Wse-MF: A weighting-based student exercise matrix factorization model

•Exercise prediction is an important pattern recognition task which can be solved by matrix factorization.•We incorporate a Student-Exercise Weighting Strategy into the loss function and apply it to the predicted data.•We apply a new optimization called SE-ALS so that we can train quickly and with low time complexity.•We explain the link between the latent factor K and the relationship between skills and exercises within a dataset.