Abstract

For the problem of content-based information filtering, this paper introduces genetic algorithm to solve this problem, because it could find optimal solutions within global context. In order to illustrate the effectiveness of this approach, it gives a new approach based on set theory, and it also gives an experiment. From the theoretical proof and the experimental results, we could see that the method is feasible and could obtain better information filtering results. With the development of information technology, the Information world provides several network information to computer users, however, people are also inevitably exposed to a lot of spam, while they enjoy the convenience of information. Hence, the network information filtering arises at this historic moment. The text information filtering (Belkin and Croft, 1992; Liddy et al., 1994; Xuan-Jing, 2003) is a process that extracts the texts from a large amount of text stream in order to meet specific user requirements.

Key words: Information filtering, content-based, feasibility analysis, convergence.