Microblog Searching Module Based on Community Detection

In this paper, we investigate the current software architecture of Twitter searching, and propose a new Microblog Searching Module (MSM) to retrieve microblog messages. MSM mainly consists of three parts. The first one is community detection with Label-Influence-Algorithm (LIA). We have conducted series of experiments in two data sets downloaded from the Sina-Microblog. And the results show that the modularity measure Q of the communities discovered by LIA is well improved. The second one is extracting microblog tags of a microblog user and the community. The last part is designing a module to expand the query word using the Hownet instead of the exact word match. The application of the Microblog Searching Module proves that the module can search the interesting topic and persons conveniently.