摘要
用户行为的可信性研究已成为当前可信计算中的一个重要内容。但是，用户行为的复杂性使得用户行为的可信研究变得非常困难。针对该问题，文中采用数据挖掘算法分析并发现用户正常行为模式和异常状态下的特征规则，并以之对用户历史行为数据进行检测，反映出用户行为的可信性，为系统安全决策提供客观的参考依据，有利于系统加强对不可信用户的安全监控，提高系统安全防范能力。
The study of user behavior trust has now become an important element in the trusted computing. However, the complexity of user behavior makes the research of its trust even more difficult. To address the problem, this paper analyses and classifies the normal user behavior patterns, and abstracts the characteristic rules in abnormal behavior by data mining algorithms. Then the historic behaviors of users are verified according to the patterns and the rules, thus to reflect the credibility of user behavior provide objective information for security decision of the system, help strengthen the monitoring of suspect users, and improve system security.
出处
《信息安全与通信保密》
2009年第8期243-245,249,共4页
Information Security and Communications Privacy
基金
现代通信国家重点实验室基金资助项目（9140C1101050706）
广西信息与通讯技术重点实验室基金资助项目（No.10908）.
关键词
可信计算
数据挖掘
行为可信
trust computing
data mining
behavior trust
