<?xml version="1.0" standalone="yes"?>
<Paper uid="W01-1510">
  <Title>Resource sharing among HPSG and LTAG communities by a method of grammar conversion from FB-LTAG to HPSG</Title>
  <Section position="4" start_page="1" end_page="1" type="intro">
    <SectionTitle>
2 Background
</SectionTitle>
    <Paragraph position="0"/>
    <Section position="1" start_page="1" end_page="1" type="sub_section">
      <SectionTitle>
2.1 Feature-Based Lexicalized Tree
Adjoining Grammar (FB-LTAG)
</SectionTitle>
      <Paragraph position="0"> LTAG (Schabes et al., 1988) is a grammar formalism that provides syntactic analyses for a sentence by composing elementary trees with two opera- null tary trees are classified into two types, initial trees and auxiliary trees (Figure 2). An elementary tree has at least one leaf node labeled with a terminal symbol called an anchor (marked with A5). In an auxiliary tree, one leaf node is labeled with the same symbol as the root node and is specially marked as a foot node (marked with A3). In an elementary tree, leaf nodes with the exception of anchors and the foot node are called substitution nodes (marked with AZ).</Paragraph>
      <Paragraph position="1"> Substitution replaces a substitution node with another initial tree (Figure 3). Adjunction grafts an auxiliary tree with the root node and foot node labeled DC onto an internal node of another tree with the same symbol DC (Figure 4). FB-LTAG (Vijay-Shanker, 1987; Vijay-Shanker and Joshi, 1988) is an extension of the LTAG formalism. In FB-LTAG, each node in the elementary trees has a feature structure, containing grammatical constraints on the node. Figure 5 shows a result of LTAG analysis, which is described not  only by derived trees (i.e., parse trees) but also by derivation trees. A derivation tree is a structural description in LTAG and represents the history of combinations of elementary trees.</Paragraph>
      <Paragraph position="2"> There are several grammars developed in the FB-LTAG formalism, including the XTAG English grammar, a large-scale grammar for English (The XTAG Research Group, 2001). The XTAG group (Doran et al., 2000) at the University of Pennsylvania is also developing Korean, Chinese, and Hindi grammars. Development of a large-scale French grammar (Abeill'e and Candito, 2000) has also started at the University of Pennsylvania and is expanded at University of</Paragraph>
    </Section>
    <Section position="2" start_page="1" end_page="1" type="sub_section">
      <SectionTitle>
Paris 7.
2.2 Head-Driven Phrase Structure
Grammar (HPSG)
</SectionTitle>
      <Paragraph position="0"> An HPSG grammar consists of lexical entries and ID grammar rules, each of which is described with typed feature structures (Carpenter, 1992). A lexical entry for each word expresses the characteristics of the word, such as the subcategorization frame and the grammatical category. An ID grammar rule represents a relation between a mother and its daughters, and is independent of lexical characteristics. Figure 6 illustrates an example of bottom-up parsing with an HPSG grammar. First, lexical entries for &amp;quot;can&amp;quot; and &amp;quot;run&amp;quot; are unified respectively with the daughter feature structures of  an ID grammar rule. The feature structure of the mother node is determined as a result of these unifications. The center of Figure 6 shows a rule application to &amp;quot;can run&amp;quot; and &amp;quot;we&amp;quot;. There are a variety of works on efficient parsing with HPSG, which allow the use of HPSG-based processing in practical application contexts (Flickinger et al., 2000). Stanford University is developing the English Resource Grammar, an HPSG grammar for English, as a part of the Linguistic Grammars Online (LinGO) project (Flickinger, 2000). In practical context, German, English, and Japanese HPSG-based grammars are developed and used in the Verbmobil project (Kay et al., 1994). Our group has developed a wide-coverage HPSG grammar for Japanese (Mitsuishi et al., 1998), which is used in a high-accuracy Japanese dependency analyzer (Kanayama et al., 2000).</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>