Towards temporally uncertain explainable AI planning

Murray, Andrew and Krarup, Benjamin and Cashmore, Michael; Bapi, Raju and Kulkarni, Sandeep and Mohalik, Swarup and Peri, Sathya, eds. (2022) Towards temporally uncertain explainable AI planning. In: Distributed Computing and Intelligent Technology - 18th International Conference, ICDCIT 2022, Lecture Notes in Computer Science. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13145 . Springer Science and Business Media Deutschland GmbH, IND, pp. 45-59. ISBN 9783030948757 (https://doi.org/10.1007/978-3-030-94876-4_3)

[thumbnail of Murray-etal-LNCS-2022-Towards-temporally-uncertain-explainable-AI-planning]
Preview
Text. Filename: Murray_etal_LNCS_2022_Towards_temporally_uncertain_explainable_AI_planning.pdf
Accepted Author Manuscript
License: Strathprints license 1.0

Download (814kB)| Preview

Abstract

Automated planning is able to handle increasingly complex applications, but can produce unsatisfactory results when the goal and metric provided in its model does not match the actual expectation and preference of those using the tool. This can be ameliorated by including methods for explainable planning (XAIP), to reveal the reasons for the automated planner’s decisions and to provide more in-depth interaction with the planner. In this paper we describe at a high-level two recent pieces of work in XAIP. First, plan exploration through model restriction, in which contrastive questions are used to build a tree of solutions to a planning problem. Through a dialogue with the system the user better understands the underlying problem and the choices made by the automated planner. Second, strong controllability analysis of probabilistic temporal networks through solving a joint chance constrained optimisation problem. The result of the analysis is a Pareto optimal front that illustrates the trade-offs between costs and risk for a given plan. We also present a short discussion on the limitations of these methods and how they might be usefully combined.