Hi,
\ni am still very confused about the label ambiguity and the many discussions. So if you want a unified solution to it, you obviously create the next discussion :)
In before: Sorry if I missed stuff and this is a stupid approach or it is already clarified somewhere.
\nI would like to propose some minimal example documents. And i would like to know the correct and complete labels for these as they would appear in an \"exhaustively\" labeled test-set (and an explanation would be great as well). I will try to propose these example documents in a way that they cover as many questionable cases as possible:
\nExample Documents:
\n\"alzheimer s disease neuroimaging initiative adni\" is a dataset is refered to by its full name and its abreveation in a single expression.
\"survey of doctorate recipients\" is a dataset that is sometimes (probably mistakenly) refered to as \"survey of doctoral recipients\".
I would update these if additional questions appear. Feel free to suggest additional example documents.
", "messages": [ { "id": "1289824", "postDate": "05/01/2021 12:06:02", "content": "Hi,
\ni am still very confused about the label ambiguity and the many discussions. So if you want a unified solution to it, you obviously create the next discussion :)
In before: Sorry if I missed stuff and this is a stupid approach or it is already clarified somewhere.
\nI would like to propose some minimal example documents. And i would like to know the correct and complete labels for these as they would appear in an \"exhaustively\" labeled test-set (and an explanation would be great as well). I will try to propose these example documents in a way that they cover as many questionable cases as possible:
\nExample Documents:
\n\"alzheimer s disease neuroimaging initiative adni\" is a dataset is refered to by its full name and its abreveation in a single expression.
\"survey of doctorate recipients\" is a dataset that is sometimes (probably mistakenly) refered to as \"survey of doctoral recipients\".
I would update these if additional questions appear. Feel free to suggest additional example documents.
", "rawMarkdown": "Hi,\ni am still very confused about the label ambiguity and the many discussions. So if you want a unified solution to it, you obviously create the next discussion :)\n\nIn before: Sorry if I missed stuff and this is a stupid approach or it is already clarified somewhere.\n\nI would like to propose some minimal example documents. And i would like to know the correct and complete labels for these as they would appear in an \"exhaustively\" labeled test-set (and an explanation would be great as well). I will try to propose these example documents in a way that they cover as many questionable cases as possible:\n\n\nExample Documents:\n1. \"alzheimer s disease neuroimaging initiative adni\" is a dataset is refered to by its full name and its abreveation in a single expression.\n\n2. \"survey of doctorate recipients\" is a dataset that is sometimes (probably mistakenly) refered to as \"survey of doctoral recipients\".\n\nI would update these if additional questions appear. Feel free to suggest additional example documents.", "votes": null }, { "id": "1289831", "postDate": "05/01/2021 12:10:50", "content": "@philculliton @julialane , could you please help me with this?
", "rawMarkdown": "philculliton @julialane , could you please help me with this?", "votes": null }, { "id": "1289859", "postDate": "05/01/2021 12:36:04", "content": "Hi - @philculliton is working on a response.
", "rawMarkdown": "Hi - @philculliton is working on a response.", "votes": null }, { "id": "1289887", "postDate": "05/01/2021 13:13:00", "content": "Hi, maybe it is a stupid question.
\nIs it right that I set the \"pub_title dataset_title dataset_label cleaned_label\" as the input and Id as output.
\nif it is right, why in test.json there are only \"section_title text\" which means there are not input and output.
\nThank you for reading
thanks for the new annotations, that helps a lot. But this question is still open for me. Could you clarify this or point me to an answer if I have overlooked something? Sorry if the answer just takes more time, just want to make sure this is not forgotten.
", "rawMarkdown": "Hi @philculliton @julialane,\n\nthanks for the new annotations, that helps a lot. But this question is still open for me. Could you clarify this or point me to an answer if I have overlooked something? Sorry if the answer just takes more time, just want to make sure this is not forgotten.", "votes": null }, { "id": "1354102", "postDate": "06/17/2021 11:36:11", "content": "\nIt has been 2 months. 6 days to go ;)
", "rawMarkdown": "Hi @philculliton @julialane,\n\nIt has been 2 months. 6 days to go ;)", "votes": null } ], "comments": [ { "id": 1289831, "author_name": "enrilohm", "author_url": "", "post_date": "05/01/2021 12:10:50", "content": "@philculliton @julialane , could you please help me with this?
", "votes": null, "replies": [] }, { "id": 1289859, "author_name": "julialane", "author_url": "", "post_date": "05/01/2021 12:36:04", "content": "Hi - @philculliton is working on a response.
", "votes": null, "replies": [] }, { "id": 1289887, "author_name": "zhangjinyuan", "author_url": "", "post_date": "05/01/2021 13:13:00", "content": "Hi, maybe it is a stupid question.
\nIs it right that I set the \"pub_title dataset_title dataset_label cleaned_label\" as the input and Id as output.
\nif it is right, why in test.json there are only \"section_title text\" which means there are not input and output.
\nThank you for reading
thanks for the new annotations, that helps a lot. But this question is still open for me. Could you clarify this or point me to an answer if I have overlooked something? Sorry if the answer just takes more time, just want to make sure this is not forgotten.
", "votes": null, "replies": [] }, { "id": 1354102, "author_name": "enrilohm", "author_url": "", "post_date": "06/17/2021 11:36:11", "content": "\nIt has been 2 months. 6 days to go ;)
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1289824": "Hi,\ni am still very confused about the label ambiguity and the many discussions. So if you want a unified solution to it, you obviously create the next discussion :)\n\nIn before: Sorry if I missed stuff and this is a stupid approach or it is already clarified somewhere.\n\nI would like to propose some minimal example documents. And i would like to know the correct and complete labels for these as they would appear in an \"exhaustively\" labeled test-set (and an explanation would be great as well). I will try to propose these example documents in a way that they cover as many questionable cases as possible:\n\n\nExample Documents:\n1. \"alzheimer s disease neuroimaging initiative adni\" is a dataset is refered to by its full name and its abreveation in a single expression.\n\n2. \"survey of doctorate recipients\" is a dataset that is sometimes (probably mistakenly) refered to as \"survey of doctoral recipients\".\n\nI would update these if additional questions appear. Feel free to suggest additional example documents.", "1289831": "philculliton @julialane , could you please help me with this?", "1289859": "Hi - @philculliton is working on a response.", "1289887": "Hi, maybe it is a stupid question.\nIs it right that I set the \"pub_title\tdataset_title\tdataset_label\tcleaned_label\" as the input and Id as output.\nif it is right, why in test.json there are only \"section_title\ttext\" which means there are not input and output.\nThank you for reading", "1296887": "Hi @philculliton @julialane,\n\nthanks for the new annotations, that helps a lot. But this question is still open for me. Could you clarify this or point me to an answer if I have overlooked something? Sorry if the answer just takes more time, just want to make sure this is not forgotten.", "1354102": "Hi @philculliton @julialane,\n\nIt has been 2 months. 6 days to go ;)" }, "source": "meta" }