<?xml version="1.0" standalone="yes"?>
<Paper uid="P96-1051">
  <Title>An Application of WordNet to Prepositional Attachment</Title>
  <Section position="2" start_page="0" end_page="0" type="metho">
    <SectionTitle>
1 Problem description
</SectionTitle>
    <Paragraph position="0"> In this paper, we address the problem of disambiguation and understanding prepositional attachment. The arguments of prepositional relations are automatically categorized into semantically equivalent classes of WordNet (Miller and Teibel, 1991) concepts. Then by applying inferential heuristics on each class, we establish semantic connections between arguments that explain the validity of that prepositional structure. The method uses information provided by WordNet, such as semantic relations and textual glosses.</Paragraph>
    <Paragraph position="1"> We have collected prepositional relations from the Wall Street Journal tagged articles of the PENN TREEBANK. Here, we focus on preposition of, the most frequently used preposition in the corpus.</Paragraph>
  </Section>
  <Section position="3" start_page="0" end_page="360" type="metho">
    <SectionTitle>
2 Classes of prepositional relations
</SectionTitle>
    <Paragraph position="0"> Since most of the prepositional attachments obey the principle of locality (Wertmer, 1991), we considered only the case of prepositional phrases preceded by noun or verb phrases. We scanned the corpus and filtered the phrase heads to create C, an ad hoc collection of sequences &lt; noun prep noun &gt; and &lt; verb prep noun &gt;. This collection is divided into classes of prepositional relations, using the following definitions: Definition 1: Two prepositional structures &lt; noun1 prep noun2 &gt; and &lt; noun3 prep noun4 &gt; belong to the same class if one of the following conditions holds: * noun1, and noun2 are hypernym/hyponym of noun3, and noun4 respectively, or * noun1, and noun2 have a common hypernym/hyponym and with noun3, and noun4, respectively. null A particular case is when noun1 (noun2) and</Paragraph>
    <Paragraph position="2"> long to the same class if one of the following conditions holds: * verb1, and noun1 are hypernym/hyponym of verb2, and noun2, respectively or * verb1, and noun1 have a common hypernym/hyponym with verb2, and noun2, respectively. null A particular case is when the verbs or the nouns are synonyms, respectively.</Paragraph>
    <Paragraph position="3"> The main benefit and reason for grouping prepositional relations into classes is the possibility to disambiguate the words surrounding prepositions. When classes of prepositional structures are identified, two possibilities arise:  1. A class contains at least two prepositional sequences from the collection g. In this case, all sequences in that class are disambiguated, because for each pair (&lt; nouni prep nounj &gt; , &lt; nounk prep nounq &gt;), nouni and nounk (and nounj and nounq respectively) are in one of the following relations: (a) they are synonyms, and point to one synset that is their meaning.</Paragraph>
    <Paragraph position="4"> (b) they belong to synsets that are in hypernym/hyponym relation.</Paragraph>
    <Paragraph position="5"> (c) they belong to synsets that have a common  hypernym/hyponym.</Paragraph>
    <Paragraph position="6"> In cases (a), (b) and (c), since words are associated to synsets, their meanings are disambiguated. The same applies for classes of prepositional sequences &lt; verb prep noun &gt;.  acquisition of company Sense 1 = { acquisition, acquiring, getting } GLOSS: &amp;quot;the act of contracting or assuming .&amp;quot;HR1 or ~possession of something&amp;quot; &amp;quot;-. HR3 { buy, purchase, take } &amp;quot;~' .</Paragraph>
    <Paragraph position="7"> GLOSS: &amp;quot;obtain by purchase~ by means of a financial transaction&amp;quot;  2. A class contains only one sequence. We disregard these classes from our study, since in  this class it is not possible to disambiguate the words.</Paragraph>
    <Paragraph position="8"> The collection C has 9511 &lt; noun of noun &gt; sequences, out of which 2158 have at least one of the nouns tagged as a proper noun. 602 of these sequences have both nouns tagged as proper nouns. Due to the fact that WordNet's coverage of proper nouns is rather sparse, only 34% of these sequences were disambiguated. Successful cases are &lt; House of Representatives &gt;, &lt; University of Pennsylvania &gt; or &lt; Museum of Art &gt;. Sequences that couldn't be disambiguated comprise &lt; Aerospaciale of France &gt; or &lt; Kennedy of Massachusetts &gt;. A small disambiguation rate of 28% covers the rest of the 1566 sequences relating a proper noun to a common noun. A successful disambiguation occurred for &lt; hundreds of Californians &gt; or &lt; corporation of Vancouver &gt;. Sequences like &lt; aftermath of Iran-Contra &gt; or &lt; acquisition of Merryl Linch &gt; weren't disambiguated. The results of the disambiguation of the rest of 7353 sequences comprising only common nouns are more encouraging. A total of 473 classes were devised, out of which 131 had only one element, yielding a disambiguation rate of 72.3%. The number of elements in a class varies from 2 to 68. Now that we found disambiguated classes of prepositional structures, we provide some heuristics to better understand why the prepositional relations are valid. These heuristics are possible inferences performed on WordNet.</Paragraph>
