pith. sign in

arxiv: 2409.20302 · v13 · submitted 2024-09-30 · 💻 cs.AI · cs.CL· cs.IR

OM4OV: Leveraging Ontology Matching for Ontology Versioning

Zhangcheng Qiang , Kerry Taylor , Weiqing Wang This is my paper

Pith reviewed 2026-05-23 20:41 UTC · model grok-4.3

classification 💻 cs.AI cs.CLcs.IR
keywords ontology matchingontology versioningsemantic webcross-reference mechanismAgent-OMontology managementversion controlalignment reuse
0
0 comments X

The pith

Ontology matching systems can be reused for ontology versioning tasks once extended with a cross-reference mechanism to fix skewed results and poor update detection.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines how ontology matching and ontology versioning overlap yet differ in practice on the Semantic Web. It formalizes an OM4OV framework that reuses existing matching systems for versioning while adding targeted support. Experiments in Agent-OM show that plain reuse produces skewed measurements, weak detection of updated entities, and weak explanations for false mappings. The authors introduce a cross-reference mechanism that reuses prior alignments to shrink the set of candidate matches and raise overall versioning accuracy. The work therefore treats versioning as an extension problem rather than a separate system-building task.

Core claim

Ontology matching systems can be effectively reused for ontology versioning tasks, but without necessary extensions produce skewed measurements, poor performance in detecting update entities, and limited explanation of false mappings; the proposed cross-reference mechanism builds on existing OM alignments to reduce matching candidates and improve overall OV performance.

What carries the argument

The OM4OV framework that adapts ontology matching for versioning, carried by the cross-reference (CR) mechanism which reuses prior alignments to limit candidate matches.

If this is right

  • Existing ontology matching tools become directly usable for versioning without complete redesign.
  • Prior alignments from matching runs can be leveraged to cut down the search space for version differences.
  • Detection of changed entities improves once the candidate pool is narrowed by cross-references.
  • False mappings become easier to explain because the mechanism records which prior alignments influenced each decision.
  • Ontology management pipelines can combine matching and versioning steps in one system.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same reuse pattern could be tested on other dynamic knowledge bases beyond OWL ontologies.
  • Integration with standard version-control systems might allow automatic detection of semantic drift in linked open data.
  • The framework suggests that versioning accuracy depends more on reuse of historical alignments than on new similarity measures.
  • Further experiments could check whether the approach scales when ontologies change at high frequency.

Load-bearing premise

That systematic comparison of ontology matching and versioning yields a framework whose extensions reliably correct the observed performance shortfalls when tested inside an existing matching system.

What would settle it

A controlled test on a new collection of ontologies in which the cross-reference mechanism shows no measurable gain in update-entity detection or overall versioning accuracy over unextended matching reuse.

read the original abstract

Due to the dynamic nature of the Semantic Web, version control is necessary to manage changes in widely used ontologies. Despite the long-standing recognition of ontology versioning (OV) as a crucial component of efficient ontology management, many approaches treat OV as similar to ontology matching (OM) and directly reuse OM systems for OV tasks. In this study, we systematically analyse similarities and differences between OM and OV and formalise an OM4OV framework to offer more advanced OV support. The framework is implemented and evaluated in the state-of-the-art OM system Agent-OM. The experimental results indicate that OM systems can be effectively reused for OV tasks, but without necessary extensions, can produce skewed measurements, poor performance in detecting update entities, and limited explanation of false mappings. To tackle these issues, we propose an optimisation method called the cross-reference (CR) mechanism, which builds on existing OM alignments to reduce the number of matching candidates and to improve overall OV performance.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper claims that direct reuse of ontology matching (OM) systems for ontology versioning (OV) leads to problems including skewed measurements, poor update-entity detection, and limited false-mapping explanations. It systematically analyzes OM-OV similarities/differences, formalizes an OM4OV framework, implements it in Agent-OM, and introduces a cross-reference (CR) mechanism that prunes matching candidates using existing alignments to improve OV performance; experiments are said to support the effectiveness of this approach.

