Personalized Recommendation of Study Materials Based on Automatic Short Answer Scoring Results

The results of automatic short answer scoring (ASAS) can be used to provide learning recommendations. However, previous studies on ASAS are focused on the quality of the scoring method, without considering giving feedback on the student performance in the exam sessions. Moreover, research on feedback in the learning activity is more concentrated on creating an adaptive learning process. Therefore, this study aims to combine the Automatic Short Answer Scoring method and an Automatic Recommendation System to create an adaptive feedback system that has the ability to personalize the recommendation of study materials for the student based on their exam result. This study uses a dataset of ten questions, including one reference answer per question. The experiment result shows the recommendation system gives the maximum similarity score at 0.8972, with the maximum average score for all questions is 0.801. Future studies may continue to examine a technique for automatically generating study materials from various resources.