AAAI 2012 Spring Symposium Series Registration
Register for the Spring Symposium!
The Association for the Advancement of Artificial Intelligence, in cooperation with Stanford University’s Department of Computer Science, is pleased to present the 2012 Spring Symposium Series, to be held Monday through Wednesday, March 26–28, 2012 at Stanford University. The titles of the six symposia are as follows:
- AI, The Fundamental Social Aggregation Challenge, and the Autonomy of Hybrid Agent Groups
- Designing Intelligent Robots: Reintegrating AI
- Game Theory for Security, Sustainability and Health
- Intelligent Web Services Meet Social Computing
- Self-Tracking and Collective Intelligence for Personal Wellness
- Wisdom of the Crowd
AI, The Fundamental Social Aggregation Challenge, and the Autonomy of Hybrid Agent Groups (SS01)
The control of autonomous humans, machines and robots working together as hybrid agent groups is an important problem for AI. Today, the war in Afghanistan has hundreds of mobile robots aloft, on land, or under the sea. But these agents are socially passive.
Current paradigms require a team of human operators for each mobile platform, with no social aggregation among nonhuman agents, precluding the formation of autonomous hybrid teams. A "social fabric" would be able to leverage heterogeneous sensing to enhance situational awareness, improving the capabilities of hybrid teams during decision-making. The literature (for example, Nature, 2011; NSF's visions for the future of social-behavioral-economic sciences) underscores the fundamental challenge of aggregation for social science. How does a collection of individuals become an autonomous group, team, or organization?
Unlike objects in physical reality, each human agent sees events and actions in social reality while embedded in different locations and under the influence of others; agents differentially collect, process, send, receive, channel and block information while they influence each other.
Uncertainty is a consideration, with two probable causes. One is based on measurement, the other on degrees of freedom (complexity). The problem of aggregation addresses the former. It reflects the physical influences of interdependence (bistability and multistability; for example, two or more sides exist to every story).
Hybrid agent teams must report on situations. Reports by humans are often reduced to ordinal data, the foundation of modern economic theory that Barzalai and colleagues believe is an error. Common examples of nonordinal interdependent effects are economic panic, real-estate bubbles, hostile mergers, and political gridlock.
Game theory initiated the mathematical study of interdependence among multiple agents. Yet, aggregating individual data into group (team) data remains unsolved. We must be able to prove mathematically that a group is different from the individuals who comprise it.
Solving the challenge of interdependence is essential for the effective and efficient engineering of autonomous multiagent teams. Once solved, hybrid teams could multitask to solve problems with firefighting, police work, reactor meltdowns, future wars, or while stationed on Mars and more. A hybrid leader would be able to efficiently control multiple autonomous robots, machines and humans as they solve problems together even under life and death situations.
Organizers
W. F. Lawless (Paine College, wlawless@paine.edu); Don Sofge (Naval Research Laboratory, Navy Center for Applied Research in Artificial Intelligence, don.sofge@nrl.navy.mil); Mark Klein (MIT Center for Collective Intelligence, m_klein@mit.edu); Laurent Chaudron (ONERA Provence Research Center, French Air Force Academy, laurent.chaudron@polytechnique.org)
For More Information
For more information, contact the organizers at the e-mails listed previously.
Designing Intelligent Robots: Reintegrating AI (SS02)
The goal of building intelligent robots has been a motivating problem for generations of AI researchers, going back at least as far as Shakey the robot in 1966. Creating such a robot is both the fully realized expression of the original impulse behind AI and an immensely rich source of research questions that address real-world problems.
However, AI is a fragmented field: well-developed and largely independent research communities exist for learning, planning, reasoning, language, perception and control. Since the challenges posted by each of these subfields are immense, most researchers have found it necessary to devote their careers to specializing in a single subfield. While immense progress has been made in each of these subfields in the last few decades, it remains unclear how they can be integrated to produce an intelligent robot. Unifying these disparate technologies will open up new avenues of research and create new application opportunities. Therefore, we believe that integration should be considered a valid research endeavor in its own right.
This symposium aims to bring together a diverse and multidisciplinary group of researchers interested in the specific objective of designing intelligent robots. The goal of the symposium is to provide common ground for their diverse interests and thereby actively encourage the integration of various AI techniques. We also hope to foster an active discussion about setting a realistic and feasible medium-term objective for integrative research so that progress can be made. The symposium will include invited talks as well as a poster session with ample time for discussion.
