Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Planning with OWL-DL Ontologies (Extended Version)

Tobias John, Patrick Koopmann

TL;DR

This work presents ontology-mediated planning (OMPS), coupling PDDL-based planning with expressive ontologies in a strongly decoupled interface. It extends planning to full OWL-DL by rewriting OMPS into PDDL and relies on justification-based reasoning to compute entailments $\operatorname{ent}(\alpha)$, enabling a generic, black-box integration with existing planners. Three optimization strategies—Basic, Concept-Based, and Schema-Leveraging—improve justification generation by reducing redundant reasoning and exploiting structural regularities. Evaluations across 116 problems from five domains show that Schema generally offers strong performance, outperforming Horn-based approaches in several cases while highlighting domain-dependent trade-offs and remaining challenges with inconsistent fluents and benchmark realism.

Abstract

We introduce ontology-mediated planning, in which planning problems are combined with an ontology. Our formalism differs from existing ones in that we focus on a strong separation of the formalisms for describing planning problems and ontologies, which are only losely coupled by an interface. Moreover, we present a black-box algorithm that supports the full expressive power of OWL DL. This goes beyond what existing approaches combining automated planning with ontologies can do, which only support limited description logics such as DL-Lite and description logics that are Horn. Our main algorithm relies on rewritings of the ontology-mediated planning specifications into PDDL, so that existing planning systems can be used to solve them. The algorithm relies on justifications, which allows for a generic approach that is independent of the expressivity of the ontology language. However, dedicated optimizations for computing justifications need to be implemented to enable an efficient rewriting procedure. We evaluated our implementation on benchmark sets from several domains. The evaluation shows that our procedure works in practice and that tailoring the reasoning procedure has significant impact on the performance.

Planning with OWL-DL Ontologies (Extended Version)

TL;DR

This work presents ontology-mediated planning (OMPS), coupling PDDL-based planning with expressive ontologies in a strongly decoupled interface. It extends planning to full OWL-DL by rewriting OMPS into PDDL and relies on justification-based reasoning to compute entailments , enabling a generic, black-box integration with existing planners. Three optimization strategies—Basic, Concept-Based, and Schema-Leveraging—improve justification generation by reducing redundant reasoning and exploiting structural regularities. Evaluations across 116 problems from five domains show that Schema generally offers strong performance, outperforming Horn-based approaches in several cases while highlighting domain-dependent trade-offs and remaining challenges with inconsistent fluents and benchmark realism.

Abstract

We introduce ontology-mediated planning, in which planning problems are combined with an ontology. Our formalism differs from existing ones in that we focus on a strong separation of the formalisms for describing planning problems and ontologies, which are only losely coupled by an interface. Moreover, we present a black-box algorithm that supports the full expressive power of OWL DL. This goes beyond what existing approaches combining automated planning with ontologies can do, which only support limited description logics such as DL-Lite and description logics that are Horn. Our main algorithm relies on rewritings of the ontology-mediated planning specifications into PDDL, so that existing planning systems can be used to solve them. The algorithm relies on justifications, which allows for a generic approach that is independent of the expressivity of the ontology language. However, dedicated optimizations for computing justifications need to be implemented to enable an efficient rewriting procedure. We evaluated our implementation on benchmark sets from several domains. The evaluation shows that our procedure works in practice and that tailoring the reasoning procedure has significant impact on the performance.
Paper Structure (19 sections, 8 theorems, 6 equations, 6 figures, 2 tables)

This paper contains 19 sections, 8 theorems, 6 equations, 6 figures, 2 tables.

Table of Contents

  1. Preliminaries
  2. Planning Problems
  3. Description Logics and Ontologies
  4. Framework
  5. Ontology-Mediated Planning
  6. Planning Procedure
  7. Optimized Generation of Justifications
  8. Basic Algorithm
  9. Concept-Based Algorithm
  10. Schema-Leveraging Algorithm
  11. Evaluation
  12. Implementation
  13. Experiments
  14. Benchmark Sets
  15. Setup
  16. ...and 4 more sections

Key Result

Theorem 1

Let $\textbf{OP}$ be an OMPS. Then, every plan for $\textbf{OP}$ can be translated into a plan in $\operatorname{rew}(\textbf{OP})$, and every plan in $\operatorname{rew}(\textbf{OP})$ can be translated into a plan in $\textbf{OP}$, by replacing each action by the corresponding action in the other s

Figures (6)

  • Figure 1: Example of ontology based planning. The interface maps ontology queries to planning predicates and atoms in the planning perspective to ABox atoms. The static part of the ontology contains information about instances (ABox) as well as general axioms (TBox). The connections between the two perspectives via the fluent (F) and query (Q) interface are shown in green.
  • Figure 2: Basic Hitting-Set-Tree algorithm
  • Figure 3: Example hitting-set trees for concept-based algorithm (A) without additional pruning and (B) with pruning.
  • Figure 4: Optimized branching for concept-based algorithm.
  • Figure 5: Cactus plot showing how many instances could be solved by the different algorithms within a given time bound. The y-axis is logarithmic.
  • ...and 1 more figures

Theorems & Definitions (20)

  • Definition 1: Ontology-Enhanced State
  • Example 1
  • Definition 2: Interface
  • Definition 3
  • Theorem 1
  • Definition 4: Justification
  • Lemma 1
  • Theorem 2
  • proof
  • Lemma 2
  • ...and 10 more