29 days ago
-
Rewrite dataset card in English and add per-split statistics
Unify README to English; document train / silver / gold / test_pool_completion counts including positives and hard negatives. -
Add test_pool_completion diagnostic splits for four dimensions
Adds 100-query pool-completion splits for emotion, intent, theme, and thought; updates dataset card, metadata, version, and checksums.
-
Add ClassicalPoetryRetrieval v1.2.0: ship 4-scale ZH annotation prompts under prompts/
-
Update ClassicalPoetryRetrieval v1.1.1: merge hard negatives into qrels (score 0–3); drop hard_negatives_* configs
-
Update ClassicalPoetryRetrieval v1.1.0: merge hard negatives into qrels (score 0–3); drop hard_negatives_* configs
-
Optimize citation BibTeX and CITATION.cff
-
Optimize citation BibTeX and CITATION.cff
-
Remove legacy-replacement note from dataset card
-
Add ClassicalPoetryRetrieval v1.0.0 (unified emotion/intent/theme/thought 4scale; train/silver/gold; graded 1–3; CC BY-NC-SA 4.0; legacy ClassicalPoetry*Retrieval untouched)
-
initial commit
