A Kind of Prediction Method of User Behaviour for Future Trustworthy Network

With the increasing development of the computer network application, network security is facing the heavy challenge. Currently, many kinds of the isolated network security measures, such as fire wall and access control, form a rampart around the protected system for defensive purposes. But these methods are not effective to deal with the network attack and destruction, due to these attacks and destructions are various, random, covert and sometimes epidemic. The international research shows that network security is on the way to trustworthy network. Apart from current security mechanism, the future trustworthy network adds network behaviour trust and user behaviour trust, so user behaviour prediction is important and significant for realization of the trustworthy network. The paper discusses how to based on past user behaviour to predict future user behaviour, including Bayesian network modeling, prediction grading, computation of prior probabilities of user behaviour, computation of prior probabilities of user behavior attribute and prediction of future user behaviour. In order to meet needs of different prediction of user behaviour for different purpose, we also discuss that how to get and store user´s past assorted behaviour statistical data. Finally, discuss how to predict user behaviour in real system and how to control user behaviour based on prediction result.