The development of a knowledge-based system supporting the diagnosis of reading and spelling problems

摘要

In this paper we argue that the quality of decision making in psychoeducational practice is likely to profit considerably from directives derived from a so-called normative diagnostic-prognostic framework, and that a computerized decision-support system is a promising tool for facilitation of the actual use of such directives. Accordingly, we describe the design and function of a knowledge-based system that is intended to support the initial stages of decision making in diagnosis of reading and spelling problems: (a) specifying the child's task performance in concrete behavioral terms, (b) identifying behavioral syndromes in these learning behaviors, and (c) generating candidate explanations, or diagnostic hypotheses, for these behaviors and syndromes. The systematic and formal foundation of the description of learning-behavior problems required for accomplishing these stages is provided by the facet definition approach. We present a mapping sentence that serves to define reading and spelling problems and that at the same time is used as a vehicle for communication between user and system. In order to show the latter function of the facet definition in the user interface of the system, a prototypical implementation of the system is described. Finally, some problems in constructing the required knowledge bases are discussed.