A decentralized employee performance appraisal framework for recruitment, performance prediction and ranking using permissioned Blockchain and ensemble learning

Abstract

Recruitment is a crucial task for Human Resource Management (HRM) and determines the creation of a competent workforce that eventually brings tangible and intangible benefits for companies. Employees are key elements in determining a company’s success and employees perform well when their skill set complements their job requirements. However, the current system fails to provide a single solution that verifies employee records and predicts employee-company compatibility. This paper proposes an recruitment system using a private permissioned blockchain architecture and ensemble learning algorithms. The paper proposes a permissioned blockchain architecture using permission protocol and smart contracts to store employee records in an immutable ledger. Development of Data and processing decentralization is inspired and in accordance with the Hyperledger Fabric system design, thus creating a decentralized data sharing system that is used to hold comprehensive employee performance records in a peer-to-peer system that allows employee data verification and retrieval by organizations in the blockchain consortium. The applicant records and previous performance appraisal records can be retrieved by a company in the consortium following smart contract rules and can be used to predict employee performance ratings based retrieved previous performance appraisal records. To predict the performance score, we used machine learning models namely supervised and ensemble learning. The system also ranks eligible candidates, based on predicted performance scores and other relevant applicant data via Multi-Criteria Decision Making Algorithm (MCDM). Finally, a Streamlit application is created where performance score predicting and ranking are done automatically with a suitable user interface for final result output.