Wiki-Based Knowledge Graph Question Answering
WikiKGQA: Wiki-Based Knowledge Graph Question Answering Challenge
In conjunction with
International Semantic Web Conference 2026
October 25-29, 2026
Nicolaus Hotel Bari
Bari, Italy
A challenge on answering natural-language questions over Wikidata and wiki-derived knowledge graphs. A challenge track, so results are directly comparable. Answering questions over Wikidata is harder than it looks because the schema is community-built and inconsistent, and the right answer often depends on which modelling convention was used. Ambiguity in the question and in the data compound, which is why apparently simple questions are often the hardest cases.
More workshops at International Semantic Web Conference 2026
- Data Management for Knowledge Graphs — Covers storage, querying, and management techniques for large-scale knowledge graphs on the web. Knowledge graphs...
- Evaluation of Language Models in Knowledge Engineering — Examines how to evaluate large language models when applied to knowledge engineering and ontology tasks....
- Graph-Enhanced LLMs for Web Data — Explores combining knowledge graphs with large language models for trustworthy management of web data. The...
- Knowledge Graphs and Model-driven Systems Engineering — Bridges knowledge graphs with model-driven engineering for building and maintaining software systems. Model-driven engineering and...
- Scientific Knowledge Representation and Discovery — Focuses on representing, discovering, and assessing scholarly and scientific knowledge as structured data. Scholarly knowledge...
- Semantic Materials Science — Applies semantic web and knowledge graph methods to materials science data and discovery. Materials data...
- Wikidata Workshop — The community workshop on Wikidata, the collaborative knowledge base, covering its data, tooling, and research...