[Paper Review] A Survey of Knowledge Tracing: Models, Variants, and Applications

Shuanghong Shen, Qi Liu+5 more|arXiv (Cornell University)|May 6, 2021

Intelligent Tutoring Systems and Adaptive Learning168 references73 citations

TL;DR

This survey comprehensively categorizes knowledge tracing (KT) models into probabilistic, logistic, and deep learning-based approaches; it reviews model variants, releases open-source KT libraries EduData and EduKTM, and discusses KT applications and future directions.

ABSTRACT

Modern online education has the capacity to provide intelligent educational services by automatically analyzing substantial amounts of student behavioral data. Knowledge Tracing (KT) is one of the fundamental tasks for student behavioral data analysis, aiming to monitor students' evolving knowledge state during their problem-solving process. In recent years, a substantial number of studies have concentrated on this rapidly growing field, significantly contributing to its advancements. In this survey, we will conduct a thorough investigation of these progressions. Firstly, we present three types of fundamental KT models with distinct technical routes. Subsequently, we review extensive variants of the fundamental KT models that consider more stringent learning assumptions. Moreover, the development of KT cannot be separated from its applications, thereby we present typical KT applications in various scenarios. To facilitate the work of researchers and practitioners in this field, we have developed two open-source algorithm libraries: EduData that enables the download and preprocessing of KT-related datasets, and EduKTM that provides an extensible and unified implementation of existing mainstream KT models. Finally, we discuss potential directions for future research in this rapidly growing field. We hope that the current survey will assist both researchers and practitioners in fostering the development of KT, thereby benefiting a broader range of students.

Motivation & Objective

• Motivate and define Knowledge Tracing (KT) and its importance for online education.
• Provide a systematic taxonomy of KT models (probabilistic, logistic, deep learning-based).
• Review variants that model learning before, during, and after the learning process.
• Highlight available datasets and open-source libraries for KT (EduData, EduKTM).
• Discuss KT applications across educational scenarios and outline future research directions.

Proposed method

• Classify KT models into three categories: probabilistic, logistic, and deep learning-based.
• Summarize foundational KT models (e.g., Bayesian Knowledge Tracing, Dynamic Bayesian Knowledge Tracing, LFA, PFA, KTM) and their technical characteristics.
• Detail deep learning KT families (DKT, memory-aware KT, exercise-aware KT, attentive KT, graph-based KT) and representative mechanisms (RNN/LSTM, memory networks, attention, transformers).
• Review variants addressing pre-learning, during-learning, and post-learning phases with richer side information and cognitive assumptions.
• Provide open-source resources (EduData, EduKTM) and summarize KT applications and future directions.

Experimental results

Research questions

• RQ1 What are the core KT models and their allocations across probabilistic, logistic, and deep learning paradigms?
• RQ2 How do KT variants capture before/during/after learning dynamics and side information?
• RQ3 What datasets and open-source tools exist to standardize KT evaluation and implementation?
• RQ4 What are the key applications of KT in diverse educational settings and what future directions are most promising?

Key findings

• The field is organized into three model families: probabilistic, logistic, and deep learning-based KT.
• A wide range of KT variants extends basic models to capture complete learning processes and side information.
• Two open-source libraries (EduData and EduKTM) provide datasets and unified implementations to support KT research.
• KT applications span adaptive learning, resource recommendation, and educational gaming across scenarios.
• The survey outlines potential future research directions to advance KT methods and applicability.