Significance. If the results hold, the work offers a concrete way to extend existing OM tools for the practical task of managing ontology evolution on the Semantic Web. The CR mechanism is a reusable optimization, and the evaluation inside a state-of-the-art system (Agent-OM) provides a useful baseline for future OV research.

major comments (2)
  1. [§5 (Evaluation)] §5 (Evaluation): The manuscript does not specify ontology-pair selection criteria, update types (simple vs. complex structural changes), or domain coverage. Because the central claim is that the CR mechanism reliably improves update-entity detection and reduces skewed measurements, the absence of these details leaves open whether the reported gains are representative or specific to the chosen test cases.
  2. [§3.3 (CR mechanism definition)] §3.3 (CR mechanism definition): The description of how CR builds on existing OM alignments to reduce candidates is high-level; it is unclear how the mechanism behaves when alignments are incomplete or when updates involve non-local structural changes, which directly affects the claimed performance gains.
minor comments (2)
  1. [Table 2] Table 2: column headers for precision/recall on update entities are not fully aligned with the text description in §4.1, making it hard to map reported numbers to the claimed improvements.
  2. [Abstract] The abstract states that experiments 'support the claims' but does not name the key metrics or datasets; adding one sentence with these would improve readability without changing the technical content.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which help strengthen the manuscript. We address each major comment point-by-point below and will incorporate revisions as indicated.

read point-by-point responses

  1. Referee: §5 (Evaluation): The manuscript does not specify ontology-pair selection criteria, update types (simple vs. complex structural changes), or domain coverage. Because the central claim is that the CR mechanism reliably improves update-entity detection and reduces skewed measurements, the absence of these details leaves open whether the reported gains are representative or specific to the chosen test cases.

    Authors: We agree that additional details on evaluation design would improve clarity and allow readers to better assess generalizability. The ontologies were drawn from OAEI benchmarks and real-world versioning scenarios (biomedical and bibliographic domains) to cover both simple and complex structural updates. We will revise §5 to include an explicit subsection on selection criteria, update classification, and domain coverage, along with a table summarizing the test cases. revision: yes

  2. Referee: §3.3 (CR mechanism definition): The description of how CR builds on existing OM alignments to reduce candidates is high-level; it is unclear how the mechanism behaves when alignments are incomplete or when updates involve non-local structural changes, which directly affects the claimed performance gains.

    Authors: The CR mechanism assumes access to prior alignments typical in versioning workflows and prunes candidates by cross-referencing them against those alignments. We acknowledge the description in §3.3 is high-level. We will expand this section with a more detailed algorithmic specification (including pseudocode) and add discussion of behavior under incomplete alignments (fallback to standard matching) and non-local changes (potential need for iterative application). revision: yes

Circularity Check

0 steps flagged

No circularity; derivation is self-contained from task analysis and external evaluation.

full rationale

The paper derives the OM4OV framework from an explicit systematic analysis of OM-OV similarities and differences, then implements and tests it inside the independent pre-existing Agent-OM system. The CR mechanism is introduced as an optimization that builds on OM alignments to prune candidates. No equations, fitted parameters, or self-citations are shown to reduce any claimed result to its own inputs by construction. The evaluation therefore supplies independent evidence rather than a renaming or self-referential loop.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The central claim rests on domain assumptions about the dynamic nature of ontologies and the value of adapting matching techniques; the paper introduces two new conceptual entities (the framework and the mechanism) without external falsifiable evidence beyond the reported experiments.

axioms (1)
  • domain assumption Ontology versioning is necessary due to the dynamic nature of the Semantic Web and is a crucial component of efficient ontology management.
    Stated as background motivation in the abstract.
invented entities (2)
  • OM4OV framework no independent evidence
    purpose: To formalize advanced support for ontology versioning by leveraging ontology matching systems.
    Newly proposed and formalized in the paper.
  • cross-reference (CR) mechanism no independent evidence
    purpose: To reduce matching candidates and improve OV performance by building on existing OM alignments.
    Newly proposed optimization method.

