With an increase in the number of massive open online course (MOOC), the amount of learning educational big data has also increased. Artificial intelligence technology is now applicable to big data analysis. This study implements a clustering system based on learning educational big data. This study uses the MOOC course offered at a university in the north of Taiwan as experimental data. Meanwhile, observing the ratio of the number of students who watched videos to that of students who finished practice exercises, we clustered students into different groups by using a K-means clustering module. Then, we use the deep learning prediction module to determine whether or not the students will change their clustering result next week. This system aims to provide teachers and students clustering results for the next week to recommend suitable learning strategies, which can facilitate the most appropriate guidance and more adaptive counseling.