A good supervisor fit is specific. If your project is about machine learning for medical imaging, the supervisor should have relevant work in medical imaging, AI, health data or a closely related field. If your project is about education policy, the supervisor should work in education research, policy, equity, curriculum, assessment, learning systems or a connected area.
Fit also includes methods. A supervisor who studies the same topic with completely different methods may still be useful, but you need to understand the difference. If you plan qualitative interviews, a supervisor with qualitative research expertise may be important. If you plan statistical modelling, data science, lab work or creative practice research, the required expertise changes.
For scholarship purposes, the proposal should look feasible at Deakin. That means Deakin should have supervision expertise, research environment, facilities, data access or intellectual community that can support the project.