Personality segmentation of users through mining their mobile usage patterns

Users' interactions with their mobile devices leave behind unique digital footprints that can reveal important information about their characteristics, including their personality. By deploying a wide range of machine learning algorithms and by analyzing patterns of mobile usage data from more than 400 users, this study examines the personality determinants of various mobile usage attributes. Considering the Big-Five personality traits (agreeableness, conscientiousness, extraversion, neuroticism, and openness), correlations between mobile usage attributes and personality traits are presented and discussed. Moreover, the study examines the possibility of predicting users' personality segments from their mobile usage attributes. Using the K-means clustering algorithm, three distinct personality segments are detected. Subsequently, different machine learning classification models are trained to predict the personality segments of users based on their mobile usage attributes. The results suggest that users' personality segments can be correctly predicted, with an overall accuracy of 76.17%. The number of contacts on the device is found to be the most significant predictor followed by the frequency and duration of outgoing calls, and then the average time spent on social media applications. Additionally, the study discusses the practical implications of the findings from the perspectives of users, service providers and mobile application providers.