Research Of The Discovery Algorithm About The Cyberspace Public Opinion's Hotspot

Now the social and the Internet environment is more and more complicated .In this situation,the Cyberspace public opinion has already caused important influence on people who are surfing on the Internet.The Cyberspace public opinion is different from the traditional one.In regard to the Cyberspace public opinion,their occurrence range is wide,spreading speed is high, and their eruption spot are difficult to be detected and controled.All above situation proved that: It is more important to effectively detect and control the Cyberspace public opinion.In this issue, we begin at the discussion of the technique used in the design and implementation of the WEB public opinion monitoring and reporting system.And then we deeply analysis how to design a effective public opinion monitoring system's architecture, at the same time, what are the proper process step of the system.In this issue, we will introduce the technique used in modules' design and implementation,and these modules are just the componets of the public opinion monitoring and reporting system. They are public opinion information collecting module, information pre-disposal module and public opinion's hotspot detection and topic tracking module.At the end of the issue, we will compare the different text classification and text clustering algorithms using the expirement results. Through the result, it is validated that the design of the public opinion's hotspot detection and topic tracking module is reasonable and necessary.In the issue, we design a WEB public opinion monitoring and reporting system having considered the need of monitoring the Cyberspace public opinion.We make our own contribution to the development of the Internet in a healthy and ordinal way.The main work we do in this issue is listed below: (1)Referencing to various related technique domestic and international, we design a WEB public opinion monitoring and reporting system's architecture;(2)In the system, there are two important assistance modules: the public opinion information collecting module and the information pre-disposal module.We analyze their design and implementation methods separately;(3)The most important core module in the system is the public opinion's hotspot detection and topic tracking module.We give emphasis to the core module's design and implementation methods;(4)We compare different data mining algorithms such as the text classification and the text clustering algorithms.And we know their advantages and disadvantages;(5)Through the result of the experiment, we have proved the superiority of the system's architecture in the system design.