Prediction Of Age, Gender And Personality Traits Using...

Prediction of Age, Gender and Personality Traits Using Facebook Data Manali Bhalgat Introduction In the last decade, social networks like Facebook [1] have emerged as popular medium of social interaction and information dissemination. From a social web data mining perspective, Facebook stores a wealth of data about people and their interests. As more and more users are creating their own content on Facebook, there is a growing interest to mine this data for use in personalized information access services, recommender systems, tailored advertisements, and other applications that can benefit from personalization. Research studies leverage on data about status updates, pages liked, number of friends, number of groups joined and other†¦show more content†¦Related Work In this section, we discuss the recent research work related to techniques for predicting age, gender and personality based on the information available on social networking sites. Most studies predict personality from the language used by a person to update status or chat in social network. According to recent works based on study of function words such as pronouns, conjunctions, articles and prepositions, the elderly use more future tense words and pronouns in their plural forms. The same studies show that males use more articles and females make heavy use of first person singular pronouns. In [7], the authors state that even with its challenges, text categorization is a reliable approach to identify the age and gender of people in social network communication. Goldbeck et al., in [3], show that people of different age groups talked about different topics. For example, those within the age group of 13 to 18 mostly discussed activities related to school, 19 to 22 year olds talked about university/college. Many efforts have been made by researchers to analyze the words used by humans to understand their psychology [2]. Public information of a group of Facebook users was collected by the authors of [6]. They were able to predict the Big-five personality traits of the users using this data within 89% accuracy. Methodology 1. Problem Definition To survey the machine learning techniques for