Artificial Intelligence Today

We discuss the application of behavioural architectures, in the robotic sense, to virtual agents. ‘Virtual Teletubbies’ are used as an example of the issues involved. we conclude that the use of such architectures has implications for the whole style in which a virtual world is modelled.

After a short analysis of the requirements that a knowledge representation language must satisfy, we introduce Description Logics, Modal Logics, and Nonmonotonic Logics as formalisms for representing terminological knowledge, time-dependent or subjective knowledge, and incomplete knowledge respectively. At the end of each section, we briefly comment on the connection to Logic Programming.

This article presents a taxonomy of strategies for fully-automated general-purpose first-order theorem proving. It covers forward-reasoning ordering-based strategies and backward-reasoning subgoal-reduction strategies, which do not appear together often. Unlike traditional presentations that emphasize logical inferences, this classification strives to give equal weight to the inference and search components of theorem proving, which are equally important in practice. For this purpose, a formal notion of search plan is given and shown to apply to all classes of strategies. For each class, the form of derivation is specified, and it is shown how inference system and search plan cooperate to generate it.

The recent advances in computer speed and algorithms for probabilistic inference have led to a resurgence of work on planning under uncertainty. The aim is to design AI planners for environments where there may be incomplete or faulty information, where actions may not always have the same results and where there may be tradeoffs between the different possible outcomes of a plan. Addressing uncertainty in AI planning algorithms will greatly increase the range of potential applications but there is plenty of work to be done before we see practical decision-theoretic planning systems. This article outlines some of the challenges that need to be overcome and surveys some of the recent work in the area.

Reasoning about stochastic dynamical systems and planning under uncertainty has come to play a fundamental role in AI research and applications. The representation of such systems, in particular, of actions with stochastic effects, has accordingly been given increasing attention in recent years. In this article, we survey a number of techniques for representing stochastic processes and actions with stochastic effects using dynamic Bayesian networks and influence diagrams, and briefly describe how these support effective inference for tasks such as monitoring, forecasting, explanation and decision making. We also compare these techniques to several action representations adopted in the classical reasoning about action and planning communities, describing how traditional problems such as the frame and ramification problems are dealt with in stochastic settings, and how these solutions compare to recent approaches to this problem in the classical (deterministic) literature. We argue that while stochastic dynamics introduce certain complications when it comes to such issues, for the most part, intuitions underlying classical models can be extended to the stochastic setting.

We survey research in the automation of deductive inference, from its beginnings in the early history of computing to the present day. We identify and describe the major areas of research interest and their applications. The area is characterised by its wide variety of proof methods, forms of automated deduction and applications.

The Word Wide Web was born as an Internet service supporting a simple distributed hypertext management system. Since its start a number of technologies have been proposed to enhance its capabilities. In this paper we describe our concept of an active Web, namely how we design the software architecture of interactive cooperative applications based on the Word Wide Web. An active Web includes agents able to use the services offered by Word Wide Web clients and servers. In an active Web both users and agents can interoperate using a set of basic mechanisms for communication and synchronization. The active Web we describe here is based on coordination technology: we explore two alternative implementations, both based on Java enriched with alternative coordination kernels.

Lifelike pedagogical agents have been the subject of increasing attention in the agents and knowledge-based learning environment communities [2, 17, 19‐21]. In parallel developments, recent years have witnessed great strides in work on cognitive models of emotion and affective reasoning [4,18, 22]. As a result, the time is now ripe for exploring how affective reasoning can be incorporated into pedagogical agents to improve students’ learning experiences.

Recently model checking representation and search techniques were shown to be efficiently applicable to planning, in particular to non-deterministic planning. Such planning approaches use Ordered Binary Decision Diagrams (OBDDs) to encode a planning domain as a non-deterministic finite automaton (NFA) and then apply fast algorithms from model checking to search for a solution. OBDDs can effectively scale and can provide universal plans for complex planning domains. We are particularly interested in addressing the complexities arising in non-deterministic, multi-agent domains. In this chapter, we present UMOP, a new universal OBDD-based planning framework for non-deterministic, multi-agent domains, which is also applicable to deterministic single-agent domains as a special case. We introduce a new planning domain description language, NADL, to specify non-deterministic multi-agent domains. The language contributes the explicit definition of controllable agents and uncontrollable environment agents. We describe the syntax and semantics of NADL and show how to build an efficient OBDD-based representation of an NADL description. The UMOP planning system uses NADL and different OBDD-based universal planning algorithms. It includes the previously developed strong and strong cyclic planning algorithms [9, 10]. In addition, we introduce our new optimistic planning algorithm, which relaxes optimality guarantees and generates plausible universal plans in some domains where no strong or strong cyclic solution exist. We present empirical results from domains ranging from deterministic and single-agent with no environment actions to non-deterministic and multi-agent with complex environment actions. Umop is shown to be a rich and efficient planning system.

