Recommendation of reviewers based on text analysis and machine learning : part a

• Main Author: Liang, Ce
• Other Authors: Lihui CHEN
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/141176
• Institution: Nanyang Technological University

When reviewing postgraduate research applications, it is very crucial for reviewers with relevant knowledge to determine applicants’ performance. At Nanyang Technological University, four examiners are paired with one applicant for background review. However, due to the variety of project scopes, the amount of applications and the high request of expertise, selecting proper reviewers can be very challenging and time-consuming. Thus, this project aims to build an examiner recommendation system to automate current manual matching process. This project approaches the goal through two methods: 1) Feature-based matching 2) Profile-based matching. In feature-based matching, similarity of features from students are calculated for pair-up. And in profile-based matching, a prediction model is set up to look for the most suitable examiners. Resumes from already paired students and reviewers are collected as raw data. And the accuracy of each method is calculated by running test cases. The feature-based matching is simple but less accurate. And the profile-based matching is generally more accurate but complicated. Through analyzing the test results, several solutions to optimize each method are proposed. This project indicates that the possibility of using recommendation system for application reviewer selection.