<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-1127">
  <Title>Novel Association Measures Using Web Search with Double Checking</Title>
  <Section position="10" start_page="1015" end_page="1015" type="concl">
    <SectionTitle>
6 Concluding Remarks
</SectionTitle>
    <Paragraph position="0"> This paper introduces five novel association measures based on web search with double checking (WSDC) model. In the experiments on association of common words, Co-Occurrence Double Check (CODC) measure competes with the model trained from WordNet. In the experiments on the association of named entities, which is hard to deal with using WordNet, WSDC model demonstrates its usefulness. The strategies of direct association, association matrix, and scalar association matrix detect the link between two named entities. The experiments verify that the double-check frequencies are reliable. null Further study on named entity clustering shows that the five measures - say, VariantDice, VariantOverlap, ariantJaccard, VariantCosine and CODC, are quite useful. In particular, CODC is very stable on word-word and name-name experiments. Finally, WSDC model is used to expand community chains for a specific personal name, and CODC measures the association of community member and the personal name. The application on personal name disambiguation shows that 9.65% and 14.22% increase compared to the system without community expansion. null</Paragraph>
  </Section>
class="xml-element"></Paper>