Dynamic Key-Value Memory Networks for Knowledge Tracing

TLDR

Knowledge Tracing tracks students’ evolving mastery of concepts during learning sequences, but existing methods like Bayesian and Deep Knowledge Tracing either treat each concept separately or cannot precisely identify which concepts a student knows or lacks. This study aims to develop a model that personalizes practice by accurately estimating each concept’s mastery level, thereby addressing limitations of prior KT methods. DKVMN uses a static key memory of concepts and a dynamic value memory that updates each concept’s mastery level, enabling joint modeling of concept relationships. Experiments demonstrate that DKVMN consistently outperforms leading KT models and can automatically uncover exercise concepts while visualizing students’ evolving knowledge states.

Abstract

Knowledge Tracing (KT) is a task of tracing evolving knowledge state of students with respect to one or more concepts as they engage in a sequence of learning activities. One important purpose of KT is to personalize the practice sequence to help students learn knowledge concepts efficiently. However, existing methods such as Bayesian Knowledge Tracing and Deep Knowledge Tracing either model knowledge state for each predefined concept separately or fail to pinpoint exactly which concepts a student is good at or unfamiliar with. To solve these problems, this work introduces a new model called Dynamic Key-Value Memory Networks (DKVMN) that can exploit the relationships between underlying concepts and directly output a student's mastery level of each concept. Unlike standard memory-augmented neural networks that facilitate a single memory matrix or two static memory matrices, our model has one static matrix called key, which stores the knowledge concepts and the other dynamic matrix called value, which stores and updates the mastery levels of corresponding concepts. Experiments show that our model consistently outperforms the state-of-the-art model in a range of KT datasets. Moreover, the DKVMN model can automatically discover underlying concepts of exercises typically performed by human annotations and depict the changing knowledge state of a student.