A novel model for hospital recommender system using hybrid filtering and big data techniques

摘要

Recommender systems help the users to get the useful information regarding their search. To overcome the disadvantages of content- based filtering and collaborative filtering, Hybrid filtering is one of the best suitable approaches. Collaborative filtering (CF) is a method which collects different kinds of information from many users and preferences about the interests of a user to make predictions automatically. The problem with CF is, as it makes predictions by considering previous rating information given by like-minded Users. So this approach would fail if no User has rated the item earlier, called cold start problem. Content-based filtering (CBF) visits user profile to get the useful information regarding with their previous searches and interests, for recommending the similar items to them. The disadvantage of CBF is the items and the attributes must be machine recognizable. This paper overcomes these issues by using Hybrid Filtering. This system fetches the information from the user and displays the nearby hospitals related to it. In this paper, the location of the user and the requirement of which type of hospital should be mentioned by the user itself. Among the detected hospitals it suggests the hospitals to users based on the user ratings. Based on the Hybrid filtering approach recommend the hospitals to the account holders. Depends on the specialty of the Hospital and user preference, the Similarity is calculated using the cosine similarity concept. In Hybrid filtering, we think about the constraint which was given by the user. The principle aim of this paper is to recommend the best hospitals to the users which will be helpful in emergency situations.