IJCAI-03 Invited Speakers

Keynote Speaker

Computer Vision: AI or Non-AI Problem
Takeo Kanade, Carnegie Mellon University, USA

Vision is one of the first areas that Artificial Intelligence tackled. Today, however, it appears as if the two fields, Computer Vision and Artificial Intelligence, have very little interaction, and the goal of developing a general vision system, such as understanding natural scenes, continues to be least understood or is almost abandoned. This talk will start with Dr. Kanade's historical perspectives on how this happened, and then present the argument that there is an opportunity to renew the tie between the two fields for the purpose of developing a capable AI-based vision system.

The Past, Present and Future of Web Information Retrieval
Mehran Sahami, Google, Inc. and Stanford University, USA

Web search engines have emerged as one of the central applications on the Internet. In fact, search has become one of the most important activities that people engage in on the Internet. Even beyond becoming the number one source of information, a growing number of businesses are depending on web search engines for customer acquisition. The first generation of web search engines used text-only retrieval techniques. Google revolutionized the field by deploying the PageRank technology - an eigenvector-based analysis of the hyperlink structure - to analyze the web in order to produce relevant results. Moving forward, Google's goal is to better allow users to obtain relevant information through the deployment of a variety of technological advances in information analysis, understanding, and retrieval. This presentation will describe some of the main challenges encountered in web information retrieval, some of the current techniques used, and will offer an overview of the search engine of the future.

Deploying Information Agents on the Web
Craig Knoblock, University of Southern California, USA

The information resources on the Web are vast, but much of the Web is based on a browsing paradigm that requires someone to actively seek information. Instead, one would like to have information agents that continuously attend to one's personal information needs. Such agents need to be able to extract the relevant information from web sources, integrate data across sites, and execute efficiently in a networked environment. In this talk, Dr. Knoblock will describe the technologies his group has developed to rapidly construct and deploy information agents on the web. This includes wrapper learning for turning online sources into agent-friendly resources, query planning and record linkage to integrate data across different sites, and streaming dataflow execution to efficiently execute agent plans. He will also describe how they applied this work within the Electric Elves project to deploy a set of agents for real-time monitoring of travel itineraries.

Web Intelligence (WI): A New Paradigm for Developing the Wisdom Web and Social Network Intelligence
Jiming Liu, Hong Kong Baptist University and Web Intelligence Consortium

Web Intelligence (WI), since it was coined in 2000, has become a new direction for scientific research and development that explores the fundamental roles as well as practical impacts of Artificial Intelligence (AI) and advanced Information Technology (IT) on the next generation of Web-empowered products, systems, services, and activities.

In this talk, Dr. Liu will show a coherent, global picture of what WI concerns and what are the top-priority research agendas. He will examine Web Intelligence (WI) as a new paradigm for developing the Wisdom Web and Web-supported social network intelligence.

While highlighting the current state of WI research and development, the talk will consist of the following five parts:
Part 1: Characterizes intelligent Web agents from various aspects.
Part 2: Considers Web mining and farming for Web intelligence as a business intelligence solution.
Part 3: Gives new directions in intelligent Web information retrieval.
Part 4: Focuses on Web knowledge management and infrastructure for Web intelligent systems.
Part 5: Concerned with Web-supported social networks intelligence.

Intelligent Systems in Travel and Tourism
Hannes Werthner, eCommerce and Tourism Research Lab (eCTRL)
ITC-irst and University of Trento, Italy

The talk will start by describing the travel and tourism domain, with its very specific features. This will also explain the importance of the sector in the e-commerce and e-business domain, where travel & tourism is the number one application domain (in b2c). Consequently, this industry depends heavily on advanced IT. The characteristics of this marketplace will also serve as the basis for presenting several application examples and running projects. Many of them, quite naturally, follow an AI based approach: travel planning and scheduling, visitor guidance systems, individual pricing or reversed auctions, product bundling and recommendation, or ontology based approaches to interoperability. In the final part the presentation will provide a future outlook, based on the European Research Vision of Ambient Intelligence and its applicability in the travel and tourism domain.

