期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2018
卷号：7
期号：6
页码：6945-6950
DOI：10.15680/IJIRSET.2018.0706087
出版社：S&S Publications
摘要：In the last few decades, several methodologies have been proposed by experts of research community in
order to enhance online shopping system. Here we made an attempt to upgrading the existing system. Our
recommender System aims at to predict future user ratings in order to estimate meaningful recommendations to a
collection of users for items that users might like. Rating is ranking or grading the item based on user’s satisfaction.
Prediction is the act of saying what product the user might like in the future based on his previous ratings. Predicting
the ratings is key task in recommender system. Recommender systems aim at finding items that are likely of interest to
a user, by exploiting different types of information sources related to both the users and the items.
This paper presents algorithms for recommending items in online shopping system. Recommendation task is
categorized as three steps: (1) grouping latent center of interest, (2) Clustering like-minded users, and (3) predict user
item ratings. We use k-means clustering algorithm to create clusters of similar users. In order to predict item, we
employ linear regression algorithm. Validate the result by employing the MAE and RMSE measures.
关键词：Collaborative Filtering; SVD++; Group recommendation and; Regression method;
