Case-Based Reasoning and Educational Technology

Processing Using Case-Based Reasoning

The CBR model includes three aspects of cognition: cases, case indices, and case processor. Cases represent the interpretations of experience and lessons learned, which assist the individual to employ episodic memory in a meaningful way. Over time, individuals develop case libraries within memory. While experts are able to generate complex and nuanced case libraries from their experience, theorists such as Roger Schank and Janet Kolodner argue that case libraries can be generated vicariously through hearing about the experience of others. Case indices enable the reasoner to determine the appropriate case applicable to the context. Although the generation of indices is often nonconscious, indices serve as cues for retrieval of related memories. A strong index enables an individual to reach beyond surface similarities.

The third aspect of CBR is the case processor, which processes results from indices, retrieves relevant cases from memory, applies lessons learned, and reindexes older cases in light of new experiences. The quality of solution generated is based upon the individual’s previous experience, ability to interpret a new situation in terms of experience, adaptation of a previous solution to a new context, and adeptness at appraisal. Expertise in a subject area is therefore determined by the ability to index knowledge, retrieve relevant memories, and apply them effectively to solve problems.

Schank argued that learning involves goals, plans, expectations, expectation failures, and explanations. Goals help to generate initial indices for retrieval. As an individual develops goals to solve problems, outcome expectations are also developed. Expectation failure represents a situation where the expectations do not meet the goal requirements. This might appear as a negative, but expectation failures incite the individual to search for explanations and abstract lessons that can be applied to future problems. If the outcome of the plan does not meet goal expectations, previous successful plans are referenced using the index and adapted to accommodate the new goal. This is another opportunity to generate new indices that are associated with the experience. Expectation failure helps set the groundwork for future reasoning and the learner’s reindexing of experiences. Lastly, explanations allow the individual to derive meaning from his or her experiences, generate relevant indices, and aid in predictions of future cases. With experience, the individual is able to make generalizations (scripts).

...