A two-stage deep neural model with capsule network for personality identification

People have different ways of thinking, feeling, and hence acting, which resulted in different personalities. Understanding one's personality and how it can be automatically identified considering the way he/she communicates to the world around can be challenging; but it can also be useful in many cases. Deep learning algorithms perform fairly well in text-based personality detection. However, many computational personality assessment models rely on limited domain knowledge. There are different personality models for classifying personality traits according to the definitions of psychologists. In this paper, we focus on the Myers-Briggs Type Indicator (MBTI) model and explain how a two-stage deep neural model for personality identification can use more information from text and therefore, have better performance in classifying input data. To this end, in the first stage, we use capsule neural networks to extract meaningful hidden patterns from word-level semantic representation to be used for calculating personality traits. Moreover, in the second stage of the proposed architecture, we benefit from contextualized document-level representation of text as well as statistical psychological features. Our experimental results on the Myers-Briggs Personality Type dataset from Kaggle which has been labeled based on the MBTI model show improvement in personality identification compared to the state-of-the-art models in the field.