ReAD: AD-Based Modular Ontology Classification

Haoruo Zhao; Bijan Parsia; Uli Sattler

doi:10.1007/978-3-030-75775-5_15

ReAD: AD-Based Modular Ontology Classification

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

For OWL ontologies, classification is the central reasoning task, and several highly-optimised reasoners have been designed for different fragments of OWL. Some of these exploit different notions of modularity, including the atomic decomposition (AD), to further optimise their performance, but this is a complex task due to ontology modules overlapping, thereby possibly causing duplication of subsumption tests.

In this paper, we use the AD to avoid both this duplication as well as other subsumption tests that can be avoided by inspecting the AD. We have designed and implemented a new AD-informed and MORe -inspired algorithm that uses Hermit and ELK as delegate reasoners, but avoids any duplicate subsumption tests between these two reasoners and further minimises these tests. We have thoroughly evaluated the effects of these two kinds of avoidance on the overall classification time on a corpus of complex ontologies.

Original language	English
Title of host publication	Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings
Editors	Wolfgang Faber, Gerhard Friedrich, Martin Gebser, Michael Morak
Pages	210-224
Number of pages	15
Volume	12678
DOIs	https://doi.org/10.1007/978-3-030-75775-5_15
Publication status	Published - 12 May 2021

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12678 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Keywords

Classification
Description logic
OWL
Reasoning

Access to Document

10.1007/978-3-030-75775-5_15

Cite this

Zhao, H., Parsia, B., & Sattler, U. (2021). ReAD: AD-Based Modular Ontology Classification. In W. Faber, G. Friedrich, M. Gebser, & M. Morak (Eds.), Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings (Vol. 12678, pp. 210-224). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12678 LNAI). https://doi.org/10.1007/978-3-030-75775-5_15

Zhao, Haoruo ; Parsia, Bijan ; Sattler, Uli. / ReAD: AD-Based Modular Ontology Classification. Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings. editor / Wolfgang Faber ; Gerhard Friedrich ; Martin Gebser ; Michael Morak. Vol. 12678 2021. pp. 210-224 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d17552d049414f7691597ee4d941a895,

title = "ReAD: AD-Based Modular Ontology Classification",

abstract = "For OWL ontologies, classification is the central reasoning task, and several highly-optimised reasoners have been designed for different fragments of OWL. Some of these exploit different notions of modularity, including the atomic decomposition (AD), to further optimise their performance, but this is a complex task due to ontology modules overlapping, thereby possibly causing duplication of subsumption tests.In this paper, we use the AD to avoid both this duplication as well as other subsumption tests that can be avoided by inspecting the AD. We have designed and implemented a new AD-informed and MORe -inspired algorithm that uses Hermit and ELK as delegate reasoners, but avoids any duplicate subsumption tests between these two reasoners and further minimises these tests. We have thoroughly evaluated the effects of these two kinds of avoidance on the overall classification time on a corpus of complex ontologies.",

keywords = "Classification, Description logic, OWL, Reasoning",

author = "Haoruo Zhao and Bijan Parsia and Uli Sattler",

note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.",

year = "2021",

month = may,

day = "12",

doi = "10.1007/978-3-030-75775-5_15",

language = "English",

isbn = "9783030757748",

volume = "12678",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "210--224",

editor = "Wolfgang Faber and Gerhard Friedrich and Martin Gebser and Michael Morak",

booktitle = "Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings",

}

Zhao, H, Parsia, B & Sattler, U 2021, ReAD: AD-Based Modular Ontology Classification. in W Faber, G Friedrich, M Gebser & M Morak (eds), Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings. vol. 12678, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12678 LNAI, pp. 210-224. https://doi.org/10.1007/978-3-030-75775-5_15

ReAD: AD-Based Modular Ontology Classification. / Zhao, Haoruo; Parsia, Bijan ; Sattler, Uli.
Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings. ed. / Wolfgang Faber; Gerhard Friedrich; Martin Gebser; Michael Morak. Vol. 12678 2021. p. 210-224 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12678 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - ReAD: AD-Based Modular Ontology Classification

AU - Zhao, Haoruo

AU - Parsia, Bijan

AU - Sattler, Uli

PY - 2021/5/12

Y1 - 2021/5/12

N2 - For OWL ontologies, classification is the central reasoning task, and several highly-optimised reasoners have been designed for different fragments of OWL. Some of these exploit different notions of modularity, including the atomic decomposition (AD), to further optimise their performance, but this is a complex task due to ontology modules overlapping, thereby possibly causing duplication of subsumption tests.In this paper, we use the AD to avoid both this duplication as well as other subsumption tests that can be avoided by inspecting the AD. We have designed and implemented a new AD-informed and MORe -inspired algorithm that uses Hermit and ELK as delegate reasoners, but avoids any duplicate subsumption tests between these two reasoners and further minimises these tests. We have thoroughly evaluated the effects of these two kinds of avoidance on the overall classification time on a corpus of complex ontologies.

AB - For OWL ontologies, classification is the central reasoning task, and several highly-optimised reasoners have been designed for different fragments of OWL. Some of these exploit different notions of modularity, including the atomic decomposition (AD), to further optimise their performance, but this is a complex task due to ontology modules overlapping, thereby possibly causing duplication of subsumption tests.In this paper, we use the AD to avoid both this duplication as well as other subsumption tests that can be avoided by inspecting the AD. We have designed and implemented a new AD-informed and MORe -inspired algorithm that uses Hermit and ELK as delegate reasoners, but avoids any duplicate subsumption tests between these two reasoners and further minimises these tests. We have thoroughly evaluated the effects of these two kinds of avoidance on the overall classification time on a corpus of complex ontologies.

KW - Classification

KW - Description logic

KW - OWL

KW - Reasoning

U2 - 10.1007/978-3-030-75775-5_15

DO - 10.1007/978-3-030-75775-5_15

M3 - Conference contribution

SN - 9783030757748

VL - 12678

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 210

EP - 224

BT - Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings

A2 - Faber, Wolfgang

A2 - Friedrich, Gerhard

A2 - Gebser, Martin

A2 - Morak, Michael

ER -

Zhao H, Parsia B , Sattler U. ReAD: AD-Based Modular Ontology Classification. In Faber W, Friedrich G, Gebser M, Morak M, editors, Logics in Artiﬁcial Intelligence - 17th European Conference, JELIA 2021, Proceedings. Vol. 12678. 2021. p. 210-224. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-75775-5_15

ReAD: AD-Based Modular Ontology Classification

Abstract

Publication series

Keywords

Access to Document

Fingerprint

Cite this