Diversifying search results using query logs

We study the problem of diversifying search results by ex- ploiting the precious knowledge stored in query logs. When an ambiguous or faceted query is submitted to a search en- gine, the user has often to reformulate it in order to better satisfy her information need by improving the perceived pre- cision of results returned. Our proposal exploits the pres- ence of different "specializations" of queries in query logs to detect the submission of ambiguous/faceted queries, and manage them by diversifying the search results returned to users, in a way that better covers the different possible inter- pretations of the query. We present an original formulation of the results diversification problem in terms of an objec- tive function to be maximized that admits the finding of an optimal solution in linear time. We evaluate our approach by using both the TREC Web diversification track method- ology, and a novel evaluation methodology based on query log analysis.