Semantic expert mining in Social Networks-A Survey release_bwvpqxjj55hohiupqcvgu6xx3u

Abstract

With the emergence of social networks such as micro-blogging services like Twitter, the expert finding has become an interesting topic. However, previous methods cannot be directly used to learn about topic experts in Twitter. Some of the new methods employ the relations among users and Twitter lists for expert finding. A probabilistic method has been developed to explore the relations (i.e. follower relation, user-list relation and list-list relation) for finding experts. To calculate the global authority of users, a Semi-Supervised Graph-based Ranking (SSGR) approach is used. Then a local relevance between users and given query is computed. By understanding the global authority and local relevance of users, all users are ranked and top-N users with highest ranking scores are retrieved which constitute the topic experts in Twitter.
In text/plain format

Archived Files and Locations