Applying Recent Innovations from NLP to MOOC Student Course Trajectory Modeling

TL;DR

This paper explores the application of NLP innovations like LSTM regularization and Transformer models to improve neural network-based predictions of student course trajectories in MOOCs.

Contribution

It introduces the adaptation of NLP techniques such as LSTM regularization and Transformers to MOOC student trajectory modeling, enhancing predictive accuracy.

Findings

LSTM with regularization improves prediction accuracy.

Transformers outperform traditional LSTM models.

NLP techniques are effective in educational data modeling.

Abstract

This paper presents several strategies that can improve neural network-based predictive methods for MOOC student course trajectory modeling, applying multiple ideas previously applied to tackle NLP (Natural Language Processing) tasks. In particular, this paper investigates LSTM networks enhanced with two forms of regularization, along with the more recently introduced Transformer architecture.