<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-2111">
  <Title>Concordances of Snippets</Title>
  <Section position="5" start_page="9" end_page="9" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> Instead of collecting ad-hoc corpora from each web-page retrieved by Google Lexware Culler builds concordances of snippets. Thanks to this limitation it is possible to look up words and phrases on the web in an instant. The quality of concordances built of snippets varies from excellent to poor dependent mainly on a search term but the goal of getting a quick glimpse of language use on the web is clearly attainable with snippets. Snippets are sufficiently long to provide disambiguating contexts. The ranking system of web search engines gives preference to the most popular web-pages, hence the prevalent language use can be expected in majority of excerpts. At the same time it is also true that extensive filtering is required in order to make accceptable concordance lines of snippets.</Paragraph>
  </Section>
class="xml-element"></Paper>