SPARQL-LLM leads multilingual question answering for knowledge graphs
SPARQL-LLM in the TEXT2SPARQL 2026 Challenge
Information Retrieval
Summary
Making computer questions to find answers in linked databases is hard because you need to know special rules and terms. The authors made a tool called SPARQL-LLM that uses advanced language AI to turn normal questions into the right database language, SPARQL. They tested it in a big contest where it did much better than other methods and worked in English, German, and Spanish. This tool is free for anyone to use and already helps biologists find data in several linked databases.
What this means in practice
- •For bioinformatics teams: Convert natural language queries into SPARQL automatically to search decentralized biological databases efficiently.
- •For multilingual software developers: Build question answering features that understand English, German, and Spanish user queries over knowledge graphs.
Authors
Panayiotis Smeros, Vincent Emonet, Ruijie Wang, Ana-Claudia Sima, Tarcisio Mendes de Farias
Abstract
Querying distributed data represented as Knowledge Graphs (KGs) via SPARQL remains a difficult task, requiring fluency in query syntax, endpoint specifications, and schema details. Knowledge Graph Question Answering (KGQA) systems address this by automatically translating natural language into SPARQL, and recent advances in Large Language Models (LLMs) have made this translation increasingly feasible. In this paper, we describe SPARQL-LLM, our open-source KGQA system, in the context of the TEXT2SPARQL 2026 challenge. In this challenge, SPARQL-LLM ranks first overall among eight competing approaches across all datasets, achieving up to 59% higher accuracy than the second-best approach and supporting multilingual question answering in English, German, and Spanish. SPARQL-LLM is publicly released at https://github.com/sib-swiss/sparql-llm and deployed in production over decentralized bioinformatics KGs at https://www.expasy.org/chat.