| Abstract |
Web search queries can be ambiguous or multi-faceted. Diversification techniques deal with this problem by giving a different array of interpretations/topics in the top search results. To perform diversification, first we need to identify the different aspects of the query in the top-k documents. Various methods and resources have been considered: topic modeling, topic extraction, query logs, etc. I will present a work-in-progress in exploring whether semantic information from Wikipedia can help in diversifying the search results. |