![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F9788695%2Fd261ca0c5741fc93f8a4c15afae865f1%2FScreenshot%202023-07-19%20at%203.49.48%20PM.png?generation=1689807495409854&alt=media\")
I've started to see some submissions that are clearly academic papers or have links to their paper via arXiv. They aren't necessarily peer reviewed, but I came across one that is attached to a github repo that has commit history going back over a year from now. Another is a paper from the Stanford department of CS and has likely been under development for some time. What are the rules regarding the previous publications? This competition has only been open since May, so submitting something that was in the works before then seems like an unfair advantage.
\nJust wanted to point this out and see if there are any concrete rules around it? If I came into this competition knowing that I was competing with academic papers that had already been in the works for almost a year, I may not have even bothered writing an essay. 🤷♂️
\nIt was brought to my attention in another post that there is a dataset containing the official submissions made by Kaggle staff. The dataset can be found here. This can help with selecting essays to review.
", "messages": [ { "id": "2338462", "postDate": "07/10/2023 19:24:49", "content": "I've started to see some submissions that are clearly academic papers or have links to their paper via arXiv. They aren't necessarily peer reviewed, but I came across one that is attached to a github repo that has commit history going back over a year from now. Another is a paper from the Stanford department of CS and has likely been under development for some time. What are the rules regarding the previous publications? This competition has only been open since May, so submitting something that was in the works before then seems like an unfair advantage.
\nJust wanted to point this out and see if there are any concrete rules around it? If I came into this competition knowing that I was competing with academic papers that had already been in the works for almost a year, I may not have even bothered writing an essay. 🤷♂️
\nIt was brought to my attention in another post that there is a dataset containing the official submissions made by Kaggle staff. The dataset can be found here. This can help with selecting essays to review.
", "rawMarkdown": "I've started to see some submissions that are clearly academic papers or have links to their paper via arXiv. They aren't necessarily peer reviewed, but I came across one that is attached to a github repo that has commit history going back over a year from now. Another is a paper from the Stanford department of CS and has likely been under development for some time. What are the rules regarding the previous publications? This competition has only been open since May, so submitting something that was in the works before then seems like an unfair advantage.\n\nJust wanted to point this out and see if there are any concrete rules around it? If I came into this competition knowing that I was competing with academic papers that had already been in the works for almost a year, I may not have even bothered writing an essay. 🤷♂️\n\n\n### UPDATE: \n\nIt was brought to my attention in another [post](https://www.kaggle.com/competitions/2023-kaggle-ai-report/discussion/422613) that there is a dataset containing the official submissions made by Kaggle staff. The dataset can be found [here](https://www.kaggle.com/datasets/paultimothymooney/list-of-public-notebooks-ai-report-competition). This can help with selecting essays to review.", "votes": null }, { "id": "2339669", "postDate": "07/11/2023 02:15:12", "content": "yes, I also saw one submission that basically copy the content of their own published paper. I am not sure but I think it should be counted as self-plagiarism
", "rawMarkdown": "yes, I also saw one submission that basically copy the content of their own published paper. I am not sure but I think it should be counted as self-plagiarism", "votes": null }, { "id": "2351205", "postDate": "07/19/2023 22:58:50", "content": "As the lead author of the \"Mini Giants\" paper which is likely one of the \"academic papers\" you are referring to, I assure you it was composed specifically for and after the release of this competition. We released a pdf version along with the notebook simply because overleaf is my preferred tool for collaboration. Our Overleaf record, with its initial edits dating back to June, serves as a proof for this statement.
\n
I support strict rules against plagiarisms. Kaggle should disqualify any paper that violates.
", "rawMarkdown": "I support strict rules against plagiarisms. Kaggle should disqualify any paper that violates.", "votes": null } ], "comments": [ { "id": 2339669, "author_name": "muhammadirfanakbar", "author_url": "", "post_date": "07/11/2023 02:15:12", "content": "yes, I also saw one submission that basically copy the content of their own published paper. I am not sure but I think it should be counted as self-plagiarism
", "votes": null, "replies": [] }, { "id": 2351205, "author_name": "mistylight", "author_url": "", "post_date": "07/19/2023 22:58:50", "content": "As the lead author of the \"Mini Giants\" paper which is likely one of the \"academic papers\" you are referring to, I assure you it was composed specifically for and after the release of this competition. We released a pdf version along with the notebook simply because overleaf is my preferred tool for collaboration. Our Overleaf record, with its initial edits dating back to June, serves as a proof for this statement.
\n
I support strict rules against plagiarisms. Kaggle should disqualify any paper that violates.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2338462": "I've started to see some submissions that are clearly academic papers or have links to their paper via arXiv. They aren't necessarily peer reviewed, but I came across one that is attached to a github repo that has commit history going back over a year from now. Another is a paper from the Stanford department of CS and has likely been under development for some time. What are the rules regarding the previous publications? This competition has only been open since May, so submitting something that was in the works before then seems like an unfair advantage.\n\nJust wanted to point this out and see if there are any concrete rules around it? If I came into this competition knowing that I was competing with academic papers that had already been in the works for almost a year, I may not have even bothered writing an essay. 🤷♂️\n\n\n### UPDATE: \n\nIt was brought to my attention in another [post](https://www.kaggle.com/competitions/2023-kaggle-ai-report/discussion/422613) that there is a dataset containing the official submissions made by Kaggle staff. The dataset can be found [here](https://www.kaggle.com/datasets/paultimothymooney/list-of-public-notebooks-ai-report-competition). This can help with selecting essays to review.", "2339669": "yes, I also saw one submission that basically copy the content of their own published paper. I am not sure but I think it should be counted as self-plagiarism", "2351205": "As the lead author of the \"Mini Giants\" paper which is likely one of the \"academic papers\" you are referring to, I assure you **it was composed specifically for and after the release of this competition**. We released a pdf version along with the notebook simply because overleaf is my preferred tool for collaboration. Our Overleaf record, with its initial edits dating back to June, serves as a proof for this statement. \n\n", "2351332": "I support strict rules against plagiarisms. Kaggle should disqualify any paper that violates." }, "source": "meta" }