How to Predict Personality from Twitter

Golbeck, J., Robles, C., Edmondson, M., & Turner, K. (2011). Predicting personality from Twitter. In Proceedings of the 2011 IEEE International Conference on Privacy, Security, Risk and Trust, and IEEE International Conference on Social Computing (pp. 149–156). IEEE Computer Society. Retrieved from https://www.demenze medicinagenerale.net/pdf/2011%20%20Predicting%20Personality%20from %20Twitter.pdf

In this paper, the authors present a method to predict a user’s personality through the information on their Twitter profile. After providing background on social networking and psychology/personality as well as the “Big Five” model of personality dimensions and its applications, the authors detail their study. To collect data, the authors developed a Twitter application to administer a 45-question version of the Big Five Personality Inventory to 50 participants, and (b) retrieve statistics about their accounts and tweets. Results were first subjected to linguistic analysis (textual processing of tweets) to predict personality traits. The text was then run in the MRC Psycholinguistic Database and performed word-by-word sentiment analysis of users’ tweets. The authors then conducted Pearson correlation and regression analyses. The results showed the authors could predict the Big Five personality traits within just over 10%; with implications that may shape future development in interface design, advertising, and recommender systems.  

‍