<?xml version="1.0" standalone="yes"?>
<Paper uid="C80-1002">
  <Title>AUTOMATIC PROCESSING OF WRITTEN FRENCH LANGUAGE</Title>
  <Section position="9" start_page="0" end_page="13" type="concl">
    <SectionTitle>
6. Conclusions
</SectionTitle>
    <Paragraph position="0"> We have presented a parsing system for French language sentences. This system is characterized by - its generality, - a good amount of flexibility, due to the fact that the system is semantically oriented, - its capability to cope with complex structures, including subordinate clauses, conjunctions and references (these last two features were not discussed in the paper).</Paragraph>
    <Paragraph position="1"> The system was designed independently of any specific application and was tested on a limited corpus (approximatively one hundred) of common french sentences. We believe that this system is applicable in all domains, provided a structured semantic dictionary. An application of our system to the medical domain is currently under development 2 6. In this domain, existing automatic processing systems do function successfully for E~glish and French pathology data 4 5 I0 12. We took up the challenge to design a system for processing patient discharge summaries in internal medicine. It is quite true that the natural language of internal medicine data does not always contain well formed sentences.</Paragraph>
    <Paragraph position="2"> This is not however a real problem. One of the main advantages of the system stems from the fact that it was designed to handle French free text, no matter how academically correct or incorrect it might be. Consequently, we expect that the system can handle all kinds of medical  diagnoses. Moreover, since it has been conceived to process questions as well as complete sentences, the system is not limited to the processing of medical data. Ultimately, it will be used to implement a complete natural language interface for a data base management system. null</Paragraph>
  </Section>
class="xml-element"></Paper>