Abstract: Microblog was used to publish important information on BBS in many universities in order to spread positive energy in campus. Studying on an actual university microblog, a system of gathering and editing information from microblog was proposed. On this basis, user interactive social behavior between microblog and BBS was analyzed, and tightness and intimacy to discover the potential friendship and the degree of attention between users on microblog were also proposed. The experiments show that it is useful to extract topic from a single microblog using special punctuation mark. At last, different methods based on emoticon and different dictionary in sentiment analysis were compared, coming to the conclusion that the method based on emoticon and cyber word dictionary performs well.