Papers by Christopher Geib
We present a method for incremental intention recognition by means of incrementally constructing ... more We present a method for incremental intention recognition by means of incrementally constructing a Bayesian Network (BN) model as more actions are observed. It is achieved based on a knowledge base of easily maintained and constructed fragments of BNs, connecting intentions to actions. The simple structure of the fragments enables to easily and efficiently acquire the knowledge base, either from domain experts or automatically from a plan corpus. We show experimental results improvement for the Linux Plan Corpus. In addition, we create a new, so-called IPD Plan Corpus, for strategies in the iterated Prisoner's Dilemma and show the experimental results for it.
Synthesis Lectures on Artificial Intelligence and Machine Learning, 2021
Plan recognition, activity recognition, and goal recognition all involve making inferences about ... more Plan recognition, activity recognition, and goal recognition all involve making inferences about other actors based on observations of their interactions with the environment and other agents. This s ynergistic a rea o f r esearch c ombines, u nites, a nd m akes u se o f t echniques a nd research from a wide range of areas including user modeling, machine vision, automated planning, intelligent user interfaces, human-computer interaction, autonomous and multi-agent systems, natural language understanding, and machine learning. It plays a crucial role in a wide variety of applications including assistive technology, software assistants, computer and network security, human-robot collaboration, natural language processing, video games, and many more. This wide range of applications and disciplines has produced a wealth of ideas, models, tools, and results in the recognition literature. However, it has also contributed to fragmentation in the field, with researchers publishing relevant results in a wide spectrum of journals and conferences. This book seeks to address this fragmentation by providing a high-level introduction and historical overview of the plan and goal recognition literature. It provides a description of the core elements that comprise these recognition problems and practical advice for modeling them. In particular, we define and distinguish the different recognition tasks. We formalize the major approaches to modeling these problems using a single motivating example. Finally, we describe a number of state-of-the-art systems and their extensions, future challenges, and some potential applications.
Implicit vs. Explicit Representation of Knowledge
Springer eBooks, 2021
This paper presents a framework for integrated plan recognition and automated planning, to produc... more This paper presents a framework for integrated plan recognition and automated planning, to produce cooperative behaviour for one agent to help another agent. By observing an "initiator" agent performing a task, the plan recogniser hypothesises how a "supporter" agent could help the initiator by proposing a set of subgoals to be achieved. A lightweight negotiation process mediates between the two agents to produce a mutually agreeable set of goals for the supporter. The goals are passed to a planner which builds an appropriate sequence of actions for satisfying these goals. The approach is demonstrated in a series of experimental scenarios.
Word sense disambiguation (WSD) has been a long-standing research objective for natural language ... more Word sense disambiguation (WSD) has been a long-standing research objective for natural language processing. In this paper we are concerned with developing graph-based unsupervised algorithms for alleviating the data requirements for large scale WSD. Under this framework, finding the right sense for a given word amounts to identifying the most "important" node among the set of graph nodes representing its senses. We propose a variety of measures that analyze the connectivity of graph structures, thereby identifying the most relevant word senses. We assess their performance on standard datasets, and show that the best measures perform comparably to state-of-the-art.
Future Directions
Springer eBooks, 2021
μCCG, a CCG-based Game-Playing Agent for μRTS
This paper presents a Combinatory Categorial Grammar-based game playing agent called μCCG for the... more This paper presents a Combinatory Categorial Grammar-based game playing agent called μCCG for the Real-Time Strategy testbed μRTS. The key problem that μCCG tries to address is that of adversarial planning in the very large search space of RTS games. In order to address this problem, we present a new hierarchical adversarial planning algorithm based on Combinatory Categorial Grammars (CCGs). The grammar used by our planner is automatically learned from sequences of actions taken from game replay data. We provide an empirical analysis of our agent against agents from the CIG 2017 μRTS competition using competition rules. μCCG represents the first complete agent to use a learned formal grammar representation of plans to adversarially plan in RTS games.
