Hot Topic Mining And Opinion Analysis Of BBS In Universities

As a traditional web applications, BBS is still one of the important platform for university students to exchange information in the network. The students on BBS can be conveniently express the demands, to show self, access to resources. Groups of students is one of the hot social events, reflecting the active exchange of positive group, but in the network information authenticity, the bad became mixed with the good, when groups of students face some things, if the schools have not pay enough attentions and timely solve, it is easy to cause discontent among the students of the group. And because of the anonymity, dissemination of information dissemination network convenient fast, easy network emergencies, therefore, the BBS has been the focus of school students public places.However, due to various universities for the network information security consciousness is not an uneven, management ability. Especially the most of the universities are mainly by artificial monitoring, to judge the information scale are not the same, it is easy to appear on an important public opinion information. At the same time, artificial monitoring often fails to find the key topic when the issue is inthe latency and the gestation period, lack of forecasting ability for emergencies, often found the key topic in the topic of the outbreak period to.Data Mining can screen out the credible, useful information which we need from the massive data, it can also find the hidden information from observed information, eventto judge the data evolution was predicted by trend. Now, data mining technology has been in the network information processing and network behavior has a lot of applications. We can use the data mining technology, based on real-time data collection, processing and analysis of the BBS information, forecast and observation on the hot topic may appear in BBS and emergencies.Methods based on the data mining, information on BBS analysis, distinguish the hot topic in our care and emergency, as a powerful reference network public opinion monitoring.The main research contents of this paper are as follows:A, data preprocessing:to analyze the information by data mining, the basic work is to preprocess the text data in BBS, including Chinese word segmentation, text feature extraction and text clustering etc..Second, calculation the heat calculation of text:as th judgment of hot topic, we need to be calculated by text browsing density and reply density, to score the BBS and each subject, whether the judgment in BBS hot topic, and then judge what specific topic belongs to the hot topic.Third, calculation the text affective computing:in the extraction of the emotional words on the basis of BBS, to appear in the theme and the reply of emotional value calculation, determine the intensity of emotion and emotional tendency of text, as emergency warning evaluation reference.