Robert Craven
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Robert Craven has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Dalal Alrajeh
Imperial College London UK
1 shared publication - Daniel Özcan
Imperial College London
1 shared publication
Latest Publications
- Open Access
- 1 November 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 22 (1) , 610-620
Reasoning About Actual Causality in Answer Set Programming
Causal models provide a formal framework for identifying and reasoning about the causes of observed phenomena, making them valuable for decision-support contexts where understanding causality is essential. Yet applying these models in practice requires automated tools for key reasoning tasks. We present an Answer Set Programming (ASP)-based tool that supports three core capabilities for all acyclic binary causal models: (1) checking whether an event is an actual cause of another; (2) finding all minimal subsets of a failed candidate that do qualify as causes; and (3) inferring all actual causes of an outcome without assuming any candidate. Our tool is the first to support all three tasks within a unified framework, guaranteeing minimal contingency sets and outperforming prior implementations in both runtime and memory. We describe the system’s design and report on an empirical evaluation using existing benchmarks.
- 1 November 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 22 (1) , 610-620
Open AccessReasoning About Actual Causality in Answer Set Programming
Causal models provide a formal framework for identifying and reasoning about the causes of observed phenomena, making them valuable for decision-support contexts where understanding causality is essential. Yet applying these models in practice requires automated tools for key reasoning tasks. We present an Answer Set Programming (ASP)-based tool that supports three core capabilities for all acyclic binary causal models: (1) checking whether an event is an actual cause of another; (2) finding all minimal subsets of a failed candidate that do qualify as causes; and (3) inferring all actual causes of an outcome without assuming any candidate. Our tool is the first to support all three tasks within a unified framework, guaranteeing minimal contingency sets and outperforming prior implementations in both runtime and memory. We describe the system’s design and report on an empirical evaluation using existing benchmarks.