Plan Recognition System Research Articles

Symbolic Plan Recognition in Interactive Narrative Environments

Interactive narratives suffer from the narrative paradox: the tension that exists between providing a coherent narrative experience and allowing a player free reign over what she can manipulate in the environment. Knowing what actions a player in such an environment intends to carry out would help in managing the narrative paradox, since it would allow us to anticipate potential threats to the intended narrative experience and potentially mediate or eliminate them. The process of observing player actions and attempting to come up with an explanation for those actions (i.e. the plan that the player is trying to carry out) is the problem of plan recognition. We adopt the framing of narratives as plans and leverage recent advances that cast plan recognition as planning to develop a symbolic plan recognition system as a proof-of-concept model of a player's reasoning in an interactive narrative environment. In this paper we outline the system architecture, report on performance metrics that demonstrate adequate performance for non-trivial domains, and discuss the implications of treating players as plan recognizers.

INTENT RECOGNITION IN MULTI-AGENT SYSTEMS: COW HERDING

In a multi-agent system, an agent may utilize its idle time to assist other agents in the system. Intent recognition is proposed to accomplish this with minimal communication. An agent performing recognition observes the tasks other agents are performing and, unlike the much studied field of plan recognition, the overall intent of an agent is recognized instead of a specific plan. The observing agent may use capabilities that it has not observed. A conceptual framework is proposed for intent recognition systems. An implementation of the conceptual framework is tested and evaluated. We hypothesize that using intent recognition in a multi-agent system increases utility (where utility is domain specific) and decreases the amount of communication. We test our hypotheses using the domain of Cow Herding, where agents attempt to herd cow agents into team corrals. A set of metrics, including task time and number of communications, is used to compare the performance of plan recognition and intent recognition. In our results, we find that intent recognition agents communicate fewer times than plan recognition agents. In addition, unlike plan recognition, when agents use the novel approach of intent recognition, they select unobserved actions to perform. Intent recognition agents were also able to outperform plan recognition agents by consistently scoring more points in the Cow Herding domain. This research shows that under certain conditions, an intent recognition system is more efficient than a plan recognition system. The advantage of intent recognition over plan recognition becomes more apparent in complex domains.

Plan and Intent Recognition in a Multi-agent System for Collective Box Pushing

AbstractIn a distributed multi-agent system, an idle agent may be available to assist other agents in the system. An agent architecture called intent recognition is proposed in this article to accomplish this with minimal communication. To assist other agents in the system, an agent performing recognition observes the tasks other agents are performing. Unlike the much-studied field of plan recognition, the overall intent of an agent is recognized instead of a specific plan. The observing agent may use capabilities that it has not observed. In this study, the key research question is: What are intent-recognition systems and how can these be used to have agents autonomously assist each other effectively and efficiently? A conceptual framework is proposed to address this question. An implementation of the conceptual framework is tested and evaluated. A set of metrics, including task time and number of communications, is used to compare the performance of plan recognition and intent recognition. This research shows that under certain conditions, an intent-recognition system is more efficient than a plan recognition system.

Recognizing Plans with Loops Represented in a Lexicalized Grammar

This paper extends existing plan recognition research to handle plans containing loops. We supply an encoding of plans with loops for recognition, based on techniques used to parse lexicalized grammars, and demonstrate its effectiveness empirically. To do this, the paper first shows how encoding plan libraries as context free grammars permits the application of standard rewriting techniques to remove left recursion and ε-productions, thereby enabling polynomial time parsing. However, these techniques alone fail to provide efficient algorithms for plan recognition. We show how the loop-handling methods from formal grammars can be extended to the more general plan recognition problem and provide a method for encoding loops in an existing plan recognition system that scales linearly in the number of loop iterations.