Data integration is a problem at the intersection of the fields of Artificial Intelligence and Database Systems. The goal of a data integration system is to provide a uniform interface to a multitude of data sources, whether they are within one enterprise or on the World-Wide Web. The key challenges in data integration arise because the data sources being integrated have been designed independently for autonomous applications, and their contents are related in subtle ways. As a result, a data integration system requires rich formalisms for describing contents of data sources and relating between contents of different sources. This paper discusses works aimed at applying techniques from Artificial Intelligence to the problem of data integration. In addition to employing Knowledge Representation techniques for describing contents of information sources, projects have also made use of Machine Learning techniques for extracting data from sources and planning techniques for query optimization. The paper also outlines future opportunities for applying AI techniques in the context of data integration.

Believability is one of the key concerns when developing synthetic characters in intelligent virtual environments. To achieve believability, the virtual characters must behave consistently with their assumed or perceived personality.

This paper describes a model for the construction of believable emotional characters with synthetic personae in intelligent virtual environments. The model assumes that personality and emotions are essentially the same mechanism and allows the definition of the characters personality by a set of emotional reactions. The system designer defines a set of concepts, which structure is inspired by Ortony, Clore and Collins’s theory of emotions, that will then automatically integrated in the IVE character cycle, inspired by Fridja’s theory of emotions. Then, the system designer implements the final IVE specific characteristics around those concepts. The methodology was applied to a real time IVE, S3A, developed in the context of the last world exposition of the century, EXPO’98, which featured two pathematic dolphins, Tristão and Isolda, who lived in the synthetic estuary of the river Sado during the four months of the EXPO’98, and were displayed to more than one million visitors.

The project showed that the followed approach is a viable solution for the creation of the synthetic personae of believable emotional agents in intelligent virtual environments.

Believable agents are autonomous agents that exhibit rich personalities. Interactive dramas take place in virtual worlds inhabited by believable agents with whom an audience interacts. In the course of this interaction, the audience experiences a story. This paper presents the research philosophy behind the Oz Project, a research group at CMU that has spent the last ten years studying believable agents and interactive drama. The paper then surveys current work from an Oz perspective.

Intelligent mobile robots need the ability to integrate robust navigation facilities with higher level reasoning. This paper is an attempt at combining results and techniques from the areas of robot navigation and of intelligent agency. We propose to integrate an existing navigation system based on fuzzy logic with a deliberator based on the so-called BDI model. We discuss some of the subtleties involved in this integration, and illustrate it on a simulated example. Experiments on a real mobile robot are under way.

Integrated project management means that design and construction planning are interleaved with plan execution, allowing both the design and plan to be changed as necessary. This requires that the right effects of change need to be propagated through the plan and design. When this is distributed among designers and planners, no one may have all of the information to perform such propagation and it is important to identify what effects should be propagated to whom, and when. We describe a set of dependencies among plan and design elements that allow such notification by a set of message-passing software agents. The result is to provide a novel level of computer support for complex projects.

Inspired by legal reasoning, this paper presents an argument-based system for defeasible reasoning, with a logic-programming-like language, and based on Dung’s argumentation-theoretic approach to the semantics of logic programming. The language of the system has both weak and explicit negation, and conflicts between arguments are decided with the help of priorities on the rules. These priorities are not fixed, but are themselves defeasibly derived as conclusions within the system.

Autonomous robots need the ability to move purposefully and without human intervention in real-world environments that have not been specifically engineered for them. These environments are characterized by the pervasive presence of uncertainty: the need to cope with this uncertainty constitutes a major challenge for autonomous robots. In this note, we discuss this challenge, and present some specific solutions based on our experience on the use of fuzzy logic in mobile robots. We focus on three issues: how to realize robust motion control; how to flexibly execute navigation plans; and how to approximately estimate the robot’s location.

This article presents the event calculus, a logic-based formalism for representing actions and their effects. A circumscriptive solution to the frame problem is deployed which reduces to monotonic predicate completion. Using a number of benchmark examples from the literature, the formalism is shown to apply to a variety of domains, including those featuring actions with indirect effects, actions with non-deterministic effects, concurrent actions, and continuous change.

After reviewing a number of Internet tools and technologies originating in the field of logic programming and discussing promissing directions of ongoing research, we describe a logic programming based networking infrastructure which combines reasoning and knowledge processing with flexible coordination of dynamic state changes and computation mobility, as well as and its use for the design of intelligent mobile agent programs.

A lightweight logic programming language, Jinni, implemented in Java is introduced as a flexible scripting tool for gluing together knowledge processing components and Java objects in networked client/server applications and thin client environments as well as through applets over the Web.

Mobile threads, implemented by capturing first order continuations in a compact data structure sent over the network, allow Jinni to interoperate with remote high performance BinProlog servers for CPU-intensive knowledge processing.

A Controlled Natural Language to Prolog translator with support of third party speech recognition and text-to-speech translation allows interaction with users not familiar with logic programming.

This paper explains methods to provide autonomous virtual humans with the skills necessary to perform stand-alone role in films, games and interactive television. We present current research developments in the Virtual Life of autonomous synthetic actors. After a brief description of our geometric, physical, and auditory Virtual Environments, we introduce the perception action principles with a few simple examples. We emphasize the concept of virtual sensors for virtual humans. In particular, we describe our experiences in implementing virtual sensors such as vision sensors, tactile sensors, and hearing sensors. We then describe knowledge-based navigation, knowledge-based locomotion and in more details sensor-based tennis.