  </Section>
  <Section position="4" start_page="360" end_page="361" type="metho">
    <SectionTitle>
3 Selectional Heuristics on WordNet
</SectionTitle>
    <Paragraph position="0"> In this section we focus on semantic connections between the words of prepositional structures. Con- null sider for example acquisition of company. Figure 1 illustrates some of the relevant semantic connections that can be drawn from WordNet when analyzing this prepositional structure.</Paragraph>
    <Paragraph position="1"> We note that noun acquisition is semantically connected to the verb acquire, which is related to the concept { buy, purchase, take}, a hypernym of { take over, buy out}. Typical objects for buy out are corporations and companies, both hypernyms of concern. Thus, at a more abstract level, we understand acquisition of company as an action performed on a typical object. Such relations hold for an entire class of prepositional structures.</Paragraph>
    <Paragraph position="2"> What we want is to have a mechanism that extracts the essence of such semantic connections, and be able to provide the inference that the elements of this class are all sequences of &lt; nounl prep nounj &gt;, with nounj always an object of the action described by nounl.</Paragraph>
    <Paragraph position="3"> Our approach to establish semantic paths is based on inferential heuristics on WordNet. Using several heuristics one can find common properties of a prepositional class. The classification procedure disambiguates both nouns as follows: the word acquisition has four senses in WordNet , but it is found in its synset number 1. The word company appears in its synset number 1. The gloss of acquisition satisfies the prerequisite of HRI: Heuristic Rule 1 (HR1) If the textual gloss of a noun concept begins with the expression the act of followed by the gerund of a verb, then the respective noun concept describes an action represented by the verb from the gloss.</Paragraph>
    <Paragraph position="4"> This heuristic applies 831 times in WordNet, showing that nouns like accomplishment, dispatch or subsidization describe actions.</Paragraph>
    <Paragraph position="5"> I\] Nr.crt. I Features for &lt; N1 &gt; of &lt; N2 &gt; Example II  Thus acquisition is a description of any of the verbal expressions contract possession, assume possession and acquire possession.</Paragraph>
    <Paragraph position="6"> The role of company is recovered using another heuristic: Heuristic Rule 2 (HR2) The gloss of a verb may contain multiple textual explanations for that concept, which are separated by semicolons. If one such explanation takes one of the forms:  then nounz and noun2 respectively are objects of that verb.</Paragraph>
    <Paragraph position="7"> Heuristic HR2 applies 134 times in WordNet, providing objects for such verbs as generalize, exfoliate or laicize.</Paragraph>
    <Paragraph position="8"> The noun company is recognized as an object of the synset {take over, buy out}, and so is corporation. Both of them are hyponyms of {business, concern, business concern}, which fills in the object role of {business, concern, business concern}. Because of that, both company and corporation from the gloss of {take over, buy out} are disambiguated and point to their first corresponding synsets. Due to the inheritance property, company is an object of any hypernyms of {take over, buy out}. One such hypernym, {buy, purchase, take} also meets the requirements of HR3: Heuristic Rule 3 (HR3) If a verb concept has another verb at the beginning of its gloss, then that verb describes the same action, but in a more specific context.</Paragraph>
    <Paragraph position="9"> Therefore, acquire is a definition of {buy, purchase, take}, that has company as an object and involves a financial transaction. These three heuristics operate throughout all the sequences of the class comprising &lt; acquisilion of company &gt;, &lt; addition of business &gt;, &lt; formalion of group &gt; or &lt; beginning of service &gt; We conclude that for this class of prepositional relations, noun2 is the object of the action described by noun1.</Paragraph>
  </Section>
  <Section position="5" start_page="361" end_page="361" type="metho">
    <SectionTitle>
4 A case study
</SectionTitle>
    <Paragraph position="0"> Table 1 illustrates the semantic relations observed in WordNet for some of the classes of prepositional relations with preposition of, when both arguments are nouns. We applied a number of 28 heuristics on 45 disambiguated classes.</Paragraph>
  </Section>
class="xml-element"></Paper>