Abstract: Nowadays identifying user attributes from their social network activities has been a common research topic. Age, gender and interest can be common user attributes which can be predicted and are essential for personalization and recommender systems. Most of the researches are based on the textual content created by user, whereas recently multimedia has gained popularity in social networks. In this paper we propose an algorithm that predicts the user gender on different networking sites.

Keywords: Social Network, Social Images, Gender Prediction, Demographics.