In (automated) negotiation systems for self-interested agents, contracts have traditionally been binding. They do not accommodate future events. Contingency contracts address this but are often impractical. As an alternative, we propose leveledcommitment contracts. The level of commitment is set by decommitting penalties. To be freed from the contract, an agent simply pays its penalty to the other contract party(ies). A self-interested agent will be ruluctant to decommit because some other contract party might decommit, in which case the former agent gets freed from the contract, does not incur a penalty, and collects a penalty from the other party. We show that despite such strategic decommitting, leveled commitment increases the expected payoffs of all contract parties and can enable deals that are impossible under full commitment. Different decommitting mechanisms are introduced and compared. Practical prescriptions for market designers are presented. A contract optimizer, ECOMMITTER, is provided on the web.

SOAR is one of the oldest and largest AI development efforts, starting formally in 1983. It has also been proposed as a unified theory of cognition (Newell 1990). Most of its current development is as an AI programming language, which was evident at the Twenty-Second SOAR Workshop held at Soar Technology near the University of Michigan in Ann Arbor on 1-2 June 2002.

Many bidders in eBay use bidding strategies that involve late bids, incremental bids, or both. Based on field evidence, we discuss the manner in which late bids are caused both by sophisticated, strategic reasoning and by irrationality and inexperience; the interaction of late bidding with incremental bidding; and the relation between market design and artificial agent design.

In this article, we first survey the three major types of computer music systems based on AI techniques: (1) compositional, (2) improvisational, and (3) performance systems. Representative examples of each type are briefly described. Then, we look in more detail at the problem of endowing the resulting performances with the expressiveness that characterizes human-generated music. This is one of the most challenging aspects of computer music that has been addressed just recently. The main problem in modeling expressiveness is to grasp the performer's "touch," that is, the knowledge applied when performing a score. Humans acquire it through a long process of observation and imitation. For this reason, previous approaches, based on following musical rules trying to capture interpretation knowledge, had serious limitations. An alternative approach, much closer to the observation-imitation process observed in humans, is that of directly using the interpretation knowledge implicit in examples extracted from recordings of human performers instead of trying to make explicit such knowledge. In the last part of the article, we report on a performance system, SAXEX, based on this alternative approach, that is capable of generating high-quality expressive solo performances of jazz ballads based on examples of human performers within a case-based reasoning (CBR) system.

Not widely known by the AI community, elevator control has become a major field of application for AI technologies. Techniques such as neural networks, genetic algorithms, fuzzy rules and, recently, multiagent systems and AI planning have been adopted by leading elevator companies not only to improve the transportation capacity of conventional elevator systems but also to revolutionize the way in which elevators interact with and serve passengers. In this article, we begin with an overview of AI techniques adopted by this industry and explain the motivations behind the continuous interest in AI. We review and summarize publications that are not easily accessible from the common AI sources. In the second part, we present in more detail a recent development project to apply AI planning and multiagent systems to elevator control problems.

Kernel methods, a new generation of learning algorithms, utilize techniques from optimization, statistics, and functional analysis to achieve maximal generality, flexibility, and performance. These algorithms are different from earlier techniques used in machine learning in many respects: For example, they are explicitly based on a theoretical model of learning rather than on loose analogies with natural learning systems or other heuristics. They come with theoretical guarantees about their performance and have a modular design that makes it possible to separately implement and analyze their components. They are not affected by the problem of local minima because their training amounts to convex optimization. In the last decade, a sizable community of theoreticians and practitioners has formed around these methods, and a number of practical applications have been realized. Although the research is not concluded, already now kernel methods are considered the state of the art in several machine learning tasks. Their ease of use, theoretical appeal, and remarkable performance have made them the system of choice for many learning problems. Successful applications range from text categorization to handwriting recognition to classification of geneexpression data.

This article is a reflection on agent-based AI. My contention is that AI research should focus on interactive, autonomous systems, that is, agents. Emergent technologies demand so. We see how recent developments in (multi-) agent-oriented research have taken us closer to the original AI goal, namely, to build intelligent systems of general competence. Agents are not the panacea though. I point out several areas such as design description, implementation, reusability, and security that must be developed before agents are universally accepted as the AI of the future.

We propose a perspective on knowledge compilation which calls for analyzing different compilation approaches according to two key dimensions: the succinctness of the target compilation language, and the class of queries and transformations that the language supports in polytime. We then provide a knowledge compilation map, which analyzes a large number of existing target compilation languages according to their succinctness and their polytime transformations and queries. We argue that such analysis is necessary for placing new compilation approaches within the context of existing ones. We also go beyond classical, flat target compilation languages based on CNF and DNF, and consider a richer, nested class based on directed acyclic graphs (such as OBDDs), which we show to include a relatively large number of target compilation languages.

We present a propositional logic to reason about the uncertainty of events, where the uncertainty is modeled by a set of probability measures assigning an interval of probability to each event. We give a sound and complete axiomatization for the logic, and show that the satisfiability problem is NP-complete, no harder than satisfiability for propositional logic.

Common wisdom has it that small distinctions in the probabilities (parameters) quantifying a belief network do not matter much for the results of probabilistic queries. Yet, one can develop realistic scenarios under which small variations in network parameters can lead to significant changes in computed queries. A pending theoretical question is then to analytically characterize parameter changes that do or do not matter. In this paper, we study the sensitivity of probabilistic queries to changes in network parameters and prove some tight bounds on the impact that such parameters can have on queries. Our analytic results pinpoint some interesting situations under which parameter changes do or do not matter. These results are important for knowledge engineers as they help them identify influential network parameters. They also help explain some of the previous experimental results and observations with regards to network robustness against parameter changes.