Enhancing precision of Markov-based recommenders using location information

Recommender systems are a real example of human computer interaction systems that both consumer/user and seller/service-provider benefit from them. Different techniques have been published in order to improve the quality of these systems. One of the approaches is using context information such as location of users or items. Most of the location-aware recommender systems utilize users' location to improve memory-based collaborative filtering techniques. However, our proposed method is based on items' location and utilizes a Markov-based approach which can be easily applied to implicit datasets. The main application of this technique is for datasets containing location information of items. Experimental results on real dataset show that performance of our proposed method is much better than the classic CF methods.