Long Term Goal Oriented Recommender Systems

Recommenders assist users to find items of their interest in large datasets. Effective recommenders enhance users satisfaction and improve customers loyalty. Current recommenders concentrate on the immediate recommendations’ value and are appraised as such but it is not adequate for long term goals. In this study, we propose long term goal recommenders that satisfy current needs of users while conducting them toward a predefined long term goal either defined by platform manager or by users. A goal is long term if it is going to be obtained after a sequence of steps. This is of interest to recommend learning objects in order to learn a target concept, and also when a company intend to lead customers to purchase a particular product or guide them to a different customer segment. Therefore, we believe it is beneficial and useful to develop a recommender algorithm that promotes goals either defined by users or platform managers. In addition, we also design methodologies to evaluate the recommender and demonstrate the long term goal recommender in different domains.