Table 9 Future direction

Increase the sample space/training dataset (Invite more learners to participate)

7

Expand or integrate the solution to other areas and courses

6

Identify, extract or create more features/observable events for better prediction

4

Educational context/theory validation through experts

4

Leverage alternative solutions to enhance the prediction/clustering accuracy

4

Enhance learners’ engagement rate

2

More personalization

2

Better data quality for AI/ML algorithm training

2

Create a dashboard to support real time monitoring and decision-making

1

Add on additional functions to bring up more learners’ interest (IDC)

1

Solution generalization to suit for other tutoring systems

1

Close loop system enhancement based on AI explanation

1

Explore more aspects of emotions (like satisfaction, not just for dropout prediction)

1

Better survey response analysis methodology

1

Better intervention mechanism

1