The concept of an AI-based expert system (ATOM) for predicting job success

Izsó, Lajos (2023) The concept of an AI-based expert system (ATOM) for predicting job success. APPLIED PSYCHOLOGY IN HUNGARY, 25 (3). pp. 5-15. ISSN 1586-7382

Preview

Text APA_2023_3_1.pdf Download (377kB) | Preview

Abstract

Background and Aims: This article offers a brief prospective exposition of the other articles published in this special issue on artificial intelligence (AI)-based expert systems for predicting job success in general. It also focuses on a concrete implementation of such a system, developed by us. Apart from providing a broader perspective on the possibilities and limitations of applying AI in various fields of human resources management (HRM), such as recruitment, selection, employment, training, performance monitoring/management, wages, labor relations, occupational safety and health (OSH), etc., this text serves as the foreword by the editor of this special issue. Recent AI applications supporting HRM are predominantly limited to recruitment; in other fields, companies still rely on traditional methods. Given the observed underperformance of HR personnel in workforce selection decisions, we have developed an AI-based system primarily for this task, known as ATOM (Artificial Intelligence for Testing Occupational success of Manpower). ATOM not only supports workforce selection but also effectively aids in recruiting and, to some extent, OSH. The core function of ATOM is to “learn” the relationship between suitable predictors (variables suitable for predicting the future job success of applicants) and relevant success criteria scores for the given job. A notable feature of ATOM that provides outstanding efficiency and flexibility is its use of multiple machine learning (ML) algorithms running concurrently, with the results of the best-performing algorithm being accepted.

Actions (login required)

Edit Item