CIRAL/ciral-corpus

CIRAL is a collection for cross-lingual information retrieval research across four (4) African languages. The collection comprises English queries and query-passage relevance judgements manually annotated by native speakers. This dataset stores passages which have been culled from news websites for CIRAL. This dataset is configured by language. An example of a passage data entry is { 'docid': 'DOCID#0#0',