Datenintegration und -abfrage durch die Zusammenführung von Ontologien und Datenbanken (DIAMOND)
Zusammenfassung der Projektergebnisse
The DFG Emmy Noether Project DIAMOND (Data Integration and Access by Merging Ontologies and Databases) has conducted foundational and applied research on the general areas of symbolic artificial intelligence and data management. The main goal was to advance the principle of ontologybased data access, which allows users to obtain more complete and more adequate answers to questions by taking background knowledge about the meaning of the underlying data into account. This underlying data can be represented as a traditional database or as a knowledge graph, which puts more emphasis on the relationships and connections of elements of the domain. The background knowledge in turn is represented as an ontology : a formal conceptual model that can be evaluated automatically to infer additional knowledge. Applying this idea to real-world tasks leads to many challenges, chiefly covered by two questions: 1. How should background knowledge be modelled (capturing all relevant information)? 2. How can query answers be computed (with the performance needed in real applications)? Project DIAMOND has addressed these questions through a variety of foundational and engineeringoriented research contributions. On the foundational side, the main topics have been the design of suitable ontology-languages, the development of methods for reasoning with them, and the study of the theoretical properties of these languages. Notable outcomes of this work have been rulebased reasoning approaches for description logic ontologies, a novel type of attributed logics that support rich knowledge graphs, and insights into the use of existential rules to simulate set-valued data. Additional theoretical contributions have been made to the theory of sub-regular languages and related automata models. On the practical side, the project has contributed to the development of the rule engine VLog, which supports reasoning in the rule-based ontology languages investigated in the project. Its specific strength and innovation is the use of memory-efficient column-based data structures, which enables VLog to handle large datasets even on common hardware. The project has made further contributions to the design of algorithms and data structures for large-scale distributed graph databases. Results of DIAMOND have influenced important applications, most notably Wikidata, the successful sister project of Wikipedia that has become the most important free knowledge graph. DIAMOND’s contributions include the first approach of encoding Wikidata in the graph database format RDF, which has become the basis for the powerful Wikidata Query Service and is answering millions of queries each day. In cooperation with the Wikimedia Foundation, DIAMOND researchers have made contributions to the design of this service, and extracted a large research data set of over 570 million real-world queries, which can help to understand the practical use of large knowledge bases. The research output of the project comprises over 70 reviewed publications, many of which have appeared at leading international venues and which have already attracted thousands of citations. The quality and visibility of the research is further recognised by two best paper awards, an honourable mention at the largest AI conference, and three best paper nominations – all at top-ranked international conferences. Principal investigator Krötzsch received the DFG Heinz Maier-Leibnitz-Preis for his contributions.
Projektbezogene Publikationen (Auswahl)
- Introducing Wikidata to the linked data web. In Proceedings of the 13th International Semantic Web Conference (ISWC 2014), volume 8796 of LNCS, pages 50–65. Springer, October 2014
Fredo Erxleben, Michael Günther, Markus Krötzsch, Julian Mendez, and Denny Vrandečić
(Siehe online unter https://doi.org/10.1007/978-3-319-11964-9_4) - Wikidata: a free collaborative knowledgebase. Commun. ACM, 57(10):78–85, 2014
Denny Vrandečić and Markus Krötzsch
(Siehe online unter https://doi.org/10.1145/2629489) - Reasonable highly expressive query languages. In Proc. 24th International Joint Conference on Artificial Intelligence (IJCAI’15), pages 2826–2832. AAAI Press, 2015 [Best Paper Award (Honourable Mention) at IJCAI 2015]
Pierre Bourhis, Markus Krötzsch, and Sebastian Rudolph
- Column-oriented datalog materialization for large knowledge graphs. In Proceedings of the 30th AAAI Conference on Artificial Intelligence, pages 258–264. AAAI Press, 2016
Jacopo Urbani, Ceriel Jacobs, and Markus Krötzsch
(Siehe online unter https://doi.org/10.1609/aaai.v30i1.9993) - Logic on MARS: ontologies for generalised property graphs. In Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI’17), pages 1188–1194. International Joint Conferences on Artificial Intelligence, August 2017
Maximilian Marx, Markus Krötzsch, and Veronika Thost
(Siehe online unter https://doi.org/10.24963/ijcai.2017/165) - Restricted chase (non)termination for existential rules with disjunctions. In Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI’17), pages 922–928. International Joint Conferences on Artificial Intelligence, 2017
David Carral, Irina Dragoste, and Markus Krötzsch
(Siehe online unter https://doi.org/10.24963/ijcai.2017/128) - Getting the most out of Wikidata: Semantic technology usage in Wikipedia’s knowledge graph. In Proceedings of the 17th International Semantic Web Conference (ISWC’18), volume 11137 of LNCS, pages 376–394. Springer, 2018 [Best Paper Award, In-Use Track ISWC 2018]
Stanislav Malyshev, Markus Krötzsch, Larry González, Julius Gonsior, and Adrian Bielefeldt
(Siehe online unter https://doi.org/10.1007/978-3-030-00668-6_23) - Chasing sets: How to use existential rules for expressive reasoning. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI’19). International Joint Conferences on Artificial Intelligence, August 2019
David Carral, Irina Dragoste, Markus Krötzsch, and Christian Lewe
- Vlog: A rule engine for knowledge graphs. In Proceedings of the 18th International Semantic Web Conference (ISWC’19) Part II, volume 11779 of LNCS. Springer, 2019 [Best Paper Award, Resources Track ISWC 2019]
David Carral, Irina Dragoste, Larry González, Ceriel Jacobs, Markus Krötzsch, and Jacopo Urbani
(Siehe online unter https://doi.org/10.1007/978-3-030-30796-7_2) - Partially ordered automata and piecewise testability. Logical Methods in Computer Science, 17(2), 2021
Tomáš Masopust and Markus Krötzsch
(Siehe online unter https://doi.org/10.23638/LMCS-17(2:14)2021)
