PRAGMATIC INTERPRETATION AND DIALOGUE MANAGEMENT IN SPOKEN-LANGUAGE SYSTEMS
PRAGMATIC INTERPRETATION AND DIALOGUE MANAGEMENT
IN SPOKEN-LANGUAGE SYSTEMS
Gert Veldhuijzen van Zanten
Institute for Perception Research
P.O.Box 513
5600 MB Eindhoven
The Netherlands e-mail: veldhvz@natlab.research.philips.com
ABSTRACT
This paper describes the design of pragmatic-interpretation and dialogue-management modules in an automatic inquiry system that can be consulted through spoken natural language over the telephone.
The system is designed around a central multi-level data structure representing the discourse that has unfolded during the dialogue. At the highest level of this discourse representation the information exchange is represented as a series of changes or updates of an information state. Conditions on the information state give rise to actions of the dialogue manager. The dialogue manager aims at achieving the user 's goal in a manner that is understandable to the user, efficient and correct. This is no trivial problem, because natural language and, in particular, speech understanding leads to many uncertainties. To deal with uncertain information, we have designed feedback and verification mechanisms and means for contextual understanding, underspecification and pragmatic inferencing.
INTRODUCTION
The understanding of natural, and in particular, spoken language involves many issues that are only partially understood and thus hard to implement on computer systems. Some of these issues are ambiguity, underspecification, contextual interpretation and pragmatic inferences. Within the NWO-TST1 priority programme on Language and Speech Technology, we are building a number of demonstrator systems,
OVIS1, OVIS2 and OVIS32, that engage in spoken1. NWO is the Netherlands Organization for Scientific Research. TST is a Dutch acronym for Language and
Speech Technology
2. OVIS is a Dutch acronym for Public Transport Information System. OVIS1 is based on a system designed at
Philips GmbH Forschungslaboratorien Aachen
IN SPOKEN-LANGUAGE SYSTEMS
Gert Veldhuijzen van Zanten
Institute for Perception Research
P.O.Box 513
5600 MB Eindhoven
The Netherlands e-mail: veldhvz@natlab.research.philips.com
ABSTRACT
This paper describes the design of pragmatic-interpretation and dialogue-management modules in an automatic inquiry system that can be consulted through spoken natural language over the telephone.
The system is designed around a central multi-level data structure representing the discourse that has unfolded during the dialogue. At the highest level of this discourse representation the information exchange is represented as a series of changes or updates of an information state. Conditions on the information state give rise to actions of the dialogue manager. The dialogue manager aims at achieving the user 's goal in a manner that is understandable to the user, efficient and correct. This is no trivial problem, because natural language and, in particular, speech understanding leads to many uncertainties. To deal with uncertain information, we have designed feedback and verification mechanisms and means for contextual understanding, underspecification and pragmatic inferencing.
INTRODUCTION
The understanding of natural, and in particular, spoken language involves many issues that are only partially understood and thus hard to implement on computer systems. Some of these issues are ambiguity, underspecification, contextual interpretation and pragmatic inferences. Within the NWO-TST1 priority programme on Language and Speech Technology, we are building a number of demonstrator systems,
OVIS1, OVIS2 and OVIS32, that engage in spoken1. NWO is the Netherlands Organization for Scientific Research. TST is a Dutch acronym for Language and
Speech Technology
2. OVIS is a Dutch acronym for Public Transport Information System. OVIS1 is based on a system designed at
Philips GmbH Forschungslaboratorien Aachen
References: Aust, H., Oerder, M., Seide, F. and Steinbiss, V. (1994). Experience with the Philips Automatic Train Timetable Information System. Proceedings. IVTTA '94, Kyoto, Japan, 67-72. Grosz, B.J. and Sidner, C.L. (1986). Attention, Intentions, and the Structure of Discourse. Computational Linguistics, Vol. 12, Nr.3, July-September 1986, 175-204 Levinson, S. (1983). Pragmatics. Cambridge University Press, Cambridge (UK). Oerder, M. and Aust, H. (1994). A Realtime Prototype of an Automatic Inquiry System. Proceedings ICSLP '94, Vol. 2, Yokohama, Japan, 703-706. Vallduví, E. (1990). The Informational Component, Ph.D. thesis, University of Pennsylvania, USA.