The Design And Implementation Of Information Theft Early Warning System Based On User Behavior

With the development of the internet, more and more people receive and send theinformation through the network. Because of the loopholes in the network system, moreand more criminals steal the user’s information through the network. According to thestatistics of domestic security sector, in the last three years tens of thousands of serversin the important sectors are controlled; more than three thousand important email wasstolen; millions of documents were theft. The theft documents investigated by thedomestic sectors in the past two years are ten times more than the files acquired throughartificial work in the past sixty years. These lead a major threat to information security.Therefore it is necessary to propose an information theft behavior analysis methodwhich can help the early warning of information theft behavior, and protect thelegitimate rights and interests.The theft warning system which is based on user behavior puts forward the methodand system. This system can discover information theft behavior and give early warningthrough the analysis of user behavior. It conductes a comprehensive research oninformation theft behavior in this paper. First, it builds the model of the user behaviorthrough support vector machine learning methods under the user’s daily internetbehavior. Then it conducts the division of the user behavior through clusteringalgorithm.This paper proposes a method to capture information theft behavior which is thecombination of the feature-based matching method and depth analysis method. Itexcavates the abnormal user behavior accurately through feature matching which iscompleted through the feature code library and scanning algorithms. Then it digs morethrough depth analysis based on user behavior model which can classify the history ofuser behavior into record category. When it’s impossible to classify, the abnormalbehavior occurs. Meanwhile this paper proposes an alert method for information thefteasy warning through the combination of SMS and email.Through the methods above, we conducte the design and implementation. Thesystem obtaines the achievement of the capture for user behavior data, the catch ofinformation theft behavior and the user’s real-time warning. It gets the defense of theinvasion effectively. And it avoides the illegal theft of user information.