Generating Examples for Use in Tutorial Explanations: The Use of a SubsumptionBased Classifier

摘要

Examples form an integral and very important part of many descriptions,especially in contexts such as tutoring and documentation generation. Previous computational work on example generation has focused on the issue of generating valid examples in different domains. However, there are a large number of examples that can be generated for a given concept, and it is important that examples intended for tutoring situations illustrate specific features that need to be communicated. There is also a strong interaction between the examples and the accompanying textual description. The number of examples to be presented, the order in which they are to be presented, and the position of the examples with respect to the text are all dependent upon the examples generated for use in the presentation. It is therefore important that the example generator be able to generate a set of suitable examples in response to a tutoring goal in a given situation. In this paper, we present one framework, based on the use of a subsumption classifier, that facilitates the generation of appropriate tutorial examples. We illustrate the working of this framework by describing the generation of examples to illustrate the syntax in the programming language LISP. Tutoring systems, Text generation, Natural Language Front-Ends.