A decentralized employee performance appraisal framework for recruitment, performance prediction and ranking using permissioned blockchain and machine learning

Abstract

Recruitment is a crucial task for Human Resource Management (HRM) and therefore determines the selection of skilled employees-who play key roles in a company's success. Employees perform well only when their skill set matches their job requirements. However, the current recruitment system fails to provide a single solution that verifies employee records and predicts employee-job compatibility. This paper proposes a recruitment system using blockchain, a machine learning model, and a multi-criteria decision analysis method. Here, blockchain technology is used to hold employee records in an encrypted manner, including their performance appraisals, in a decentralized system that allows data verification by supported organisations. QR code is used to verify encrypted employee records which are then decrypted within the system and used to predict their performance for the hiring company, using a classification model. Here, after comparing the performance of different classification models, the best-performing model, SVM Classification, has been used. Finally, the system ranks eligible candidates, based on employee records including their predicted performance appraisal rating, using a TOPSIS algorithm. The SVM classification model and TOPSIS algorithm are deployed in a Streamlit application, which predicts employee performance appraisal ratings and ranks candidates by generating scores using TOPSIS. The scores are then sorted from highest to lowest, therefore ranking candidates from most suitable to least suitable for a job role.