Research Of Hot-topic Discovery About Network Public Opinion

The influence of network public opinions to the stability of social life is growing.In the past,major Internet public opinion events made people realized that Internet played a major role in social supervision.At the same time,network management is the development direction of network public opinions in recent years.If the Internet public opinion emergency cannot be handled property and the network cannot be achieved timely,it’s likely to cause social public opinions,trigger the bad mood and cause the vigorous behavior.At last pose a threat to network and social order.Therefore,how to find hot topics of the network public opinions is of important theoretical and realistic significance.In view of this,we base on the network public opinions as the breakthrough point,make web news links as the research object,and build the network public opinion hot topic discovery model: 1)Reading relevant materials and summing up the technology and theory of network public opinion monitoring to make it the theoretical support for this article.2)In view of the network public opinion,we use the multi Agent theory of artificial intelligence to design model.3)Using the Page Rank algorithm to design the Collecting-Agent,sorting pages and extracting valuable information through the number of hyperlinks.4)Using space vector text representation model to design the Processing-Agent,and making the information standard and using vector to express it.5)Using LDA-SinglePass topic identification method to design the Analyzing-Agent,and identifying hot topics.6)According to the indicators of topics,designing effective heat calculation model and building the Deciding-Agent,then determining the hot topics.