Organizing Committee
George Konidaris (Massachusetts Institute of Technoloty), Byron Boots (Carnegie Mellon University), Stephen Hart (GM), Todd Hester (University of Texas, Austin), Sarah Osentoski (Bosch Research and Technology Center), David Wingate (Massachusetts Institute of Technology).
For More Information
For more information, see the supplementary symposium web site.
Game Theory for Security, Sustainability and Health (SS03)
There is a large and growing interest in applying game theory to security, sustainability, and health; which are grand challenges for engineering in the 21st century. In fact, the last five years have seen game theory based systems developed and applied to real-world domains. For example, software assistants have been developed for randomized patrol planning for the Los Angeles International Airport police, the Federal Air Marshal Service and the United States Transportation Security Administration. Also game theory has been utilized for decentralized control, operation and management of future generation electricity.
While there has been significant progress, there still exist many major challenges facing the design of effective approaches to deal with the difficulties in security, sustainability, and health. Addressing these challenges requires collaboration from different communities including artificial intelligence, game theory, operations research, social science, and psychology. This symposium is structured to encourage a lively exchange of ideas between members from these communities. Topics of interest that may be discussed include but are not limited to:
- Game theory foundations
- Algorithms for scaling to very large games
- Human factors and intelligent user interfaces
- Agent/human interaction for preference elicitation and optimization
- Risk analysis
- Decision making under uncertainty
- Multiagent simulation
- Software development
- Modeling and Evaluation
- Distributed control in energy systems
This symposium will feature presentations for all accepted papers. There will be invited talks and a panel discussion by experts from a variety of relevant fields.
Organizing Committee
Bo An (University of Southern California, boa@usc.edu), Vincent Conitzer (Duke University, conitzer@cs.duke.edu), Manish Jain (University of Southern California, manishja@usc.edu), Sarit Kraus (Bar-Ilan University, sarit@cs.biu.ac.il), Kevin Leyton-Brown (University of British Columbia, Vancouver, kevinlb@cs.ubc.ca) Sarvapali Ramchurn (University of Southampton, sdr@ecs.soton.ac.uk), Milind Tambe (University of Southern California, tambe@usc.edu)
For More Information
For more information, see the supplementary symposium web site.
Intelligent Web Services Meet Social Computing (SS04)
Development of web services faces significant challenges concerning quality of design, development costs, endorsement of services by the community, integration and interoperability of services from different domains and effective sharing of services among users and developers. This spring symposium will bring together two lines of research whose combination can help in dealing with these issues, namely intelligent web services and social computing research. Social Computing is a promising approach that can help to understand user and community behaviour and related computational challenges around Web services development.
The topics of the symposium include the following:
- Social and technical requirements for collaborative web service development
- Platforms and user interfaces for crowdsourcing web service development, and verification
- Techniques for contextualized reviewing and rating of web services
- Methods to incentivize, boost, and influence community participation throughout the lifecycle of web services
- Methods to define and mashup service descriptions with Linked Data vocabularies
- Systems and techniques for context- and social-based recommendation of web services
- Methods for collaborative authoring of semantic annotations (for example RDFa, SAWSDL)
- Argumentation frameworks and norms for reaching consensus on service implementation, description, and integration
- Trust in collaborative web service construction
- Mining, monitoring and analysis of behaviour and activities of web service online communities (such as ProgrammableWeb, and Seekda)
- Analysis of web service usage patterns and associated social and technical parameters
- Extraction of web service descriptions from tags
- Case studies for use of social computing to construct and manage web services
The symposium will include presentations of short and long papers, research proposals, demonstrations, invited speakers, panel and open discussions. Invited speakers include John Musser (ProgrammableWeb), Steffen Staab (University of Koblenz, Germany), and Fausto Giunchiglia (University of Trento, Italy). Researchers working in areas of semantic web, web services, linked data, intelligent agents, social networks, social computing or web science will be encouraged to participate in the symposium.
Chairs
Tomas Vitvar (Czech Technical University), Harith Alani (Knowledge Media Institute, Open University, UK), David Martin (Apple)
For More Information
For more information, see the supplementary symposium web site.
