Predicting Academician Publication Performance using Decision Tree.
Mohd Zakree Ahmad Nazri1, Rohayu Abd Ghani2, Salwani Abdullah3, Mas Ayu4, R Nor Samsiah.5
1Mohd Zakree Ahmad Nazri, Mohd Center for Professional and Leadership Development, UniversitiKebangsaan Malaysia.
2Rohayu Abd Ghani, Mohd Center for Professional and Leadership Development, Universiti Kebangsaan Malaysia.
3Salwani Abdullah, Mohd Center for Artificial Intelligence, Faculty of Information Science & Technology, Universiti Kebangsaan Malaysia.
4Mas Ayu, Mohd Center for Artificial Intelligence, Faculty of Information Science & Technology, Universiti Kebangsaan Malaysia.
5Nor Samsiah, Mohd Center for Artificial Intelligence, Faculty of Information Science & Technology, Universiti Kebangsaan Malaysia.
Manuscript received on 25 June 2019 | Revised Manuscript received on 08 July 2019 | Manuscript Published on 17 July 2019 | PP: 180-185 | Volume-8 Issue-2S July 2019 | Retrieval Number: B10260782S19/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This research focuses on predicting academician performance in terms of publication rate and investigate the factors that affect academicians’ achievement. This study investigates how scientific publication rate by individual is influenced by factors such as gender, age, number of research grant and academic position of the researchers. Having a decision rules, university leaders can understand upcoming trends with respect to leadership requirements and academicians needs. It is also helping university managements understand challenges and therefore can deploy the right strategies for human resource management interventions. This paper describes the development of the predictive model using a data mining technique. Previous studies have shown that there are many important variables when analysing academicians’ productivity at the individual level. What is unusual with our approach is that this study is using Decision Tree to identify the patterns for predicting next year’s performance. Decision Tree, C4.5, J48 and PART is a common predictive method for prediction as there are other methods that are better suited for predictive analytics such as regression or metaheuristic algorithms. However, with finding knowledge among the attributes obtained from the university’s databases, we can predict the performance of an academician staff. To find strong and valid rules, different measures like min Interest, lift, leverage and conviction are considered. The study, involving almost 3000 university lecturers, shows a number of interesting patterns that can be used for predicting individual performance.
Keywords: Higher Education Institution, Predictive Analytics, Tree to Rule Induction.
Scope of the Article: Smart Learning and Innovative Education Systems