{ "id": 235297, "title": "What is \"the mention of datasets\"", "url": "/competitions/coleridgeinitiative-show-us-the-data/discussion/235297", "author_name": "", "post_date": "2021-04-28T18:30:56.124649900Z", "votes": 8, "comment_count": 7, "views": 0, "content": "

The purpose of this competition, according to the Data Description Page, is as follows
\nThe objective of the competition is to identify the mention of datasets within scientific publications.

\n

There have been some similar discussions in the discussions, but I haven't really understood them yet!

\n

Question: HOW TO ANNOTATE
\nEx1
\nThe ADNI-related labels in the existing dataset include

\n\n

However, if you want to take a closer look, the following should be added

\n\n

(Since the score was higher when the \"alzheimers disease neuroimaging initiative\" and \"adni\" were also labeled than when only \"adni\" was labeled, it is unlikely that a single label will cover the same data set with different names.)

\n

Ex2
\nOne of the NSF-NCSES related labels is

\n\n

However, in practice, because it can come out as follows

\n
\n

National Science Foundation, National Center for Science and Engineering Statistics, Survey of Doctorate Recipients, 2010, and National Survey of College Graduates

\n
\n

So the following are also candidates for labels

\n\n

To what extent should we predict dataset names (longest match, exhaustive, or partial)?

", "messages": [ { "id": "1287152", "postDate": "04/28/2021 18:30:56", "content": "

The purpose of this competition, according to the Data Description Page, is as follows
\nThe objective of the competition is to identify the mention of datasets within scientific publications.

\n

There have been some similar discussions in the discussions, but I haven't really understood them yet!

\n

Question: HOW TO ANNOTATE
\nEx1
\nThe ADNI-related labels in the existing dataset include

\n\n

However, if you want to take a closer look, the following should be added

\n\n

(Since the score was higher when the \"alzheimers disease neuroimaging initiative\" and \"adni\" were also labeled than when only \"adni\" was labeled, it is unlikely that a single label will cover the same data set with different names.)

\n

Ex2
\nOne of the NSF-NCSES related labels is

\n\n

However, in practice, because it can come out as follows

\n
\n

National Science Foundation, National Center for Science and Engineering Statistics, Survey of Doctorate Recipients, 2010, and National Survey of College Graduates

\n
\n

So the following are also candidates for labels

\n\n

To what extent should we predict dataset names (longest match, exhaustive, or partial)?

", "rawMarkdown": "The purpose of this competition, according to the Data Description Page, is as follows\n`The objective of the competition is to identify the mention of datasets within scientific publications.`\n\nThere have been some similar discussions in the discussions, but I haven't really understood them yet!\n\n**Question: HOW TO ANNOTATE**\nEx1\nThe ADNI-related labels in the existing dataset include\n* adni\n* alzheimer s disease neuroimaging initiative adni\n* alzheimers disease neuroimaging initiative\n\nHowever, if you want to take a closer look, the following should be added\n* alzheimers disease neuroimaging initiative adni\n* alzheimer s disease neuroimaging initiative\n\n(Since the score was higher when the \"alzheimers disease neuroimaging initiative\" and \"adni\" were also labeled than when only \"adni\" was labeled, it is unlikely that a single label will cover the same data set with different names.)\n\nEx2\nOne of the NSF-NCSES related labels is\n* National Center for Science and Engineering Statistics, Survey of Doctorate Recipients\n\nHowever, in practice, because it can come out as follows\n> National Science Foundation, National Center for Science and Engineering Statistics, Survey of Doctorate Recipients, 2010, and National Survey of College Graduates\n\nSo the following are also candidates for labels\n* survey of doctorate recipients\n* survey of doctorate recipients 2010\n* survey of doctorate recipients 2010 and national survey of college graduates\n* national science foundation national center for science and engineering statistics survey of doctorate recipients\n* national science foundation national center for science and engineering statistics survey of doctorate recipients 2010\n* national science foundation national center for science and engineering statistics survey of doctorate recipients 2010 and national survey of college graduates\n\n**To what extent should we predict dataset names (longest match, exhaustive, or partial)?**", "votes": null }, { "id": "1287266", "postDate": "04/28/2021 20:59:28", "content": "

The answer to this question is: nobody knows 😔

", "rawMarkdown": "The answer to this question is: nobody knows 😔", "votes": null }, { "id": "1287356", "postDate": "04/29/2021 01:34:43", "content": "

Oh My God!!!

", "rawMarkdown": "Oh My God!!!", "votes": null }, { "id": "1287388", "postDate": "04/29/2021 03:19:35", "content": "

@philculliton said that he is discussing with the host and will provide detailed answers for these questions (including score calculation when short/long labels are used).

