AAAI-12: Tutorial Forum

AAAI-12 Tutorial Forum

The Tutorial Forum of the Twenty-Sixth AAAI Conference on Artificial Intelligence (AAAI-12) will be held July 22-23 in Toronto. The Tutorial Forum provides an opportunity for junior and senior researchers to spend two days each year freely exploring exciting advances in disciplines outside their normal focus. We believe this type of forum is essential for the cross fertilization, cohesiveness, and vitality of the AI field. We all have a lot to learn from each other; the Tutorial Forum promotes the continuing education of each member of the AAAI.

To encourage full participation by technical conference registrants, no separate fee will be charged for admittance to the Tutorial Forum in 2012.

Sunday, July 22, 9:00 AM – 1:00 PM

• SA1: Agentpalooza: Rapid Creation and Deployment of Embodied Conversational Agents
  Andrew M. Olney, Patrick Hays, and Whitney L. Cade
• SA2: Trading Agents
  Michael Wellman and Amy Greenwald
• SA3: Traffic Management and AI
  Biplav Srivastava and Anand Ranganathan

Sunday, July 22, 2:00 PM – 6:00 PM

• SP1: Collective Intelligence
  Haym Hirsh
• SP2: Probabilistic Matrix/tensor Block Models for Two-way/Multi-way Network Modeling
  Zenglin Xu and Alan Qi
• SP3: Probabilistic Planning with Markov Decision Processes
  Andrey Kolobov and Mausam

Monday, July 23, 9:00 AM – 1:00 PM

• MA1: A Game Theoretic Approach to Social Networks
  Ramasuri Narayanam and Amit A. Nanavati
• MA2: Heuristic Search: The Basics and Beyond
  Wheeler Ruml and Jordan Thayer
• MA3: Search-Based Planning: Toward High Dimensionality and Differential Constraints
  Mihail Pivtoraiko, Maxim Likhachev, and Sven Koenig

Monday, July 23, 2:00 PM – 6:00 PM

• MP1: Entity Resolution: Theory, Practice, and Open Challenges
  Lise Getoor and Ashwin Machanavajjhala
• MP2: Text Mining from User Generated Content
  Ronen Feldman and Lyle Ungar
• MP3: Theory and Practice of Answer Set Programming
  Esra Erdem, Joohyung Lee, and Yuliya Lierler

SA1: Agentpalooza: Rapid Creation and Deployment of Embodied Conversational Agents

Andrew M. Olney, Patrick Hays, and Whitney L. Cade

Embodied conversational agents are virtual characters that engage users in conversation with appropriate speech, gesture, and facial expression. The high cost of developing embodied conversational agents has led to a recent increase in open-source agent platforms. In this tutorial, we present a framework for the rapid creation and deployment of embodied conversational agents. The tutorial will introduce novices to the basic concepts involved in creating ECAs. The tutorial will further describe a content creation pipeline that allows users to use a mixture of low cost and free tools to create customized virtual humans, animate them using motion capture, and deploy them as intelligent interfaces via the XNA Framework. The material covered in the tutorial is suitable both as a starting point for the development of a more advanced system or as a teaching tool for AI curricula.

The tutorial will include (1) an introduction to embodied conversational agents (2) an overview of common 3D art and animation technologies; (3) a simple content creation pipeline for ECAs; (4)a simple motion capture pipeline for animating ECAs; and (5) integrating ECAs into intelligent interfaces.

The tutorial requires no prerequisite knowledge for ECA creation. However, for integration with an intelligent interface, some familiarity with Visual Studio will be useful.

Andrew Olney is presently an assistant professor in the Department of Psychology at the University of Memphis and the associate director of the Institute for Intelligent Systems. His primary research interests are in natural language interfaces. Specific interests include vector space models, dialogue systems, grammar induction, robotics, and intelligent tutoring systems.

Patrick Hays is a research assistant at the University of Memphis. He is a recent graduate with a BA in Psychology. Patrick's work focuses on 3D animation, 3D modeling, graphic arts, and human-computer interaction.

