AAAI 2012 Fall Symposium Descriptions

AAAI 2012 Fall Symposium Descriptions

AI for Gerontechnology

The aging population, the increasing cost of formal health care, caregiver burden and the importance that older adults place on living independently in their own homes motivate the need for the development of patient-centric technologies that promote safe independent living. These patient-centric technologies need to address various aging related physical and cognitive health problems such as heart disease, diabetes, deterioration of physical function, falling, wandering, strokes, and memory problems, lack of medication adherence, cognitive decline and loneliness. Advances in the sensor and computing technology that allow for ambient unobtrusive and continuous home monitoring have opened new vistas for the development of such technologies.

AI is central to such systems as it deals with the process of transforming raw sensor data into human interpretable abstractions and innovating new human computer interfaces for the older adults. AI can help in decision making and analyzing the sheer volume of captured data from a variety of sensing technologies for understanding the physical activities, nighttime behaviors, medication taking, socialization and ongoing physiological changes in the older adults. As the availability of longitudinal data increases, we have an unprecedented opportunity to discover new early predictors of clinically significant events. This is a challenging research area that has seen increasing interest among the research community due to the need of the hour.

The goal of this symposium is to understand the role of AI in current research through presentation of case studies and work in progress research and to brainstorm novel technologies that can be innovated and improved by AI research. The symposium will bring together researchers working on various artificial intelligence aspects of gerontechnology in order to highlight the current challenges, as well as to identify the future grand challenges.

Topics

Topics of discussion at the symposium include (but are not restricted to) application of novel AI techniques in the following areas:

Rehabilitation systems
Persuasive technology for health aging
Cognitive orthotics systems
Physiological change/anomaly detection
Socialization and emotional wellbeing in aging population
Medication adherence
Activity and behavior monitoring
Emergency situation detection
Privacy preservation
Evaluation mechanisms
Position papers on challenges faced by the caregiving and nursing community.

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.

Submissions

We invite authors to submit full and short papers on original and unpublished work. Papers may be submitted to ckn@eecs.wsu.edu. Submitted papers should be organized in AAAI proceedings format and should not exceed 6 pages. Submissions will be judged on their technical merit and the potential to generate discussion and collaborations. Submission of a paper indicates the willingness of at least one author to register for the symposium and present the paper.

Organizing Committee

Diane J Cook (Washington State University), Narayanan C Krishnan (Washington State University), Parisa Rashidi (University of Florida), Marjorie Skubic (University of Missouri-Columbia) and Alex Mihalidis (University of Toronto)

For More Information

For more information, please see the supplementary symposium website.

Artificial Intelligence of Humor

Human ability to communicate is incomplete without the use of humor. If a computational system is ever to approximate human communication ability or act as a competent partner in a conversation with a human, humor must be accounted for. The general goal of the proposed Symposium is to advance the state of the art in the direction of developing an AI system ("the system") capable of understanding the mechanism of a joke at a level sufficient for providing a punch line to a human generated setup (even if unintentional) and conversely, for computer reacting competently to a human generated punch line that follows a setup, generated by either participant. The effort is multidisciplinary in nature, and the participants from all of the contributing disciplines, viz., computational semantics, knowledge representation, computational psychology, AI theory, humanoid robotics, human-computer interface, human factors, to name just a few, are invited to participate.

Topics

Topics include (but are not limited to) the following:

humor detection and generation
semantic representation of jokes
reasoning within jokes
priming and saliency in jokes
humor preferences
humor ontology
modeling humor competence
modeling humor performance
humor in humanoid robotics
social computing with humor
detecting humor trends
computational humor for education

Format

The symposium will include a few invited papers, submitted research papers, and encourage the participants to suggest panels, round table discussions, mini-symposia, special sessions, etc. There will be a clear focus on discussion: in fact, to the extent possible, we will encourage the electronic circulation of invited and accepted presentations prior to the symposium, so that most of the time allotted to each paper be spent on its discussion.

