Monday, 21st of July, 2008
morning:
Knowledge Compilation: A Sightseeing Tour
Pierre Marquis,
CRIL Universitι Lille-Nord de France, Artois, France
Abstract: Pioneered two decades ago, knowledge compilation has been for a few years acknowledged as an important research topic in AI. Knowledge compilation is concerned with the pre-processing of some pieces of information in order to improve the computational efficiency of some tasks. In AI such tasks typically amount to inference or decision making.
Knowledge compilation gathers a number of research lines focusing on different problems, ranging from theoretical ones (where the key issue is the compilability one, i.e., determining whether computational improvements can be guaranteed via pre-processing) to more practical ones (mainly the design of compilation algorithms for some specific tasks, like clausal entailment). The (multi-criteria) choice of a target language for knowledge compilation is another major issue of knowledge compilation.
In this tutorial, I will review a number of the most common results of the literature on knowledge compilation in the propositional case. The tour will include some attractions, especially algorithms for improving clausal entailment and other forms of inference; a visit of the compilability district and a promenade following the knowledge compilation map are also planned.
.
afternoon:
Temporal and Numeric Planning
Maria Fox
Department of Computer and Information Sciences, University of Strathclyde, Glasgow, UK
Abstract: …
Tuesday, 22nd of July, 2008
morning:
Reasoning about Cooperative Games
Michael J. Woolridge
University of Liverpool, UK
Abstract: Cooperation is, of course, an issue at the very heart of multi-agent systems research. When trying to understand how and why cooperation occurs, multi-agent systems researchers have looked to models from game theory, and in particular, cooperative games have proved to be an enormously useful and influential framework both for building multi-agent systems and understanding their underlying principles. However, from a computational perspective, cooperative games present a number of challenges, chief among them being how they can be succinctly represented and how to reason efficiently with such representations. In this tutorial, I survey work on the representation of cooperative games in multi-agent systems. I assume a basic knowledge of AI principles (eg rule-based knowledge representation, very basic logic), but no knowledge of game theory or cooperative games. I introduce the basic models used in cooperative game theory, and the relevant solution concepts. I then describe the key computational issues surrounding such models, and survey the main approaches developed over the past decade for representing and reasoning about cooperative games in AI and computer science generally.
afternoon:
Dynamic Trust, Reputation and Recommender Systems for Wed-based Social Networks
John S. Baras
Institute for Systems Research, and Departments of Electrical and Computer Engineering and Computer Science, University of Maryland College Park.
Abstract: We describe models for web-based social networks and the associated concepts of trust and reputation. We focus on dynamic distributed models for the same and emphasize the significance of time variations in connectivity, user population and user behavior. We develop dynamic systems representations for reputation systems and the associated recommender systems. The underlying quantitative methods include dynamical systems on graphs, collaborative games, constrained coalitional games, auctions and mechanisms, on-line distributed monitoring and filtering, iterative learning in both dynamic graphs and in games. We describe methods for evaluating the integrity of recommender systems, the effects of topology, mechanisms for strengthening the robustness and resiliency of such systems against colluding agents. We close with a description of promising future directions for both research and for developing practical implementable such reputation and recommender systems.
|
