Research On The Model And Improvement Of Text Information Filtering In The Internet

Network text information filtering is a branch of information filtering. It mainly acts to search user's interested information in the dynamic text flow according to user's demanded model as on the Internet the basis of information manifestations is the text information flow. Network text information filtering involves a wide range of knowledge, which combines the theory of knowledge, artificial intelligence, natural language understanding and other fields of knowledge. As a new field of study, network text information filtering has been gradually deepening. The technology is increasingly difficult, and its development is also increasing from simulation to reality. With the development of computer applications from vision to reality and the increasing popularity of the Internet today, the network text filtering can effectively solve the problem of information lost in the information services and management aspect, playing an increasingly larger role. In this paper, the network text information filtering model is the key research. It is mainly on the Web filtering model of users and user modeling technology for matching. Then the network text information filtering model is given based on general information filtering application model. The paper also analyses the important components and features of the network information filtering model such as user profiles, document model, user feedback. The commonly used model of the three matching technology which includes Boolean logic model, probability model and vector space model is focused. Finally, the paper analyses the two technical defects and a modified text filtering model is given. The improved text filtering model is based on the vector space model and Hownet dictionary, using VSM which says the text of the flow of information and expression of the needs of users and furthering expansion of the concept to improve user profile and text flow characterization. Finally text filtering model use quality evaluation methods to filter again on the first filter-based text information related to the relevant, enhancing the value of content filtering. Along with the information redundancy and increased demand for information services, text information filtering technology and its application model will be more and more emphasized by people.