Submissions

We invite regular research paper submissions (around 8 pages), short research papers (around 4 pages), proposals for panels, round-table discussions, special sections, listing the rationale and committed participants (1 page). Please submit to the symposium managing cochair, Julia M. Taylor, at jtaylor1@purdue.edu

Organizing Committee

Chair: Victor Raskin, vraskin@purdue.edu, LING/CERIAS, Purdue University, 217 Recitation Hall, 656 Oval Drive, West Lafayette, Indiana 47907-2086 USA

Managing Cochair: Julia M. Taylor, ;jtaylor1@purdue.edu, CIT/CERIAS, Purdue University, 253 Knoy Hall, 401 N. Grant Street, West Lafayette, Indiana 47907-2021, USA

Members: Anton Nijholt, anijholt@cs.utwente.nl, CS, University of Twente, Postbus 217, 7500 AE Enschede, The Netherlands; Willibald Ruch, w.ruch@psychologie.uzh.ch, PSYCH, University of Zurich, Department of Psychology, Personality and Assessment, Binzmühlestr. 14/7, CH-8050 Zürich, Switzerland

For More Information

For more information, please see the supplementary symposium website.

Discovery Informatics: The Role of AI Research in Innovating Scientific Processes

Addressing the ambitious research agendas put forward by many scientific disciplines requires meeting a multitude of challenges in intelligent systems, information sciences, and human-computer interaction. Many aspects of the scientific discovery process are often largely manual and could be automated, improved, or made more efficient. Better interfaces for collaboration, visualization, and understanding would significantly improve scientific practice. Scientific data, publications, and tools could be published in open formats with appropriate semantic descriptions and metadata annotations to improve sharing and dissemination. Opportunities for broader participation in well-defined scientific tasks enable human contributors to provide large amounts of data, annotations, or complex processing results that could not otherwise be obtained. Improvements and innovations across the spectrum of scientific processes and activities will have a profound impact on the rate of scientific discoveries.

This symposium will provide a forum for researchers interested in understanding the role of AI techniques in improving or innovating scientific processes. We seek submissions that: (1) report on success stories that illustrate the potential of future research in this field; (2) discuss lessons learned in the process of addressing challenging aspects of the scientific process; (3) analyze the impact of a particular technique in an area of science and reflect on its potential for broader applicability in other sciences; and (4) propose future concepts grounded in lessons learned and an understanding of the challenges in the scientific discovery process.

Topics

Topics of interest include but are not limited to the following:

Ontologies and knowledge bases that model particular areas of scientific knowledge
Semantic representations of metadata for all aspects of scientific processes
Techniques for organizing scientific literature
Workflow systems to manage complex data analysis processes
Knowledge discovery techniques that are embedded in the context of scientific investigations
Integrative approaches of machine learning and scientific model induction
Automated systems for experiment design, data analysis, and hypothesis generation and refinement
User-centered design of intelligent systems that partner with scientists to perform complex tasks
Integrated approaches to visualizing data, models, and the connections between them to foster new insights
Cognitive-centered design of scientist aids
Social computing systems that let novice participants contribute to scientific tasks

Submissions

Submissions should be up to 6 pages, using the AAAI style files. Submissions should be uploaded to the symposiusm submission site.

Symposium cochairs: Will Bridewell (Stanford University), Yolanda Gil (University of Southern California), Haym Hirsh (Rutgers University), Kerstin Kleese van Dam (Pacific Northwest National Laboratory), Karsten Steinhaeuser (University of Minnesota)

For More Information

For more information, please see the supplementary symposium website.

Human Control of Bio-Inspired Swarms

Robotic systems composed of a large number of robots, often called robot swarms, are envisioned to play an increasingly important role in applications such as search, rescue, surveillance, and reconnaissance operations. Nowadays, many mobile robots that are deployed for such applications are still teleoperated by a single or multiple operators. While these platforms are individually very capable, the development of cheaper hardware allows the consideration of swarm systems composed many more robots but with each individual being far less powerful. To control such systems is a considerable challenge due to the limitations of each individual robot and the sheer number of robots that need to be coordinated to successfully complete a mission. Autonomous algorithms provide an opportunity to mitigate some of the complexity an operator faces in controlling such swarms. But it is not clear which tasks will ultimately fall to the operator and which should rather be solved by the autonomy.

