Graph-based Massive Open Online Course (MOOC) Dropout Prediction using Clickstream Data in Virtual Learning Environment
| Autor: | Masafumi Shingu, Masaru Todoriki, Arseny Tolmachev, Satoshi Nakashima, Ryo Ishizaki, Koji Maruhashi, Izumi Nitta |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
business.industry
Computer science Massive open online course Graph based Physics::Physics Education Machine learning computer.software_genre Field (computer science) ComputingMilieux_COMPUTERSANDEDUCATION Virtual learning environment Artificial intelligence business computer Dropout (neural networks) Clickstream Transformer (machine learning model) Interpretability |
| Zdroj: | ICCSE |
| Popis: | Although the field of massive open online course (MOOC) is expanding, it faces the challenge of high dropout rate. To ensure continued learning by students, it is important to conduct an analysis based on a dropout prediction model that utilizes student behavior history data. In this paper, we propose a dropout prediction model using graph-based machine learning involving graph-structured relationships between various actions taken by students. The dropout prediction model is constructed using a graph-based machine learning technique which is based on tensor decomposition and transformer approaches. The performance of the proposed model is comparable to that of graph convolutional networks. Furthermore, we consider the interpretability of the proposed model based on the examples of student behavior graphs. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|