<?xml version="1.0" standalone="yes"?>
<Paper uid="C92-1009">
  <Title>Feature Structure Based Semantic Head Driven Generation</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> PracticM generation systems must lnwe linguistic knowledge of both specilic expressions like idioms and generM grammatical constructions, ;rod ttmy should efgtciently produce sm'face strings applying that knowledge \[\[\]\[2\].</Paragraph>
    <Paragraph position="1"> In order to satisfy the first requirement, our system employs a set of trees annotated with fe,~ture structures to represent generation knowledge. l:;ach tree represents a t?agment of a syntactic strncture, and is paired with a semantic feature structure. We can describe idiomatic eonstructions, by making a tree which cont~tins lexical specifications and is paired with a specilie rather than general semautic structure. Because feature structures allow partial speeiiicatiom we can encode generation knowledge r;mgiug over multiple levels of generality in a. uniform way.</Paragraph>
    <Paragraph position="2"> llowever, notice that this property will be restricted if we use DCG or (tixed arity) term notation 1 Suppose there is a generation knowledge structure whose syntactic part is &amp;quot;go on foot&amp;quot;. 'rim feat, tu'e structure notation of its semantic part will be sonmthing like: ~The flexibility of structure notation colnpated Lo tetln notation is also discussed il~ \[4\].</Paragraph>
    <Paragraph position="3"> \[ \[Rein GO\] \[Agent ?agent \[\] \] \[Instrument FOOT\]\].</Paragraph>
    <Paragraph position="4"> while the term notation is :</Paragraph>
    <Paragraph position="6"> These two notations seem to be equivalent, but there is a cruciN diflerence. A generation knowledge structure containing the fe~tture-based selnantics will still be unifiable even if the semantic input to be unified contains additional material. Thus the knowledge structure will be discovered and its syntactic information can he used for generation. By contrast, a term-based input with additiona.1 elements would not unify with the term-based semantic structure shown above. It would thus be necessary to create additional generation structures containing distinct (though partly overlN)ping) term-based semantic structures. Such additional structures are redundant ~tn(l cause superfluous output.</Paragraph>
    <Paragraph position="7"> For example, consider the a,ugmented feature structure (3).</Paragraph>
    <Paragraph position="8">  it will indeed nnify with (1) above. But term-based input semantic structure (4) will not unify with term-based semantic structure (2).</Paragraph>
    <Paragraph position="9"> instrument(time(go(ken), 10:00am), foot).</Paragraph>
    <Paragraph position="10"> (4) To unifv (2), semantic ,structure (5) would a.lso be required.</Paragraph>
    <Paragraph position="11"> time(instzument(go(ken), foot), 10:00ma).</Paragraph>
    <Paragraph position="12"> (5) AcrEs DE COLING-92. NANTES. 23 28 AOt~q&amp;quot; 1992 3 2 PROC. OI; COLING 92. NANTES. AUG. 23 28. 1992 For this reason, our generation knowledge consists of trees represented as feature structures. A tree can be substituted for a leaf node of asother tree to form a larger structure. Thus, tile tree can be regarded as a rule in a context-free feature-structure-based unification grammar.</Paragraph>
    <Paragraph position="13"> The second requirement for a generation system is efficient creation of syntactic structures. This is the main topic of this paper. Our system is based upon Semantic }lead Driven Generation \[6\], which is an efficient algorithm for unilication based formalisms. However, this algorithm requires some additional mechanisms to efficiently retrieve relevant generation knowledge, because feature structures can not be easily indexed.</Paragraph>
    <Paragraph position="14"> The algorithm presented here uses a nmltiple index network of feature structures to efficiently choose relevant generation knowledge from the knowledge base. The algorithm &amp;quot;also uses an hypothetical node so as to efficiently maintain ambiguous structures during generation.</Paragraph>
  </Section>
class="xml-element"></Paper>