Affordable Access

An Automated Implementation of Academic Staff Performance Evaluation System based on Rough Sets Theory

  • Ojokoh, Bolanle
  • Akinsulire, Victor
  • Isinkaye, Folasade O
Publication Date
Oct 14, 2019
Australasian Journal of Information Systems
External links


The essence of evaluating employees’ performance in any tertiary institution is to realize the goals of the institution by measuring the contribution of each employee. Effective human resource evaluation is paramount to the development of any organization. An automated method is needed to remove the limitations and facilitate the duties of human resource management. In this paper, rough set theory, a mathematical technique that deals with vagueness and uncertainty of imperfect data analysis is adopted for the evaluation of academic staff profile for promotion, grants and other academic purposes. The entire appraisal process of academic staff was translated into a web-based application where every user can fill, edit, update, and submit the annual performance evaluation report form. The indiscernible property of rough set approach is a unique factor in assessing every academic staff under the department and faculty/school by the head of department and dean respectively. With this, the system generates an information table handling all the necessary conditions for promoting academic staff and the corresponding decisions taken. A model for rating publications was proposed to reduce the sentiments involved in manual rating. Reports were generated as output of each evaluation procedure. One hundred (100) dataset of academic staff of the Federal University of Technology, Akure, Nigeria was used in the experiment to evaluate the performance of the system. The results of the system obtained score were compared with the institution standard and it was found that the system scores were above standard, the average precision of the system shows 60% effectiveness which showed that the proposed system is efficient for academic performance evaluation process.

Report this publication


Seen <100 times