The contributions of the paper are summarized as follows.

• We propose a knowledge tracing model that extends the DKT model to consider both a learning sequence and the forgetting behavior by explicitly modeling the forgetting behavior using multiple features.
• We have conducted experiments showing that our proposed model outperforms conventional methods in terms of the predictive performance on the knowledge tracing datasets.
• We have also examined how the combination of multiple types of forgetting information influences the performance.