Reasoning About the Unsaid: Misinformation Detection with Omission-Aware Graph Inference

TL;DR

This paper introduces OmiGraph, a novel omission-aware framework for misinformation detection that models and leverages omitted content and relations to improve detection accuracy.

Contribution

It pioneers the explicit modeling of omission-based deception using an omission-aware graph and relation inference, advancing misinformation detection methods.

Findings

Achieves +5.4% F1 score improvement on benchmarks.

Demonstrates effectiveness of omission modeling in misinformation detection.

Introduces a new framework for omission-aware relation inference.

Abstract

This paper investigates the detection of misinformation, which deceives readers by explicitly fabricating misleading content or implicitly omitting important information necessary for informed judgment. While the former has been extensively studied, omission-based deception remains largely overlooked, even though it can subtly guide readers toward false conclusions under the illusion of completeness. To pioneer in this direction, this paper presents OmiGraph, the first omission-aware framework for misinformation detection. Specifically, OmiGraph constructs an omission-aware graph for the target news by utilizing a contextual environment that captures complementary perspectives of the same event, thereby surfacing potentially omitted contents. Based on this graph, omission-oriented relation modeling is then proposed to identify the internal contextual dependencies, as well as the dynamic omission intents, formulating a comprehensive omission relation representation. Finally, to extract omission patterns for detection, OmiGraph introduces omission-aware message-passing and aggregation that establishes holistic deception perception by integrating the omission contents and relations. Experiments show that, by considering the omission perspective, our approach attains remarkable performance, achieving average improvements of +5.4% F1 and +5.3% ACC on two large-scale benchmarks.