Combining Candidate and Document Models for Expert Search

We describe our participation in the TREC 2008 Enterprise track and detail our lan- guage modeling-based approaches. For document search, our focus was on query expansion us- ing profiles of top ranked experts and on docu- ment priors. We found that these techniques result in small, but noticeable improvements over our baseline method. For expert search, we combine candidate- and document-based models, and also bring in web evidence. We found that the com- bined models significantly and consistently out- performed our very competitive baseline models.