Personalized rankings of educational institutions

Education is one of the important research domains and building intelligent systems is one of the most exciting applications of Artificial Intelligence. A challenging problem in this domain, which involves a lot of potential users and data, is to find a suitable educational establishment that can match the particular preferences and needs of people. Although there exists a number of known national and international ranking lists, however, almost all of these rankings are non-personalized and offer the same lists of schools to absolutely different people. This thesis addresses this problem and presents the design, implementation and evaluation of a personalized recommendation system in the education domain. The system is capable of eliciting preferences from its users, learn from the preferences, and intelligently generate a personalized ranking list of educational institutes for each target user. The quality of the suggested ranking lists has been evaluated in a real user study, and measured in terms of accuracy, diversity, novelty, satisfaction and capability to understand the particular preferences of different users. The results have shown that the users have been satisfied with the quality of the personalized ranking lists and assessed the system as a usable system.