Design And Implementation Of The Internet Public Opinion Monitoring System

With the rapid development of Internet technology,the information on the Internet is growthing explosively,public opinion information monitoring will play a very important role,which can find out vulgar,malicious and inductive information from the Internet.Because of the huge amount of information,it has been difficult for the traditional form of manual supervision to have effective coverage on the Internet public opinion,which have brought hidden dangers to the Internet security and even social stability,so the Internet public opinion monitoring system has become imminent.In this thesis,we designed a Internet public opinion monitoring system to process and analyze public opinion within the specified range of Internet according to the content that users concerned,then provide users with public opinion condition and trends analysis on the application interface.We choose the most used cross-platform programming language-Java,relying on SSH framework(Struts +Spring + Hibernate)to develop this system,the database architecture that we used is MongoDB plus MySQL which is fresh,that not only improves the read-write speed of data while taking into account the system's transaction processing capabilities.We use WebCollector which is an open source crawler framework to develop web crawler that can crawl the content of websites,we use Aho-Corasick algorithm to pretreat the content that crawled from websites and use the open source Word segmentation component to participle the public opinion information.We do polar emotion analysis on the public opinion information with the method that relying on emotion dictionary,finally we display the result of processing and analysis through the application interface to users.The rapid development of the Internet makes the vocabulary of the system in the process of running the system will become bigger and bigger,commonly used string matching algorithm has been unable to meet the requirements of the normal operation of the system.In this thesis,we improved the efficiency of the matching through improving the existing string matching algorithm to meet the performance requirements of the system.The monitoring system in this thesis mainly includes public opinion information statistics;addition and deletion of monitoring condition;the display of details of public opinion;comprehensive analysis of public opinion trend and so on.The system is stable and the interface is simple and friendly.