Learning Tree Automata with Term Rewriting

2026-05-08

AI Summary

This work addresses the high query complexity of traditional tree automaton learning algorithms and their inability to exploit structural prior knowledge inherent in tree languages. The authors propose a novel approach that integrates term rewriting systems with Angluin-style active learning, leveraging rewriting rules to capture semantic properties of the target tree language—such as permutation invariance of subtrees—and employing deductive reasoning to automatically answer a subset of membership and equivalence queries. By embedding term rewriting into the active learning framework for tree automata—a first in the field—the method substantially reduces interactive query costs. Empirical evaluation on multiple tree languages with natural structural constraints demonstrates its effectiveness, achieving significant reductions in the number of required queries.

Abstract

We present an extension of the Angluin-style learning algorithm for tree automata that incorporates deductive inference. The learning algorithm is provided with a term rewriting system that specifies properties of the target tree language (e.g., the order of subtrees under a symbol f is irrelevant). This term rewriting system is used to infer answers to some queries, which reduces the query complexity of the learning algorithm. We present examples of rewrite systems that express natural properties of tree-structured data, which yield a significant reduction in the number of queries.

Problem

• tree automata
• term rewriting
• query complexity
• Angluin-style learning
• deductive inference

Innovation

• tree automata
• Angluin-style learning
• term rewriting
• deductive inference
• query complexity