Plan Recognition is the task of identifying the goals and plans of an agent by observing its behavior within the environment. The problem has been extensively studied in the context of planning, in particular bringing forward stochastic techniques dealing with probability distributions over the possible agent goals, under the assumption that observations are reliable. More recently, a connection between this problem and process mining techniques has been established, paving the way towards the application of alignment-based conformance checking techniques from process mining to tackle plan recognition problems in a setting where observations may be faulty. In this work, we reconcile these two lines of research in a unified framework that deals at once with uncertainty over the goals and the faithfulness of observations. Instead of using ad-hoc techniques to solve this problem, we cast it as a probabilistic trace alignment problem, trading off between the similarity of observations and plans, and the likelihood that the agent is performing those plans. We assess the effectiveness of our approach by conducting a comparative experimental evaluation on state-of-the-art benchmarks.

Assessing whether an agent has abandoned a goal or is actively pursuing it is important when multiple agents are trying to achieve joint goals, or when agents commit to achieving goals for each other. Making such a determination for a single goal by observing only plan traces is not trivial, as agents often deviate from optimal plans for various reasons, including the pursuit of multiple goals or the inability to act optimally. In this article, we develop an approach based on domain independent heuristics from automated planning, landmarks, and fact partitions to identify sub-optimal action steps—with respect to a plan—within a fully observable plan execution trace. Such capability is very important in domains where multiple agents cooperate and delegate tasks among themselves, such as through social commitments , and need to ensure that a delegating agent can infer whether or not another agent is actually progressing towards a delegated task. We demonstrate how a creditor can use our technique to determine—by observing a trace—whether a debtor is honouring a commitment. We empirically show, for a number of representative domains, that our approach infers sub-optimal action steps with very high accuracy and detects commitment abandonment in nearly all cases.