Educational Group Recommendations By Learning Group Expectations

Abstract

Recommender systems (RS) can assist users by producing a list of recommended items tailored to user preferences. It has been used in the area of educations as one of the technology-enhanced learning techniques, e.g., recommending learning materials or suggesting after-school programs. Group learning has been one of the popular ways in learning and teaching, while the group recommender system (GRS) is able to recommend items to a group of users, instead of a single user. In this paper, we propose a novel group recommendation model for educational recommendations. The model is able to learn the group expectations by taking advantage of the multicriteria preferences. The proposed method is general enough, and we can reuse the existing aggregating strategies to produce the top-N group recommendations. Our results demonstrate the effectiveness of our proposed group recommendation models by using a real-world data set.