Classifying Team Players in Software Industry Based on Personality Traits: An Expert-Guided Machine Learning Approach

In contemporary team dynamics in the information technology (IT) industry, the effective formation of teams goes beyond the mere consideration of technical prowess; it also relies on a nuanced understanding of human traits. This effort proposes a machine learning approach guided by expert insights to identify the team player potential of IT professionals. This study introduces a systematic taxonomy for team formation to effectuate this, identifying the most paramount factors that gauge team player dynamics. This approach signifies the extent of the team player role in individuals grounded on real-world inputs. This endeavour develops an automated framework based on experts' criteria in response to the nuanced understanding of human traits. The five most influential traits were identified through a comprehensive review of the experiences of experts in the field and relevant studies focused on team building. In building this framework through machine learning, three algorithms were mainly explored they are Random Forest, Support Vector Machine (SVM), and Neural Network algorithms, where The Random Forest algorithm, which showcases the best performance and accuracy, was selected. By exploring machine learning methods, the aim is to minimise the potential biases inherent in human judgment by establishing a fair process for classifying individuals' team player roles, leading to effective team building. Using this machine learning model, an application can be developed that can estimate the extent of specific traits in an employee and predict their level of teamwork. The results of this study lay the preliminaries for future work characterisation, providing valuable perception into the development of predictive frameworks for team player traits in the IT industry.