Webis Query-Task-Mapping Corpus 2019 (Webis-QTM-19)

The Webis Query-Task-Mapping Corpus 2019 (Webis-QTM-19) comprises three benchmark datasets on the query-task-mapping problem, which consists of finding the correct task for a new query in a given task-split background query log.

It comprises three subdatasets in separate CSV files, each of which has three columns:

• Query. The query string.
• Source. The source of the query. In all datasets, a source field with value  'google' or 'bing' indicates that the query was derived from query suggestions  from the respective search engine; otherwise, the query is from one of the underlying base corpora:
  • 'lucc' : lucchese:2011
  • 'webis' : stein:2013b
  • 'trc'  : stein:2016a
  • 'trec'  : various collections of TREC queries
  • 'wikihow' : based on titles of wikiHow questions
• Task. The ID of the ground-truth task for the corresponding query.

Further details can be found in reference:
Michael Völske, Ehsan Fatehifar, Benno Stein, and Matthias Hagen. Query-Task Mapping. In 42nd International ACM Conference on Research and Development in Information Retrieval (SIGIR 2019), July 2019. ACM.
http://doi.acm.org/10.1145/3331184.3331286