Optimality of Collective Choice in Social Insects and Social Robots
Jean-Louis Deneubourg, University Libre du Bruxelles, Belgium

Data Integration: Successes and Challenges
Alon Halevy, University of Washington, USA

Integration of data from multiple sources is one of the longest standing problems facing the AI and Database research communities. In addition to being a problem in large enterprises, research on this topic has been fueled by the promise of integrating data on the WWW. In the past few years, we have made very significant progress on data integration, from the conceptual and algorithmic aspects, to the systems and product aspects. This talk will briefly review our successes in data integration, and will describe some significant current challenges. In particular, I will discuss the problem of trying to semi-automatically find a semantic mapping between a pair of schemas/ontologies, and argue that AI techniques are crucial in this context. I describe an approach to schema matching that is based on analyzing a large corpus of database schemas and learning properties of how terms are used in database structures. Finally, this talk will argue that such a corpus offers other exciting opportunities for AI research.

Constraint Satisfaction, Databases, and Logic
Phokion G. Kolaitis, University of California, Santa Cruz, USA

Constraint satisfaction problems constitute a broad class of algorithmic problems that are ubiquitous in several different areas of artificial intelligence and computer science. Indeed, constraint satisfaction problems encompass Boolean satisfiability, graph colorability, and numerous other problems in database theory, temporal reasoning, machine vision, and belief maintenance.

In their full generality, constraint satisfaction problems are NP-complete and, thus, presumed to be algorithmically intractable. For this reason, extensive research efforts have been devoted to the pursuit of "islands of tractability" of constraint satisfaction, that is, special cases of constraint satisfaction problems for which efficient algorithms exist.

The aim of this talk is to present an overview of recent advances in the investigation of the computational complexity of constraint satisfaction with emphasis on the connections of this area of research with database theory and logic.

Self-reconfiguring Robots: Challenges and Successes
Daniela Rus, Dartmouth University, USA

We wish to create versatile robots by using self-reconfiguration: hundreds of small modules autonomously organize and reorganize as geometric structures to best fit the terrain on which the robot has to move, the shape of the object the robot has to manipulate, or the sensing needs for the given task. Self-reconfiguration allows large collections of small robots to actively organize as the most optimal geometric structure to perform useful coordinated work. A self-reconfiguring robot consists of a set of identical modules that can dynamically and autonomously reconfigure in a variety of shapes, to best fit the terrain, environment, and task. Self-reconfiguration leads to versatile robots that can support multiple modalities of locomotion and manipulation. Self-reconfiguring robots constitute large scale distributed systems. Because the modules change their location continuously they also constitute ad-hoc networks. This talk will discuss the challenges and successes of creating self-reconfiguring robots, ranging from designing hardware capable of self-reconfiguration to developing distributed controllers and planners for such systems that are scalable, adaptive, and support real-time behavior.

Automated Verification = Graphs, Automata and Logic
Moshe Vardi, Rice University, USA

In automated verification one uses algorithmic techniques to establish the correctness of the design with respect to a given property. Automated verification is based on a small number of key algorithmic ideas, tying together graph theory, automata theory, and logic. In this self-contained talk, Dr. Vardi will describe how this "holy trinity" gave rise to automated-verification tools.

New Trends in Automated Reasoning
Andrei Voronkov, Manchester University, UK

This talk presents an overview of the past and present of automated reasoning in first-order logic and tries to predict how it will develop in the next several years. Dr. Voronkov considers the theory of automated reasoning, implementation and currently available systems, and applications. The presentation will be centered around two main motives: (1) efficiency and (2) usefulness for the existing and possible future applications.

User Interfaces: An AI Challenge
Daniel S. Weld, University of Washington, USA

Today's computer interfaces are one size fits all. Users with little programming experience have very limited opportunities to customize an interface to their task and work habits. Furthermore, the overhead induced by generic interfaces will be proportionately greater on small form-factor PDAs, embedded applications and wearable devices. Searching for a solution, researchers argue that productivity can be greatly enhanced if interfaces anticipated their users, adapted to their preferences, and reacted to high-level customization requests. But realizing these benefits is tricky, because there is an inherent tension between the dynamism implied by automatic interface adaptation and the stability required in order for the user to predict the computer's behavior and maintain control. This talk will list challenges for the field, describe principles governing effective adaptation, and present new algorithms for data mining user action traces and dynamically transforming interfaces.

Quantum Information: Fundamentals and Applications
Anton Zeilinger, Vienna University, Austria

The fundamental concepts in the emerging field of quantum information are quantum superposition, entanglement and the objective randomness of the individual event. Based on these concepts the possible experimental realizations and possibilities are discussed. These include quantum cryptography, teleportation, quantum communication and quantum gates for future quantum computers. All material will be presented in a way suitable for an audience not familiar with quantum physics.