Locating query-oriented experts in Microblog search

In this paper, we propose an approach to locating query- oriented experts in Microblog. We first define the experts by social inuence and content relevance. Then, we adopt the BM25 model to calculate the content relevance of each ac- count. For the social inuence, we present a global-ranking algorithm as GUserRank and a topic-ranking algorithm as TUserRank after applying the LDA topic model. After that, we output the ranking expertise degree of each candi- date for evaluation. Our experimental results show that the proposed approach is effective and promising. Especially, the topic-ranking algorithm achieves an improvement with 40.11% over the baseline. Furthermore, our approach does not rely on the data sets such that it can be duplicated in many fields.