Daehee Han


2024

pdf bib
KorSmishing Explainer: A Korean-centric LLM-based Framework for Smishing Detection and Explanation Generation
Yunseung Lee | Daehee Han
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track

To mitigate the annual financial losses caused by SMS phishing (smishing) in South Korea, we propose an explainable smishing detection framework that adapts to a Korean-centric large language model (LLM). Our framework not only classifies smishing attempts but also provides clear explanations, enabling users to identify and understand these threats. This end-to-end solution encompasses data collection, pseudo-label generation, and parameter-efficient task adaptation for models with fewer than five billion parameters. Our approach achieves a 15% improvement in accuracy over GPT-4 and generates high-quality explanatory text, as validated by seven automatic metrics and qualitative evaluation, including human assessments.
Search
Co-authors
Venues