The Design And Implementation Of Network Content And Public Opinion Monitoring System

Currently, the Internet, as an important tool for information exchange, shouldering the responsibilities of global information transmission. With the rapid development of the Internet, Internet public opinion monitoring technology has become an important part of the research. However, domestic researches on network monitoring public opinion is mostly on the monitoring level and few on the control of the spread. Based on Nanjing research project-mobile Internet network monitoring and security management system (Item number TJ21303), the thesis designs and implements a network public opinion monitoring system which includes monitoring, control, local backtracking and statistical analysis.The thesis briefly introduces the basic theories related to network public opinion monitoring system, including network data capture technology, public opinion analysis mining and web development technology. According to user needs, the thesis puts forward deployment and overall structural design of the system.Then, the thesis designs each module, implementing data acquisition, denoising, text extraction and Chinese word segmentation. This thesis focuses on the topic clustering, negative public sentiment warning and hot spots monitoring. After convert text feature vector model by TF-IDF, improved single-pass clustering algorithm achieves topic monitoring. Matching negative public opinion library achieves negative public opinion monitoring. In the thesis, Libnids achieves reduction of text. Iptables achieves policy control. Three-tier architecture model to achieve control and management front end. Echarts show public opinion data through charts.Finally, the thesis describes the development environment, functional testing and interface display of the system, verifying the availability of the design.