Prediction Learning Style Based on prior Knowledge for Personalized Learning

Abstract

The current online learning model still lacks the personalization that is able to accommodate the needs of learners. One of the learning needs is that the learning model should be in accordance with the learning style. Currently the automated detection of learning style have two approaches, namely data driven and literature based. Both approaches focus on retrieving data from the interaction of learners with the system. The interaction of the learners with the system becomes the initial data to determine the learning style. However, the learner interaction approach to this system only has an accuracy value below 80 %. One of the causes of lack of accuracy in learning style detection is that it does not take into account the possession of prior knowledge of each learner. Prior knowledge is the knowledge and skills of the learners who are expected to have the basis of determining learning styles. This is because prior knowledge has four levels: Knowledge of Fact, Knowledge of Meaning, Integration of Knowledge and Application of Knowledge that are closely related to learning style. This study proposes a smart learning model using prior knowledge to detect learning styles. There are three stages used in this research: 1. Generating prior knowledge by using the assessment with Weight Cosine Coefficient (WCC) algorithm 2. Measuring the prior knowledge of each learner, where the result of this measurement is named Level of Knowledge (LOK) and the last is prediction of learning style possessed by mapping.