<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-0833">
  <Title>Simple Features for Statistical Word Sense Disambiguation</Title>
  <Section position="3" start_page="0" end_page="0" type="relat">
    <SectionTitle>
2 Related Work
</SectionTitle>
    <Paragraph position="0"> In 1950, Kaplan carried out one of the earliest WSD experiments and showed that the accuracy of sense resolution does not improve when more than four words around the target are considered (Ide and V eronis, 1998). While researchers such as Masterman (1961), Gougenheim and Michea (1961), agree with this observation (Ide and V eronis, 1998), our results demonstrate that this does not generally apply to all words. A large context window provides domain information which increases the accuracy for some target words such as bank.n, but not others like di erent.a or use.v (see Section 3). This con rms Mihalcea's observations (Mihalcea, 2002). In our system we allow a larger context window size and for most of the words such context window is selected by the system.</Paragraph>
    <Paragraph position="1"> Another trend consists in de ning and using semantic preferences for the target word.</Paragraph>
    <Paragraph position="2"> For example, the verb drink prefers an animate subject in its imbibe sense. Boguraev shows that this does not work for polysemous verbs because of metaphoric expressions (Ide and V eronis, 1998).</Paragraph>
    <Paragraph position="3"> Furthermore, the grammatical structures the target word takes part in can be used as a distinguishing tool: \the word 'keep', can be disambiguated by determining whether its object is gerund (He kept eating), adjectival phrase (He kept calm), or noun phrase (He kept a record)&amp;quot; (Rei er, 1955). In our second system we approximate the syntactic structures of a word, in its di erent senses.</Paragraph>
    <Paragraph position="4"> Mooney (Mooney, 1996) has discussed the e ect of bias on inductive learning methods.</Paragraph>
    <Paragraph position="5"> In this work we also show sensitivity of Na ve Bayes to the distribution of samples.</Paragraph>
  </Section>
class="xml-element"></Paper>