Automatic identification of personality traits using X contents based on Myers Briggs index
Subject Areas : هوش مصنوعی و رباتیکMelika Kalhor 1 , Mitra Mirzarezaee 2 , Touraj BaniRostam 3
1 - Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran
2 - Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran
3 - Computer Engineering Department, Islamic Azad University, Central Tehran Branch, Tehran, Iran
Keywords: Social Networks, Personality Traits, Myers Briggs, Deep learning, Bert,
Abstract :
Personality traits are a collection of behaviors and way of thinking in social environment and daily life. Psychologists usually use personality questionnaires to determine personality types. The increasing availability of online social data have stimulated researchers' interest in the study of advanced computational methods for text-based personality identification. Personality identification by computational and machine learning models based on text is an emerging phenomenon with high potential. In this article, the personality of people based on the Myers Briggs index has been investigated and analyzed using Twitter data. To investigate how text-based digital footprints can be used to predict users' personalities, a multi-model deep learning architecture along with three pre-trained language models Bert , Roberta, and XLNET, which are relatively new architectures in language processing are employed for extracting features from people's posts in X social media for identification of personality traits. the final decision is make via model averaging. The proposed model generates a predictive model for each personality trait. The results obtained shows an accuracy 88.5% and an F1-measure score of 0.882. This work provides a basis for the development of a personality trait identification system that can help the organizations to recruit and select the right personnel and improve its business by knowing the personality and preferences of its customers.