We describe the ZAROFF system, a plan-based controller for the players in a game of hide and seek... more We describe the ZAROFF system, a plan-based controller for the players in a game of hide and seek. The system features visually realistic human figure animation including realistic human locomotion. We discuss the planner's interaction with a changing environment to which it has only limited perceptual access. A hierarchical planner translates the game's goals of finding hiding players into locomotion goals, assisted by a special-purpose search planner. We describe a system of parallel finite state machines for controlling the player's locomotion. Neither path-planning nor explicit instructions are used to drive locomotion; agent control and apparent complexity are the result of the interaction of a few relatively simple behaviors with a complex (and changing) environment. Comments
The intentional planning system
This thesis presents the Intentional Planning System (ItPlanS) a hierarchical planner that is des... more This thesis presents the Intentional Planning System (ItPlanS) a hierarchical planner that is designed for domains in which plans are required as rapidly as possible, with limited knowledge. These constraints argue for an incremental approach to the planning that does not require the system to make commitments beyond the information that is currently available.
International Conference on Artificial Intelligence Planning Systems, Jun 13, 1994
This paper describes the Intentional Planning System(ItPlanS) an incremental, hierarchical planne... more This paper describes the Intentional Planning System(ItPlanS) an incremental, hierarchical planner that uses a series of experts to develop plans. This system takes seriously tile concept of the context sensitivity of actions while working in a resource bounded framework. GEm 55
This paper presents an architecture for agents that search for and manipulate objects. It is demo... more This paper presents an architecture for agents that search for and manipulate objects. It is demonstrated in the SodaJack system, a system that animates a human working at a soda fountain. The system is constructed as a set of three interacting planners. Two of these planners are special-purpose modules which contribute context-specific plans for the tasks of searching for and manipulating objects. The search planner is used to convert knowledge acquisition goals into goals of searching locations. An object specific reasoner is used to build object sensitive plans for manipulating specific objects. Finally, an incremental hierarchical planner is used to integrate these two special purpose planners into a complete system which interleaves planning and acting.
The high degree of autonomy that is being built into today's control systems, coupled with their ... more The high degree of autonomy that is being built into today's control systems, coupled with their vulnerability to attack at the computer network level, mandates a recasting of the man-machine social contract. We now have the technology and the capability to enable these control systems to entertain "a healthy skepticism" about the validity of the commands being given then and the identity of the issuers.
Partial Observability in Grammar Based Plan Recognition
Plan Recognition (Dagstuhl Seminar 11141)
Dagstuhl Reports, 2011
This Dagstuhl seminar brought together researchers with a wide range of interests and backgrounds... more This Dagstuhl seminar brought together researchers with a wide range of interests and backgrounds related to plan and activity recognition. It featured a substantial set of longer tutorials on aspects of plan and activity recognition, and related topics and useful methods, as a way of establishing a common vocabulary and shared basis of understanding. Building on this shared understanding, individual researchers presented talks about their work in the area. There were also panel discussions which addressed questions about how to best foster progress in the field --- specifically how to improve our ability to compare different plan and activity recognition algorithms --- and address the question of whether to assume rationality in the modeled agents (a question that is of great concern in many fields at this time). This report presents a summary of the talks and discussions at the seminar.
Artificial Intelligence, Jul 1, 2009
We present the PHATT algorithm for plan recognition. Unlike previous approaches to plan recogniti... more We present the PHATT algorithm for plan recognition. Unlike previous approaches to plan recognition, PHATT is based on a model of plan execution. We show that this clarifies several difficult issues in plan recognition including the execution of multiple interleaved root goals, partially ordered plans, and failing to observe actions. We present the PHATT algorithm's theoretical basis, and an implementation based on tree structures. We also investigate the algorithm's complexity, both analytically and empirically. Finally we present PHATT's integrated constraint reasoning for parametrized actions and temporal constraints.
