Abstract | ||
---|---|---|
An algorithm is described for ordering by probability of relevance overlapping document subsets from which a searcher should choose the next document. The algorithm produces the ordering without assumptions of index term independence, improves its performance with increasing feedback, and estimates most accurately the probabilities of relevance of the subsets most likely to be relevant. The efficiency and effectiveness of the algorithm are analyzed theoretically. |
Year | DOI | Venue |
---|---|---|
1991 | 10.1016/0306-4573(91)90029-L | Inf. Process. Manage. |
Keywords | Field | DocType |
ranking large document collection,state space search,humanities,engineering,relevance information retrieval,social sciences,state space,computer science,feedback,algorithms,probability,indexing,information needs | Data mining,Information needs,Information retrieval,Ranking,Computer science,Search engine indexing,State space search,Relevance (information retrieval),Probabilistic logic,Index term,State space | Journal |
Volume | Issue | ISSN |
27 | 1 | Information Processing and Management |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Michael D. Gordon | 1 | 1051 | 99.36 |