| With the development of computer networks and network openness, sharing and interconnection of the degree of expansion, the Internet is increasingly becoming the primary means of information exchange, At the same time, some new Internet business continues to rise, such as e-commerce (Electronic Commerce), digital Monetary (Digital Cash), Internet banking (Network Bank), etc., on a higher network security requirements, makes Internet security Secrets of becoming a new hotspot. Important measure to ensure that network security is one of the user authentication, its purpose is to prevent user identity by some illegal users have stolen. Research for authentication is now a lot, but most of the studies focus on the first authentication of security research, When the user first authenticated, no further authentication, For unauthorized users to steal user passwords and user password login to the system this situation can not do any processing.To solve this problem, this paper proposes a personalized authentication point of view. After the user login gives users the minimum permissions, and to continue to monitor the user to collect the user's personalized information. After the specified time, according to information gathered and then a second user authentication, so that the user identity is the illegal use of this situation, and so the appropriate treatment, reducing identity theft after the user harm.This thesis introduces the Web usage mining techniques, and how to use Web usage mining techniques from the client's log files, cookies and server log files to obtain the user's personalized information, that is, the user browse Web system left behind by the personal habits data Then the decision tree algorithm classification algorithm, personalized information to users to analyze order to establish a user can determine the identity of the decision tree, the next time the user login, you can use the decision tree generated by their secondary authentication. |
PDF Full Text Request