<?xml version="1.0" standalone="yes"?>
<Paper uid="C00-2128">
  <Title>A Statistical Approach to the Processing of Metonymy</Title>
  <Section position="4" start_page="885" end_page="885" type="metho">
    <SectionTitle>
3 Information Source
</SectionTitle>
    <Paragraph position="0"> \Y=e use a large corpus to extract nouns which can be syntactically related to the exl)licit term of a metonylny. A large corpus is vahmble as a source of such nouns (Church and Hanks, 1990; Brown et al., 1992).</Paragraph>
    <Paragraph position="1"> We used Japanese noun phrases of the fornl A no B to extract nouns that were syntactically related to A. Nouns in such a syntactic relation are usually close semantic relatives of each other (Murata et al., 1999), and occur relatively infrequently. We thus also used an A near B relation, i.e. identifying tile other nouns within the target sentence, to extract nouns that may be more loosely related to A, trot occur more frequently. These two types of syntactic relation are treated differently by the statistical nleasure which we will discuss in section 4.</Paragraph>
    <Paragraph position="2"> The Japanese noun phrase A no B roughly corresponds to the English noun phrase B of A, lint it has a nmch broader range of usage (Kurohashi and Sakai, 1999). In fact, d no B can express most of the possible types of semmltic relation between two nouns including metonymic 2~Ford' is spelled qtSdo' ill Japanese. We have used English when we spell Japanese loan-words from English for the sake of readability.</Paragraph>
    <Paragraph position="3"> concepts such as that the name of a container can represent its contents and the name of an artist can imply an art~brnl (container for contents and artist for artform below).a Examples of these and similar types of metonymic concepts (Lakoff and Johnson, 1980; Fass, 1997) are given below.</Paragraph>
    <Paragraph position="4">  These exalnt)les suggest that we can extract semantically related nouns by using tile A no B relation.</Paragraph>
  </Section>
  <Section position="5" start_page="885" end_page="887" type="metho">
    <SectionTitle>
4 Statistical Measure
</SectionTitle>
    <Paragraph position="0"> A nletonymy 'Noun A Case-Marker R, Predicate V' can be regarded as a contraction of 'Noun A Syntactic-Relation (2 Noun B Case-Marker R Predicate V', where A has relation Q to B (Yamamoto et al., 1998). For example, Shakc.spcare wo yomu (read) (read Shakespeare) is regarded as a contraction of Shakespeare no .sakuhin (works) 'wo yomu (read the works of Shakespeare), where A=Shake.spcare, Q=no, B=.sakuhin, R=wo, and V=yomu.</Paragraph>
    <Paragraph position="1"> Given a metonymy in the fbrln A R 17, the appropriateness of noun B as an interpretation of the metonymy under the syntactic relation Q is defined by LQ(BIA,/~, V) - Pr(BIA, (2, 1~, V), (2) ayamamoto et al. (\]998) also used A no /3 relation to interpret metonymy.</Paragraph>
    <Paragraph position="2">  where Pr(.-.) represents l)robal/ility and Q is either an A no B relation or an A near \]3 relation. Next;, the appropriateness of noun \]3 is defined by M(BIA, Ie, V) -nlaxLc~(BIA, l~,V ). (3) O We rank nouns 1)y at)plying the measure 214. Equation (2) can be decomposed as follows:</Paragraph>
    <Paragraph position="4"> where (A, O) and {\]~,, V} are assumed to l)e indel)endent of each other.</Paragraph>
    <Paragraph position="5"> Let f(event)1)e the frequen(:y of an cve'nt and Classc.s(\])) be the set of semantic (:lasses to which B belongs. 'l'he expressions in Equation (4) are then detined t)y 4</Paragraph>
    <Paragraph position="7"> We onfitted Pr(H,, 17) fi'om Equation (4) whell we calculated Equation (3) in the experiment de, scribed in section 5 for the sake of simplicit&gt;</Paragraph>
    <Paragraph position="9"> inition for the sake of simplicity. This simplification has little effect on the tilml results because ~--;c'cc~ ........ (m Pr(l~lC)f(C,I~', V) &lt;&lt; I will usually hohl. More Sol)histieated methods (M;mning ml(t Schiitze, 1999) of smoothing f)robability distribution m~y I)e I)eneticial. itowever, al)l)lying such methods and comparing their effects on the interpretation of metonymy is beyond the scope of this l)aper.</Paragraph>
    <Paragraph position="10"> This treatment does not alter the order of the nouns ranked by the syst;em because l?r(H., V) is a constant for a given metonymy of the form AR V.</Paragraph>
    <Paragraph position="11"> Equations (5) and (6) difl'er in their treatment of zero frequency nouns. In Equation (5), a noun B such that f(A, Q, B) = 0 will l)e ignored (assigned a zero probal)ility) because it is unlikely that such a noml will have a close relationshii / with noun A. In Equation (6), on the other hand, a noun B such that f(B, R, V) = 0 is assigned a non-zero probability. These treatments reflect the asymmetrical proper~y of inetonymy, i.e. ill a nletonylny of the form A 1{ 1~ an implicit term 13 will have a much tighter relationship with the explicit term A than with the predicate V. Consequently, a nouil \]3 such that f(A,Q, B) &gt;&gt; 0 A f(B, JR, V) = 0 may be appropri~te as an interpretation of the metonymy.</Paragraph>
    <Paragraph position="12"> Therefore, a non-zero t)robat)ility should be assign(;d to Pr(l~., VI1X ) ev~,n it' I(B, 2e, V) ; (). ~ Equation (7) is the probability that noun J3 occurs as a member of (::lass C. This is reduced to  fU~) if13 is not ambiguous, i.e. IC/a,~,sc.,s,(/3)\[ = f(c) 1. If it is ambiguous, then f(B) is distributed equally to all classes in Classes(B).</Paragraph>
    <Paragraph position="13">  The frequency of class C is ol)tained similarly: null</Paragraph>
    <Paragraph position="15"> where 13 is a noun which belongs to the class C.</Paragraph>
    <Paragraph position="16"> Finally we derive</Paragraph>
    <Paragraph position="18"> In summary, we use the measure M as defined in Equation (3), and cah:ulated by applying Equation (4) to Equation (9), to rank nouns according to their apl)ropriateness as possible interpretations of a metonymy.</Paragraph>
    <Paragraph position="19"> Example Given the statistics below, bottle we akeru (open) (open a bottle) will be interpreted 5The use of Equation (6) takes into account a noun/3 such that J'(l:~, l{, V) = 0. But, Stlch &amp; llOtlll is usually ignored if there is another noun B' such that f(13', H., V) &gt; 0 be~,~,,se. Eo'~ct ....... U~)P, USIO)J'(C,~e.,V) &lt;&lt; a &lt; J'(lY, H,, V) will usually hokl. This means thai the co-occurrence 1)rol)al)iliW between implicit terms and verbs are also important in eliminating inapl)rol)riate nomls.</Paragraph>
    <Paragraph position="20">  as described in the fbllowing t)aragraphs, assuming that cap and rcizSko (refl'igerator) are the candidate implicit terms.</Paragraph>
    <Paragraph position="21"> Statistics:</Paragraph>
    <Paragraph position="23"> and rcizSko are not close semantic relatives of each other. This shows the effectiveness of using A no B relation to filter out loosely related words.</Paragraph>
    <Paragraph position="25"> Since M &gt; M we conclude that cap is a more appropriate imt)licit term than rcizSho. This conclusion agrees with our intuition.</Paragraph>
  </Section>
class="xml-element"></Paper>