Evaluating Prediction Accuracy, Developmental Challenges, and Issues of Recommender Systems

Abstract

Modern ways of living have made the people to depend on internet services for everything. The mounting information from various sources like social media, implicit and explicit information, user's geographical location, and the internet of things had increased the need of a recommender system. From e-governance to e-shopping, a recommender system helps people in finding the needed item or information and also boosts sales in the market of those items. Though many studies elaborate about recommendation systems, challenges in developing the recommendation systems, prevailing issues of recommendation systems and discussions on prediction accuracy are not detailed in any of the earlier works. Therefore, in this article, in order to increase the accuracy of the recommender system, the developmental challenges and issues in constructing recommender systems and for evaluation metrics in prediction accuracy are identified and detailed.

A Novel Rule based Data Mining Approach towards Movie Recommender System

“…In today's digital world, one of the major day-to-day challenges is to find the appropriate information from the vast amount of data that is available online. There are various personalization techniques available in the market to overcome this problem; one aspect of personalization is recommendation system which is explained in detail by Moses and Babu [1]. The primary objective of a recommender UDC 004.62:004.891:791.…”