An Online Personality Traits Mining Approach Based on Cluster Analysis

Mining personality traits online plays an important role in obtaining user personality characteristics, which can indirectly affect online behavior, especially in online learning context. However, traditional researches on personality traits are mainly based on self-assessment, with the disadvantages of being subjective, not scalable and requiring lots of workload. This paper proposed an online personality traits mining approach on Sina microblog social platform by using Gaussian Mixture Model algorithm for data mining and cluster analysis. Through analysis and comparison of online user text features and behavioral characteristics, a 14-cluster personality traits model was constructed and unified in the Big Five-Factor Model by quantifying the correlation between personality traits and characteristics. With the advantages of accuracy, objectivity and strong scalability, the approach proposed in this paper provides new ideas for online personality traits mining and can be used in multiple application scenarios.