ibm-research/

StatCanDialogueRetrieval

This dataset is part of a Table + Text retrieval benchmark. Includes queries and relevance judgments across test split(s), with corpus in 1 format(s): corpus.

	Configs




	

Config Description Split(s)

default Relevance judgments (qrels): qid, did, score test

queries Query IDs and text test_queries

corpus Plain text corpus: _id, title, text corpus

	TableIR Benchmark Statistics




	

Dataset Structured #Train #Dev #Test #Corpus… See the full description on the dataset page: https://huggingface.co/datasets/ibm-research/StatCanDialogueRetrieval.