Student essays representing their individual reflections on a collaborative web-based course in International Business are computationally analyzed according to a classification scheme based on a set of a priori fuzzy categories. This classification method enables the identification of themes and trends in the student responses that can be used to illustrate an overall evaluation of the personal learning experiences for this course. By processing the classification results using a computational neural network, we can depict the clustering intensity of thematic elements and illustrate the strength of dependencies between classification attribute values topologically using a self-organizing map (SOM), which provides a pattern recognition visualization. The resulting SOM can then be used to compare successive depictions for future iterations of new thematic data from student self-evaluations.
Sallis, Philip J. and Masi, Catherine A., "A Rule-Based Method for Determining the Degree of Student Satisfaction of a Web-Based Learning Environment" (2000). AMCIS 2000 Proceedings. 3.