<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1196">
  <Title>Understanding Students' Explanations in Geometry Tutoring</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
5 Conclusions
</SectionTitle>
    <Paragraph position="0"> We present a natural language understanding system that combines unification-based syntactic processing with logic-based semantics. The system is used in conjunction with a Geometry Cognitive Tutor to help students better understand geometry. The approach we take allows for an elegant solution to the problem of determining equivalence between various ways to express the same meaning. Study results show that the system works reasonably well on classifying students' explanations on a grid of about 200 fine-grained categories, although there is space for further improvement. One particular problem is robustness in the face of ungrammaticality. Also the question of whether natural language explanations improve students' understanding of geometry still waits for a definitive answer.</Paragraph>
  </Section>
class="xml-element"></Paper>