Text Meets Topology: Rethinking Out-of-distribution Detection in Text-Rich Networks

2025-08-25

AI Summary

Out-of-distribution (OOD) detection in text-rich networks remains challenging due to coupled distributional shifts across textual and structural domains. Method: This paper proposes TextTopoOOD, a unified framework featuring the TNT-OOD model, which systematically models distribution shifts at three levels—textual content, topological structure, and latent topic semantics. It introduces a cross-attention mechanism to integrate local structural context with textual representations and employs a HyperNetwork to generate node-specific transformation parameters, enabling joint, adaptive modeling of semantic and topological features. Contribution/Results: Extensive experiments across 11 benchmark datasets under four controlled OOD scenarios—text augmentation, embedding perturbation, edge rewiring, and semantic linkage shift—demonstrate that TNT-OOD consistently outperforms state-of-the-art methods, validating the effectiveness and generalizability of text-graph joint OOD detection.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Graph-based Machine Learning
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Out-of-distribution (OOD) detection remains challenging in text-rich networks, where textual features intertwine with topological structures. Existing methods primarily address label shifts or rudimentary domain-based splits, overlooking the intricate textual-structural diversity. For example, in social networks, where users represent nodes with textual features (name, bio) while edges indicate friendship status, OOD may stem from the distinct language patterns between bot and normal users. To address this gap, we introduce the TextTopoOOD framework for evaluating detection across diverse OOD scenarios: (1) attribute-level shifts via text augmentations and embedding perturbations; (2) structural shifts through edge rewiring and semantic connections; (3) thematically-guided label shifts; and (4) domain-based divisions. Furthermore, we propose TNT-OOD to model the complex interplay between Text aNd Topology using: 1) a novel cross-attention module to fuse local structure into node-level text representations, and 2) a HyperNetwork to generate node-specific transformation parameters. This aligns topological and semantic features of ID nodes, enhancing ID/OOD distinction across structural and textual shifts. Experiments on 11 datasets across four OOD scenarios demonstrate the nuanced challenge of TextTopoOOD for evaluating OOD detection in text-rich networks.

Problem

• Detecting out-of-distribution nodes in text-rich networks
• Addressing intertwined textual and topological feature shifts
• Handling attribute, structural, and label-based distribution shifts

Innovation

• Cross-attention module fuses local structure
• HyperNetwork generates node-specific transformation parameters
• Models interplay between text and topology