Multimodal learning recommendation - Using adaptive neuron-fuzzy inference system for microlearning

Abstract

This paper presents a multimodal learning strategy based on learners’ preferences to improve adaptive learning processes. At present, the use of fragmented content is considered a more effective learning method for online learners. Access to an idea or news in various ways makes it easier for students to understand and retain information. However, this challenges learners’ pace of learning, because too much digital information can interfere with their goals. To predict learners’ preferences for what they are learning, this study first uses the neuro-fuzzy reasoning method to diagnose learners’ activities related to their styles in a microlearning learning environment. Then, based on the results, a recommendation model is developed to help learners participate in adaptive learning activities and digital contents. This study makes the tracking of online learning activities consistent with the multimodal learning mode. Our results show that the identification model divided the sample of 154 learners into 9 learning categories. Learning activities corresponding to learning preferences enables learners to obtain answers quickly through the recommendation mode. The study also demonstrates that the effectiveness of the learning mechanism can facilitate the teaching process, generate rich learning guidance, and help teachers design a better, well-structured course. Therefore, learners can easily achieve their learning goals when the recommended content is received.