GRISP: Guided Recurrent IRI Selection over SPARQL Skeletons

2026-04-22

AI Summary

This work addresses the challenge of error-prone entity and predicate linking in mapping natural language questions to SPARQL queries over knowledge graphs. The authors propose an end-to-end approach based on a fine-tuned small language model that first generates a natural language representation of a SPARQL query skeleton and then iteratively re-ranks candidate IRIs under knowledge graph constraints to replace placeholders. By jointly training the skeleton generation and listwise re-ranking tasks, the model achieves guided, precise IRI selection. Experimental results on standard benchmarks over Wikidata and Freebase demonstrate that the proposed method significantly outperforms existing state-of-the-art systems, achieving new record accuracy in question answering.

Technology Category

• Natural Language Processing: Question Answering
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Description Logics

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We present GRISP (Guided Recurrent IRI Selection over SPARQL Skeletons), a novel SPARQL-based question-answering method over knowledge graphs based on fine-tuning a small language model (SLM). Given a natural-language question, the method first uses the SLM to generate a natural-language SPARQL query skeleton, and then to re-rank and select knowledge graph items to iteratively replace the natural-language placeholders using knowledge graph constraints. The SLM is jointly trained on skeleton generation and list-wise re-ranking data generated from standard question-query pairs. We evaluate the method on common Wikidata and Freebase benchmarks, and achieve better results than other state-of-the-art methods in a comparable setting.

Problem

• SPARQL
• knowledge graph
• question answering
• IRI selection
• natural language query

Innovation

• GRISP
• SPARQL skeleton
• small language model
• IRI selection
• knowledge graph question answering