Ai Magazine, Dec 1, 2010
Interactive entertainment has become a dominant force in the entertainment sector of the global e... more Interactive entertainment has become a dominant force in the entertainment sector of the global economy. In 2000, John Laird and Michael van Lent justified interactive entertainment as a domain of study in AI when they posited that computer games could act as test beds for achieving human-level intelligence in computers, leveraging the fidelity of their simulations of realworld dynamics. There is an additional perspective on AI for games: increasing the engagement and enjoyment of the player. This perspective is consistent with the perspective of computer game developers. For them, AI is a tool in the arsenal of the game to be used in lieu of real people when no one is available for a given role. Examples of such roles are opponents, companions, and nonplay characters (NPCs) in roles that are not fun to play such as shopkeepers, farmers, and victims; cinematographer; dungeon master; plot writer; or game designer. Reports
National Conference on Artificial Intelligence, 2010
Previous work on the YAPPR plan recognition system provided algorithms for translating convention... more Previous work on the YAPPR plan recognition system provided algorithms for translating conventional HTN plan libraries into lexicalized grammars and treated the problem of plan recognition as one of parsing. To produce these grammars required a fixed bound for any loops within the grammar and a presented a problem for optional actions within HTN plans. In this work we show that well known transformations from formal language theory can be used to rewrite the plan grammars to remove these limitations on the plan libraries. Related Work Graph Based Methods: Many pieces of previous work (Kautz 1991; Avrahami-Zilberbrand and Kaminka 2005), did not explicitly discuss the problems of recognizing plans with optional actions or loops. However systems that viewed
Plan, Activity, and Intent Recognition: Theory and Practice
Morgan Kaufmann eBooks, Mar 10, 2014
Plan recognition is the task of predicting an agent’s top-level plans based on its observed actio... more Plan recognition is the task of predicting an agent’s top-level plans based on its observed actions. It is an abductive-reasoning task that involves inferring plans that best explain observed actions. Most existing approaches to plan-recognition and other abductive-reasoning tasks either use first-order logic (or subsets of it) or probabilistic graphical models. While the former cannot handle uncertainty in the data, the latter cannot handle structured representations. To overcome these limitations, we explore the application of statistical–relational models that combine the strengths of both first-order logic and probabilistic graphical models to plan recognition. Specifically, we introduce two new approaches to abductive plan recognition using Bayesian logic programs (BLPs) and Markov Logic Networks (MLNs). Neither of these formalisms is suited for abductive reasoning because of the deductive nature of the underlying logical inference. In this chapter, we propose approaches to adapt both these formalisms for abductive plan recognition. We present an extensive evaluation of our approaches on three benchmark datasets on plan recognition, comparing them with existing state-of-the-art methods.
To be effective, current intrusion detection systems (IDSs) must incorporate artificial intellige... more To be effective, current intrusion detection systems (IDSs) must incorporate artificial intelligence methods for plan recognition. Plan recognition is critical both to predicting the future actions of attackers and planning appropriate responses to their actions. However network security places a new set of requirements on plan recognition. In this paper we present an argument for including plan recognition in IDSs and an algorithm for conducting plan recognition that meets the needs of the network security domain.
International Conference on Automated Planning and Scheduling, Sep 14, 2008
This document formalizes and discusses the implementation of a new, more efficient probabilistic ... more This document formalizes and discusses the implementation of a new, more efficient probabilistic plan recognition algorithm called Yet Another Probabilistic Plan Recognizer, (Yappr). Yappr is based on weighted model counting, building its models using string rewriting rather than tree adjunction or other tree building methods used in previous work. Since model construction is often the most computationally expensive part of such algorithms, this results in significant reductions in the algorithm's runtime. Problem Intuition All probabilistic plan recognition systems based on weighted model counting function in roughly the same way. These systems compute the exclusive and exhaustive set of models that can explain a given set of observations. They
Uploads
Papers by Christopher Geib