Private Traits and Attributes are Predictable from Digital Records of Human Behaviour

You are here

This article argues that an individual’s traits and attributes can be predicted using their digital records, given the abundance of personal information found online. According to the authors, digitally mediated behaviors can easily be recorded and analyzed, allowing for the development of a number of new services, including personalized search engines, recommender systems, and targeted online marketing. However, this widespread availability of individual behavior, together with the desire to learn more about customers and citizens, presents serious challenges related to privacy and data ownership. Kosinski, Stillwell and Graepel’s analysis is based on a dataset of over 58,000 volunteers who provided their Facebook Likes, detailed demographic profiles, and the results of several psychometric tests. The proposed model uses dimensionality reduction for preprocessing the Likes data, which are then entered into logistic/linear regression to predict individual psychodemographic profiles.


The authors’ findings are a valuable contribution to the field of interpreting and understanding online behaviour and tools used to identify these patterns. For practitioners working to identify patterns of internet usage by potential extremists or the implications of big data collection this article is required reading.

Michael Kosinski, David Stillwell, and Thore Graepel