Deep Neural Network Based Student Response Modeling With Uncertainty, Multimodality and Attention
Year: 2020
Type: article
Abstract: In this thesis, I investigate deep neural network based student response modeling, more specifically Knowledge Tracing (KT). Knowledge Tracing allows Intelligent Tutoring Systems to infer which topics or skills a student has mastered, thus adjusting curriculum accordingly. Deep neural network based knowledge tracing models like Deep Knowledge Tracing (DKT) and Dynamic Key-Value Memory Network (DKVMN) have achieved significant improvements compared with conventional probabilistic models. There are mainly two goals in this thesis: 1) To have a better understanding of existing deep neural network based models and their predictions through visualization and through incorporating uncertainties. 2) To improve the performance of student response modeling with multimodality and attention mechanisms. In this thesis, I will first introduce the background and show why deep neural network based knowledge tracing models might have less depth than anticipated through visualization. Next, I propose a more practical way of alleviating the concerns of these deep models by incorporating uncertainty for each prediction. Then, I will discuss how adding more modalities and attention mechanisms might help improve model performance.
Author: Xinyi Ding
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W3188932427