The number of recruitment postings on digital recruitment hiring platforms has increased since the COVID-19 pandemic. However, the weak surveillance and operations of these platforms, combined with the fact that most job seekers have relatively low vigilance and strong desire for recruitment offers, enable scammers to easily deceive job seekers for their money and confidential information. In this work, we combine prevailing text mining techniques (i.e., ChatGPT with prompting engineering and supervised machine learning) with interpersonal deception theory (IDT) from social science to design an interpretable IT system to predict fraudulent recruitment posting on digital recruitment-hiring platforms. We compare our designed framework with the state-of-the-art general-purpose algorithms to demonstrate the efficacy of our system using two testbeds. To further confer intuitive and human understandable operational guidance of IDT-driven design for the fraudulent recruitment phenomenon, we perform instance-agnostic and instance-specific explanation analysis based on the aggregated marginal contributions of IDT-driven contextualized features. A between-subjects user experiment empirically shows that IDT-driven explanations enhance users’ trust, understanding, and perceived usefulness. Using an illustrative example, we further quantify the economic value of IDT-driven design via a cost-revenue analysis. We conclude the academic contributions and practical implications of our work to job seekers, recruiters, and third-party recruitment-hiring platforms.
Wang et al. (Tue,) studied this question.