Self-Tracking and Collective Intelligence for Personal Wellness (SS05)
How can we quantify our health? How can our health data be integrated into personalized medicine, improved wellness, and contributions to scientific discovery? These are the significant questions to improve our daily life. To tackle this issue, our symposium aims to explore the answers to the above questions by integrating two approaches, individual and collective viewpoints, in improving personal wellness. The approach from the individual viewpoint focuses on recently developed self-tracking technologies for monitoring personal health conditions such as sleep, diet, exercise, and vital signs data, and for analyzing personal medical data and personal genome data. The approach from the collective viewpoint focuses on collective intelligence as a potential resource for finding useful knowledge for personal wellness from the knowledge of other individuals and groups. The role of artificial intelligence and other technologies is examined in helping to create value in our future personal wellness. This symposium will bring together an interdisciplinary group of researchers to discuss possible solutions for personal wellness.
The scope of our interests include, but are not limited to the following:
- Self-tracking for personal wellness
(Sleep monitoring, diet monitoring, vital data monitoring, personal medicine, personal genome, new types of self-tracking devices, portable mobile tools) - Collective intelligence for personal wellness
(Data mining for scientific discovery on collective data, biomedical informatics and systems biology, data visualization) - Field study for personal wellness
(life log analyses such as vital data analyses, lifestyle related disease improvement experiment such as metabolic syndrome or diabetes, sleep improvement experiment) - Application for personal wellness
(life log applications, wellness service application, medical recommendation system, care support system for aged persons, web service for personal wellness, games for health and happiness) - Community platform for personal wellness
(Citizen science platform, do it yourself (DIY) trials, quantified self business model)
The symposium is organized by the invited talks, presentations, and posters and interactive demos.
Invited Speakers
Invited speakers at the symposium will include Atul J. Butte (Stanford University), Yukiko Shiki (University of Kansai), Rollin McCray (HeartMath Research Center), Sudheendra Hangal (Stanford University), and Emiliana Simon-Thomas (Stanford University).
Chairs
Takashi Kido, cochair (Riken Genesis Company Ltd. (Japan); kido.takashi@gmail.com) and Keiki Takadama (The University of Electro-Communications (Japan); keiki@inf.uec.ac.jp)
For More Information
For more information, see the supplementary symposium web site.
Wisdom of the Crowd (SS06)
Crowdsourcing provides a convenient and increasingly popular method for gathering large amounts of data and annotations. Amazon's Mechanical Turk and CrowdFlower, games such as the ESP Game, and requests for free annotation help such as LabelMe are just a few examples of crowdsourcing efforts. These attempts have taught us many lessons and brought up yet more questions. How can we most effectively elicit the information we need from a distant and potentially anonymous workforce? What kind of workforce is required for different tasks such as user studies and data set labeling? How can we train and evaluate workers?
This symposium brings together researchers from robotics, user interfaces, games, computer vision, and other disciplines exploring the core scientific research challenges of crowdsourcing. The symposium will facilitate interaction among researchers and work toward formulating a set of guidelines for future crowdsourcing endeavors.
Topics of interest include, but are not limited to the following:
- Applications for crowdsourcing: data set annotation, user studies, search relevance, content authoring, and integration of crowdsourcing and AI
- Reward strategies: no direct compensation (the LabelMe data set), low (Mechanical Turk) or high per-task compensation, and making the task fun, such as by using games (the ESP Game).
- Methods for selecting an appropriate workforce: recruiting experts, creating experts and trusted workers, learning worker expertise, changing compensation models, and requiring workers to pass tests.
- Methods for efficiently evaluating results: no evaluation, evaluating each task by hand, allowing workers to evaluate each other, and automated evaluation.
The symposium will combine a variety of activities to facilitate interaction among participants from different communities and discussion of key challenges, including invited talks, individual technical presentations by researchers to serve as case studies, open panel discussions and brainstorming around different applications and modalities, and working groups to create basic guidelines and evaluation strategies, providing a common starting point for future development and evaluation.
Invited speakers include Ed Chi (Google Inc.) and Rob Miller (Massachusetts Institute of Technology).
Chairs
Caroline Pantofaru (Willow Garage), Sonia Chernova (Worcester Polytechnic Institute), Alexander Sorokin (CrowdFlower)
For More Information
For more information, see the supplementary symposium web site.