Design And Implementation Of Book Recommendation System Based On Mahout

As the social network is more and more deeply into the user's life,the various social networking platform appeared at home and abroad.Based on this,considering the importance of reading in people's life,we put forward the establishment of the social network platform based on the book,based on the books of social can make users manage their books and make form their own clubs.To make users have better social experience,we put forward the applications of personalized recommendation service to book reading and social management platform,so that it can help users find like-minded friends and books that they might be interested in.In book reading and management of social platform project,we hope that by providing personalized recommendation service,the user can take connect with other users who have the similar interests.The user not only can manage his books through the establishment of his own library,but also can find his like-minded friends.Also,book publishers can run advertisements in the platform.we hope that use books which user likes or user's friends like to find the books that user might be interested in.In this way,users are more likely to find books they may interested in.By using personalized recommendation service,users will have a better social experience.This thesis mainly completed the design and implementation of book recommendation system based on the Mahout,by using and expanding Mahout framework,we can provide recommendation service for the book reading and management social platform.Around this target,we put forward a method that combine two kinds of mainstream recommendation algorithm to recommend.In this paper,we use the Apache Mahout to build book recommendation system,and completed the design and implementation of the book recommendation system,finally,designed test case to test the main function of the system.The main work has been completed as follows:(1)In terms of book information data preprocessing,this paper use Apache Solr to create index for book information in the database,and then use Solr MoreLikeThis(MLT)to achieve finding similar books.(2)In terms of personalized recommendation,this paper build a recommendation system based on Mahout,using user-based collaborative filtering recommendation algorithm to recommend In order to solve the problem of cold-start of the collaborative filtering recommendation algorithm,we combine collaborative filtering recommendation algorithm and content-based recommendation to recommend.(3)In terms of recommender system,we integrate the recommendation system into the social network platform based on the book reading and management,so it is more convenient to provide personalized recommendation service for the social platform.On the basis of the research and analysis on the user's needs,this paper designe and implement the book recommendation System based on Mahout.