Details
| Original language | English |
|---|---|
| Title of host publication | Knowledge Engineering and Knowledge Management |
| Subtitle of host publication | 24th International Conference, EKAW 2024, Proceedings |
| Editors | Mehwish Alam, Marco Rospocher, Marieke van Erp, Laura Hollink, Genet Asefa Gesese |
| Publisher | Springer Science and Business Media Deutschland GmbH |
| Pages | 471-489 |
| Number of pages | 19 |
| ISBN (electronic) | 978-3-031-77792-9 |
| ISBN (print) | 9783031777912 |
| Publication status | Published - 2025 |
| Event | 24th International Conference on Knowledge Engineering and Knowledge Management, EKAW 2024 - Amsterdam, Netherlands Duration: 26 Nov 2024 → 28 Nov 2024 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 15370 LNAI |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 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 subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
- General Computer Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Knowledge Engineering and Knowledge Management: 24th International Conference, EKAW 2024, Proceedings. ed. / Mehwish Alam; Marco Rospocher; Marieke van Erp; Laura Hollink; Genet Asefa Gesese. Springer Science and Business Media Deutschland GmbH, 2025. p. 471-489 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 15370 LNAI).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
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 -