A Systematic Study of Approaches used to Address the Long Tail Problem

E-commerce relies heavily on the Recommendation System. The primary purpose of the Recommendation System is to assist users in selecting appropriate items. Many Recommendation Systems only recommend popular items to users and ignore unpopular ones. These unpopular items have very few ratings. These unpopular items create a long tail in the Recommendation System. A recommendation system refers to this type of problem as the Long Tail Problem. These unpopular or tail items, if comes into the mainstream for recommendations to the users can increase the sales of items along with popular or head items, by providing users an opportunity for one-stop shopping for both their mainstream and niche tastes. This paper provides an overview of various techniques used to address the Long Tail problem of the Recommendation System. A comparative analysis of all the methods used to address the Long Tail Problem, their evaluation criteria, and their results is also included in the paper. Researchers and academicians will benefit from a comprehensive study of the Long Tail Problem in Recommendation Systems by better understanding existing work and developing future directions.