<?xml version="1.0" standalone="yes"?>
<Paper uid="J88-3004">
  <Title>RECOGNIZING AND RESPONDING TO PLAN-ORIENTED MISCONCEPTIONS</Title>
  <Section position="19" start_page="0" end_page="0" type="relat">
    <SectionTitle>
7 RELATED WORK
</SectionTitle>
    <Paragraph position="0"> Two approaches have been used to detect and correct misconceptions. The first approach is used by many intelligent tutoring systems (Anderson, Boyle, and Yost 1985; Brown and Burton 1978; Burton 1982; Stevens, Collins, and Goldin 1982). These systems locate mistaken beliefs in a data base of error-explanation pairs and provide the associated explanation. A basic problem with this approach is that, because there is no information about the underlying causes of the errors, these systems can handle only those misconceptions known in advance.</Paragraph>
    <Paragraph position="1"> The other approach avoids the difficulty inherent in enumerating all possible misconceptions within a domain by using strategies that address an entire class of misconceptions. The user's misconception is classified according to the abstract reasoning error likely to have led to it. This approach shares many features with recognizing abstract thematic situations (such as irony) in narratives, where such situations are defined in terms of abstract planning errors made by the narrative characters (Dyer 1983; Dyer, Flowers, and Reeves 1987; Dolan and Dyer 1986). Once an appropriate strategy is found, it can be used to generate advice (in narratives, this advice may be in the form of adages). In advisory systems, this approach has been applied to both objectand plan-oriented misconceptions.</Paragraph>
  </Section>
class="xml-element"></Paper>