pith-pipeline@v0.9.0 · 5691 in / 1458 out tokens · 45113 ms · 2026-05-23T20:41:06.632035+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

23 extracted references · 23 canonical work pages · 1 internal anchor

  1. [1]

    Qwen Models , 2024

    Alibaba Qwen Team . Qwen Models , 2024. URL: https://qwenlm.github.io

    work page 2024

  2. [2]

    Claude Models , 2024

    Anthropic . Claude Models , 2024. URL: https://docs.anthropic.com/en/docs/about-claude/models/all-models

    work page 2024

  3. [3]

    Brick: Towards a Unified Metadata Schema For Buildings

    Bharathan Balaji, Arka Bhattacharya, Gabriel Fierro, Jingkun Gao, Joshua Gluck, Dezhi Hong, Aslak Johansen, Jason Koh, Joern Ploennigs, Yuvraj Agarwal, Mario Berges, David Culler, Rajesh Gupta, Mikkel Baun Kj rgaard, Mani Srivastava, and Kamin Whitehouse. Brick: Towards a Unified Metadata Schema For Buildings . In Proceedings of the 3rd ACM International ...

    work page doi:10.1145/2993422.2993577 2016

  4. [4]

    McGuinness, Peter F

    Sean Bechhofer, Frank van Harmelen, Jim Hendler, Ian Horrocks, Deborah L. McGuinness, Peter F. Patel-Schneider, and Lynn Andrea Stein. OWL Web Ontology Language Reference , 2004. URL: https://www.w3.org/TR/owl-ref/

    work page 2004

  5. [5]

    Construction and exploitation of an historical knowledge graph to deal with the evolution of ontologies

    Silvio Domingos Cardoso, Marcos Da Silveira , and Cédric Pruski. Construction and exploitation of an historical knowledge graph to deal with the evolution of ontologies . Knowledge-Based Systems , 194:105508, 2020. https://doi.org/10.1016/j.knosys.2020.105508 doi:10.1016/j.knosys.2020.105508

    work page doi:10.1016/j.knosys.2020.105508 2020

  6. [6]

    Semantic precision and recall for ontology alignment evaluation

    J \'e r \^o me Euzenat. Semantic precision and recall for ontology alignment evaluation . In Proceedings of the 20th international joint conference on Artificial intelligence , pages 348--353, 2007

    work page 2007

  7. [7]

    Google Gemma Models , 2024

    Google Gemma Team and Google DeepMind . Google Gemma Models , 2024. URL: https://ai.google.dev/gemma

    work page 2024

  8. [8]

    What Is an Ontology? , pages 1--17

    Nicola Guarino, Daniel Oberle, and Steffen Staab. What Is an Ontology? , pages 1--17. Springer, 2009. https://doi.org/10.1007/978-3-540-92673-3_0 doi:10.1007/978-3-540-92673-3_0

    work page doi:10.1007/978-3-540-92673-3_0 2009

  9. [9]

    Are we better off with just one ontology on the Web? Semantic Web , 11(1):87–99, 2020

    Armin Haller and Axel Polleres. Are we better off with just one ontology on the Web? Semantic Web , 11(1):87–99, 2020. https://doi.org/10.3233/SW-190379 doi:10.3233/SW-190379

    work page doi:10.3233/sw-190379 2020

  10. [10]

    The RealEstateCore Ontology

    Karl Hammar, Erik Oskar Wallin, Per Karlberg, and David H \"a lleberg. The RealEstateCore Ontology . In The Semantic Web - ISWC 2019 , pages 130--145, 2019. https://doi.org/10.1007/978-3-030-30796-7_9 doi:10.1007/978-3-030-30796-7_9

    work page doi:10.1007/978-3-030-30796-7_9 2019

  11. [11]

