Text-Based Personality Assessment Using Deep Learning Techniques

Personality classification from textual data has increased recently in wide range of applications. Personality appears as a critical component impacting an individual’s eligibility for certain employment in the world of e-recruitment. The deep learning-based ensemble model which consists of Long short-term Memory (LSTM) combined with Gated Recurrent Unit (GRU), typically trained from scratch for text classification and machine learning-based stacking classifier model are explored specially for better performance along with various computational methodologies by diving into well-known personality framework “The Big Five” personality traits. The study analyzes both deep learning models and machine learning methodologies in depth, providing a full evaluation of their mechanics, benefits, and limits. The accuracy, F-1 score, recall, precision are the evaluation metrics used to calculate optimal performance among the models and LSTM+GRU is the effective model with high accuracy.