Seminar in Assisted Cognition

CSE Home

About Us

Search

Contact Info

Links  Readings and Calendar  AC Project Home Page  Assisted Cognition Mailing List

Time: Fridays 1:30-2:30 Note time/date/room change! Place:  EE1-003 (except May 24th, meet in EE1-042) Credit:  1 (higher credit by special arrangement.)  Auditors also welcome. Instructor:  Henry Kautz  (Assoc. Prof., CSE)  http://www.cs.washington.edu/homes/kautz      (206) 543-1896       Sieg Hall 417 Other course contact: Don Patterson (Grad. Student, CSE) <djp3@cs.washington.edu> This is the reading seminar on Assisted Cognition.  This seminar may be taken for credit by CSE graduate students, and by other UW students with permission of the instructor.  We also welcome participation from anyone in the greater UW community who is interested in the subject.  This reading seminar will also provide good background for students potentially interested in joining the Assisted Cognition Project at UW.  We will use the general mailing list for the Assisted Cognition Project as the mailing list for this seminar; please use the links at left to sign up. What is Assisted Cognition?  An emerging cross-disciplinary effort combining work in artificial intelligence, ubiquitous computing, and medicine, that aims to create computer systems that help people with diminished or damaged cognitive functioning (such as Alzheimer's patients).  While work in AI has traditionally focused on expert-level problem-solving (e.g., medical diagnosis, chess, control of industrial equipment) we will be looking at ways that AI systems can help people perform the ordinary tasks of daily living -- thus increasing their independence and quality of life.  In this seminar we will read and discuss a broad array of papers that provide background for work in this area.  Topics include: Behavior recognition - tracking a person's actions on the basis of low-level data gathered from a variety of sensing devices - GPS, IR badges, sound, etc. Plan recognition - fusing behavior data with a model of the tasks of daily living in order to recognize a person's goals and intentions. Using machine learning techniques to develop models that are personalized to the individual. Intervention strategies - Determining when a person is having difficulty performing a task or is endangered, and deciding how the system should intervene in a truly helpful manner. Caregiving - What lessons have been learned in the nursing and medical communities about effective caregiving strategies for people with cognitive disabilities?  How should this impact ideas for computer-based support for such individuals?

	Department of Computer Science & Engineering University of Washington Box 352350 Seattle, WA  98195-2350 (206) 543-1695 voice, (206) 543-2969 FAX [comments to kautz]