Study On Agent-based For Network Information Content Monitoring

The paper studies the structure,functions and characteristics of network information system for monitoring in the current.It puts forward the structure of network information content supervising based on Agent.The paper elaborated on the principles and techniques that is about the technology of network information monitoring for content getting and content restore.At the same time,the paper introduced the text data mining,text classification and related knowledge.Then it analyses the related theory of information filter.Text classification is important basis for information content analysis.The paper describes the content and procedure of text classification,text pretreatment,text expression model and classification algorithms and other key technologies.And then we designed and implemented an information system.Experimental results applied to text categorization problem show that our the method has high accuracy rate and recall rate.At last,the paper gives a series of analysis and improvement proposal.