Design And Implementation Of Internet Opinions Monitoring And Analyzing System

A sharp development of domestic Internet products has accured for the several decades years,as its advantages such as high transmission speed,huge user base,wide coverage and strong social mobilization ability,the Internet has opened up a new place of public opinion.But some lawless elements spread falsehoods and reactionary statements with this approach even has caused some extremely adverse social impacts.These are the problems coming out with the transformation of information dissemination.Accordingly,to strengthen the supervision of the network public opinion and to enhance the government's ability to control the public opinion correctly is of vital importance to the healthy development of our country in the new period.This paper mainly introduces the design and implementation of Internet public opinion monitoring and analyzing system.Users can access multi-dimensional analysis of custom defined events,such as the proportion of emotional polarities,the development of events.They can also get public opinion reports,public opinion warning and other functions.To do as much as possible to the overall supervision of public opinion,a large number of data acquisition is essential,which is also the basic demand of public opinion system.The data sources of the system include the news website,mobile news client,forums and other Internet public information,data acquisition module achieves custom expansion based on the Heritrix crawler framework.Crawler module has the ability to collect nearly one thousand sites,and produce standardized files for the use of text analysis module.Persistent data storage employs HBase database in which massive data exists.High performance systems need to feedback the user's desired data accurately in the shortest time that is one of the most major factors affecting user experience in general,these rely on an efficient search engine to help.In addition,this paper will introduce the application of search engine Solr in text search and massive data statistics.Solr is an excellent tool in data retrievaling,and will take a very important role in the whole Internet public opinion monitoring and analyzing system.After searching on data acquisition and search engine at home and abroad,learning from the exceptional text sentiment analysis algorithm,and exercising the basic idea of modern software engineering management,the paper gets a general solution that meets needs relevantly.The successful commercial operation of this public opinion monitoring and analyzing system has brought a convenient and efficient internet public opinion monitoring tool for the government departments at all levels,combated the crime of disrupting social stability,promoted the dissemination of positive energy information and made great positive contribution to purifying the network environment and inhibiting the adverse events.