Evaluation of Language Models in Knowledge Engineering
ELMKE 2026: 4th Workshop on Evaluation of Language Models in Knowledge Engineering
In conjunction with
International Semantic Web Conference 2026
October 25-29, 2026
Nicolaus Hotel Bari
Bari, Italy
Examines how to evaluate large language models when applied to knowledge engineering and ontology tasks. Evaluation is the bottleneck: language models produce ontology and knowledge-engineering output that looks correct and often is not, and existing benchmarks were not designed to catch that. Expect methodology rather than new models. Ontology work has correctness criteria that a fluent but wrong answer fails silently, which makes automated evaluation particularly awkward.
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...
- 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...
- Wiki-Based Knowledge Graph Question Answering — A challenge on answering natural-language questions over Wikidata and wiki-derived knowledge graphs. A challenge track,...