<?xml version="1.0" standalone="yes"?>
<Paper uid="E85-1036">
  <Title>A RULE-BASED APPROACH TO EVALUATING IMPORTANCE IN DESCRIPTIVE TEXTS</Title>
  <Section position="11" start_page="248" end_page="249" type="concl">
    <SectionTitle>
5. CONCLUSION
</SectionTitle>
    <Paragraph position="0"> The importance evaluator described in the previous sections is written in Franz Lisp and it is presently running in a prototype version on a SUN-2 workstation. Much experimental work is currently ongoing on this prototype in order to assess its operation, enlarge its knowledge base, and test its performance with a sufficiently large set of sample texts.</Paragraph>
    <Paragraph position="1"> The major contribution of the work reported in the paper can be found in the novel proposed approach to importance evaluation that, according to the results so far achieved, proved to be viable and appropriate both from the cognitive and the computational points of view.</Paragraph>
    <Paragraph position="2">  The research has disclosed several new directions for future work. Among these we mention: extending the importance rule base to cover the rhetoric and stylistic aspects of the text; introducing meta-rules to deal with the problems of rule activation scheduling, and of conflict resolution among rules; improving the goal matching techniques in order to implement a flexible mechanism for interpreting the content of encyclopedia frames according to the current goal; giving the evaluator the capability of changing the goal during the evaluation process, depending on the content of the processed text.</Paragraph>
  </Section>
class="xml-element"></Paper>