A hybrid framework for enhancing correlation to solve cold-start problem in recommender systems

The online shopping is becoming a trend in the age of digital technology. By using intelligent recommendations, the online shops or online retailers directly approach and meet customers´ demand easier than the physical stores. However, the online shopping still has its drawbacks, among a variety of diverse product types, sizes and design, customers need to browse and filter from a wide range of sub-categories to find the suitable products. That is why the justice system that collects customer information and products to make appropriate suggestions for each user is raised encouraged using on the commercial website. The purpose of this work aims at proposing a hybrid framework for enhancing correlation to solve cold-start problem in recommender systems. Experiments are performed using MovieLens dataset to make a realistic methodology.