Knowledge-Tracing-Driven Exercise Recommendation¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The survey's first application area uses traced knowledge states to choose learning resources automatically. "Given the inferred knowledge states, one common strategy is selecting the next exercise that will best advance students’ knowledge acquisition." Examples include the BKT-sequence algorithm, which returns the exercise whose predicted score is closest to the expected score, a DKVMN extension recommending exercises in SPOCs, and a BKT video model for evaluating video resources.

Design Implications¶

Context¶

Requirements¶

• A KT model estimating each student's current knowledge state per knowledge concept
• Predicted scores for candidate exercises, matched against an expected score that depends on the current knowledge state

Constraints¶

• The survey says existing solutions mainly assign non-mastered exercises, which it calls reasonable but too broad to advance learning effectively

Target Learners¶

• Students in intelligent tutoring systems and Small Private Online Courses

Target Learning Goals¶

• Knowledge acquisition and learning efficiency through appropriately difficult exercises

Related Strategies¶

• Adaptive Difficulty
• Knowledge Tracing Learner Modeling Task

Examples¶

-

Key Sources¶

• Shuanghong Shen, Qi Liu, Zhenya Huang, Yonghe Zheng, Minghao Yin, Minjuan Wang, and Enhong Chen. (2021). A Survey of Knowledge Tracing: Models, Variants, and Applications. https://arxiv.org/abs/2105.15106