RESEARCH++: An Academic Social Networking Research Community Portal for Profiling and Expertise Classification
Jonardo R. Asor, Marco Antonio T. Subion · 2018 International Seminar on Research of Information Technology and Intelligent Systems (ISRITI) · 2018
Institutional research gives so much contribution on the current situation of the world especially in introducing new products. Developing institutional research paper should have a clear and competent basis. In the Philippines, the students were required to produced and institutional research before graduation as it is the main requirement in completing particular program. Each of them will have a particular professor who will act as their adviser and help them throughout the research process. It would be a brilliant idea to have an adviser who is identified expert on the topic that the student was working on. One of the methods that the student was doing is to check all the past research works of specific instructor to classify its expertise. This study aims to develop an expertise model for research portal that is capable of classifying the expertise of the instructor through predictive analytics. With the total accuracy mean rating of 82.06% and 0.98 F1-score, Naïve Bayes shows an exemplary performance in classifying expertise of particular individuals based on its past research works.