Design And Implementation Of Personalized Book Recommendation System Based On Mahout

As an auxiliary user selection,recommendation system has quickly attracted the attention of relevant researchers and the Internet market.At the same time,personalized recommendation service has also been widely used in various fields.This paper takes Hadoop,a distributed storage and computing platform,as the operating environment of the system,and USES Mahout's collaborative filtering algorithm and top-10 algorithm as the core recommendation algorithm.A personalized book recommendation system based on Mahout is implemented.The main work of this thesis is as follows:Firstly,The current mainstream recommendation algorithms such as collaborative filtering,content-based and hybrid recommendation related principles and application scenarios are analyzed and studied.Combine the characteristics of the system and the wide application of the algorithm.A recommended algorithm suitable for the system was selected.Combine the characteristics of the system and the wide application of the algorithm.A recommended algorithm suitable for the system was selected.Secondly,Use JavaWeb technology to build a personalized book recommended platform website.This system uses the SSH framework and MySQL development commonly used in Java development to realize data interaction between the background,foreground and database.Thirdly,Based on the characteristics of the system,the recommendation algorithm selects the recommendation based on collaborative filtering and the recommende based on Top-10.These two are used for personalized recommendations for individual users and Top-10 popular recommendations for mass users,which are also used for cold start issues for new users.Finally,A personalized book recommendation system based on Mahout framework based on Hadoop is designed and implemented.