Posts tagged Topic modeling

Trade-off between diversity and precision

Result diversification based on query-specific cluster ranking

0
1 year ago
by in ,

Result diver­si­fi­ca­tion is a retrieval strat­egy for deal­ing with ambigu­ous or multi-faceted queries by pro­vid­ing doc­u­ments that cover as many facets of the query as pos­si­ble. We pro­pose a result diver­si­fi­ca­tion frame­work based on query-specific clus­ter­ing and clus­ter rank­ing, in which diver­si­fi­ca­tion is restricted to doc­u­ments belong­ing to clus­ters that More >

TREC

Heuristic Ranking and Diversification of Web Documents

0
2 years ago
by in ,

We describe the par­tic­i­pa­tion of the Uni­ver­sity of Amsterdam’s Intel­li­gent Sys­tems Lab in the web track at TREC 2009. We par­tic­i­pated in the adhoc and diver­sity task. We find that spam is an impor­tant issue in the ad hoc task and that Wikipedia-based heuris­tic opti­miza­tion approaches help to boost More >

TREC

Topical Diversity and Relevance Feedback

0
2 years ago
by in ,

We describe the par­tic­i­pa­tion of the Uni­ver­sity of Amsterdam’s Intel­li­gent Sys­tems Lab in the rel­e­vance feed­back track at TREC 2009. Our main con­clu­sion for the rel­e­vance feed­back track is that a top­i­cal diver­sity approach pro­vides good feed­back doc­u­ments. Fur­ther, we find that our rel­e­vance feed­back algo­rithm seems to help More >

Go to Top