Research Challenges

How can humans influence swarms following "baked-in" control laws?
What are the characteristics of models of interagent influence?
What metaphors are most effective for humans
Biomemetic
Physicomemetic
Is the human role to assist the swarm (break it out of local minima) or to direct the swarm based on things it cannot sense?
What are the most effective mechanisms for selecting members of the swarm to be influenced?
How can optimization for autonomy be balanced with optimization for collaboration and cooperation?
How can controls and displays be designed to support central control of a distributed system?
Adversary response: how to detect when a swarm has been compromised?
How can a human infer the intent and "situation awareness" of a swarm
Can quorum sensing be understood by a human controller?

Format

The symposium will consist of presentations of relevant current work and position papers. We will have invited talks from leaders in the field and a panel to foster a general discussion of issues. The symposium is intended to serve as a springboard to take the research on this very important and useful problem forward.

Submissions

We solicit papers describing original work, either in-progress or finished. Papers should follow the AAAI formatting style, with a page-limit of 6 pages. To submit a paper, authors are asked to send an e-mail to Michael Lewis at ml@sis.pitt.edu providing their affiliation details and paper title in order to obtain a username and password and access the upload page.

Organizing Committee

Michael Lewis, Michael Goodrich, Andreas Kolling, Paul Scerri, Marc Steinberg, and Katia Sycara

For More Information

For more information, please see the supplementary symposium website.

Information Retrieval and Knowledge Discovery in Biomedical Text

The amount of biological and medical literature has grown exponentially within the last decade. This data may be in the form of journal citations in PubMed, in the form of clinical summaries in healthcare institutions or in the form of blogs and user comments that express personal opinions on the different healthcare topics such as drug adverse effects or disease treatments. This material, be it expressed by researchers, medical professionals or medical care receivers, is of significant importance in terms of the wealth of information that it possesses. However it is only valuable if efficient and reliable ways of accessing and analyzing that information are available.

In this symposium we would like to address novel research on computational techniques for information retrieval and knowledge discovery from biomedical and clinical texts, with a focus on machine learning or natural language processing, as well as novel applications of existing techniques to the open problems in text processing in biomedical domain. We will invite several speakers from the biomedical text processing community who will present current research problems in this field, and we will invite contributed talks on novel learning approaches that can improve the analysis and retrieval of biological and medical information.

We solicit two types of submissions to the symposium: (1) contributed talks or posters and (2) open problems.

Contributed Talks or Posters

We invite submissions that address new algorithmic and methodological contributions to the spectrum of problems in biomedical text analysis, where textual resources can include semi-structured and unstructured biomedical text, clinical text, social media and any other healthcare related text media.

Open Problems

For open problems, we request the authors to submit a one page description that motivates and explains an existing open problem in text analysis and information retrieval in biomedical domain. The main goal here is to foster active discussion.

Papers for the symposium will be collected and made into an AAAI technical report, which will be distributed to attendees on CD and included in the AAAI Digital Library. The issuance of technical report allows the work to be cited.

Submissions

The full papers submitted to the symposium should be at most eight pages long following the AAAI paper format. The extended abstracts should be one page long in the same format. The full-paper submissions will be considered and may be accepted for either an oral or a poster presentation, the extended abstracts may be accepted for the poster presentation only.

Submission Site.

Organizing Committee

Lana Yeganova (Staff Scientist, National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health); Rezarta Islamaj Dogan (Research Fellow, National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health); Vahan Grigoryan (Associate, Cloud Analytics Group, Booz Allen Hamilton); Mark Dredze (Assistant Research Professor, Department of Computer Science and the Human Language Technology Center of Excellence, Johns Hopkins University)

For More Information

For more information, please see the supplementary symposium website. If you have any questions, do not hesitate to direct your questions to Rezarta.Islamaj@nih.gov or Lana.Yeganova@nih.gov

Machine Aggregation of Human Judgment

The AAAI 2012 Fall Symposium on Machine Aggregation of Human Judgment focuses on combining human and machine inference. For unique events and data-poor problems, there is no substitute for human judgment. Even for data-rich problems, human input is needed to account for contextual factors. For example, textual analysis is data rich, but context and semantics often make automated parsing unusable. However, humans are notorious for underestimating the uncertainty in their forecasts and even the most expert judgments exhibit well-known cognitive biases. The challenge is therefore to aggregate expert judgment such that it compensates for the human deficiencies.

