<?xml version="1.0" standalone="yes"?>
<Paper uid="W03-2008">
  <Title>Natural Language Analysis of Patent Claims</Title>
  <Section position="6" start_page="1" end_page="1" type="concl">
    <SectionTitle>
5 Status and Future Work
</SectionTitle>
    <Paragraph position="0"> The analyzer is in the late stages of implementation as of May 2003. The static knowledge sources have been compiled for the domain of patents about apparatuses. The morphological analysis and syntactic chunking are operational and well tested.</Paragraph>
    <Paragraph position="1"> The case-role dependency detection is being currently tested and updated. The compatibility of the analyzer and fully operational generator has been proved and tested. First experiments have been done to use the analyzer for such applications as machine translation and improving claim readability. We have not yet made a large-scale evaluation of our analysis module. This leaves the comparison between other parsers and our approach as a future work. The preliminary results show a reasonably small number of failures, mainly due to the incompleteness of analysis rules that are being improved and augmented with larger involvement of predicate knowledge.</Paragraph>
    <Paragraph position="2"> We intend to a) add an optional interactive module to the analyzer (that would allow for human interference into the process of analysis to improve its quality), and complete the integration of the analyzer into a machine translation system and an application for improving claim readability.</Paragraph>
    <Paragraph position="3"> Another direction of work is developing applications in a variety of languages (software localization); b) develop a patent search and extraction facility on the basis of the patent sublanguage and our parsing strategy.</Paragraph>
  </Section>
class="xml-element"></Paper>