With the rampant usage of social media, fraudsters try to employ malevolent social bots that tend to generate counterfeit tweets and try to establish relationships with other users on the social media by acting like followers or try to generate multiple counterfeit accounts that get involved in malevolent activities. They also tend to post malevolent URLs that are used to navigate genuine users to malevolent web servers. Thus it is very essential to differentiate the bot accounts from genuine accounts. It is observed that bots can be identified by analyzing the profile based featured and URL features that they post such as redirected URL, spam data, frequency of URL sharing etc than social features. In this project, we suggest a novel approach using Deep Learning techniques that uses profile based features for exposing pernicious bots on social networks.