Awesome Paper 2016-06-1

Input: Query and All related documents process:

  1. Tokenize all documents and query into a vectors. input: all the query and documents output: the vectors corresponding to query and documents.

  2. Evaluate how extend every document satisfies query by BM25. input: all vectors from 1 step. output: the p(R=1| d,q) of every document.

  3. Rank every document by qPRP and PT. input: the p(R=1| d,q) output: the rank level of every document.

  4. evaluate the order input: the rank from step 3 and how much a document satisfies the query in reality. output: the precise rate.

output: the order and the evaluate based how much a document satisfies the query in reality.

Phone

07318457661

Address

National University of Defense Tecnology
Changsha, Hunan 410073
China