Graduates Employability Analysis using Classification Model: A Data Mining Approach

Abstract

The employability of graduates serves as a measure of the success of every program offered within a Higher Education Institution.  The employment assessment and evaluation of employment status allow improvement using data mining to analyze a vast amount of data in various areas.  This study builds graduate’s employment model using classification tasks in data mining, compares several data-mining approaches such as the Bayesian method and the Tree method with visualization, and explores the Association Rule using Apriori. The experiment used a classification task in Waikato Environment for Knowledge Analysis (WEKA) and compared the results of each algorithm, where several classification models were used.  The experiment was conducted using accurate data sets from 1,489 graduate students for three years.  The study provides valuable information about the graduate employment status, forecasting, visualization, and the exploration of classifiers algorithm to analyze the graduate employability in government, non-government organizations, self-employed, and unemployed.  It is recommended to relate graduate employability to curriculum assessment and performance evaluation to identify measures and policies to improve students' performance.