", "rawMarkdown": "philculliton said that [he is discussing with the host](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/228322#1270869) and will provide detailed answers for these questions (including score calculation when short/long labels are used).", "votes": null }, { "id": "1287398", "postDate": "04/29/2021 03:35:50", "content": "

That's my reaction too…

", "rawMarkdown": "That's my reaction too...", "votes": null }, { "id": "1287935", "postDate": "04/29/2021 14:41:23", "content": "

Oh, Good discussion. I had overlooked this.
\nI hope he will answer it.

", "rawMarkdown": "Oh, Good discussion. I had overlooked this.\nI hope he will answer it.", "votes": null }, { "id": "1295985", "postDate": "05/06/2021 21:13:13", "content": "

That's definitely the best answer I could find… please also have a look here to try to fix it

", "rawMarkdown": "That's definitely the best answer I could find... please also have a look [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972) to try to fix it", "votes": null }, { "id": "1296101", "postDate": "05/07/2021 01:28:08", "content": "

Good suggestion. I was thinking that would be better too.

\n
\n

only ONE label for EACH mention.

\n
", "rawMarkdown": "Good suggestion. I was thinking that would be better too.\n> only ONE label for EACH mention.", "votes": null } ], "comments": [ { "id": 1287266, "author_name": "hannes82", "author_url": "", "post_date": "04/28/2021 20:59:28", "content": "

The answer to this question is: nobody knows 😔

", "votes": null, "replies": [ { "id": 1287356, "author_name": "ht5brer", "author_url": "", "post_date": "04/29/2021 01:34:43", "content": "

Oh My God!!!

", "votes": null, "replies": [] }, { "id": 1287398, "author_name": "leecming", "author_url": "", "post_date": "04/29/2021 03:35:50", "content": "

That's my reaction too…

", "votes": null, "replies": [] }, { "id": 1295985, "author_name": "eguidotti", "author_url": "", "post_date": "05/06/2021 21:13:13", "content": "

That's definitely the best answer I could find… please also have a look here to try to fix it

", "votes": null, "replies": [] }, { "id": 1296101, "author_name": "ht5brer", "author_url": "", "post_date": "05/07/2021 01:28:08", "content": "

Good suggestion. I was thinking that would be better too.

\n
\n

only ONE label for EACH mention.

\n
", "votes": null, "replies": [] } ] }, { "id": 1287388, "author_name": "sirishks", "author_url": "", "post_date": "04/29/2021 03:19:35", "content": "

@philculliton said that he is discussing with the host and will provide detailed answers for these questions (including score calculation when short/long labels are used).

", "votes": null, "replies": [ { "id": 1287935, "author_name": "ht5brer", "author_url": "", "post_date": "04/29/2021 14:41:23", "content": "

Oh, Good discussion. I had overlooked this.
\nI hope he will answer it.

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1287152": "The purpose of this competition, according to the Data Description Page, is as follows\n`The objective of the competition is to identify the mention of datasets within scientific publications.`\n\nThere have been some similar discussions in the discussions, but I haven't really understood them yet!\n\n**Question: HOW TO ANNOTATE**\nEx1\nThe ADNI-related labels in the existing dataset include\n* adni\n* alzheimer s disease neuroimaging initiative adni\n* alzheimers disease neuroimaging initiative\n\nHowever, if you want to take a closer look, the following should be added\n* alzheimers disease neuroimaging initiative adni\n* alzheimer s disease neuroimaging initiative\n\n(Since the score was higher when the \"alzheimers disease neuroimaging initiative\" and \"adni\" were also labeled than when only \"adni\" was labeled, it is unlikely that a single label will cover the same data set with different names.)\n\nEx2\nOne of the NSF-NCSES related labels is\n* National Center for Science and Engineering Statistics, Survey of Doctorate Recipients\n\nHowever, in practice, because it can come out as follows\n> National Science Foundation, National Center for Science and Engineering Statistics, Survey of Doctorate Recipients, 2010, and National Survey of College Graduates\n\nSo the following are also candidates for labels\n* survey of doctorate recipients\n* survey of doctorate recipients 2010\n* survey of doctorate recipients 2010 and national survey of college graduates\n* national science foundation national center for science and engineering statistics survey of doctorate recipients\n* national science foundation national center for science and engineering statistics survey of doctorate recipients 2010\n* national science foundation national center for science and engineering statistics survey of doctorate recipients 2010 and national survey of college graduates\n\n**To what extent should we predict dataset names (longest match, exhaustive, or partial)?**", "1287266": "The answer to this question is: nobody knows 😔", "1287356": "Oh My God!!!", "1287388": "philculliton said that [he is discussing with the host](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/228322#1270869) and will provide detailed answers for these questions (including score calculation when short/long labels are used).", "1287398": "That's my reaction too...", "1287935": "Oh, Good discussion. I had overlooked this.\nI hope he will answer it.", "1295985": "That's definitely the best answer I could find... please also have a look [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972) to try to fix it", "1296101": "Good suggestion. I was thinking that would be better too.\n> only ONE label for EACH mention." }, "source": "meta" }