Inproceedings3927: Unterschied zwischen den Versionen
Xi5455 (Diskussion | Beiträge) K |
Xi5455 (Diskussion | Beiträge) K |
||
Zeile 25: | Zeile 25: | ||
}} | }} | ||
{{Publikation Details | {{Publikation Details | ||
− | |Abstract=In order to transform a Knowledge Graph (KG) into a low dimensional vector space, it is beneficial to preserve as much semantics as possible from the different components of the KG. Hence, some link pre- diction approaches have been proposed so far which leverage literals in addition to the commonly used links between entities. However, the pro- cedures followed to create the existing datasets do not pay attention to literals. Therefore, this study presents a set of KG completion benchmark datasets extracted from Wikidata and Wikipedia, named LiterallyWiki- data. It has been prepared with the main focus on providing benchmark datasets for multimodal KG Embedding (KGE) models, specifically for models using numeric and/or text literals. Hence, the benchmark is novel as compared to the existing datasets in terms of properly handling liter- als for those multimodal KGE models. LiterallyWikidata contains three datasets which vary both in size and structure. Benchmarking exper- iments on the task of link prediction have been conducted on Liter- allyWikidata with extensively tuned unimodal/multimodal KGE mod- els. The datasets are available at https://doi.org/10.5281/zenodo. 4701190 | + | |Abstract=In order to transform a Knowledge Graph (KG) into a low dimensional vector space, it is beneficial to preserve as much semantics as possible from the different components of the KG. Hence, some link pre- diction approaches have been proposed so far which leverage literals in addition to the commonly used links between entities. However, the pro- cedures followed to create the existing datasets do not pay attention to literals. Therefore, this study presents a set of KG completion benchmark datasets extracted from Wikidata and Wikipedia, named LiterallyWiki- data. It has been prepared with the main focus on providing benchmark datasets for multimodal KG Embedding (KGE) models, specifically for models using numeric and/or text literals. Hence, the benchmark is novel as compared to the existing datasets in terms of properly handling liter- als for those multimodal KGE models. LiterallyWikidata contains three datasets which vary both in size and structure. Benchmarking exper- iments on the task of link prediction have been conducted on Liter- allyWikidata with extensively tuned unimodal/multimodal KGE mod- els. The datasets are available at https://doi.org/10.5281/zenodo.4701190 |
|ISBN=978-3-030-88361-4 | |ISBN=978-3-030-88361-4 | ||
|ISSN=0302-9743, 1611-3349 | |ISSN=0302-9743, 1611-3349 | ||
+ | |Download=2021-Gesese-Alam-Sack-LiterallyWikidata.pdf | ||
|Link=https://link.springer.com/chapter/10.1007%2F978-3-030-88361-4_30 | |Link=https://link.springer.com/chapter/10.1007%2F978-3-030-88361-4_30 | ||
|DOI Name=10.1007/978-3-030-88361-4_30 | |DOI Name=10.1007/978-3-030-88361-4_30 | ||
|Forschungsgruppe=Information Service Engineering | |Forschungsgruppe=Information Service Engineering | ||
}} | }} |
Aktuelle Version vom 10. November 2022, 13:28 Uhr
LiterallyWikidata - A Benchmark for Knowledge Graph Completion Using Literals
LiterallyWikidata - A Benchmark for Knowledge Graph Completion Using Literals
Published: 2021
September
Herausgeber: Andreas Hotho, Eva Blomqvist, Stefan Dietze, Achille Fokoue, Ying Ding, Payam Barnaghi, Armin Haller, Mauro Dragoni, Harith Alani
Buchtitel: The Semantic Web – ISWC 2021 – 20th International Semantic Web Conference, ISWC 2021, Virtual Event, October 24–28, 2021, Proceedings
Nummer: 12922
Reihe: Lecture Notes in Computer Science (LNCS)
Seiten: 511–527
Verlag: Springer International
Organisation: 20th International Semantic Web Conference (ISWC 2021)
Referierte Veröffentlichung
BibTeX
Kurzfassung
In order to transform a Knowledge Graph (KG) into a low dimensional vector space, it is beneficial to preserve as much semantics as possible from the different components of the KG. Hence, some link pre- diction approaches have been proposed so far which leverage literals in addition to the commonly used links between entities. However, the pro- cedures followed to create the existing datasets do not pay attention to literals. Therefore, this study presents a set of KG completion benchmark datasets extracted from Wikidata and Wikipedia, named LiterallyWiki- data. It has been prepared with the main focus on providing benchmark datasets for multimodal KG Embedding (KGE) models, specifically for models using numeric and/or text literals. Hence, the benchmark is novel as compared to the existing datasets in terms of properly handling liter- als for those multimodal KGE models. LiterallyWikidata contains three datasets which vary both in size and structure. Benchmarking exper- iments on the task of link prediction have been conducted on Liter- allyWikidata with extensively tuned unimodal/multimodal KGE mod- els. The datasets are available at https://doi.org/10.5281/zenodo.4701190
ISBN: 978-3-030-88361-4
ISSN: 0302-9743, 1611-3349
Download: Media:2021-Gesese-Alam-Sack-LiterallyWikidata.pdf
Weitere Informationen unter: Link
DOI Link: 10.1007/978-3-030-88361-4_30
Information Service Engineering