Birte Schmitz

Knowledge-based disambiguation for machine translation

The resolution of ambiguities is one of the central problems for Machine Translation. In this paper we propose a knowledge-based approach to disambiguation which uses Description Logics (dl) as representation formalism. We present the process of anaphora resolution implemented in the Machine Translation systemfast and show how thedl systemback is used to support disambiguation.The disambiguation strategy uses factors representing syntactic, semantic, and conceptual constraints with different weights to choose the most adequate antecedent candidate. We show how these factors can be declaratively represented as defaults inback. Disambiguation is then achieved by determining the interpretation that yields a qualitatively minimal number of exceptions to the defaults, and can thus be formalized as exception minimization.

Dialogue acts in automatic dialogue interpreting

In this paper we demonstrate that for an adequate translation of an utterance spoken in a dialogue the dialogue act it performs has to be determined. We introduce an approach that automatically assigns types of dialogue acts to utterances on the basis of both micro- and macro-structural information. Technically, this assignment is realized by modeling preference rules as weighted defaults in the Description Logic system FLEX. The dialogue-act type of an utterance is determined by qualitatively minimizing the exceptions to these defaults. The results described here have been developed within the VERBMOBIL project, a project concerned with face-to-face dialogue interpreting funded by the German Federal Ministry of Education, Science, Research and Technology (BMBF). We present the rather positive results of a first evaluation of this implementation showing the accuracy of dialogue act assignment.

The VERBMOBIL domain model version 1.0

This report describes the domain model used in the German Machine Translation project VERBMOBIL. In order make the design principles underlying the modeling explicit, we begin with a brief sketch of the VERBMOBIL demonstrator architecture from the perspective of the domain model. We then present some rather general considerations on the nature of domain modeling and its relationship to semantics. We claim that the semantic information contained in the model mainly serves two tasks. For one thing, it provides the basis for a conceptual transfer from German to English; on the other hand, it provides information needed for disambiguation. We argue that these tasks pose different requirements, and that domain modeling in general is highly task-dependent. A brief overview of domain models or ontologies used in existing NLP systems confirms this position. We finally describe the different parts of the domain model, explain our design decisions, and present examples of how the information contained in the model can be actually used in the VERBMOBIL demonstrator. In doing so, we also point out the main functionality of FLEX, the Description Logic system used for the modeling. The VERBMOBIL Domain Model Version 1.0