Abstract— Admission recommender system provides a way to finding suitable course in universities that best fit students’ merits and interests. Iraqi universities are using a complex admission process which is not supporting recommender systems, and it is only based on the candidates’ Grade Point Average (GPA). In addition, applicants don’t have enough information on available courses and their available majors. In this paper, we prepare a new admission recommender system that is developed by using a hybrid method of Neural Network (NN), Decision Tree (DT), and Our Proposed Algorithm (OPA). We using DT to classify the applicant to 10 groups, each group has special properties, but NN used to apply the applicant to the available courses that can apply, our algorithm mixed with NN to finding best course. Our algorithm considers GPA, test score, candidates’ interest and their desire jobs as decision parameters. The aim of this paper to solve the complex systems that used more than one criteria to apply more than one courses. The result of the system is better and more reliable outcome comparing to other available admission systems; furthermore our system can find a suitable course for each applicant. Key

Index Terms— E-recommender system, admission system, neural network, decision tree.

Karzan Wakil is with Fine Arts Institute in Sulaimaniyah (FAIS)-Iraq, Rizgari, Sulaimaniyah, KRG, Iraq (email: karzanwakil@gmail.com).
Ban Akram and Ngein Kamal are with University of Human Development (UHD), Sarchia, Sulaimaniyah, KRG, Iraq.
Amirhossein Safi is with University Technology Malaysia (UTM), Skudai, Johor, Malaysia.

Cite: Karzan Wakil, Ban Akram, Ngein Kamal, Amirhossein Safi, " Web Recommender System for Private Universities’ Admission in Iraq: UHD Case Study," International Journal of e-Education, e-Business, e-Management and e-Learning vol. 4, no. 5, pp. 329-340, 2014.