What's Left Unsaid? Detecting and Correcting Misleading Omissions in Multimodal News Previews
Misleading omissions can shift readers from the full article context.Abstract
This paper studies omission-based misleadingness in multimodal news previews, where a factually correct image and headline pair can still cause interpretation drift by leaving out context. It introduces the MM-Misleading benchmark for evaluation and OMGuard for interpretation-aware detection and rationale-guided correction.
Type
Publication
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Fanxiao Li
CSC Visiting Student (Sep ‘25)
Visiting student; interests include Multimodal Misinformation, Large Vision-Language Models.
Herun Wan
CSC Visiting Student (Oct ‘25)
Visiting student; interests include Online Malicious Content Analysis such as Misinformation Detection and Social Bot Detection.
Min-Yen Kan
Associate Professor
WING lead; interests include Digital Libraries, Information Retrieval and Natural Language Processing.