INTEGRATED TECHNIQUES
FOR GENERATION AND INTERPRETATION

Richmond H. Thomason

University of Pittsburgh
Intelligent Systems Program
University of Pittsburgh
Pittsburgh, PA 15260

CONTACT INFORMATION

Paper Mail: Intelligent Systems Program, University of Pittsburgh, Pittsburgh, PA 15260,

Electronic Mail: thomason@isp.pitt.edu

Office Telephone: (412) 624-5791

FAX: (412) 624-6130

WWW PAGE

http://www.pitt.edu/~thomason/thomason.html

PROGRAM AREA

Speech and Natural Language Understanding.

KEYWORDS

Natural-language-generation, natural-language-generation, discourse-planning, discourse-simulation, cooperative-dialog, abduction

PROJECT SUMMARY

The project's long-range goal is to create a unified architecture for interactive discourse, incorporating both interpretation and generation. The immediate goal is to develop an environment for simulating and testing dialog in cooperative planning environments.

Traditional computational approaches to the problems of discourse have treated the problems of interpretation and generation separately. The chief problem of interpretation is how to fill in missing information; the chief problem of generation is how to leave out redundant or reconstructible information. We seek to develop a unified architecture that will enable us to test hypotheses about the knowledge structures that these processes share, and the interactions between them. The project will yield an environment that simulates cooperative dialog in a task domain, as well as an analyzed corpus of collected data. Technology transfer based on the FASTUS discourse interpretation system will be carried out by SRI International.

The project is premised on the use of simulated discourse to collect data about discourse, and to test specific hypotheses about discourse interpretation and generation. Note that a discourse simulator can be tested (a) by generating entirely simulated dialogues in which discourse strategies and conditions are varied, or (b) by testing the simulations' interactions with human subjects. And (c) the simulation architecture can also be used as an environment for gathering data concerning conversations between human subjects in the task domain. The project will use all these forms of testing.

The basis of our approach to interpretation is the abductive model of Jerry Hobbs and his associates at SRI. (Jerry Hobbs, Mark Stickel, Douglas Appelt, and Paul Martin, "Interpretation as abduction," Artificial Intelligence 63, 1993, pp. 69-142.) The basisf of our approach to generation is the plan-based approach of Johanna Moore. (See Johanna D. Moore, Participating in Explanatory Dialogues, MIT Press, 1995.) The theoretical work of the project consists largely of unifying these approaches. The discourse simulation will be based on this unification.

PROJECT REFERENCES

Pamela W. Jordan and Richmond H. Thomason, "Empirical Methods in Discourse: Limits and Prospects." AAAI Spring Symposium, 1995.

Pamela W. Jordan and Megan Moser "Multi-level Coordination in Computer-mediated Conversation." GEORGETOWN.

Pamela W.Jordan and Richmond H.Thomason, "Empirical Methods in Discourse: Limits and Prospects." AAAI Spring Symposium, 1995.

AREA BACKGROUND

We are trying to build on work in natural language generation and in natural language understanding to understand how the two processes interact and model each other in conversation. And we are interested in exploring simulation as an empirical technique in discourse. If we and others with similar ideas are successful, a new subdiscipline of computational linguistics will emerge, drawing on work such as the following.

AREA REFERENCES

Jerry Hobbs, Mark Stickel, Douglas Appelt, and Paul Martin. "Interpretation as abduction," Artificial Intelligence 63, 1993, pp. 69-142.)

Johanna Moore. Participating in Explanatory Dialogues: Interpreting and Responding to Questions. MIT Press, Cambridge, MA, 1995.

Herbert Clark. Arenas of language use. University of Chicago Press, Chicago, 1992.

Richmond Thomason. "Accommodation, meaning, and implicature interdisciplinary foundations for pragmatics." In P. Cohen, J. Morgan, and M. Pollack, eds., Intentions in communication. MIT Press, Cambridge, Massachusetts, 1990, pp. 325-363.

Marilyn Walker. Informational redundancy and resource bounds in dialog. Ph.D. Dissertation, University of Pennsylvania, Philadelphia, 1993.

RELATED PROGRAM AREAS

Virtual Environments.

Adaptive Human Interfaces.

Usability and User-Centered Design.

Intelligent Interactive Systems for Persons with Disabilities.