Research Article Open Access

Lecturer Performance Analysis using Multiple Classifiers

Hawraz Ahmad1 and Tarik Rashid1
  • 1 Salahaddin University-Erbil, Iraq
Journal of Computer Science
Volume 12 No. 5, 2016, 255-264

DOI: https://doi.org/10.3844/jcssp.2016.255.264

Submitted On: 26 May 2015 Published On: 13 June 2016

How to Cite: Ahmad, H. & Rashid, T. (2016). Lecturer Performance Analysis using Multiple Classifiers. Journal of Computer Science, 12(5), 255-264. https://doi.org/10.3844/jcssp.2016.255.264

Abstract

Lecturer performance analysis has enormous influence on the educational life of lecturers in universities. The existing system in universities in Kurdistan-Iraq is conducted conventionally, what is more, the evaluation process of performance analysis of lecturers is assessed by the managers at various branches at the university andin view of that, in some cases the outcomes of this process cause a low level of endorsement among staffs who believe that most of these cases are opinionated. This paper suggests a smart and an activesystem in which both unique and multiple soft computing classifier techniques are used to examine performance analysis of lecturers of college of engineering at Salahaddin University-Erbil (SUE). The dataset collected from the quality assurancedepartment at SUE. The dataset composes of three sub-datasets namely: Student Feedback (FB), Continuous Academic Development (CAD) and lecturer’s portfolio (PRF). Each of the mentioned sub-datasets is classified with a different classifier technique. FB uses Back-Propagation Neural Network (BPNN), CAD uses Naïve Bayes Classifier (NBC) and the third sub-dataset uses Support Vector Machine (SVM) as a classifier technique. After implementing the system, the results of the above sub-datasets are collected and then fed as input data to BPNN technique to obtain the final result and accordingly, the lectures are awarded, warned or punished.

  • 1,003 Views
  • 1,046 Downloads
  • 1 Citations

Download

Keywords

  • Lecturer Performance
  • Soft Computing
  • Multiple Classifiers