Combining object detection and causality mining for efficient development of augmented reality-based on-the-job training systems in hotel management

Abstract

ABSTRACT The purpose of this study is to propose a methodology for establishing an augmented reality (AR)-based model for efficient OJT through object detection and causality mining, a novel text analysis method. Articles on hotel management published in the last decade and useful for OJT were collected, information on the causal relationships between them extracted, and related rules saved to a rule base. Using the same data, we detected various objects through SSD (Single Shot Multibox Detector), a real-time object detection system. Then we matched sets of causalities and displayed them to trainees wearing AR devices. This methodology reduces development and maintenance costs required to operate OJT programmes. Trainees are immersed in the training environment, which improves the effectiveness of the training. To show the feasibility of the proposed method, we developed a prototype AR-OJT system for hotel management training, automatically extracting knowledge on hotel management from articles according to the proposed method. The results demonstrate that the AR-OJT group shows better performance in terms of learning motivation and self-regulated learning than the control OJT group. No significant difference in learning performance was found between the two groups, which implies that traditional OJT can be substituted with AR-OJT.