End-to-End Verification for Subgraph Solving

Research output: Contribution to journalConference articleResearchpeer-review

Documents

  • Fulltext

    Final published version, 317 KB, PDF document

  • Stephan Gocht
  • Ciaran McCreesh
  • Magnus O. Myreen
  • Nordström, Jakob
  • Andy Oertel
  • Yong Kiam Tan

Modern subgraph-finding algorithm implementations consist of thousands of lines of highly optimized code, and this complexity raises questions about their trustworthiness. Recently, some state-of-the-art subgraph solvers have been enhanced to output machine-verifiable proofs that their results are correct. While this significantly improves reliability, it is not a fully satisfactory solution, since end-users have to trust both the proof checking algorithms and the translation of the high-level graph problem into a low-level 0-1 integer linear program (ILP) used for the proofs. In this work, we present the first formally verified toolchain capable of full end-to-end verification for subgraph solving, which closes both of these trust gaps. We have built encoder frontends for various graph problems together with a 0-1 ILP (a.k.a. pseudo-Boolean) proof checker, all implemented and formally verified in the CAKEML ecosystem. This toolchain is flexible and extensible, and we use it to build verified proof checkers for both decision and optimization graph problems, namely, subgraph isomorphism, maximum clique, and maximum common (connected) induced subgraph. Our experimental evaluation shows that end-to-end formal verification is now feasible for a wide range of hard graph problems.

Original languageEnglish
JournalProceedings of the AAAI Conference on Artificial Intelligence
Volume38
Issue number8
Pages (from-to)8038-8047
Number of pages10
ISSN2159-5399
DOIs
Publication statusPublished - 2024
Event38th AAAI Conference on Artificial Intelligence, AAAI 2024 - Vancouver, Canada
Duration: 20 Feb 202427 Feb 2024

Conference

Conference38th AAAI Conference on Artificial Intelligence, AAAI 2024
CountryCanada
CityVancouver
Period20/02/202427/02/2024
SponsorAssociation for the Advancement of Artificial Intelligence

Bibliographical note

Publisher Copyright:
Copyright © 2024, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

ID: 390581269