Datasets:
Tasks:
Text Classification
Formats:
parquet
Sub-tasks:
multi-label-classification
Languages:
English
Size:
100K - 1M
License:
License tag note and TextGCN 23 class slice
#7
by AxelNoun - opened
Factual note on the cc-by-nc-4.0 tag: it comes from the 2020 loading script (huggingface/datasets PR 1537) filling the template LICENSE field (the TODO is still in the source) and a YAML card. TREC, NLM, and Hersh do not grant CC BY NC. NLM terms apply (courtesy line, no endorsement, data are not current MEDLINE; abstracts may be publisher copyrighted).
Separately we published the 7400 doc TextGCN single label slice. It is not a replacement for this full 348k IR collection: https://huggingface.co/datasets/AxelNoun/ohsumed-23-single-label