Multi-label Learning with a Cone-Based Geometric Model

Mena Leemhuis; Özgür L. Özçep; Diedrich Wolter

doi:10.1007/978-3-030-57855-8_13

Multi-label Learning with a Cone-Based Geometric Model

Mena Leemhuis^*, Özgür L. Özçep, Diedrich Wolter

^*Corresponding author for this work

University of Bamberg

Abstract

Recent approaches for knowledge-graph embeddings aim at connecting quantitative data structures used in machine learning to the qualitative structures of logics. Such embeddings are of a hybrid nature, they are data models that also exhibit conceptual structures inherent to logics. One motivation to investigate embeddings is to design conceptually adequate machine learning (ML) algorithms. This paper investigates a new approach to embedding ontologies into geometric models that interpret concepts by closed convex cones. As a proof of concept this cone-based embedding was implemented in a ML algorithm for weak supervised multi-label learning. The system was tested with the gene ontology and showed a performance similar to comparable approaches, but with the advantage of exhibiting the conceptual structure underlying the data.

Original language	English
Title of host publication	ICCS 2020: Ontologies and Concepts in Mind and Machine
Editors	Mehwish Alam, Tanya Braun, Bruno Yun
Number of pages	9
Volume	12277 LNAI
Publisher	Springer, Cham
Publication date	10.09.2020
Pages	177-185
ISBN (Print)	978-3-030-57854-1
ISBN (Electronic)	978-3-030-57855-8
DOIs	https://doi.org/10.1007/978-3-030-57855-8_13
Publication status	Published - 10.09.2020
Event	25th International Conference on Conceptual Structures - Bolzano, Italy Duration: 18.09.2020 → 20.09.2020 Conference number: 245219

Research Areas and Centers

Centers: Center for Artificial Intelligence Luebeck (ZKIL)
Research Area: Intelligent Systems

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-030-57855-8_13

Cite this

@inproceedings{83c01109a39f49a891c7f85ac4fd33d5,

title = "Multi-label Learning with a Cone-Based Geometric Model",

abstract = "Recent approaches for knowledge-graph embeddings aim at connecting quantitative data structures used in machine learning to the qualitative structures of logics. Such embeddings are of a hybrid nature, they are data models that also exhibit conceptual structures inherent to logics. One motivation to investigate embeddings is to design conceptually adequate machine learning (ML) algorithms. This paper investigates a new approach to embedding ontologies into geometric models that interpret concepts by closed convex cones. As a proof of concept this cone-based embedding was implemented in a ML algorithm for weak supervised multi-label learning. The system was tested with the gene ontology and showed a performance similar to comparable approaches, but with the advantage of exhibiting the conceptual structure underlying the data.",

author = "Mena Leemhuis and {\"O}z{\c c}ep, {{\"O}zg{\"u}r L.} and Diedrich Wolter",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.; 25th International Conference on Conceptual Structures , ICCS 2020 ; Conference date: 18-09-2020 Through 20-09-2020",

year = "2020",

month = sep,

day = "10",

doi = "10.1007/978-3-030-57855-8_13",

language = "English",

isbn = "978-3-030-57854-1",

volume = "12277 LNAI",

series = "Lecture Notes in Computer Science",

publisher = "Springer, Cham",

pages = "177--185",

editor = "Alam, {Mehwish } and Braun, {Tanya } and Yun, {Bruno }",

booktitle = "ICCS 2020: Ontologies and Concepts in Mind and Machine",

}

Leemhuis, M, Özçep, ÖL & Wolter, D 2020, Multi-label Learning with a Cone-Based Geometric Model. in M Alam, T Braun & B Yun (eds), ICCS 2020: Ontologies and Concepts in Mind and Machine. vol. 12277 LNAI, Lecture Notes in Computer Science, vol. 12277 LNAI, Springer, Cham, pp. 177-185, 25th International Conference on Conceptual Structures , Bolzano, Italy, 18.09.20. https://doi.org/10.1007/978-3-030-57855-8_13

Multi-label Learning with a Cone-Based Geometric Model. / Leemhuis, Mena; Özçep, Özgür L.; Wolter, Diedrich.
ICCS 2020: Ontologies and Concepts in Mind and Machine. ed. / Mehwish Alam; Tanya Braun; Bruno Yun. Vol. 12277 LNAI Springer, Cham, 2020. p. 177-185 (Lecture Notes in Computer Science; Vol. 12277 LNAI).

Research output: Chapters in Books/Reports/Conference Proceedings › Conference contribution › peer-review

TY - GEN

T1 - Multi-label Learning with a Cone-Based Geometric Model

AU - Leemhuis, Mena

AU - Özçep, Özgür L.

AU - Wolter, Diedrich

N1 - Conference code: 245219

PY - 2020/9/10

Y1 - 2020/9/10

N2 - Recent approaches for knowledge-graph embeddings aim at connecting quantitative data structures used in machine learning to the qualitative structures of logics. Such embeddings are of a hybrid nature, they are data models that also exhibit conceptual structures inherent to logics. One motivation to investigate embeddings is to design conceptually adequate machine learning (ML) algorithms. This paper investigates a new approach to embedding ontologies into geometric models that interpret concepts by closed convex cones. As a proof of concept this cone-based embedding was implemented in a ML algorithm for weak supervised multi-label learning. The system was tested with the gene ontology and showed a performance similar to comparable approaches, but with the advantage of exhibiting the conceptual structure underlying the data.

AB - Recent approaches for knowledge-graph embeddings aim at connecting quantitative data structures used in machine learning to the qualitative structures of logics. Such embeddings are of a hybrid nature, they are data models that also exhibit conceptual structures inherent to logics. One motivation to investigate embeddings is to design conceptually adequate machine learning (ML) algorithms. This paper investigates a new approach to embedding ontologies into geometric models that interpret concepts by closed convex cones. As a proof of concept this cone-based embedding was implemented in a ML algorithm for weak supervised multi-label learning. The system was tested with the gene ontology and showed a performance similar to comparable approaches, but with the advantage of exhibiting the conceptual structure underlying the data.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85091575147&origin=inward&txGid=ef966a3390f529f08d1e7022a32a6b09

UR - https://www.mendeley.com/catalogue/6f1ceb82-4736-3ad8-84f8-77a90626553b/

U2 - 10.1007/978-3-030-57855-8_13

DO - 10.1007/978-3-030-57855-8_13

M3 - Conference contribution

SN - 978-3-030-57854-1

VL - 12277 LNAI

T3 - Lecture Notes in Computer Science

SP - 177

EP - 185

BT - ICCS 2020: Ontologies and Concepts in Mind and Machine

A2 - Alam, Mehwish

A2 - Braun, Tanya

A2 - Yun, Bruno

PB - Springer, Cham

T2 - 25th International Conference on Conceptual Structures

Y2 - 18 September 2020 through 20 September 2020

ER -

Multi-label Learning with a Cone-Based Geometric Model

Abstract

Research Areas and Centers

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this