 

Whitney Cade is a graduate student in the Department of Psychology at the University of Memphis. Her research interests include intelligent tutoring systems, expert tutors, pedagogical strategies, 3D agents, and machine learning.

SA2: Trading Agents

Michael Wellman and Amy Greenwald

Bidding in electronic markets is pervasive throughout the modern economy, from finance (equity, currency, and commodity markets), to electronic commerce (Amazon, eBay, and others), to supply chains (B2B marketplaces), to Internet advertising (search and display ad exchanges). The tutorial frames and motivates the problem of developing automated trading strategies for such markets. We use illustrative examples from several application domains to motivate the problems, which we then address at the more abstract level of generic auction environments, which can be seen as canonical trading games. We survey the state-of-the-art in analyzing strategies for basic market games, and introduce complex market scenarios beyond current theoretical understanding. To deal with such rich applications, we present a general methodology (empirical and game-theoretic) for trading agent design and analysis.

The tutorial is designed for those interested in trading domains as an application area for AI methods, for either research or practice. It will be broadly accessible to an AI audience. Exposure to basic concepts of game theory (such as Nash equilibrium) and auction theory (such as revenue equivalence) is helpful, but not assumed.

Please note that although the tutorial is relevant to algorithmic trading in financial markets, it does not cover techniques for predicting price movements in financial markets.

Michael P. Wellman is a professor of computer science and engineering at the University of Michigan. His research focuses on computational markets and strategic reasoning, with applications to electronic commerce. Wellman served as chair of ACM SIGecom and as executive editor of JAIR. He is a Fellow of AAAI and ACM.

 

Amy Greenwald is an associate professor of computer science at Brown University. She studies game-theoretic and economic interactions among computational agents, with applications in domains like dynamic pricing and autonomous bidding. She was awarded a Sloan Fellowship, a Fulbright Scholarship, and a PECASE.

SA3: Traffic Management and AI

Biplav Srivastava and Anand Ranganathan

The aim of the tutorial is to make early and experienced researchers aware of the traffic management area, provide an insightful overview of the current efforts in Research and in practice (real world pilots) that include AI techniques, and whet interest for newer efforts on important open issues.

The intended audience is a beginner to medium experienced researcher who is curious about AI in general and problems related to transportation and traffic management in particular. The tutorial will expect the audience to have a basic computer science background consisting of data structures, algorithms, databases and search. Traffic management is a pressing problem for cities around the world. Moreover, it is a highly visible perspective of a city's life affecting all aspects of its citizens' economic and personal activities. Consequently, there is significant academic and commercial interest in addressing this problem. Although a long preserve of urban planners and civil engineers, traffic management had traditionally attracted limited attention in AI with major efforts in multiagent system, simulation and optimization. As the transportation industry adopts the intelligent transportation system vision, computer scientists in general, and AI researchers in particular, are looking at traffic management issues with renewed vigor employing more AI techniques. Furthermore, although there are existing techniques that are applicable to fairly structured traffic in developed countries (like USA, Europe, and Japan); there is paucity of methods that apply to chaotic traffic in developing countries (like India, Vietnam).

The topics we will cover include (1) the traffic management problem, which is often assumed but not well formalized. The proposers have done some latest work on this and presented at major traffic forum.

We will then cover (2) sensing traffic data from the street. This topic will include important technologies for collecting raw data and converting them to traffic data (speed, volume) including (a) inductive-loop detectors as data source; (b) GPS-based vehicles as data source (c) Noisy methods for chaotic traffic (including call data record with telcos as data source and people-as-sensors); and (d) privacy issues in data collection.

We then cover (3) traffic state estimation (approaches, traffic flow models, fundamental diagrams); (4) spatiotemporal analysis for traffic and user mobility patterns; (5) simulators for traffic analysis; (6) end-user analytics (for city traffic prediction; for city congestion estimation; for citizens route recommendation; for citizens multimode travel recommendation using public transportation); (7) traffic standards; and (8) practical considerations for pilot.

