Study On The Personalized Recommendation Based On Product Attribute

With the rapid development of Internet technology,wed applications can offer more and more information and services than ever before.As a result,it's difficult for users to find useful information among this “over-load information stream”.A recommendation system is a personalized technique for solving the above issues.Recommendation system,also known as the personalized recommendation system,makes personalized recommendations of information,products or services for users by different technologies.Collaborative filtering is one of the most widely used and successful methods for recommendation,which has attracted great attention from all walks of life.It is also a good way to solve the problem of over-loading of the Internet information.However,as an increasing number of people begin to use the Internet for online shopping,which produce a huge amount of online shopping data.As a result,the efficiency of collaborative recommendation algorithm has been reduced.How to find useful data among this “over-load” information stream,how to make collaborative recommendation algorithm efficiency again and how to improve user experience become the focus of this study.Obviously,people would show the same interest when they buy similar kind of goods.Therefore,we propose a personalized recommendation model based on product attributes.Firstly,we will build a commodity attribute set by reading and analyzing lots of related literatures.Then we will apply Clustering Algorithm Based on Sparse Feature Vector to deal with the set.Performing collaborative filtering algorithm within the commodity group which includes the target user's item for personalized recommendation.At last,we will obtain data through the questionnaire and simulate this model.Our experiments suggest that the model which based on product attributes can reduce the computational work of finding neighboring users,while at the same time providing better quality than the traditional collaborative filtering.