Research on Crime Degree of Internet Speech Based on Machine Learning and Dictionary

Abstract

Intelligent security technology provides important clues and basis for case detection, however, the traditional intelligent security technology can not alarm a crime's happening in advance. By researching the relationship between criminal psychology and speech feature, we proposed a crime degree theory of internet speech which can alarm a crime's happening in advance by taking advantage of the internet speech. The theory which was based on criminal psychology, using multiple analytical methods such as machine learning and emotional dictionary, establishing the mathematical model and theoretical framework, gave the preliminary implementation method of the real system. Experimental result has shown that the pre-alarming system based on our theory has good pre-alarming capability against crime.