    Dynamic ontologies on the web

    Jeff Heflin and James Hendler. Dynamic ontologies on the web . In Proceedings of the AAAI Conference on Artificial Intelligence , pages 443--449, 2000

    work page 2000

  12. [12]

    Ontology Versioning and Change Detection on the Web

    Michel Klein, Dieter Fensel, Atanas Kiryakov, and Damyan Ognyanov. Ontology Versioning and Change Detection on the Web . In Knowledge Engineering and Knowledge Management: Ontologies and the Semantic Web , pages 197--212. Springer, 2002. https://doi.org/10.1007/3-540-45810-7_20 doi:10.1007/3-540-45810-7_20

    work page doi:10.1007/3-540-45810-7_20 2002

  13. [13]

    Ontology versioning on the Semantic Web

    Michel CA Klein and Dieter Fensel. Ontology versioning on the Semantic Web . In Proceedings of the First International Conference on Semantic Web Working , page 75–91, 2001

    work page 2001

  14. [14]

    Meta Llama Models , 2024

    Meta . Meta Llama Models , 2024. URL: https://www.llama.com

    work page 2024

  15. [15]

    ASHRAE Standard 223P , 2024

    National Renewable Energy Laboratory . ASHRAE Standard 223P , 2024. URL: https://www.energy.gov/eere/buildings/ashrae-standard-223p

    work page 2024

  16. [16]

    OpenAI Models , 2024

    OpenAI . OpenAI Models , 2024. URL: https://platform.openai.com/docs/models

    work page 2024

  17. [17]

    Rodríguez-Martínez, and Alma Gómez-Rodríguez

    Lorena Otero-Cerdeira, Francisco J. Rodríguez-Martínez, and Alma Gómez-Rodríguez. Ontology matching: A literature review . Expert Systems with Applications , 42(2):949--971, 2015. https://doi.org/10.1016/j.eswa.2014.08.032 doi:10.1016/j.eswa.2014.08.032

    work page doi:10.1016/j.eswa.2014.08.032 2015

  18. [18]

    Ontology change detection using a version log

    Peter Plessers and Olga De Troyer. Ontology change detection using a version log . In International Semantic Web Conference , pages 578--592, 2005

    work page 2005

  19. [19]

    Agent-OM: Leveraging LLM Agents for Ontology Matching , 2024

    Zhangcheng Qiang, Weiqing Wang, and Kerry Taylor. Agent-OM: Leveraging LLM Agents for Ontology Matching , 2024. https://arxiv.org/abs/2312.00326 arXiv:2312.00326

    work page arXiv 2024

  20. [20]

    RDF 1.2 Schema , 2024

    RDF-star Working Group . RDF 1.2 Schema , 2024. URL: https://www.w3.org/TR/rdf12-schema/

    work page 2024

  21. [21]

    Supporting ontology adaptation and versioning based on a graph of relevance

    Najla Sassi, Wassim Jaziri, and Saad Alharbi. Supporting ontology adaptation and versioning based on a graph of relevance . Journal of Experimental & Theoretical Artificial Intelligence , 28(6):1035--1059, 2016

    work page 2016

  22. [22]

    OWL: A Systematic Approach to Temporal Versioning of Semantic Web Ontologies

    Abir Zekri, Zouhaier Brahmia, Fabio Grandi, and Rafik Bouaziz. OWL: A Systematic Approach to Temporal Versioning of Semantic Web Ontologies . Journal on Data Semantics , 5(3):141--163, 2016. https://doi.org/10.1007/s13740-016-0066-3 doi:10.1007/s13740-016-0066-3

    work page doi:10.1007/s13740-016-0066-3 2016

  23. [23]

    ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools

    Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Dan Zhang, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, Jing Zhang, Jingyu Sun, Juanzi Li, Lei Zhao, Lindong Wu, Lucen Zhong, Mingdao Liu, Minlie Huang, Peng Zhang, Qinkai Zheng, Rui Lu, Shuaiqi Duan, Shudan Zhang,...

    work page internal anchor Pith review Pith/arXiv arXiv 2024