Proactive Mediation in Plan-Based Narrative Environments

In interactive plan-based narrative environments, users' actions must be monitored to ensure that conditions necessary for the execution of narrative plans are not compromised. In the Zocalo system, management of user actions has been performed on a reactionary basis by a process called mediation. In this paper, we describe an extension to this approach, proactive mediation, which calculates responses to user input in an anticipatory manner. A proactive mediation module accepts as input a plan describing the actions being performed by the user (generated by a plan recognition system) and identifies portions of that plan that jeopardize the causal structure of the overall narrative. Once these portions are identified, proactive mediation generates modifications to the narrative plan structure that avoid the unwanted interaction between user and story. This extension to the original mediation algorithm provides more responses to a user's actions and generates responses that are tailored to the user's activity.

Model and algorithmic framework for detection and correction of cognitive errors

This paper outlines an approach that we are taking for elder-care applications in the smart home, involving cognitive errors and their compensation. Our approach involves high level modeling of daily activities of the elderly by breaking down these activities into smaller units, which can then be automatically recognized at a low level by collections of sensors placed in the homes of the elderly. This separation allows us to employ plan recognition algorithms and systems at a high level, while developing stand-alone activity recognition algorithms and systems at a low level. It also allows the mixing and matching of multi-modality sensors of various kinds that go to support the same high level requirement. Currently our plan recognition algorithms are still at a conceptual stage, whereas a number of low level activity recognition algorithms and systems have been developed. Herein we present our model for plan recognition, providing a brief survey of the background literature. We also present some concrete results that we have achieved for activity recognition, emphasizing how these results are incorporated into the overall plan recognition system.

Controlling inference in plan recognition

An algorithm based on an assessment of the completeness of an explanation can be used to control inference in a plan recognition system: If the explanation is complete, inference is stopped. If the explanation is incomplete, inference is continued. If it cannot be determined whether the explanation is complete, then the system weighs the strength of its interest in continuing the analysis against the estimated cost of doing so. This algorithm places existing heuristic approaches to the control of inference in plan recognition into a unified framework. The algorithm rests on the principle that the decision to continue processing should be based primarily on the explanation chain itself, not on external factors. Only when an analysis of the explanation chain proves inconclusive should outside factors weigh heavily in the decision. Furthermore, a decision to discontinue chaining should never be final; other components of the system should have the opportunity to request that an explanation chain be extended. An implementation of the algorithm, called PAGAN, demonstrates the usefulness of this approach.

A meta-rule approach to flexible plan recognition in dialogue

Although a number of researchers have demonstrated that reasoning on a model of the user's plans and goals is helpful in language understanding and response generation, current models of plan inference cannot handle naturally occurring dialogue. This paper argues that model building from less than ideal dialogues has a great deal in common with processing ill-formed input. It defines well-formedness constraints for information-seeking dialogues and contends that strategies for interpreting ill-formed input can be applied to the problem of modeling the user's plan during an ill-formed dialogue. It presents a meta-rule approach for hypothesizing the cause of dialogue ill-formedness, and describes meta-rules for relaxing the plan inference process and enabling the consideration of alternative hypotheses. The advantages of this approach are that it provides a unified framework for handling both well-formed and ill-formed dialogue, avoids unnatural interpretations when the dialogue is proceeding smoothly, and facilitates a nonmonotonic plan recognition system.

Classifying and detecting plan-based misconceptions for robust plan recognition

In order to maintain a truly cooperative dialogue, an intelligent natural language interface must be able to recognize misconceptions in the user's plans. This involves solving two problems: determining what sorts of mistakes people make when they reason about plans, and figuring out how to recognize these mistakes when they occur. There are sixteen distinct ways that a plan can be improperly constructed. Each of these corresponds to one type of plan-based misconception. But it is not necessary to consider all sixteen classes; the classification can be simplified to ten categories of detectable, distinguishable plan-based misconception. This classification can be used to develop a robust plan recognition algorithm, which is capable of recognizing both correct plans and plans that contain misconceptions. Probabilistic methods can be used to handle the three main difficulties of robust plan recognition: dealing with the hard classes of misconceptions, controlling the combinatorial explosion that results from the ambiguity inherent in faulty plans, and selecting the most likely plan among multiple competing explanations. The theory that is developed here is implemented in a program called Pathfinder, which is a probability-based plan recognition system based on the A$\sp\*$ best-first search algorithm that has been extended to recognize novel plan failures in an intelligent interface environment.