The AI topics we will cover are (1) learning; (2) filtering and state estimation; (3) graph and path analytics; (4) planning; (5) scheduling; (6) ontology; and (7) data mining.

Biplav Srivastava (Biplav) is a senior researcher at IBM Research — India, based out of New Delhi. He is an expert in AI Planning, Services and the emerging Sustainability area with focus on Traffic. He has interests in planning, scheduling, policies, learning and information representation/ ontology. He has applied these techniques to applications in cyberphysical systems (Smarter Cities projects), semantic web, web services, services science, autonomic computing and bioinformatics. He is currently involved in two main efforts: (1) build semantic models to streamline interagency collaboration in cities, and (2) novel sensing techniques for traffic, water and public safety.

Srivastava has more than 75 research papers in refereed conferences and journals, 14 US issued patents and more than 30 other US applications. He is an IBM Master Inventor and represents IBM at the W3C work group on Government Linked Data. He received Ph.D. in 2000 and M.S. in 1996 from Arizona State University, USA and B.Tech. in 1993 from IT-BHU, India, all in computer science.

Anand Ranganathan (Anand) is a research staff member at the IBM TJ Watson Research Center. He is part of the IBM Infosphere Streams research team and has been involved in applications of stream processing technologies in a variety of Smarter Planet projects, including transportation and energy management. His transportation projects have been piloted in cities like Dublin and Stockholm, and he has experience working on traffic related issues in a number of other cities. A key theme of his research has been in exploring the knowledge engineering, software engineering and user interaction challenges in developing modular, reusable, component-based applications. He completed his PhD at the Department of Computer Science in the University of Illinois at Urbana-Champaign in 2005. He received his BTech in computer science and Engineering from the Indian Institute of Technology in Chennai in 2000. His broad research interests include data management, web 2.0, ubiquitous computing, distributed systems, services computing, the semantic web, artificial intelligence and software engineering. He is an IBM Master Inventor.

SP1: Collective Intelligence

Haym Hirsh

"Collective intelligence" refers to ways that information and communications technologies are bringing people and computing together to achieve outcomes that were previously beyond our individual capabilities or expectations. Collective intelligence makes contact with AI in three ways. First, AI scholars and practitioners are using collective intelligence as an element in their work, such as to create corpora in computational linguistics or computer vision or to evaluate results in user interfaces or information retrieval. Second, AI knowledge, such as in machine learning and computational linguistics, has become a key enabler of many examples of collective intelligence, such as mining consumer behaviors and product review sentiments to facilitate product recommendation. Third, collective intelligence offers a provocative phenomenon to consider by those seeking computationally tractable understandings of intelligence. This tutorial will survey the state of the art in collective intelligence from an AI perspective. First, it will discuss examples of collective intelligence in which people explicitly act and participate in ways to achieve desired outcomes, such as editing Wikipedia articles, submitting or rating reviews at Amazon.com, or identifying astronomical objects in GalaxyZoo. Second, it will present examples in which collectively intelligent outcomes arise through the computationally distilled wisdom of the behaviors and products of individuals rather than through their explicit participation, as exhibited by Google's page ranking algorithm and Amazon's recommendation system. The tutorial will conclude with a discussion of prospects for the future, including ways that multiple forms of collective intelligence are being successfully integrated into larger systems, as well as examples where collective intelligence is being used in numerous undesirable ways.

Haym Hirsh is a professor of computer science at Rutgers University. His research has focused on foundations and applications of machine learning, data mining, information retrieval, and artificial intelligence, especially targeting questions that integrally involve both people and computing. Most recently these interests have turned to crowdsourcing, human computation, and collective intelligence. From 2006-2010 he served as Director of the Division of Information and Intelligent Systems at the National Science Foundation, and from 2010-2011 he was a Visiting Scholar at MIT's Center for Collective Intelligence at the Sloan School of Management. Haym received his BS from the Mathematics and Computer Science Departments at UCLA and his MS and PhD from the Computer Science Department at Stanford University.

