Heuristic Approaches for Goal Recognition in Incomplete Domain Models
This addresses a critical limitation for AI systems in real-world domains where domain knowledge is often incomplete or incorrect, though it appears incremental as it builds on prior work relaxing assumptions.
The paper tackles the problem of goal recognition in incomplete and possibly incorrect domain models, which is a limitation of existing methods that assume complete and correct domain theories. The result is the development of new techniques that demonstrate efficiency and accuracy on a large dataset of goal and plan recognition problems.
Recent approaches to goal recognition have progressively relaxed the assumptions about the amount and correctness of domain knowledge and available observations, yielding accurate and efficient algorithms. These approaches, however, assume completeness and correctness of the domain theory against which their algorithms match observations: this is too strong for most real-world domains. In this paper, we develop goal recognition techniques that are capable of recognizing goals using \textit{incomplete} (and possibly incorrect) domain theories. We show the efficiency and accuracy of our approaches empirically against a large dataset of goal and plan recognition problems with incomplete domains.