A Study of Recommender Systems Using Markov Decision Process

As the number of items increase in every domain, users find it particularly difficult to decide which item is most suitable for them. Due to the availability of such large pool of data, it is crucial to automate the process. For carrying out this task, recommender systems are used. But conventional recommender systems adopt a static view of data. Hence, reinforcement learning is used to make a dynamic recommendation of items to the user. In the presented paper, research work done in recommender systems by using the Markov Decision Process, which is a reinforcement learning model, has been discussed.