Author(s)

Yanchen Cui, Peng Zhang, Lizhi Wu

Corresponding Author

Peng Zhang

ABSTRACT

This paper aims to adopt the means and methods of emotional analysis and subject mining to analyze and judge the relation between emotional trend of online public opinions and evolution and change of the situation, so as to provide reference to monitoring and handling of public opinions for the government. The emotional dictionary construction for public opinions of fire services is conducted with PMI-IR algorithm in way of domain dictionary construction, and the subject extraction and emotional analysis are conducted for topic comments related to “nanny arson” in micro-blogs and forums in combination with the public opinion features of domestic fire control. The emotional trend and the focus of netizens can be promptly mastered in the time dimension with the methods of emotional dictionary construction and subject extraction adopted in the paper, thus providing certain guidance to prompt monitoring of public opinions and the perspective selection for information issuance.

KEYWORDS

Data mining, subject extraction, emotional analysis, computer simulation