<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1502">
  <Title/>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> This paper is concerned with the semi-automatic grammar development of real-scale grammars.</Paragraph>
    <Paragraph position="1"> For natural language syntax, lexicalised TAGs are made of thousands of trees, carrying an extreme structural redundancy. Their development and their maintenance is known to be cumbersome as the size of the grammar raises significantly.</Paragraph>
    <Paragraph position="2"> To counter the lack of generalisations inherent to strong lexicalisation, various proposals for semi-automatic grammar development have been carried out: lexical rules or meta-rules (Becker, 2000) and metagrammars: (Candito, 1999; Gaiffe et al., 2002; Xia, 2001). The aim of these frameworks is twofold: expressing general facts about the grammar of a language and factorising the information to avoid redundancy.</Paragraph>
    <Paragraph position="3"> The metagrammar path adopts a different perspective from the lexical rule based grammar development: instead of describing how a derived tree is different from a canonical one, grammatical description mainly consists of combining fragmentary tree descriptions or building blocks.</Paragraph>
    <Paragraph position="4"> The paper is structured as follows. We start in section 2 by providing motivations and background information on the framework we are using. Section 3 shows that the metagrammar framework may be viewed as an offline system allowing to express high level well-formedness constraints on elementary grammatical structures while preserving TAG computational and formal properties. Section 4 shows how to implement efficiently this constraint-based approach with logic programming techniques and finally section 5 provides an idea of the performance of the implemented system.</Paragraph>
  </Section>
class="xml-element"></Paper>