Development and Evaluation of a Large-Scale Agent-Based System for Information Literacy Education - Improving the Automatic Collection of Learning Results through Template Matching

Freeware learn-to-touch-type software, which is available for download without charge, caters for individual learning style but, in many cases, is not equipped with the features to manage the entire learning of the students. Using this type of training software for individual learning styles, our research developed a learning management system for the automatic collection of learning results in a large-scale training session. This system uses template matching to convert the image data on the achievement results screen into text data, and stores the results in a database. Our system was applied in teaching 640 students, and learning was shown clearly to be effective by including a comparison with another faculty (764 students) that did not use the system. In addition, an improvement to the image processing for the automatic collection of learning results was developed.