<?xml version="1.0" standalone="yes"?>
<Paper uid="C94-2189">
  <Title>ROBUST METHOD OF PRONOUN RESOLUTION USING FULL-TEXT INFORMATION</Title>
  <Section position="5" start_page="7158" end_page="7158" type="evalu">
    <SectionTitle>
4 Results
</SectionTitle>
    <Paragraph position="0"> We. examined 112 third-person pronouns in 1904 consecutive sentences fl'om eight chapters of two different computer manuals. One \[1\] is a typical computer manual for computer experts such as programmers and system operators, and the other \[2\] is a primer for novice users of a computer.</Paragraph>
    <Paragraph position="1"> In this experiment, we excluded instances in which it pronmninalizes a sentence, as in do it, those in which it refers to a syntactically recoverable that-clmlse or to-infinitive clause, and those in which ~In the sentences contained in Figure 1, the underlining and tile change of font for the target pronoun it were done by tile attthor.</Paragraph>
    <Paragraph position="2"> 2Noun phrases with the same lemma referred to in the preceding sentences are indicated by underlines.</Paragraph>
    <Paragraph position="3">  All four of the cursor movement keys are typematic; they kee t) repeating as long ~s they are held down. The Cursor Up key moves the cnrsor up one line.</Paragraph>
    <Paragraph position="4"> Like the other cnrsor movement keys, this key moves the cursor one line or many lines depending on how long you hold down the key.</Paragraph>
    <Paragraph position="5"> The Cursor Right key moves the cursor to the right.</Paragraph>
    <Paragraph position="6"> Hoht the key down.</Paragraph>
    <Paragraph position="7"> When the cursor reaches the right end of the line, it goes off the screen and reappears on the left side, one line bdow the line it was on. If the cursor is on the bottom line of the screen and is run M1 the way to the right, it goes off the screen and reappears in the upper left corner. The Cursor Left key mow's the cursor one position to the left.  it occurs in a time or weather construction. When an identical pronoun is included in the candidate list, the system assmnes that these pronouns share tile same antecedent. For example, we assumed that M1 the instances of it in When it reaches the left end of the line, it goes off the screen and reappears on the right, one line above the line it was on have the same antecedent.</Paragraph>
    <Paragraph position="8"> As a result of our strategy of enlarging the scope for selection of cmtdidates, tile average nuinber of candidate noun phrases was 4.1.</Paragraph>
    <Paragraph position="9"> Our algorithm chose a correct antecedent in 105 (:a.ses, giving a success ratio of 93.8%. In 28 of those 112 cases, there was among the candidates an identical t)ronoun that; referred to the same antecedent; thus, in 84 cases, antecedents were selected by evaluating the syntactic position, frequency of repetition, and collocation pattern of each candidate noun phrase.</Paragraph>
    <Paragraph position="10"> As shown in Table 1, without any information on repetition or collocation patterns, the success rate of selection based only on syntactic position was 82.7%, while tile success rate for selection based only on frequency of repetition was 60.7%. This result indicates that pronominMized noun phrases were actually repeated more than twice within ten consecutive sentences in over 60% of the cases. Thus, preference according to the frequency of repetition contributed to the selection of the correct antecedent. In 16 of the 22 cases in which this information preferred a wrong candidate noun phrase, the preference value was overridden by the negative preference value caused by a syntactic position far from tile sentence of the pronoun, or by a larger preference assigned to some other candidate with an identicM collocation pattern.</Paragraph>
    <Paragraph position="11"> Identical collocation patterns were found within the same chapter in 22 of 84 cases in which the preference value was ev',duated in selecting all antecedent. Although this is only 26.2% of the cases, collocation preference (lid not support any wrong candidates. Moreover, in 50.0% of the 22 cases, another preference value, either syntactic position or repetition, supported a wrong candidate. Therefore, preference according to collocation pattern contributed to the selection of the correct antecedent.</Paragraph>
    <Paragraph position="12"> Table 2 shows the distances and directions of sell- null tences in which a collocation pattern supporting a candidate noun phrase to modify the modifiee of tile pronoun was found. The results indicate that such information was extracted from a relatively small area of a text. In addition, relative collocation patterns were extracted from both previous and following sentences. null Out of the 37 cases in which the identical collocation patterns were found, synonym relations were used in seven cases (18.9%).</Paragraph>
  </Section>
class="xml-element"></Paper>