Predicting Student Actions in a Procedural Training Environment

TL;DR

This paper introduces a collective student model derived from log data to predict student actions in a virtual lab, enhancing personalized tutoring feedback in an intelligent tutoring system.

Contribution

It presents a novel collective student modeling approach using clustered logs and extended automata for improved action prediction in procedural training environments.

Findings

Model provides reasonably good predictions.

Supports more personalized tutoring feedback.

Validated with student logs from a 3D virtual biotechnology lab.

Abstract

Data mining is known to have a potential for predicting user performance. However, there are few studies that explore its potential for predicting student behavior in a procedural training environment. This paper presents a collective student model, which is built from past student logs. These logs are firstly grouped into clusters. Then an extended automaton is created for each cluster based on the sequences of events found in the cluster logs. The main objective of this model is to predict the actions of new students for improving the tutoring feedback provided by an intelligent tutoring system. The proposed model has been validated using student logs collected in a 3D virtual laboratory for teaching biotechnology. As a result of this validation, we concluded that the model can provide reasonably good predictions and can support tutoring feedback that is better adapted to each student type.