SP2: Probabilistic Matrix/tensor Block Models for Two-way/Multi-way Network Modeling

Zenglin Xu and Alan Qi

Network analysis has been becoming an important research area in a number of disciplines, such as bioinformatics, system biology, economics, sensor networks, and social sciences. In this tutorial, we present an overview on principled Bayesian network modeling techniques with a focus on stochastic block models. We will cover basic theories of matrix- and tensor- variate distributions and processes and describe how to apply these theories to model network interactions, such as user-movie-time relations. Unlike classical bilinear or multilinear matrix/tensor factorization methods, Bayesian approaches can handle missing data, represent prediction uncertainty and yields interpretable results.

Zenglin Xu is currently a postdoctoral researcher in computer science at Purdue University. Before this, he worked at Max Planck Institute for Informatics in 2009. He obtained his PhD in the Chinese University of Hong Kong. He has been a visiting scholar in Michigan State University.

 

Alan Qi is an assistant professor of computer science and Statistics at Purdue University. He obtained his PhD from MIT in 2005 and then worked as a postdoctoral researcher. He received the Newton Breakthrough Research Award from Microsoft Research (2008), the Purdue Interdisciplinary Award (2010), and the NSF CAREER award (2011).

SP3: Probabilistic Planning with Markov Decision Processes

Andrey Kolobov and Mausam

Markov Decision Processes (MDPs) are a powerful tool for sequential decision-making under uncertainty, a core subfield of Artificial Intelligence. Their applications include planning advertising campaigns, putting together scientific agenda of a Mars Rover, playing games, and many others.

This tutorial takes an algorithmic approach to covering MDPs. It starts by describing the basic MDP types: finite, total discounted-reward, and stochastic shortest path MDPs. It then proceeds to techniques for solving them, from the oldest optimal ones (Value Iteration, Policy Iteration), to heuristic search algorithms (LAO*, LRTDP, etc.), to state-of-the-art approximation approaches such as RFF -- the special emphasis area of this tutorial. The lecture will conclude by discussing extensions to the established MDP classes and promising research directions.

The tutorial attendees are not expected to have any prior familiarity with MDPs, and will walk away with the knowledge of this field sufficient for conducting research in it.

Andrey Kolobov is a Ph.D. student in the Department of Computer Science at the University of Washington, Seattle, advised by Daniel S. Weld and Mausam. His main line of research concentrates on mathematical properties of various MDP classes and practically efficient algorithms for solving them. His other research interests include machine learning and applications of game theory to designing laws.

Mausam is a research assistant professor in the Department of Computer Science at the University of Washington, Seattle. His research interests span various sub-fields within Artificial Intelligence, including AI planning under uncertainty, heuristic search, Web- scale natural language processing, machine learning and AI applications to crowd-sourcing. Relevant to this tutorial, Mausam's salient contributions include extension of MDPs for concurrent durative actions; optimal MDP solvers — both external and internal memory; and recently, state of the art approximate solvers for large MDPs.

MA1: A Game Theoretic Approach to Social Networks

Ramasuri Narayanam and Amit A. Nanavati

This tutorial provides the conceptual underpinnings of the use of game theoretic models in social network analysis and brings out how these models supplement and complement existing approaches for social network analysis. In the first part of the tutorial, we provide rigorous foundations of relevant concepts in game theory, mechanism design and social network analysis. In the second part of the tutorial, we bring out how game theoretic approach helps analyze social networks better and also how to apply the game theoretic concepts to problem solving in a rigorous way. In particular, we present a comprehensive study of a few contemporary and pertinent problems in social networks such as social network formation, social network monetization, design of incentive mechanisms, and economics of networks.

