Recommender system is a very young area of machine learning & Deep Learning research. The basic goal of the recommender system is to create a relationship between items and consumers. The relationship provides recommendations based on user interest. content-based, collaborative, demographic, hybrid filtering, knowledge-based, utility-based, classification model are well-known recommender models. The model uses an item's specifications in content-based filtering to suggest other objects with similar features. Collaborative filtering takes into the user's previous activity which means the user has previously viewed or purchased, as well as ratings Provided by the user to those items and similar conclusions reached by other users' item lists. View user profile data such as age category, gender, education, and living area to detect commonalities with other profiles.[31] All three filtering techniques are used in hybrid filtering. In the process of recommendations, various challenges are faced by the system. So, this paper lists various solutions by researchers in recent days.