There are fundamental theoretical reasons to expect aggregated estimates to outperform individual forecasts. These theoretical results are borne out by a robust empirical literature demonstrating the superiority of opinion pools and prediction markets over individual forecasts, and of ensemble forecasts over those of top models. While weighted forecasts are theoretically optimal, among human experts unweighted forecasts have been hard to beat.

This symposium focuses on methods with the potential to come closer to the theoretical optimum. While a number of methods have shown promise individually, there is potential for significant advancement from combining them into structured, efficient, repeatable elicitation and aggregation protocols. Benefits of improved aggregation methods include substantial increases in the quality and reliability of expert judgments, removing misunderstanding, illuminating context dependence of forecasts, and reducing overconfidence and motivational bias in forecasts. On the other hand, there's some skepticism that statistical models can outperform experts most of the time. Machine reasoning lacks the context to know when the models no longer apply, or in cases like natural language, simply lack sufficient context to be reliable in open-world or novel problems. This symposium considers powerful hybrid techniques using humans to help aggregate computer models.

A broad range of researchers in the AI community and other application fields such as econometrics, sociology, political science, and intelligence analysis will find this symposium interesting and useful. Bringing these disciplines together to the venue also greatly facilitates the research endeavors.

Topics

Topics include but are not limited to the following:

Reasoning under uncertainty
Ensembles and aggregation
Information fusion
Crowdsourcing techniques and applications
Information elicitation and presentation
Performance evaluation: scalability and accuracy
Prediction markets
Collective intelligence

Submissions

The symposium will accept a number of regular papers (6–8 pages), and short papers/extended abstracts (2 pages). In addition to oral presentations, we intend to provide several poster sessions for more interactions. Further, invited talks by leading researchers in the fields and/or domain experts will be arranged. We will also reserve substantial time for questions and discussions after talks.

Submissions should be done through the EasyChair submission site. Authors, who do not have accounts on EasyChair, will be directed to create a new account before they can make a submission.

Organizing Committee

Kathryn B. Laskey and Wei Sun (George Mason University), John Irvine (Draper Laboratory), Dirk B. Warnaar (Applied Research Associates, Inc.), H. Van Dyke Parunak (Jacobs Technology Inc.)

For More Information

For more information, please see the supplementary symposium website.

Robots Learning Interactively from Human Teachers

To harness robots' full capabilities, human end-users should be able to customize their robots' behaviors through natural teaching methods. Due to their accessibility to non-expert users, interactive learning methods have attracted widespread attention in recent years. Endowing machines with such learning capabilities enables users to teach robots interactively and intuitively, as they would teach other humans. These natural teaching interfaces can also indirectly guide learning algorithms; for instance, via aiding feature selection, action abstraction, and natural language interpretation.

The goal of this symposium is to increase awareness and interest in interactive learning methods, and foster inter-disciplinary collaboration by bringing researchers across many disciplines together to discuss and exchange ideas on the current and potential future research directions.

Topics

We seek broad participation from researchers in the areas including, but not limited to the following:

Imitation learning
Learning from demonstration
Developmental psychology
Teachable agents
Cognitive models
Human-robot interaction
Long-term learning through interaction
Unimodal and multi-modal media for efficient human-robot interaction
Caregiver modeling
Adaptive systems
Transfer learning
Sliding autonomy
Learning performance evaluation

Program

Most of the symposium program will consist of oral presentations and poster/interactive sessions (depending on the number of submissions). We also aim to have 2–3 plenary speakers. The symposium will additionally include organizer-led discussion sessions to exchange ideas in a less formal setting.

Submissions

We welcome regular papers, extended abstracts, and late breaking results (2–8 pages in PDF AAAI format) as well as proposals for demos and panels. Submissions will be evaluated on both their technical merit along with their potential to generate discussion and create community collaboration.

Submission website.

