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

TL;DR

This paper introduces a comprehensive framework and novel models for out-of-distribution detection in text-rich networks, addressing the complex interplay of textual and topological features across diverse OOD scenarios.

Contribution

It proposes the TextTopoOOD framework for evaluating OOD detection and the TNT-OOD model that fuses text and topology using cross-attention and HyperNetworks, advancing the understanding of OOD in complex networks.

Findings

Effective detection across multiple OOD scenarios

Enhanced ID/OOD distinction through text-topology fusion

Demonstrated on 11 datasets with diverse shifts

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.