I will describe research on learning and reasoning about computer users' intentions, preferences, and attention, and highlight opportunities and challenges for machine learning at the human-computer interface. I will illustrate key ideas in the context of representative projects at Microsoft Research focusing on automation, user assistance, communications, and mixed-initiative interaction. After reviewing concepts from the Coordinate, Lookout, Notification Platform, Seer, and Swish projects, I will discuss the prospect for new kinds of adaptation based on design-time and real-time learning. I will conclude with a discussion of challenge problems for learning in human-computer interaction.
· E. Horvitz and J. Apacible. Learning and
Reasoning about Interruption. Proceedings of the Fifth International
Conference on Multimodal Interfaces, November 2003,
· E. Horvitz, A. Jacobs, D. Hovel. Attention-Sensitive Alerting, Proceedings of UAI '99, Conference on Uncertainty and Artificial Intelligence, July 1999, pp. 305-313.
· E. Horvitz. Principles of
Mixed-Initiative User Interfaces. Proceedings of CHI '99, ACM SIGCHI
Conference on Human Factors in Computing Systems,
· N. Oliver and E. Horvitz. Selective Perception Policies for Guiding Sensing and Computation in Multimodal Systems: A Comparative Analysis, Proceedings of the Fifth International Conference on Multimodal Interfaces, November 2003, ACM Press, Vancouver, BC, Canada.
· N. Oliver, E. Horvitz, and A. Garg. Layered Representations for Recognizing Office Activity, Proceedings of the Fourth IEEE International Conference on Multimodal Interaction (ICMI 2002), Pittsburgh, PA, pp.3-8
· E. Horvitz and T. Paek. Harnessing Models of Users' Goals to Mediate Clarification Dialog in Spoken Language Systems. Proceedings of the Eighth International Conference on User Modeling, July 2001.
· E. Horvitz. Uncertainty, Action, and Interaction: In Pursuit of Mixed-Initiative Computing. Intelligent Systems, Sept./ October Issue, IEEE Computer Society.
· D. Azari, E. Horvitz, S. Dumais, E. Brill. A Decision Making
Perspective on Web Question Answering, Proceedings of the Nineteenth
Conference on Uncertainty in Artificial Intelligence,
· K. Toyama and E. Horvitz. Bayesian Modality
Fusion: Probabilistic Integration of Multiple Vision Algorithms for Head
Tracking. Proceedings of ACCV 2000, Fourth Asian Conference on Computer
Vision, January 2000.
· E. Horvitz and T. Paek, DeepListener:
Harnessing Expected Utility to Guide Clarification Dialog in Spoken Language
Systems, 6th International Conference on Spoken Language Processing (ICSLP 2000),
· T. Paek, E. Horvitz, E. Ringger, Continuous Listening for Unconstrained Spoken Dialog, 6th International Conference on Spoken Language Processing (ICSLP 2000), Beijing, November 2000.
· T. Lau and E. Horvitz, Patterns of
Search: Analyzing and Modeling Web Query Refinement. Proceedings of the
Seventh International Conference on User Modeling,
· D. Heckerman and E. Horvitz. Inferring Informational Goals from Free-Text Queries. Proceedings of the Fourteenth Conference on Uncertainty in Artificial Intelligence, July 1998.
· E. Horvitz and M. Barry. Display of Information for Time-Critical Decision Making. Proceedings of the Eleventh Conference on Uncertainty in Artificial Intelligence, August 1995.