基于复杂网络方法的舆情热点挖掘
黄敏;胡学钢
【期刊名称】《计算机仿真》 【年(卷),期】2011(028)009
【摘要】This paper focused on the Internet public opinion hot spot mining based on the complex network theory and methods. Internet public opinion analysis techniques can be divided into content-based analytical methods and analyses based on data mining methods. Both are traditional technology and have no relation with the network. This paper gave the idea of public opinion analyses and proposed a network-based approach to resolve network problems. The public opinion web pages could be represented as nodes of complex networks and the sides were links of those pages. The public opinion network meets the complex network characteristics. Based on Hits algorithm and PageRank algorithm, we can find out the most popular public opinions in the network, which showed a hot public opinion. With a data set from Wikipedia, experimental results show that both methods can identify hot spots of public opinion networks. Because PageRank method focuses on the impact of link nodes and Hits algorithm focus on the core degree of the nodes, the mining results have some difference. We can combine the mining results of the two methods according to the field characteristics to get the Internet public opinion
基于复杂网络方法的舆情热点挖掘
