PersoNet: A Novel Framework for Personality Classification-Based Apt Customer Service Agent Selection

Abstract

Personality classification has garnered significant interest in psychology, computational social science, and Machine Learning (ML) due to its wide-ranging applications. This paper presents PersoNet, an innovative framework developed to identify personality types using the Myers-Briggs Type Indicator (MBTI), aimed at enhancing customer service experiences by matching customers with suitable support agents. PersoNet employs a Bidirectional Long Short-Term Memory (BiLSTM) neural network architecture and has achieved an impressive classification accuracy of over 93.98%. Our extensive experiments with the MBTI dataset reveal that the BiLSTM architecture effectively captures both temporal dependencies and semantic subtleties in textual data, contributing to this high level of accuracy. Consequently, PersoNet can accurately select customer service agents who match customer personalities, achieving a Customer Satisfaction Rate (CSR) of over 97.82%-a notable improvement of 20.25% in CSR based on our experimental data. These results establish PersoNet as a cutting-edge tool in personality classification, surpassing existing methods in both accuracy and computational efficiency and markedly enhancing customer service quality.