Research On Personalized Book Recommendation System Based On Hadoop Platform

Nowadays,the types of library resources are becoming more and more abundant,and the number of books is more and more.In the massive library resources,how to find out the books which users are interested in is the focus of this paper.Hadoop platform under the personalized book recommendation technology to better solve this problem.Personalized recommendation technology to a large number of user information data to learn,analyze and dig out the user interested in the information recommended to the user.Can be based on large data platform Hadoop to achieve the parallel analysis of book data analysis,through the Mahout framework under the user clustering collaborative filtering recommendation algorithm to the user personalized recommendation.Based on the collaborative filtering recommendation of user clustering,clustering for users can improve the recommended accuracy and efficiency.This paper first introduces the research status of the recommendation system at home and abroad,and summarizes several classic recommendation algorithms,including user-based collaborative filtering and content-based collaborative filtering and hybrid recommendation.Then the implementation of the proposed algorithm in the open source machine learning framework Mahout is introduced.The collaborative filtering recommendation based on user clustering is used and compared with the traditional recommendation algorithm.Finally,by building the Hadoop distributed platform,using the Mahout framework to build the client book recommendation processing center,and use the data migration tool Sqoop as HDFS and MySQL bridge,the recommended results are migrated to the MySQL database.The recommended data stored in the MySQL database can be displayed on the Web page through the Web recommendation engine.Web side of the recommended engine is built through the Maven to Tomcat as a Web container,using the SSH framework to achieve a Hadoop and Mahout based on personalized book recommendation system.In general,this article in the study of Hadoop platform under the personalized book recommendation process to complete the main contents include the following three points:(1)The analysis of the Mahout composition of the open source machine learning algorithm is completed,and the working principle of the Taste recommendation engine is studied,and the implementation of the recommended algorithm in Mahout is discussed.(2)The collaborative filtering recommendation algorithm based on user clustering is recommended and compared with the traditional recommendation algorithm.(3)Using Hadoop platform,through Mahout and SSH framework and other technologies,to build a Web-based personalized book recommendation system.