Data Driven Decision Support to Fund Graduate Studies in Abroad Universities

Each year, many undergraduate students from developing countries try to continue their graduate studies in foreign universities. Admission is not easy, especially for those who seek positions with full funding and scholarship. Chance of acceptance and getting fund is dependent on many factors like GPA, GRE, IELTS scores, and the field of study, university name and number of papers. Since the process of application is cost and time consuming, students should just apply for universities with a high chance of acceptance and funding. Students usually reach out to those who applied before and use their experience to have a smarter choice. In some countries like Iran, there are portals and websites in which previously admitted students share their experience and information. In this paper, we use the data provided by these students to build models for predicting the chance of a student for getting fund from different universities. After cleaning and preprocessing, we build decision trees which take the person data as input and calculate the probability of getting financial support from various universities. This model also helps us to find out the most important factors in succeeding to achieve funding. Also, admission seekers can follow the provided rules to estimate their chance of getting fund and obtain ideas about how to improve their profiles to increase their chances. Additionally, we use a k-nearest neighbor algorithm to find k most similar records to user's profile. These similar records are used to predict the chance of acceptance and getting fund. Undoubtedly these models1 are beneficial for students who have profound desire as well as students who are trying to pursue higher study abroad with financial support.