Rapid Network Public Opinion Prediction Method Based On Principal Component Analysis

Nowadays, all sectors of the community pay much attention to their network public opinions because of the Internet’s growing influences. Unlike the traditional communication methods, the Internet has more openness, anonymity, this makes people prefer to share their ideas over it. With the increasing number of netizens, network public opinions begin to carry influence and spread quickly, widely, but our country lacks relevant regulations and supervisions. In order to guarantee the Internet safety, alleviate the social contradiction, this paper adopts a forecast method which can predict and early warning the network public opinions. The main work of this paper is as follows:(1) Introduce the public opinion and the public network opinion, analyze their characteristics, and explain the significance of predict.(2) According to the characteristics of Internet information dissemination and public opinion on the internet, and on the basis of analyzing the information of network public opinion, Network crawler using heuristic strategy to collection of network public opinion data. To pre process the data of network public opinion, determine the scope of its collection and set up the characteristics of its text information.(3) The indicators system of the internet public opinion is an important basis for making decision to assess network development trend of public opinion. These indicators’ number is large, their structure is complex, their date often exist redundant information, and the calculation of index is large. A method to solve the above mentioned problems based on principal component analysis was presented. The principal component analysis method can combine multiple linear network public opinion indicators into one or several comprehensive indicators which are independent of each other. The quantity of integrated indicators is less than the original index. But they still retain most of the information of original index. Therefore, this method can ensure the accuracy of the premise and improve the efficiency of network public opinion prediction at the same time.