University of Michigan - Ann Arbor

- Introduction
- Related Work
- Preliminaries
- Proposed Approach
- Experiments
- Conclusion

Augmenting Knowledge Tracing by Considering Forgetting Behavior

- CAEs that use KT are very widespread, even more so BKT
- However, many students have forgetting which is typically not accounted for in most BKT models
- In this study, they focus on modeling and predicting a student’s
knowledge considering the forgetting behavior from large amounts
of learning logs to support providing adaptive learning materials in
CAE systems.

Augmenting Knowledge Tracing by Considering Forgetting Behavior (Introduction)

Augmenting Knowledge Tracing by Considering Forgetting Behavior (Preliminaries)

The proposed approach uses the Deep Knowledge Tracing (DKT) model as its base model because DKT is a deep neural network that can easily incorporate multiple input sources and capture nonlinear dynamics among them. The approach extends DKT so that it also accounts for student forgetting, adapting its prediction of student performance using three forgetting-related features: 1) repeated time gap: the elapsed time between the current interaction and the previous interaction on the same skill id; 2) sequence time gap: the elapsed time between the current interaction and the previous interaction in the sequence, regardless of skill id; and 3) past trial count: the number of times the student has previously answered questions with the same skill id.

Augmenting Knowledge Tracing by Considering Forgetting Behavior (Proposed Approach)

- two datasets: ASSISTments and slepemapy.cz
- five-fold cross validation in which each dataset was split based on students
- the combination of concatenation and multiplication (hybrid) achieved the best AUC on both the ASSISTments and slepemapy.cz datasets, with values of of 0.7309 and 0.8046, respectively
- more features used for modeling, the better the predictive performance

Augmenting Knowledge Tracing by Considering Forgetting Behavior (Experiments)

The research on augmenting knowledge tracing by considering forgetting behavior concludes with the following key findings:
- **DKT Extension:** Extending the Deep Knowledge Tracing (DKT) model to incorporate forgetting behavior is highly feasible and effective.
- **Superior Performance:** The extended model outperforms the standard DKT model in predicting student performance on two real-world datasets (ASSISTments and slepemapy.cz).
- **Feature Combination:** Combining multiple features that influence forgetting (such as repeated and sequence time gaps) yields the most significant improvements in predictive performance.

Learn Before

Related

Related