<?xml version="1.0" standalone="yes"?>
<Paper uid="P98-1017">
  <Title>An Efficient Kernel for Multilingual Generation in Speech-to-Speech Dialogue Translation</Title>
  <Section position="2" start_page="0" end_page="110" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> In this paper we present core aspects of the multilingual natural language generation component VM-GECO 1 that has been integrated into the research prototype of Verbmobil (Wahlster, 1993; Bub et al., 1997), a system for spontaneous speech-to-speech dialog translation.</Paragraph>
    <Paragraph position="1"> In order to achieve multilinguality as elegantly as possible we found that a clear modular separation between a language-independent general kernel generator and language-specific parts which consist of syntactic and lexical knowledge sources was a very promising approach. Accordingly, our generation component 1VM-GECO is an acronym for &amp;quot;VerbMobil GEneration COmponents.&amp;quot; consists of one kernel generator and language-specific knowledge sources for the languages used in Verbmobih German and English with current work on Japanese.</Paragraph>
    <Paragraph position="2"> Additionally, the kernel generator itself can be modularized furthermore into two separate components. The task of the so-called mieroplanning component is to plan an utterance on a phrase- or sentence-level (Hovy, 1996) including word-choice (section 2). It generates an annotated dependency structure which is used by the syntactic generation component to realize an appropriate surface string for it (section 3). The main goal of this further modularization is a stepwise constraining of the search-space of alternative linguistic realizations, using abstracted views on different choice criteria.</Paragraph>
    <Paragraph position="3"> Multilingual generation in dialog translation imposes strong requirements on the generation module. A very prominent problem is the nonwellformedness (incorrectness, irrelevance, and inconsistency) of spontaneous input. It forces the realization of robust generation to be able to cope with erroneous and incomplete input data so that the quality of the generated output may vary between syntactically correct sentences and semantically understandable utterances. On the level of knowledge sources this is achieved by using a highly declarative HPSG grammar which very closely reflects the latest developments of the underlying linguistic theory (Pollard and Sag, 1994) and covers phenomena of spoken language. This HPSG is compiled into a TAG grammar in an offtine pre-Processing step (Kasper et al., 1995) which keeps the declarative nature of the grammar intact (section 3).</Paragraph>
    <Paragraph position="4"> Maybe the most important requirement on the generation module of a speech-to-speech translation system is real-time processing. The  above mentioned features of VM-GECO contribute to the efficiency of the generation component. The TAG-formalism is well known for the existence of efficient syntactic generation algorithms (Kilger and Finkler, 1995).</Paragraph>
    <Paragraph position="5"> In general, all knowledge sources of all modules are declarative. The main advantage is that this allows for an easier adaptation of the generation component to other domains, languages and semantic representation languages besides the easier extendability of the current system. The feasibility of the language adaptation was proved in the Verbmobil project itself where the (originally English) generator was recently extended to cover German and is currently adapted for Japanese. The adaptation to another domain and also to another specification language for intermediate structures was shown in another translation project which uses in contrast to Verbmobil an interlingua based approach (section 4.1).</Paragraph>
  </Section>
class="xml-element"></Paper>