On the Semantics of Generative SPARQL

2026-06-22

AI Summary

This work addresses the challenge of seamlessly integrating the generative capabilities of language models into SPARQL while preserving its formal semantics and resolving semantic mismatches between generated values and RDF terms. To this end, the authors propose GenOp, an extension to SPARQL that enables invoking language models to produce typed result mappings, and formally define its semantics under a fixed dataset assumption. The core contributions include the first formal semantic framework for generative SPARQL, the introduction of a compatibility relation generalizing equality-based matching, the characterization of fixpoint semantics for acyclic and stratified fragments, and the establishment of semantics-preserving algebraic rewriting rules. Under assumptions of deterministic bounded generation and finite candidate coverage, the approach is executable with data and combined complexity matching that of standard SPARQL.

Abstract

We extend SPARQL with a generative query construct, called \tx{GenOp}, whose evaluation calls a language model and produces typed solution mappings. We define the semantics of the GenOp in the query in a way that maintains the fixed-dataset assumption, on which formal semantics of SPARQL build, and extend solution mappings with values generated by the language model. We formalize the semantics of the extended language over these mappings using a compatibility relation that generalizes equality and supports similarity-based matching between RDF terms and generated values. We analyze the semantic consequences of generative query patterns, focusing on mapping-level recursion induced by the reuse of generated bindings. Under deterministic bounded generation and finite candidate coverage assumptions, we characterize acyclic and stratified fragments with fixpoint semantics, establish algebraic equivalence and semantics-preserving rewrite rules, and provide an executable evaluation method; and we show that data and combined complexity coincide with those of standard SPARQL.

Problem

• Generative SPARQL
• language model integration
• query semantics
• solution mappings
• RDF

Innovation

• Generative SPARQL
• GenOp
• language model integration
• compatibility relation
• fixpoint semantics