Paper Title
Content-Based Filtering of Undesired Hindi Messages from Online Social Network Users Wall

Online Social Networks (OSN) are popularly used for sharing data, information or knowledge among the people having similar interests. Today’s OSN like, Facebook classifies the messages based on sender’s relationship with the receiver. The challenge is to give OSN users the ability to filter a message posted on their own private space based on its content, like blocking a political message. In this paper, we propose Filtering Undesired Hindi Messages (FUHM) Algorithm that allows users to define their own filtering criteria for messages posted on their wall. This is attained by a rule based system and a text classifier. Time taken in classification process adds to the efficiency of overall system. Proposed algorithm is time efficient as compare to previously available methods as we use multiclass classification, classifying message into different categories in one step. Keywords - Content based Filtering, Message Filtering, Naive Baye’s Classifier, Online Social Network, Text Classification.