Social Implications of OCEAN Personality: An Automated BERT-Based Approach

Chaitanya Chakka, Bathini Sai Akash, Aruna Malapati

Abstract:

The importance of personality within society is paramount, as it profoundly influences individual and collective behaviors, interpersonal interactions, and the overall functionality of societies. However, for a long
time, personality detection from online social texts has been lacking in performance. This is due to the limited
data availability and constrained supervised learning frameworks over small labeled datasets. In this work, we
present a novel approach to personality prediction utilizing BERT in conjunction with two notable datasets,
achieving proficient accuracy across the OCEAN traits. The research also extracts linguistic cues that do not
require supervision. Finally, we perform extensive empirical analysis to conclude over four research questions
that deal with the social implications of personality. The approach provides pragmatic results, making use of
the designed automatic personality prediction pipeline. The code has also been made open source to facilitate
enhanced innovation and research benefits (https://github.com/LearningLeopard/personality-prediction).