Research On Multi-Agent And Swarm Intelligence Tibetan Network Public Opinion Management

Internet Network Information Center report shows the number of China’s Internet users and the Internet penetration rate increased rapidly in recent years, public participation is becoming more common in the Internet life. They are increasingly willing to through the network media to express themselves in work, study and side issues. It is often through the news comment, vote, reproduced form reflected a period of time such as centers of public opinion. Faced to the emergence of Internet big data, rely solely on artificial methods to detect hot public opinion is obvious not enough, it needs a hot topic of intelligent network discovery mechanism for public opinion analysis and data analysis, feedback, and management, in order to achieve public opinion monitoring and the network public opinion to guide. The article is intended to support Tibetan, Chinese and English corpus hot topic discovery and analysis of public opinion, to facilitate information queries on relevant topics to decision-makers.Swarm intelligence is a biological groups emerged relatively strange phenomenon, the biological groups of intelligent monomers between communication, the emergence of a feature collaborate highly self-organization, this paper apply this feature to Tibetan text clustering algorithm, it described Tibetan text clustering method based on swarm intelligence (abbreviated SCAST) in detail. This paper introduces two classic swarm intelligence algorithm of ant colony optimization algorithm, reference to a number of research and theoretical results on the ant colony optimization algorithm was a new application. Combined with multi-agent technology, intelligent group of Tibetan text proposed clustering algorithm of distributed multi-agent (referred DSCAST), this distributed, high performance improved text clustering capabilities. To optimize process of Tibetan text clustering, it improve the efficiency and quality of clustering.Finally, this paper designed and implemented based on multi-agent and swarm intelligence public opinion Tibetan network management system, the system collects Tibetan, Chinese and English portal site data, it clustering to preprocessed text data and identify topics by single-pass clustering algorithm, combined with public opinion analysis algorithms, graphing and show public opinion.