<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1102">
  <Title>Detecting Transliterated Orthographic Variants via Two Similarity Metrics</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We propose a method to detect transliterated orthographic variants. The method is marked by the use of string similarity and contextual similarity via contextual vectors. The method achieved a 0.889 F-measure in an open test. The results showed that detection of short word variants is very dif cult, and a dictionary raised the precision for such short words. However, contextual similarity did not contribute as expected to the detection of orthographic variants.</Paragraph>
  </Section>
class="xml-element"></Paper>