems and testing out for a leak is a really challenging task. But, the recent Airbus problem just had fragments of images and this competition had opened data freely available to look at.
\n\nI believe that causation of such obvious leak might be the communication gap between competition setters and the hosts. It's at an edge and the time that we need a leak-free and smooth going competitions (a few of them too). Too many competitions spoil the broth!
\n\nPeace.
", "votes": -1, "replies": [] }, { "id": 402246, "author_name": "S D", "author_url": "", "post_date": "2018-10-11T12:13:25.750000", "content": "The problem with option 1 is I suspect the data will take us to 0 error...
", "votes": 1, "replies": [] }, { "id": 402257, "author_name": "Allen", "author_url": "", "post_date": "2018-10-11T12:28:08.643000", "content": "Even if they reset the competition, first they would have to plug up the leak, right?
", "votes": 0, "replies": [] }, { "id": 402390, "author_name": "Erik Bruin", "author_url": "", "post_date": "2018-10-11T16:17:21.113000", "content": "Really??? I always suspected this, but did not want to be the person messing up this competition! See screenshot in my EDA that shows 1.8M sessions in Google Analytics while we have 1.7 in the train+test. I thought/hoped they took a random sample somehow.....
", "votes": 0, "replies": [] } ] }, { "id": 402099, "author_name": "beluga", "author_url": "", "post_date": "2018-10-11T07:21:10.130000", "content": "LOL I am glad that I chose doodle recognition :)
", "votes": 6, "replies": [] }, { "id": 402103, "author_name": "mezoganet", "author_url": "", "post_date": "2018-10-11T07:23:55.533000", "content": "Fully agree about your opinion concerning Jerome : integrity and honesty.
\n\nThe question is now : what Kaggle will do within the next hours/days ?
", "votes": 1, "replies": [ { "id": 402121, "author_name": "", "author_url": "", "post_date": "2018-10-11T07:44:50.637000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 402128, "author_name": "Youhan Lee", "author_url": "", "post_date": "2018-10-11T07:53:13.913000", "content": "I agree with you. He saved my time!
", "votes": 2, "replies": [] }, { "id": 403659, "author_name": "Antonio García Morte", "author_url": "", "post_date": "2018-10-14T09:05:10.743000", "content": "Congratulations! :-)
", "votes": 0, "replies": [] }, { "id": 402911, "author_name": "NelsonGon", "author_url": "", "post_date": "2018-10-12T14:54:11.850000", "content": "I was getting my skills up so as to try this as my first featured competition. It's sad that people out there just want to cop paste/ find ways of cheating. Thanks Jerome and I hope Kaggle find a way to solve all this.
", "votes": 0, "replies": [ { "id": 402915, "author_name": "Y Assous", "author_url": "", "post_date": "2018-10-12T15:08:39.757000", "content": "I'm not sure people want to cheat in a malicious way.. This is part of Data Science, finding all the data they can gather, wherever it is (in a legal way, of course...), to achieve a goal (in Kaggle, to have the highest rank for some, for others like you and me, to learn).
", "votes": 0, "replies": [] } ] }, { "id": 402384, "author_name": "Aleksandr Kosolapov", "author_url": "", "post_date": "2018-10-11T16:09:18.027000", "content": "I see a rally at the top LB now for several other participants. Is this just the beginning?)
", "votes": 0, "replies": [ { "id": 402391, "author_name": "Regan Lui ", "author_url": "", "post_date": "2018-10-11T16:19:47.723000", "content": "quite disappointed now. I spent hours on feature engineering to reach a top position but now others can easily achieve an exceptionally high score just by exploiting the leak. I hope the kaggle admins can give an official response asap and handle the situation in a fair and just manner.
", "votes": 4, "replies": [] }, { "id": 402396, "author_name": "Aleksandr Kosolapov", "author_url": "", "post_date": "2018-10-11T16:29:38.907000", "content": "Don't worry. I am sure that the situation will be corrected and everyone will continue to fight honestly.
", "votes": 2, "replies": [] }, { "id": 402656, "author_name": "Andy Li", "author_url": "", "post_date": "2018-10-12T04:19:01.397000", "content": "Kaggle hopefully will give it their best to carry on the competition! Keep it up!
", "votes": 0, "replies": [] } ] }, { "id": 402113, "author_name": "Subrahmanyam V", "author_url": "", "post_date": "2018-10-11T07:38:23.373000", "content": "I haven't fully understood your comment about the leak. \nHowever I see that Jerome's submission is a way ahead of everyone else's.
", "votes": 0, "replies": [ { "id": 402152, "author_name": "Ilya Khristoforov", "author_url": "", "post_date": "2018-10-11T08:46:10.303000", "content": "Jerome actually found a leak of data and made it public. The leak found, in fact, makes the whole contest under a question and the competition could be stopped/reset or conditions could be changed.
\n\nSee the discussions below:\nhttps://www.kaggle.com/c/ga-customer-revenue-prediction/discussion/68235#401880
", "votes": 2, "replies": [] }, { "id": 402219, "author_name": "Subrahmanyam V", "author_url": "", "post_date": "2018-10-11T10:53:00.653000", "content": "Thank you. I followed the thread and now I understand the context.
", "votes": 0, "replies": [] } ] }, { "id": 402727, "author_name": "", "author_url": "", "post_date": "2018-10-12T07:56:21.373000", "content": "", "votes": 0, "replies": [] }, { "id": 402105, "author_name": "savproga", "author_url": "", "post_date": "2018-10-11T07:25:05.807000", "content": "I am very sad =( But, thank you Jerome
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "402078": "Although I always have a tendency to moan about where I think Kaggle gets it wrong - on the whole the community is second to none.\n\nJerome is a great example of this - integrity and honesty in showing the leak that he could have sat on until the last day.\n\nIf this competition is reset I would like to think some Kaggle Swag will be on its way to Jerome", "402110": "Thanks @Scirpus & @mezoganet I appreciate.\n\n> ...that he could have sat on until the last day... <\n\n\nOmg, don't make me regret :)))", "402197": "Be a man like Jerome", "402112": "A big thank you to Jérôme who investigated, released this issue and accepted to submit a leaky submission to prove data was freely available on Google Analytics !\n\nOptions are (1) the available data is published as a Public Dataset, (2) the competition is reset or (3) the competition is cancelled\n\nI suspect option 1 will be chosen.", "402099": "LOL I am glad that I chose doodle recognition :)", "402103": "Fully agree about your opinion concerning Jerome : integrity and honesty.\n\nThe question is now : what Kaggle will do within the next hours/days ? ", "402128": "I agree with you. He saved my time!", "403659": "Congratulations! :-)", "402911": "I was getting my skills up so as to try this as my first featured competition. It's sad that people out there just want to cop paste/ find ways of cheating. Thanks Jerome and I hope Kaggle find a way to solve all this. ", "402384": "I see a rally at the top LB now for several other participants. Is this just the beginning?)", "402113": "I haven't fully understood your comment about the leak. \nHowever I see that Jerome's submission is a way ahead of everyone else's. ", "402727": "", "402105": "I am very sad =( But, thank you Jerome" } }