Framework for user identification using writeprint approach

There has been an abrupt increase in the use of social networking sites from the past decade. This has given rise to various issues of social networking sites. The ever rising popularity of social networking site has lured the cyber criminals. The misuse of online messages for illegal purposes has become a serious issue. Thus, it has led to a great increase in threat to social networks. Hence, user identification is necessary. This paper proposes a framework for detecting user identification for social networking website using a writeprint approach. The writeprint uses the behavior type of biometric information for analysis and detection. In writeprint approach the stylometric characteristics is written text of the user, which is used to detect user identity. Firstly, the text is downloaded and features are extracted using text mining or web mining technique. The most of the features are extracted from this text using natural language processing tool. Multi layered perceptron model, a supervised machine learning technique is used for user identity detection.