Abstract:
Nowadays, large scale spreading of the pornographic and rumour content on the Internet has been a serious problem of network security. The authors give a survey of the main network information content audit techniques, which usually sniffers network packets from key point of network, filters packets to find the contents against security policy, and could effectively prevent the spreading of harmful contents. With the viewpoint from global to local and from bottom to top, the key issues in content auditing are discussed. Firstly, the current works of the auditing model are introduced and the main deficiencies of existing models are presented. Secondly, the techniques for auditing data capturing and load balancing are analyzed. Thirdly, the development of content analysis technologies are introduced, such as, pattern matching, text semantics analysis, hot topic extraction, malicious images recognition, etc. Fourthly, the methods of evaluation and prediction of the content security situation, and technologies of data online processing and blocking are discussed. Through analyzing the open problems and issues, which include streaming video content auditing, key words dynamic updating, analysis of dynamic information flow characteristics, and dynamic propagation prosess of network information, the directions and suggestions for further research are proposed as the concluding remarks.