1 / 8

NRRC Summer ‘02 Workshop Proposal: Habitability

NRRC Summer ‘02 Workshop Proposal: Habitability. Christy Doran, MITRE Joe Marks, MERL cdoran@mitre.org marks@merl.com. December 5, 2001. MITRE. Problem. Basic premise: Assume an imperfect Q&A system -- what can we do to make it more usable?

Download Presentation

NRRC Summer ‘02 Workshop Proposal: Habitability

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. NRRC Summer ‘02Workshop Proposal:Habitability Christy Doran, MITRE Joe Marks, MERLcdoran@mitre.org marks@merl.com December 5, 2001 MITRE

  2. Problem • Basic premise: Assume an imperfect Q&A system -- what can we do to make it more usable? • Our proposal (courtesy of Bill Ogden, NMSU) is to enhance the habitability of Q&A systems “Habitability, as the term will be used here, is obtained when a significant proportion of the users utterances are interpreted in such a way that the user in some meaningful sense is carried closer towards accomplishing the task at hand.’’ Ivan Bretan citing: W.C. Watt, "Habitability," American Documentation, July, 1968, pp. 338--351.

  3. What might this mean for Q&A systems? • Representing system expertise • With the goal of imbuing the system with an ability to articulate the extent of its knowledge • Presenting query results (or lack thereof) • Communicating data and metadata efficiently • Explaining how the system works • Only an issue when it doesn’t work well? • Incremental question/answer refinement • Human-computer collaboration • Allowing teamwork (human) • Human-human collaboration

  4. Approach • Habitability task not sufficiently well defined at this time • No obvious silver bullet • Interaction needed between at least three different communities: • HCI (Visualization, Groupware) • NLP (Dialogue, Q&A) • Target users • Our proposal: Workshop that would bring these communities together to further explore the topic and identify approaches to improving habitability of Q&A systems

  5. Details of the workshop • 2.5-day workshop • Meeting format Day 1 AM: All participants to run and comment on a Q&A system, and to experiment with other representative tools • IBM Q&A system (Roukos & Ittycheriah) Day 1 AM: Historical motivation by Bill Ogden Day 1 PM: Brief presentations from each subarea for establishing common appreciation of issues & challenges Day 2: Structured discussion based on first day’s activities to identify most promising approaches for improving habitability of Q&A systems Day 3 AM: Decide on best way(s) to pursue the approaches identified at a follow-on event

  6. Proposed workshop participants Bold = preliminary commitment • Goal is 23 participants, including Doran, Marks, and Ogden • Natural language processing (NLP) • Dialogue: Lyn Walker (AT&T), Diane Litman (Pittsburgh), Candy Sidner (MERL), David Traum (USC ICT) • Q&A: James Pustejovsky (LingoMotors), Liz Liddy (Syracuse), Marc Light (MITRE), Sanda M. Harabagiu (SMU) • Human-computer interaction (HCI) • Visualization: Marti Hearst (UC Berkeley), Kent Wittenburg (MERL),Jock Mackinlay (Xerox PARC),Stephen Eick (Visual Insights) • Groupware: Kori Inkpen (Dalhousie), Terry Winograd (Stanford), Mark Ackerman (Michigan),Saul Greenberg (Calgary) • Q&A system users (members of NRRC Specialists Group) • Mark Zimmerman (USGC), John Donelan (USGC),Kelcy Allwein (USGC),Jean-Michel Pomarede (USGC)

  7. Results of the workshop • Meeting tentatively scheduled for late February • Workshop to be documented in an article or report or roadmap to engage relevant communities in Q&A research • Another possible outcome is a proposal for a follow-on event, e.g., a AAAI Symposium or a conference workshop

  8. Resources and costs • Workshop • Travel and honoraria: $34K • Site expenses: @MITRE $10K, more compelling location $25K • Lodging and local transportation • Transportation and set-up of demo equipment • Estimated total: cheaper $44K, less cheap $59K • Follow-on activity: $50K • This could be a second workshop, or other activity selected by the group

More Related