Intranet Threat Detection Technology Based on Non-frequent Pattern

The existing network information security is mainly focused on the classification of gateway or network perimeter and so on. The lack of measures relevant to security threat from intranet hosts leads to information divulging with great loss. The intranet should be of high security. And it´s a latent threat when the intranet users visit the website with illegal motives. In view of these, the security threat mainly from the intranet is analyzed in the paper. Kripke structure is adopted to describe the mathematical model for intranet security system. Non-frequent pattern discovery algorithm based on matrix is designed. And a new intranet-threat monitoring model, which is suitable for dense network, is proposed. Thus the problem of effective monitoring and warning for intranet security is solved.