SGOOD: Substructure-enhanced Graph-Level Out-of-Distribution Detection

Zhihao Ding, Jieming Shi, Shiqi Shen, Xuequn Shang, Jiannong Cao, Zhipeng Wang, Zhi Gong

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Graph-level representation learning is important in a wide range of applications. Existing graph-level models are generally built on i.i.d. assumption for both training and testing graphs. However, in an open world, models can encounter out-of-distribution (OOD) testing graphs that are from different distributions unknown during training. A trustworthy model should be able to detect OOD graphs to avoid unreliable predictions, while producing accurate in-distribution (ID) predictions. To achieve this, we present SGOOD, a novel graph-level OOD detection framework. We find that substructure differences commonly exist between ID and OOD graphs, and design SGOOD with a series of techniques to encode task-agnostic substructures for effective OOD detection. Specifically, we build a super graph of substructures for every graph, and develop a two-level graph encoding pipeline that works on both original graphs and super graphs to obtain substructure-enhanced graph representations. We then devise substructure-preserving graph augmentation techniques to further capture more substructure semantics of ID graphs. Extensive experiments against 11 competitors on numerous graph datasets demonstrate the superiority of SGOOD, often surpassing existing methods by a significant margin. The code is available at https://github.com/TommyDzh/SGOOD.

Original languageEnglish
Title of host publicationCIKM 2024 - Proceedings of the 33rd ACM International Conference on Information and Knowledge Management
PublisherAssociation for Computing Machinery
Pages467-476
Number of pages10
ISBN (Electronic)9798400704369
DOIs
StatePublished - 21 Oct 2024
Event33rd ACM International Conference on Information and Knowledge Management, CIKM 2024 - Boise, United States
Duration: 21 Oct 202425 Oct 2024

Publication series

NameInternational Conference on Information and Knowledge Management, Proceedings
ISSN (Print)2155-0751

Conference

Conference33rd ACM International Conference on Information and Knowledge Management, CIKM 2024
Country/TerritoryUnited States
CityBoise
Period21/10/2425/10/24

Keywords

  • graph classification
  • out-of-distribution detection
  • reliability
  • trustworthy model

Fingerprint

Dive into the research topics of 'SGOOD: Substructure-enhanced Graph-Level Out-of-Distribution Detection'. Together they form a unique fingerprint.

Cite this