The tutorial deals with both theory and practice of the game theoretic models with applications to social network analysis. In particular, we first present in-depth understanding of the tools and techniques in game theory and mechanism design, and then present the practical utility of these theoretical concepts by addressing important problems associated with strategic and economic aspects of social networks. In this process, we also present case studies of a few relevant and deployed multiagent systems. It is important to note the primary content of this tutorial comes from game theory, mechanism design, and networked multiagent systems in general.

Ramasuri Narayanam is a researcher in IBM Research — India. His research interests are game theory, mechanism design, and social networks. Prior to joining IBM Research, he received his Ph.D. degree from Department of Computer Science and Automation, Indian Institute of Science (IISc), Bangalore, India. His Ph.D. work is on designing game theoretic models for social network analysis. He received an honorable mention award from Yahoo! Labs Key Scientific Challenges, 2010 for his work in Ph.D. He is also a recipient of Microsoft Research Ph.D. Fellowship for the duration 2007-2010. He is the coauthor of a Game Theoretic Problems in Network Economics and Mechanism Design Solutions.

Amit A. Nanavati is a senior researcher and manages Telecom Solutions Research at IBM Research, India. He has been working on Telecom social network analysis (The SNAzzy project) since its inception in 2006. He is particularly interested in applications of graph theory in various domains. He is also a "spoken web" evangelist — trying to promote the vision of a world-wide spoken web hosted in the Telecom network, which does not require an Internet connection or the ability to read and write. He also dabbles with speech in mobile and pervasive environments. He has published several papers and patents. He was recently named a Master Inventor at IBM Research. Prior to joining IBM Research, he was with Netscape Communications. He received the PhD degree in computer science from Louisiana State University, Baton Rouge. Before completing his PhD, he spent a summer in the Jet Propulsion Laboratory, California Institute of Technology.

MA2: Heuristic Search: The Basics and Beyond

Wheeler Ruml and Jordan Thayer

Heuristic search is seeing widespread use in applications from video games and robot motion planning to parsing and sequence alignment. This tutorial provides an introductory survey covering both classical and contemporary algorithms in shortest-path graph search. Both theoretical and implementation issues will be discussed. We will start with the setting of searching for optimal solutions (for example, A*) and move on to searching for any solution, searching for solutions under cost bounds, and searching under time bounds. In addition to classical lower bounds on solution cost, we will discuss how to incorporate estimates of solution cost and length in order to speed up search. Our emphasis will be on distinguishing different problem settings, identifying algorithms that apply to each, and developing intuitions for when each algorithm will perform well or poorly.

Wheeler Ruml is an assistant professor of computer science at the University of New Hampshire. His research interests include heuristic search and optimization, with a current emphasis on time-aware planning. Before joining the University of New Hampshire in 2007, he managed the Embedded Reasoning Area at the Palo Alto Research Center.

Jordan Thayer is a PhD candidate at the University of New Hampshire, where his work focuses on heuristic search under cost and time constraints. He has published papers at IJCAI, ICAPS, and SoCS, and has been a visiting researcher at University of Freiburg and Ben-Gurion University of the Negev.

MA3: Search-Based Planning: Toward High Dimensionality and Differential Constraints

Mihail Pivtoraiko, Maxim Likhachev, and Sven Koenig

Many real-world planning problems push the limits of traditional AI solutions. Competent reasoning systems, such as robots, must make split-second decisions, while facing considerable challenges in their environment. Systematic search is a traditional planning tool due to a number of beneficial features, yet it does not scale well to many realistic planning problems. This four-hour tutorial attempts to demonstrate that a large class of real-world hard problems — involving high dimensionality and differential constraints — can be solved efficiently using systematic graph search enabled with algorithmic and representation improvements.

It is our goal to make the tutorial a stand-alone discussion and not to place strict prerequisites on the background of the audience. In particular, we only assume that the audience will be familiar with the rudimentary concepts in planning and search. We will strive to make the tutorial as interactive as possible, and we will review any concepts as needed. The discussion will be illustrated with plentiful examples, including recent fielded robotics systems such as nonholonomic mining trucks, autonomous cars (including Boss, the winner of the DARPA Urban Challenge), high-DOF mobile manipulators, autonomous aerial vehicles and quadruped robots, among others.

