Enhancing Job Recruitment Prediction through Supervised Learning and Structured Intelligent System: A Data Analytics Approach

Abstract

Personnel recruitment processes in various government agencies, ministries, boards, and parastatals encounter challenges in effectively selecting candidates who meet specified requirements for job placement on time. Moreover, human resource (HR) managers face the additional burden of appeasing top government officials while also mitigating issues of nepotism and bias during recruitment. The success or failure of any organization heavily relies on the recruitment and retention of its workforce. Consequently, the decision to select suitable candidates for job positions is of utmost importance to management in every organization. This work develops a structured intelligent system that selects the best machine learning (ML) classification model for predicting applicants’ employability based on their attributes using the industry job selection criteria. A dataset of 16240 applicants’ records collected from Akwa State Universal Basic Education Board (AKSUBEB) was used to train and test the performance of the ML models. Naïve Bayes (NB), Logistic Regression (LR), Support Vector Machines (SVM), Random Forest (RF), and Decision Tree (DT) classifiers were deployed where results indicate that DT emerged the most effective classifier with a 98% prediction accuracy followed by RF with accuracy of 97.59% while LR recorded the least accuracy of 79.43%.  This outcome indicates that tree-based ML structures can significantly help HR personnel to efficiently select suitable candidates for given job positions with reduced overhead in the recruitment process.