Abstract:
In this paper we have considered mixed language model that is used for experts finding in areas such as social network analysis and information retrieval, and proposed an adaptation of this model for the social network Twitter. We also have reviewed Twitter popularity metrics and proposed Twitter users' popularity estimation approach based on expert finding, which allows to rank users according to the probability of user being an expert in given query, and have implemented a prototype for data collection and popularity estimation, based on our approach.
Keywords:social network analysis, information retrieval, data mining, expert finding, popularity analysis.
Fulltext:
PDF file (603 kB)