In today's Online social networks (OSNs) there are basically some fundamental issues occurs while using the social networking sites people have lot of friend on social sites and some of them post unwanted word on there private wall and it sometime disturbed them. We propose a system allowing OSN user have a direct control on the messages posted on their walls. This is achieved through the LDA algorithm which is used in the system, that allows users to customize the filtering criteria to be applied to their walls, and a Machine Learning based LDA algorithm automatically filter the word and doesn't post on the user wall and user get block who post unwanted word on the wall. User have authority to unblock the block user if he/she want to unblock. Index Term - Online Social Networks, Information Filtering, Short Text, Classification Policy-Based Personalization, Facebook API's.