Social Popularity based SVD++ Recommender System

Abstract

Recommender systems have shown a lot of awareness in the past decade. Due to their great business value, recommender systems have also been successfully deployed in business, such as product recommendation at flipkart, HomeShop18, and music recommendation at Last.fm, Pandora, and movie recommendation at Flixstreet, MovieLens, and Jinni. In the past few years, the incredible growth of Web 2.0 web sites and applications constitute new challenges for Traditional recommender systems. Traditional recommender systems always ignore social interaction among users. But in our real life, when we are asking our friends or looking opinions, reviews for recommendations of Mobile or heart touching music, movies, electronic gadgets, restaurant, book, games, software Apps, we are actually using social information for recommendations. In this paper social popularity factor are incorporated in SVD++ factorization method as implicit feedback to improve accuracy and scalability of recommendations.

TWIN: Personality-based Intelligent Recommender System

“…For example, the MapReduce methodology [6] provides a way to collect information from various distributed sources. Dimensionality reduction strategies such as the matrix factorization method of singular value decomposition (SVD) [7] help to draw efficient generalities from the compressed data to improve recommendations. Babak et al [8] also exploit cross-domain collaborative filtering techniques.…”

A Novel Embedding Model for Relation Prediction in Recommendation Systems

“…None of users and items overlaps between source domain and target domain, so the CLFM and the PCLF cannot work in this case. We compare our proposed model with two approaches: Matrix Factorization (MF) [14] and SVD++ [15]. As far as we know, they are the only two approaches which can fit this case.…”

“…Baseline. For rating prediction, we compare our proposed model with the memory-based collaborative filtering methods User-KNN [23], Item-KNN [28] and two typical model-based collaborative filtering recommendation approaches: MF [2], [13], SVD++ [15], and an SVM-based multi-label classification algorithm CLR [8]. For user-KNN and item-KNN, we selected the number of neighbors K among {5,10,20}.…”

“…Maybe not. To cope with the problem, most existing approaches [14], [15], [23], [28] have been proposed to recommend the users personal significant and interesting items on e-commence websites, by estimating unknown rating which the user may rate the unrated item, i.e., rating prediction as (user,rating=?,item). Basically, the more high rating the user may rate the item, the more possible that the item would be recommended to the user.…”

“…may be 5. numbers and learn nothing about the ratings' embeddings (Embedding denotes the latent semantic representation.) in the training process [14], [15]. User prediction refers to the problem how to predict unknown users with given rating and item, i.e., (user=?,rating,item), while item prediction is similarly defined.…”