The Research And Application Of Network Public Opinion Monitoring And Control System Design

In addition to the characteristics of social public opinion, such as instant,concealment, randomness, divergenty and permeability, Internet also has its ownunique characteristics: sudden, interactivity, richness and deviation. Recently, thenumber of information has increased exponentialy. How to detecting a influence topicon events from massive data, and how to tracking these hot topics of the networkpublic opinion effectively control have become very important.In this paper, the key technology of the network public opinion monitoringmanagement system is studied, including network public opinion informationacquisition, Webpage text preprocessing, topic detection and topic tracking, and thenetwork public sentiment orientation analysis. Among them, the informationacquisition of network public opinion is extracting the loosely structured information,including web crawler and the Meta search engine technology. Webpage textpreprocessing is to noise in information filtering Webpage, including denoising,segmentation, and feature extraction. Topic detection and topic tracking is a processof text clustering and text classification in essence. Network public opinionorientation analysis includes two kinds of text orientation research technologis,semantic-based research and machine learning-based research.Based on the above several key technology research, this paper proposed aframework for public opinion monitoring system, introduced the prototype systemuses key technologies and development tools, and designed a prototype system.Completed the overall design and security system design is based on B/S networkpublic opinion monitoring and management systems and realized the real-timemonitoring of the emergent event and sensitive information on the Internet. And at theend, this paper summarized the main content of this paper studied and areas forimprovement.