Organizing Committee

Cetin Mericli (chair), Carnegie Mellon University, (cetin@cmu.edu) Carnegie Mellon University 5000 Forbes Avenue, Pittsburgh, PA 15213, USA Phone: 412-2687663 Fax: 412-2685576

Brenna Argall (Northwestern University, brenna.argall@northwestern.edu); Maya Cakmak (Georgia Institute of Technology, maya@cc.gatech.edu); W. Bradley Knox (University of Texas at Austin, bradknox@cs.utexas.edu); Tekin Mericli (Boğaziçi University, Turkey, tekin.mericli@boun.edu.tr)

Steering Committee

Chad Jenkins (Brown University, cjenkins@cs.brown.edu); Peter Stone (University of Texas at Austin, pstone@cs.utexas.edu); Andrea Thomaz (Georgia Institute of Technology, athomaz@cc.gatech.edu); Manuela Veloso (Carnegie Mellon University, veloso@cmu.edu)

For More Information

For more information, please see the supplementary symposium website.

Social Networks and Social Contagion

With the emergence of computational social science as a field of collaboration between computer scientists and social scientists, the study of social networks and processes on these networks (social contagion) has been gaining interest. Many topics of traditional sociological interest (such as the diffusion of innovation, emergence of norms, identification of influencer) can now be studied using detailed computational models and extensive simulation. The advent and popularity of online social media also allows the creation of massive data sets, which can inform models and underlying sociological theory. The ubiquity of "smart devices" (such as smart phones) also provides opportunities to gather extensive data on the behaviors and interactions of humans in "real space".

The goal of this symposium is to bring together a community of researchers interested in addressing these issues and to encourage interdisciplinary approaches to these problems. Papers are invited on topics including, but not limited to, the following:

Social Contagion
the spread of ideas or beliefs
emotion contagion
diffusion of information
the spread of changes in language
diffusion of innovations
emergence of norms
interventions to prevent contagion
influence maximization
complex contagion
virtual agents, agent-human contagion
Social Networks
collaborative tagging
collaborative filtering
community structure
social capital in networks
correlating demographics and structure
Game Theory in Social Networks and Social Contagion
influence maximization
influence blocking maximization game
other game-theoretic approaches
Network Evolution
homophily and heterophily
relation between structure and dynamics
Network Formation
models based on sociological theory, for example, structural balance
organizations, formal versus informal networks
network structure inference and label inference
network generation models
Human Data Elicitation
expression of attitudes/personality from online sources (such as Twitter and Facebook)
using social media for tracking social contagion, developing social networks, etc.
crowdsourcing as a means to learn about humans
Massively Multiplayer Online Games (MMOG) as "virtual laboratories" to study social contagion
reality mining for social networks

The symposium will consist of two and a half days of events, including oral presentations of accepted papers, invited talks, and a poster session.

Submissions

Submissions can be in two forms: a paper of up to eight pages in AAAI format for consideration for an oral presentation, or an extended abstract of up to two pages in the AAAI format for consideration for a poster presentation. Submission will be through EasyChair.

Organizing Committee

Samarth Swarup (Symposium Chair), Network Dynamics and Simulation Science Laboratory, Virginia Bioinformatics Institute, Virginia Tech, 1880 Pratt Drive (0477), Research Building XV, Suite 1100, Blacksburg, VA 24061, Ph: 1-540-231-1248, email: swarup@vbi.vt.edu Madhav Marathe (Network Dynamics and Simulation Science Laboratory, Virginia Bioinformatics Institute, Virginia Tech, mmarathe@vbi.vt.edu); Kiran Lakkaraju (Cognitive Science and Modeling Group, Sandia National Laboratories, klakkar@sandia.gov); Noshir Contractor (Department of Industrial Engineering and Management Sciences, Northwestern University, nosh@northwestern.edu); Milind Tambe (Computer Science and Industrial and Systems Engineering Departments, University of Southern California, tambe@usc.edu); Winter Mason (Howe School of Technology Management, Stevens Institute of Technology, wmason@stevens.edu)

For More Information

For more information, please see the supplementary symposium website.

AAAI 2012 Fall Symposia Descriptions