A Social Stability Analysis System Based on Web Sensitive Information Mining

Researches on domestic social stability analysis mainly focus on construction of social stability theory, architecture and index, while few pay attention on quantitative analysis. In this paper, a social stability supervising framework is proposed based on sensitive Web information mining, semantic pattern matching and quantitative calculating. A sensitive information knowledge base is constructed by analyzing sensitive information about social environment, national harmonious and happy index of people’s live in natural language online news texts from Internet, and recognizing hot keywords as well as the event trends led by the keywords. A social stability index theoretic model and a quantitative calculating model are proposed to evaluate social stability quantitatively. Parameters of the calculating model are determined by employing social investigations and an iterative feedback learning method. A prototype system is built on proposed framework and experiments are conducted on six frontier provinces, e.g., Xinjiang and Tibet. The result of an average accurate of 73.29 % shows the effectiveness of the proposed model.