Mihail Pivtoraiko is a post-doctoral fellow at the GRASP Laboratory, University of Pennsylvania. He obtained his Ph.D. (2012) and M.S. degrees (2005) at the Robotics Institute, Carnegie Mellon University. Pivtoraiko spent over a year at the NASA/Caltech Jet Propulsion Laboratory investigating autonomous rover operations in the context of flight missions.

Maxim Likhachev is a research assistant professor with the Robotics Institute and National Robotics Engineering Center (NREC), both part of School of Computer Science, Carnegie Mellon University. He is also an adjunct faculty at the Computer and Information Science department at University of Pennsylvania and a member of the GRASP laboratory.

Sven Koenig is a professor in computer science at the University of Southern California. He received his Ph.D. degree in computer science from Carnegie Mellon University. He also holds M.S. degrees from the University of California at Berkeley and Carnegie Mellon University.

MP1: Entity Resolution: Theory, Practice and Open Challenges

Lise Getoor and Ashwin Machanavajjhala

Entity resolution (ER), the problem of extracting, matching and resolving entity mentions in structured and unstructured data, is a long-standing challenge in artificial intelligence, statistics, information retrieval, and database management. It has been approached using a variety of techniques, including constraint-based methods, statistical methods, and methods which perform probabilistic inference. Accurate and fast entity resolution has huge practical implications in a wide variety of commercial, scientific and security domains.

This tutorial gives an overview of practical aspects and theoretical approaches to the problem of entity resolution. It summarizes ER research from AI, machine learning, database and information retrieval communities. In addition to giving attendees a thorough understanding of existing ER models, algorithms and evaluation methods, the tutorial will cover important research topics such as scalable ER, active and lightly supervised ER, and query-driven ER. The tutorial will also include a hands-on segment, where attendees will have the opportunity to try existing algorithms on sample datasets. The tutorial will close with a discussion of larger, related topics such as identity management and privacy.

Lise Getoor is an associate professor in the Computer Science Department at the University of Maryland, College Park. She received her PhD from Stanford University in 2001. Her current work includes research on link mining, statistical relational learning and representing uncertainty in structured and semistructured data. She has also done work on social network analysis and visual analytics. She has published numerous articles in machine learning, data mining, database, and artificial intelligence forums. She was awarded an NSF Career Award, is an action editor for the Machine Learning Journal, is a JAIR associate editor, has been a member of AAAI Executive council, and has served on a variety of program committees including AAAI, ICML, IJCAI, ISWC, KDD, SIGMOD, UAI, VLDB, and WWW.

Ashwin Machanavajjhala is a senior research scientist in the Knowledge Management group at Yahoo! Research. His primary research interests lie in data privacy and security, big-data management and statistical methods for information integration. Machanavajjhala graduated with a Ph.D. from the Department of Computer Science, Cornell University. His thesis work on defining and enforcing privacy was awarded the 2008 ACM SIGMOD Jim Gray Dissertation Award Honorable Mention. Machanavajjhala is currently leading a project on building a Web-scale distributed unsupervised information extraction system for automatically creating structured databases of entities from semistructured web pages, and is involved in building a large-scale distributed entity deduplication system at Yahoo! as part of the Web-Of-Concepts initiative. His research specialty is in developing highly scalable statistical methods for information extraction and entity matching on distributed systems. He has developed solutions for extracting structured entities from semi-structured web-pages, blocking, clustering in entity resolution, and integrating temporally changing data. He has published in WWW, WSDM, VLDB and SIGMOD.

