Predicting Career Trajectories of Information System Graduates Using AdaBoost Algorithm

The undergraduate unemployment rate in Indonesia is increasing, necessitating universities to produce quality graduates. In alignment with global technological advancements, universities are committed to enhancing the quality of their graduates through curriculum evaluation. In this study, we investigated the Department of Information Systems at the University of Riau and developed an alum career analysis system using the AdaBoost algorithm. The data used in the study was collected via an online survey from graduates. Relevant features were selected from features to preprocess the data. The experiments conducted using the SAMME.R boosting method, with a learning rate of 0.1, 100 estimators, and cross-validation on the five-fold sampling method, resulted in the highest accuracy of 83.3%. The decision tree of the AdaBoost algorithm explains the factors influencing alums to pursue careers in their field, with English proficiency, learning methods, alum cohorts, and alum's activeness in job searching being the main determinants of success in their careers. The alum career analysis system proved reliable for evaluating the curriculum and producing quality graduates in line with technological advancements. Future work includes expanding the system to include graduates from other departments and implementing findings to improve the curriculum.