Predicting Faculty Research Productivity using J48 Decision Tree Algorithm

Abstract

Research productivity is seen as a shared issue present in most academic institutions, as research output requires skills, time and patience. In the Philippines, higher education institutions are mandated to undertake research and other similar investigations in various academic areas since they are one of the key institutions who play major role in the generation and dissemination of knowledge. This study generally aimed to come up with prediction model for faculty research productivity. Specifically, it tried to seek solutions to the following: identification of necessary attributes that have significant relationship to faculty research productivity, generate computational model for predicting research productivity of faculty through J48 Decision Tree Algorithm and validate the performance of computational model using confusion matrix analysis, precision, recall and f-measure. Results showed that attributes such as paper proposal, length of service, age, teaching loads, academic ranks, designation/s, civil status, academic qualification, sex and status of appointment has vital influence to research productivity of faculty. The computational model generated had an acceptable computed accuracy, precision, recall and f-measure results. Based on the data model performance results, the system can be used and can be implemented in an actual working system.