Trec CAR Dataset
Information RetrievalEnglish
Trec CAR Dataset is a information retrieval dataset in English from Dietz et al. with ~285,000 records in CBOR format.
About Trec CAR Dataset
Dataset contains topics, outlines, and paragraphs that are extracted from English Wikipedia (2016 XML dump). Wikipedia articles are split into the outline of sections and the contained paragraphs.
Details
- Task
- Information Retrieval
- Language
- English
- Format
- CBOR
- Rows / instances
- ~285,000
- Creator
- Dietz et al.
- Year
- 2019