Integrating Temperament and Personality Traits into Recruitment Frameworks using Adaptive Neuro-Fuzzy Inference system

Abstract

Recruitment decisions continue to rely on structured interviews, academic credentials, and cognitive testing, tools that capture technical competence but consistently overlook the behavioural and dispositional qualities that determine how a new hire actually performs once in role. Decades of organisational research link stable temperament and personality traits to job performance, stress tolerance, teamwork, and retention, yet most intelligent recruitment systems still treat behavioural evaluation as an afterthought. This paper proposes an Adaptive Neuro-Fuzzy Inference System (ANFIS) that integrates applicant temperament, operationalised through the Big Five personality dimensions, with conventional recruitment attributes to produce a continuous, interpretable measure of candidate suitability. Grounded in Trait Theory, Person-Job Fit Theory, and Computational Intelligence Theory, the framework fuzzifies nine input variables, five temperament dimensions and four recruitment attributes, using Gaussian membership functions, and combines them through a first-order Sugeno inference architecture whose rule base is derived through subtractive clustering rather than exhaustive enumeration. A theoretical review of five candidate personality frameworks confirms the Big Five as the most empirically validated and computationally tractable basis for the model, with Conscientiousness and Emotional Stability emerging as the strongest cross-occupational predictors. The resulting nine-input, single-output architecture produces a Job Performance Suitability Index (JPSI) that maps applicants onto four ordinal placement categories. The paper argues that combining the interpretability of fuzzy reasoning with the adaptive learning capacity of neural networks offers a transparent, bias-reducing alternative to both rigid rule-based screening and opaque black-box predictive models, and outlines the training, validation, and comparative evaluation work required to bring the framework to full deployment.