Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autorschaft

  • Michalis Mitsios
  • Dharmen Punjani
  • Sara Abdollahi
  • Simon Gottschalk
  • Eleni Tsalapati
  • Elena Demidova
  • Manolis Koubarakis

Organisationseinheiten

Externe Organisationen

  • University of Athens
  • Université Jean Monnet Saint-Étienne
  • Deutsche Akademie der Technikwissenschaften (acatech)
  • Athens Technology Center S.A.
  • Rheinische Friedrich-Wilhelms-Universität Bonn
  • Lamarr-Institut für Maschinelles Lernen und Künstliche Intelligenz
  • Archimedes/Athena RC
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des SammelwerksKnowledge Engineering and Knowledge Management
Untertitel24th International Conference, EKAW 2024, Proceedings
Herausgeber/-innenMehwish Alam, Marco Rospocher, Marieke van Erp, Laura Hollink, Genet Asefa Gesese
Herausgeber (Verlag)Springer Science and Business Media Deutschland GmbH
Seiten471-489
Seitenumfang19
ISBN (elektronisch)978-3-031-77792-9
ISBN (Print)9783031777912
PublikationsstatusVeröffentlicht - 2025
Veranstaltung24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024 - Amsterdam, Niederlande
Dauer: 26 Nov. 202428 Nov. 2024

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band15370 LNAI
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Abstract

Most studies on semantic question answering (QA) are predominantly focused on encyclopedic knowledge graphs like DBpedia and Wikidata. These studies cover, if at all, the spatial and temporal characteristics of geospatial entities in isolation, not addressing them simultaneously. In this paper, we introduce a pipeline for creating question answering datasets for evaluating the reasoning capabilities of QA models in the context of geographic changes over time. This pipeline generates questions, GeoSPARQL queries, and corresponding answers by leveraging subgraph and query template extraction techniques. We exemplify this pipeline with the creation of the GeoChangesQA dataset with questions over a knowledge graph of US counties and states and their changes from 1629 to 2000. By evaluating GeoChangesQA using a Transformer-based model, we demonstrate that historical geospatial questions pose a substantial challenge for semantic question answering.

ASJC Scopus Sachgebiete

Zitieren

Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph. / Mitsios, Michalis; Punjani, Dharmen; Abdollahi, Sara et al.
Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings. Hrsg. / Mehwish Alam; Marco Rospocher; Marieke van Erp; Laura Hollink; Genet Asefa Gesese. Springer Science and Business Media Deutschland GmbH, 2025. S. 471-489 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 15370 LNAI).

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Mitsios, M, Punjani, D, Abdollahi, S, Gottschalk, S, Tsalapati, E, Demidova, E & Koubarakis, M 2025, Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph. in M Alam, M Rospocher, M van Erp, L Hollink & GA Gesese (Hrsg.), Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bd. 15370 LNAI, Springer Science and Business Media Deutschland GmbH, S. 471-489, 24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024, Amsterdam, Niederlande, 26 Nov. 2024. https://doi.org/10.1007/978-3-031-77792-9_28
Mitsios, M., Punjani, D., Abdollahi, S., Gottschalk, S., Tsalapati, E., Demidova, E., & Koubarakis, M. (2025). Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph. In M. Alam, M. Rospocher, M. van Erp, L. Hollink, & G. A. Gesese (Hrsg.), Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings (S. 471-489). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 15370 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-77792-9_28
Mitsios M, Punjani D, Abdollahi S, Gottschalk S, Tsalapati E, Demidova E et al. Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph. in Alam M, Rospocher M, van Erp M, Hollink L, Gesese GA, Hrsg., Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings. Springer Science and Business Media Deutschland GmbH. 2025. S. 471-489. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-77792-9_28
Mitsios, Michalis ; Punjani, Dharmen ; Abdollahi, Sara et al. / Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph. Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings. Hrsg. / Mehwish Alam ; Marco Rospocher ; Marieke van Erp ; Laura Hollink ; Genet Asefa Gesese. Springer Science and Business Media Deutschland GmbH, 2025. S. 471-489 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Download
@inproceedings{f6f9ab36ae914310a249e767f914dd43,
title = "Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph",
abstract = "Most studies on semantic question answering (QA) are predominantly focused on encyclopedic knowledge graphs like DBpedia and Wikidata. These studies cover, if at all, the spatial and temporal characteristics of geospatial entities in isolation, not addressing them simultaneously. In this paper, we introduce a pipeline for creating question answering datasets for evaluating the reasoning capabilities of QA models in the context of geographic changes over time. This pipeline generates questions, GeoSPARQL queries, and corresponding answers by leveraging subgraph and query template extraction techniques. We exemplify this pipeline with the creation of the GeoChangesQA dataset with questions over a knowledge graph of US counties and states and their changes from 1629 to 2000. By evaluating GeoChangesQA using a Transformer-based model, we demonstrate that historical geospatial questions pose a substantial challenge for semantic question answering.",
author = "Michalis Mitsios and Dharmen Punjani and Sara Abdollahi and Simon Gottschalk and Eleni Tsalapati and Elena Demidova and Manolis Koubarakis",
note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024 ; Conference date: 26-11-2024 Through 28-11-2024",
year = "2025",
doi = "10.1007/978-3-031-77792-9_28",
language = "English",
isbn = "9783031777912",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "471--489",
editor = "Mehwish Alam and Marco Rospocher and {van Erp}, Marieke and Laura Hollink and Gesese, {Genet Asefa}",
booktitle = "Knowledge Engineering and Knowledge Management",
address = "Germany",

