Predicting Career Decisions Among Graduates of Tafseer and Hadith

Abstract

The overall aim of the research was to identify influential factors that best predict career decisions or job choice among graduates of the Department of Tafseer and Hadith at the Universitas Islam Negeri Sultan Syarif Kasim Riau. Instead of using a longitudinal, cohort study using statistical analysis, machine learning techniques such as Decision Tree and Naïve Bayes was applied to search for unknown patterns or rules. This study compares the performance of the machine learning methods in discovering hidden patterns or factors that influenced the alumni career decisions. One of the primary concern of the university is whether their career choice after graduation are relevant or match to their field of studies. Our studies show that CGPA, cohort, additional expertise, and gender are the main factors that influenced alumni career success. We found that a cohort of graduate students was unable to find relevant professions in their field of studies. The experimental result shows that Naïve Bayes outperforms Decision Tree with the best accuracy score of 97.1 % and 92.6% subsequently. Thus, it can be concluded that the prediction model and analysis using Naïve Bayes have the potential to be used effectively. Despite its low performance, Decision Tree able to extract the main factors that influenced an alumnus career efficiently. These findings are valuable and useful both for the institution to better understand and improve the quality of its program and graduates, and also the community of machine learning in understanding the techniques behaviors with small datasets.