Deep Reinforcement Learning-based Authentic Dialogue Generation To Protect Youth From Cybergrooming

Anonymous

Abstract: Cybergrooming is defined as a crime towards potential victims, especially teens, by building close personal relationships with them with the purpose of sexual exploitation via online media. Cyber or online sexual grooming has been recognized as a serious cyber crime. However, there have been insufficient programs to proactively protect the youth from cybergrooming. In this work, we present a generative chatbot framework, called SERI (Stop cybERgroomIng), that can generate simulated conversations between a perpetrator chatbot and a potential victim chatbot. To realize the simulation of authentic conversations in the context of cybergrooming, we take deep reinforcement learning (DRL)-based dialogue generation for authentic simulation of the conversations between a potential victim and a perpetrator (i.e., cybergroomer). The design of the SERI is motivated to ensure a safe and authentic environment to strengthen the youth's precautionary awareness of cybergrooming while any unnecessary ethical issues (e.g., the potential misuse of the SERI) are removed or minimized. We developed the SERI as a preliminary platform that can deploy the perpetrator chatbot to interact with human users (i.e., youth) to observe youth users' responses to strangers or acquaintances and collect the reactions when the youth users are asked for private or sensitive information by the perpetrator. We evaluated the quality of conversations generated by the SERI based on open-source, referenced, unreferenced metrics, and human evaluation.

Paper Type: long

Consent To Share Data: yes