A Single Approach to Decide Chase Termination on Linear Existential Rules

Michel Leclère (1) , Marie-Laure Mugnier (1) , Michaël Thomazo (2) , Federico Ulliana (1)

1 GRAPHIK - Graphs for Inferences on Knowledge
2 VALDA - Value from Data

Existential rules, long known as tuple-generating dependencies in database theory , have been intensively studied in the last decade as a powerful formalism to represent ontological knowledge in the context of ontology-based query answering. A knowledge base is then composed of an instance that contains incomplete data and a set of existential rules, and answers to queries are logically entailed from the knowledge base. This brought again to light the fundamental chase tool, and its different variants that have been proposed in the literature. It is well-known that the problem of determining, given a chase variant and a set of existential rules, whether the chase will halt on any instance, is undecidable. Hence, a crucial issue is whether it becomes decidable for known subclasses of existential rules. In this work, we consider linear existential rules, a simple yet important subclass of ex-istential rules that generalizes inclusion dependencies. We show the decidability of the all instance chase termination problem on linear rules for three main chase variants, namely semi-oblivious, restricted and core chase. To obtain these results, we introduce a novel approach based on so-called derivation trees and a single notion of forbidden pattern. Besides the theoretical interest of a unified approach and new proofs, we provide the first positive decidability results concerning the termination of the restricted chase, proving that chase termination on linear existential rules is decidable for both versions of the problem: Does every fair chase sequence terminate? Does some fair chase sequence terminate?

Domaines: Intelligence artificielle [cs.AI]

Michel Leclère, Marie-Laure Mugnier, Michaël Thomazo, Federico Ulliana. A Single Approach to Decide Chase Termination on Linear Existential Rules. [Research Report] arXiv:1810.02132. 2018. ⟨lirmm-01892375⟩