Adaptive Interfaces for Destination Advice


The goal of this subproject is to develop an interactive system that gives advice to drivers about likely destinations, such as restaurants and gasoline stations, that meet the driver's criteria. This will take the form of a conversational interface that carries out a dialog, in spoken language, with the user that helps him transform an initially vague query, like ``Where should I have lunch today?'', into a specific location. The system should also have the ability to adapt to the user's preferences, so that the human-machine dialog becomes more efficient over time.

We have developed an initial adaptive interface that operates in the restaurant domain and that takes written text as input. The current system stores previous interactions with a users in a case library, which it uses to select the order of questions and to recommend options the user will find attractive. We have designed and started to implement the successor system, which will support spoken interaction and a more flexible model of the dialog process.

This research was funded by DaimlerChrysler Research and Technology.


Contributors to the Project

  • Professor Renée Elio

  • Dr. Mehmet Goker

  • Dr. Afsaneh Haddadi

  • Professor Pat Langley

  • Amy Perfors

  • Dr. Jeff Shrager

  • Professor Stanley Peters

  • Professor Cynthia Thompson

  • Lei Wang

  • Related Papers

    Thompson, C. A., Göker, M., & Langley, P. (2002). A personalized system for conversational recommendations (Technical Report UUCS-02-013). School of Computer Science, University of Utah, Salt Lake City. Submitted to Journal of Artificial Intelligence Research.

    Langley, P., Thompson, C., Elio, R. & Haddadi, A. (1999). An adaptive conversational interface for destination advice. Proceedings of the Third International Workshop on Cooperative Information Agents (pp. 347-364). Uppsala, Sweden.



    For more information, please send email to langley@csli.stanford.edu .


    Return to research overview page.