}

Download

TY - GEN

T1 - Generating a Question Answering Dataset About Geographic Changes in a Knowledge Graph

AU - Mitsios, Michalis

AU - Punjani, Dharmen

AU - Abdollahi, Sara

AU - Gottschalk, Simon

AU - Tsalapati, Eleni

AU - Demidova, Elena

AU - Koubarakis, Manolis

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

PY - 2025

Y1 - 2025

N2 - Most studies on semantic question answering (QA) are predominantly focused on encyclopedic knowledge graphs like DBpedia and Wikidata. These studies cover, if at all, the spatial and temporal characteristics of geospatial entities in isolation, not addressing them simultaneously. In this paper, we introduce a pipeline for creating question answering datasets for evaluating the reasoning capabilities of QA models in the context of geographic changes over time. This pipeline generates questions, GeoSPARQL queries, and corresponding answers by leveraging subgraph and query template extraction techniques. We exemplify this pipeline with the creation of the GeoChangesQA dataset with questions over a knowledge graph of US counties and states and their changes from 1629 to 2000. By evaluating GeoChangesQA using a Transformer-based model, we demonstrate that historical geospatial questions pose a substantial challenge for semantic question answering.

AB - Most studies on semantic question answering (QA) are predominantly focused on encyclopedic knowledge graphs like DBpedia and Wikidata. These studies cover, if at all, the spatial and temporal characteristics of geospatial entities in isolation, not addressing them simultaneously. In this paper, we introduce a pipeline for creating question answering datasets for evaluating the reasoning capabilities of QA models in the context of geographic changes over time. This pipeline generates questions, GeoSPARQL queries, and corresponding answers by leveraging subgraph and query template extraction techniques. We exemplify this pipeline with the creation of the GeoChangesQA dataset with questions over a knowledge graph of US counties and states and their changes from 1629 to 2000. By evaluating GeoChangesQA using a Transformer-based model, we demonstrate that historical geospatial questions pose a substantial challenge for semantic question answering.

UR - http://www.scopus.com/inward/record.url?scp=85210834700&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-77792-9_28

DO - 10.1007/978-3-031-77792-9_28

M3 - Conference contribution

AN - SCOPUS:85210834700

SN - 9783031777912

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

SP - 471

EP - 489

BT - Knowledge Engineering and Knowledge Management

A2 - Alam, Mehwish

A2 - Rospocher, Marco

A2 - van Erp, Marieke

A2 - Hollink, Laura

A2 - Gesese, Genet Asefa

PB - Springer Science and Business Media Deutschland GmbH

T2 - 24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024

Y2 - 26 November 2024 through 28 November 2024

ER -

Von denselben Autoren