Cold Start Problem in Recommendation System: A Solution Model Based on Clustering and Association Rule Techniques

Generally cold start problem refers to the new user who are coming in big data as well it also related to new items which are included in any data set. Problem arises when companies or policy makers don’t have the information about new user/item in recommendation of products. Recommendation system shows the user who is interested in particular product implementing contents based, collaborative based or hybrid approaches. Meanwhile recommender system faces the cold start issue which means the recommendation system is not recognizing the new product or new user. In another words the system doesn’t have information about preferences in order to make recommendation. To overcome the cold start problem, in this paper we are suggesting a solution by combining clustering techniques and association rule. The study is based on extracts of different approached studies. This paper also lets you understand the proposed models to overcome the problem of cold start.