Research And Design On Wireless Internet Negative Information Filter System

With the rapid development of mobile communications, people's demand for mobile communications is no longer the only voice communications, more and more people need to acquire information through mobile telephone terminals as well as wireless communication network anytime and anywhere. Against this background, wireless Internet arises and come into fast-growing, WAP websites also have sprung up. People can obtain relevant information by visiting these WAP websites using mobile telephones anytime and anywhere. However, while people are enjoying the convenience brought about by the WAP Internet, they are also troubled with such WAP sites, which filled with a lot of negative information. The negative information not only disturbs people acquiring normal information, but also brings consumers along the deduction Trap, which leading to the loss of consumers' interests. And the negative information is much more a fact of erosion for the forming course of young people's World view and outlook of life.Therefore, the negative information filtering specifically for wireless Internet is extremely urgent. However, the research and application of negative information filtering is mainly concentrated on the traditional Internet, when it comes to the wireless field, it's relatively rare. On the other hand, the approach to current wireless Internet is comparatively unitary, and the access network is also under control and managed by the mobile telecom operators, so from a practical perspective, it is more effective and easily attainable for telecom operators to build the negative information filtering system in their network.For this purpose, this paper focuses on the topic of negative information filtering for wireless Internet. It considers the actual mobile network structure, designs an integrated structure of negative information filtering system for wireless Internet and its deployment solutions in the mobile network. At the same time, it studies the models and theories of text filtering, and chooses the suitable text classification algorithms for this negative information filtering. Also it carries out some improvements to these algorithms, and put forwards the class-center-vector classification algorithm based on feedback learning. The experiment shows that the feedback learning does improve the performance of classification algorithm. This paper also proposes and designs the WAP websites crawler system, being used to discover negative information websites initiatively, and this crawler system is considered as the complement of the negative information filtering system. Finally, this paper detailedly analyzes and discusses the key technology of realizing the negative information filtering system and the WAP network crawler system. Also, it develops a prototype system and makes some tests.