The research of sina malicious comments detection based on semantic information and stop-words table

The research of malicious comments in sina weibo is very important. Because a large number of malicious comments seriously undermine the user experience in sina weibo. Based on the malicious comments detection technology named semantic information, this paper gives a different technology which improves the process of malicious dictionary construction and the process of malicious comments detection determination. In order to improve the accuracy of malicious comments detection, we use a stop-words table which fits sina environment. In addition, we propose a simple recovery mechanism based on vocabulary use frequency to avoid the explosive growth of the dictionary.