Explanation-Based Knowledge Acquisition of Electronics.
Abstract
This is the final report in a project that examined how knowledge of practical electronics could be acquired from materials similar to that appearing in electronics training textbooks, from both an artificial intelligence perspective an an experimental psychology perspective. Practical electronics training materials present a series of basic circuits accompanied by an explanation of how the circuit performs the desired function. More complex circuits are then explained in terms of these basic circuits. This material thus presents schema knowledge for individual circuit types in the form of explanations of circuit behavior. Learning from such material would thus consist of first instantiating any applicable schemas, and then constructing a new schema based on the circuit structure and behavior described in the explanation. If the basic structure of the material is an effective approach to learning, learning about a new circuit should be easier if the relevant schemas are available than not. This result was obtained for both an artificial intelligence system that used standard explanation-based learning mechanisms and with human learners in a laboratory setting, but the benefits of already having the relevant schemas were not large in these materials. The close examination of learning in this domain, and the structure of knowledge, should be useful to future cognitive analyses of training in technical domains.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 30, 1992
- Accession Number
- ADA255069
Entities
People
- David Kieras
Organizations
- University of Michigan