<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1603">
  <Title>Preliminary Lexical Framework for English-Arabic Semantic Resource Construction</Title>
  <Section position="3" start_page="0" end_page="0" type="relat">
    <SectionTitle>
2 Related Work
</SectionTitle>
    <Paragraph position="0"/>
    <Section position="1" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.1 Arabic-English dictionary combination
</SectionTitle>
      <Paragraph position="0"> As pointed out previously, translation plays an important role in CLIR. Most of the CLIR systems participating in the (Arabic) Cross-Language Information Retrieval track1 at the Text REtrieval Conference (TREC)2 used a query translation dictionary-based approach where each source query term was looked up in the translation resource and replaced by all or a subset of the available translations to create the target query (Larkey, Ballesteros, and Connell, 2002), (Gey and Oard, 2001), (Oard and Gey, 2002). The four main sources of translation knowledge that have been applied to CLIR are ontologies, bilingual dictionaries, machine translation lexicons, and corpora.</Paragraph>
      <Paragraph position="1"> Research shows that combining translation resources increases CLIR performance (Larkey et al., 2002) Not only does this combination increase translation coverage, it also refines translation probability calculations. Chen and Gey used a combination of dictionaries for query translation and compared retrieval performance of this dictionary combination with machine translation (Chen and Gey, 2001). The dictionaries outperformed MT. Small bilingual dictionaries were created by Larkey and Connell (2001) for place names and also inverted an Arabic-English dictionary to English-Arabic. They found that using dictionaries that have multiple senses, 1 There have been two large scale Arabic information retrieval evaluations as part of TREC. These Arabic tracks took place in 2001, and 2002 and had approximately 10 participating teams each.</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>