[Paper Review] A Simple Recommender Engine for Matching Final-Year Project Student with Supervisor

Mohammad Hafiz Ismail, Tajul Rosli Razak+2 more|arXiv (Cornell University)|Aug 8, 2019

Educational Technology and Assessment3 references8 citations

TL;DR

This paper proposes a simple recommender engine that matches final-year project students with supervisors using the Euclidean distance algorithm on questionnaire-derived interest data. Implemented in Java, the system demonstrates feasibility through initial testing, showing promise for solving the challenge of student-supervisor matching in academic projects.

ABSTRACT

This paper discusses a simple recommender engine, which can match final year project student based on their interests with potential supervisors. The recommender engine is constructed based on Euclidean distance algorithm. The initial input data for the recommender system is obtained by distributing questionnaire to final year students and recording their response in CSV format. The recommender engine is implemented using Java class and application, and result of the initial tests has shown promises that the project is feasible to be pursued as it has the potential of solving the problem of final year students in finding their potential supervisors.

Motivation & Objective

• To address the challenge of final-year students struggling to find suitable project supervisors.
• To develop a lightweight, practical solution for student-supervisor matching in academic settings.
• To evaluate the feasibility of a recommender system using basic similarity metrics like Euclidean distance.
• To provide a scalable and implementable tool for academic institutions to improve project supervision matching.

Proposed method

• Collect student interests via a structured questionnaire distributed to final-year students.
• Store responses in CSV format for structured data processing.
• Implement a recommender engine using the Euclidean distance algorithm to compute similarity between student interests and supervisor expertise.
• Use Java as the programming language to build and deploy the application.
• Rank potential supervisor-student matches based on minimum Euclidean distance scores.
• Validate the system through initial testing to assess feasibility and performance.

Experimental results

Research questions

• RQ1 Can a simple recommender system effectively match final-year project students with suitable supervisors based on interest alignment?
• RQ2 How well does the Euclidean distance algorithm perform in measuring similarity between student and supervisor interest profiles?
• RQ3 What is the feasibility of implementing a lightweight, non-complex system for student-supervisor matching in academic institutions?
• RQ4 Can questionnaire-based interest data be effectively used to generate reliable recommendations?

Key findings

• The recommender engine successfully computes similarity scores between student interests and supervisor profiles using the Euclidean distance algorithm.
• Initial testing results indicate the system is feasible for implementation and shows promise in solving the student-supervisor matching problem.
• The use of structured questionnaire data in CSV format enables efficient data processing and integration into the recommendation pipeline.
• The system is implemented in Java, demonstrating practicality and ease of deployment in academic environments.
• The approach is lightweight and does not require complex machine learning models, making it accessible for small to medium-sized institutions.
• The study confirms that basic similarity metrics can yield meaningful recommendations in academic supervision matching contexts.