A Study of Recommender Systems on Social Networks and Content-based Web Systems

Abstract

Everybody rely on recommendations in everyday life from other people either orally or by reviews printed in newspapers or websites. Recommender systems are a subfamily of information filtering systems that explore to predict the 'rating' or 'preference' that user would give to an item. These systems are best known for their use in e-commerce websites where they use input about a customer's interest to generate a list of recommended items. Many recommender systems explicitly rate to represent customer's interest by using only the items that the customers purchase, but can also use other attributes, including items viewed, subject interests and demographic data. They direct users towards those items that meet their needs by reducing unwanted information spaces. To perform recommendation a number of techniques have been proposed, including content-based, collaborative, and hybrid techniques. To improve performance and to outweigh the drawbacks of individual recommendation techniques, these techniques are sometimes combined to form hybrid recommenders. This paper is categorized into seven sections. Section-I presents the introduction related to the recommendation systems used in the social networks and online Web systems, section-II critically analyzed the related literature work about collaborative recommendation, contentbased recommendation, and hybrid recommendation, section-III describes the business aspects of recommender systems, section-IV describes various ways of displaying recommendations to a customer, section-V investigates the various recommendations techniques and their limitations and section-VI provides the conclusion of the recommender systems. In this paper the efforts are made to review and discover the techniques to investigate the proper usage of recommender systems in the e-commerce applications.

Prediction of Movie Quality via Adaptive Voting Classifier

“…Aforementioned features are further used by classifiers such as K-Nearest Neighbor for generating recommendations [62]. Content-based filtering techniques use item features for recommendations [33], [63], [64]. The items with similar properties become the candidates for recommendation results [65].…”

A hybrid user-item-based collaborative filtering model for e-commerce recommendations

“…A large number of approaches have been developed for non-personalized and personalized RSs such as product association approach (Poriya et al, 2014), content-based, information filtering (Hooda et al, 2014;Woldan et al, 2020), collaborative filtering (memory-based -user-based (Resnik et al, 1994;Li & Li, 2020) and item-based (Sarwar et al, 2001;Ujkani et al, 2020;Ajaegbu, 2021), model-based (Kane, 2018)), knowledge-based (Aggarwal, 2016), multi-criteria recommender systems (Shambour, 2021;Smirnov & Ponomarev, 2020), etc.…”