The higher education sector has been growing at a significant rate demanding a variety of skills and knowledge from both sides: instructors and learners.
Thesis supervision is one of the main challenges as this activity combines research and teaching practices. To ensure the success of a student’s thesis project, it is vital to accommodate the student’s demands and expectations with the supervisor’s availability, experience and knowledge.
This article provides an automated process for supervisors’ allocation using a machine learning technique based on the current procedure adopted at the Engineering Institute of Technology (EIT), Perth, Australia.
The automated process has great potential considering that large numbers of thesis students require supervisors every semester in most institutions.
The key to achieve the most suitable student-supervisor match within a short timeframe is assessing certain key factors from both supervisors’ and students’ sides efficiently.
The DecisionTreeClassifier in Python is used for the training of a classification model, as human experience can be translated to a decision tree.
The methodology includes the quantifying of supervisor selection criteria, the cleaning of the data, the training and testing of the decision tree model.
A case study is conducted to demonstrate the application of the automated process and to validate the efficiency of the automatin