ORCID
https://orcid.org/0009-0008-3913-7836
Date of Award
Summer 2024
Language
English
Embargo Period
7-23-2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
College/School/Department
Department of Computer Science
Program
Computer Science
First Advisor
Sherry Sahebi
Committee Members
Mei-Hwa Chen, Ming-Ching Chang, Hassan Khosravi
Keywords
Educational data mining, Student Knowledge tracing, Student behavior Modeling, Transfer Learning, Multi-task Learning
Subject Categories
Artificial Intelligence and Robotics | Other Computer Sciences
Abstract
Online education systems have grown in popularity over the past few years, providing abundant opportunities for students to learn. As the number of students using these systems grows, it promotes the development of the Educational Data Mining (EDM) field, which leverages statistical, machine learning, and data mining technologies to explore large-scale educational data and develop methods to better understand student learning.
In this dissertation, we investigate two essential topics in EDM: Student Knowledge Tracing (KT) and Behavior Modeling (BM). KT aims to quantify and model student knowledge gained from learning activities, while BM focuses on tasks such as modeling student choices and preferences for future learning materials. Accurately tracing student knowledge and modeling student behavior can help us better understand students learning. This understanding can be applied to recommend useful learning materials to students, detect knowledge gaps, and better plan study schedules, to improve learning efficiency.
Online education systems provide students with access to diverse types of learning materials, such as video lectures, textbooks, and interactive questions. Students interact with and learn from these materials in different ways, leading to various types of learning activities (multi-activity). Additionally, in many online education systems, students can follow their preferences when choosing learning materials to study rather than following a predefined sequence set by a teacher or instructor, leading to varied behavior patterns. Every type of learning activity and each one individually can contribute to the student learning process, affecting both their knowledge and behavior. We argue that student knowledge and student behavior, or the choice of learning materials can be associated with each other. Therefore, it is essential to understand student knowledge and behavior throughout the learning trajectory involving multiple types of learning activities. Thus, the goals of this dissertation include studying student knowledge and behavior, as well as their association, and developing and improving student knowledge and behavior modeling when students interact with multiple types of learning materials.
We treat student learning activities involving multiple types of learning materials as multi-activity learning sequences and suggest adopting transfer learning to address these multi-activity sequence modeling problems. In this dissertation, we aim to investigate the following tasks: (1) Student knowledge acquisition from multiple learning material types. (2) Knowledge transfer modeling between different learning material types. (3) Simultaneous knowledge and behavior modeling.
We hypothesize that modeling student learning from different types of activities and behavior related to learning material choices can effectively improve the understanding of student knowledge and behavior.
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Zhao, Siqian, "Multi-activity Student Knowledge and Behavior Modeling via Transfer Learning" (2024). Electronic Theses & Dissertations (2024 - present). 24.
https://scholarsarchive.library.albany.edu/etd/24