MP2: Text Mining from User Generated Content

Ronen Feldman and Lyle Ungar

The proliferation of documents available on the Web and on corporate intranets is driving a new wave of text mining research and application. Earlier research addressed extraction of information from relatively small collections of well-structured documents such as newswire or scientific publications. Text mining from the other corpora such as the web requires new techniques drawn from data mining, machine learning, NLP and IR. Text mining requires preprocessing document collections (text categorization, information extraction, term extraction), storage of the intermediate representations, analysis of these intermediate representations (distribution analysis, clustering, trend analysis, association rules, etc.), and visualization of the results. In this tutorial we will present the algorithms and methods used to build text mining systems. The tutorial will cover the state of the art in this rapidly growing area of research, including recent advances in unsupervised methods for extracting facts from text and methods used for web-scale mining. We will also present several real world applications of text mining. Special emphasis will be given to lessons learned from years of experience in developing real world text mining systems, including recent advances in sentiment analysis and how to handle user generated text such as blogs and user reviews.

Ronen Feldman is one of the world's most recognized experts in the field of text mining, link analysis and the semantic analysis of data. In 1997, he founded ClearForest, a Boston-based business intelligence company later acquired by Reuters. He coined the term "text mining" in 1995, and his textbook, The Text Mining Handbook: Advanced Approaches in Analyzing Unstructured Data is considered the world's premier authority on this complex topic. He currently serves as the head of the Information Systems Department at the Business School of the Hebrew University of Jerusalem and was an adjunct professor at New York University Stern Business School. He has given over 30 tutorials on text mining and information extraction and has written numerous scholarly papers on these topics. He received his Ph.D. in computer science from Cornell University and his B.Sc. in math, physics and computer science from the Hebrew University of Jerusalem. He began his career in the military assigned to the elite Talpiot Group, and served for eight years as an officer in the Israel Air Force.

Lyle H. Ungar is an associate professor of computer and information science at the University of Pennsylvania. Ungar received a B.S. from Stanford University and a Ph.D. from the Massachusetts Institute of Technology. He directed Penn's Executive Masters of Technology Management (EMTM) program for a decade, and is currently the associate director of the Penn Center for BioInformatics (PCBI). Ungar has published over 100 articles and holds eight patents. His current research focuses on developing scalable machine learning methods for data mining and text mining.

MP3: Theory and Practice of Answer Set Programming

Esra Erdem, Joohyung Lee, and Yuliya Lierler

Answer set programming (ASP) is a declarative programming paradigm oriented towards knowledge-intensive tasks and combinatorial search problems. Its main idea is to reduce the given search problem to computing stable models of a logic program, and to use an ASP solver to perform search. ASP is particularly suited for modeling incomplete, inconsistent and dynamic domains. It has become a major knowledge representation formalism and has been applied to several areas in AI including planning, diagnosis, information integration, and bioinformatics. Wide applications of ASP motivated various extensions to the language and implementations, including integrations of ASP with constraint satisfaction and description logics. Recently, the stable model semantics, a mathematical basis of ASP, is shown to be closely related to classical logic.

The tutorial will provide an interactive, educational session to introduce the current state of the art in declarative problem solving via ASP. The audience will walk away with an understanding of the mathematical foundation of ASP, algorithms and systems for computing answer sets, recent applications of ASP including biomedical query answering systems and cognitive robotics, and the applicability of ASP to problems of their own interests.

Prerequisite knowledge: Basic knowledge of first-order logic will be assumed.

 

Esra Erdem is a faculty member at Sabanci University. She received her Ph.D. in computer sciences at the University of Texas at Austin (2002), and visited University of Toronto and Vienna University of Technology for postdoctoral research (2002-2006). Her research is in the area of knowledge representation and reasoning.

 

Joohyung Lee is an assistant professor at Arizona State University. He received his Ph.D. in computer science from the University of Texas at Austin in 2005. His research interests are in knowledge representation, logic programming, computational logics, and security. His 2004 AAAI paper received an Honorable Mention for the Outstanding Paper Award.

Yuliya Lierler is a Computing Innovation Fellow Postdoc at the Computer Science Department at the University of Kentucky. She completed her Ph.D. in computer science at the University of Texas at Austin in 2010. She is interested in knowledge representation, automated reasoning, declarative problem solving, and natural language understanding.

Registration

Information