Expertise Retrieval Using Search Engine Results

Expertise retrieval has been largely explored on a few collections crawled from the intranets of organizations. In contrast, only limited external information has been used and studied. In this paper, we have a research on the approaches and effectiveness of expertise retrieval using search engine results. Using appropriate queries, we search for each expert his or her relevant information from the internet and create collections that are quite different from the intranet ones. On such basis, different search queries are compared for the effectiveness of their results. Further, we try on different fields of the results and make a comparison between their effects. Besides, results inside and outside the organization are experimented separately to make clear their different effects. In our experiments, the language modeling approach of expertise retrieval still works well with search engine results. To conclude, we suggest that search engine is an effective source of expertise information and can produce considerable performance in expertise retrieval.