Design And Implementation Of The Information Processing Module In Microblog-based Public Opinion Analysis System

Microblog, as a rapid rise of the emerging social network, for the right and false information are intermingled, somebody using microblog to spread rumours, or to disturb the social order occured occasionally.Nowadays, microblog release millions tweets per hour, monitoring and analysing so many tweets only by manual work is impossible, so develop a microblog public opinion analysis system rely on modern automatic text analysis technology for early warning is imminent.The information processing module for a microblog-based public opinion analysis system is designed and developed. First, the overall frame design of microblog public opinion analysis system is introduced, include underlying index, information collection, word segmentation module, open source software and related technology. The analysis method of this system was conducted by keywords, for its use of latent semantic analysis (LSA) also made the corresponding introduction.The overall design, architecture, detailed design and coding of information process module is followed, mainly include real-time statistics, custom statistics, synonyms classfication, microblog influence analysis functions. System to accomplish the main features include:1) Real-time statistics of microblog and early warning;2) All kinds of complex and precise custom statistics;3) Synonyms classfication;4) Tweet analysis and user analysis;5) Providing analysis API function;This system will also play an important role in the future, to provide information support and data sharing for the existing application system of the DZG Platform.