Design And Implementation Of E-commerce Recommendation System Based On Big Data Technology

Since the birth of information technology,it has been known to maintain a trend of rapid development,and the development of Internet technology has made the relationship between people closer.As time goes by,the data in the network becomes more and more complicated,and the problem of information overload becomes more and more serious.How to obtain useful information from huge amounts of data has become a problem that people pay attention to.The recommendation system was invented to solve this problem.After years of development,the recommendation system has had relatively successful experience in areas including movie recommendation,music recommendation,social networking,and e-commerce.The recommendation algorithm is the cornerstone of the recommendation system.At present,researchers have proposed a variety of different recommendation algorithms to adapt to different recommendation scenarios.However,no recommendation algorithm can solve all the problems.Mixed recommendation can effectively alleviate the defects of a single recommendation method by combining the strengths of multiple recommendation methods and taking advantage of the shortcomings.Based on a mixed recommendation idea,this paper designs and implements an e-commerce recommendation system based on big data technology.The work content is divided into two parts: design and implementation.First,the system needs to be analyzed,and the overall architecture of the system is designed.Then specifically explained the design and implementation of each recommendation module: This article combines a demographic-based recommendation algorithm,content-based recommendation algorithm,ALS-based collaborative filtering recommendation algorithm,and model-based real-time recommendation algorithm to form a hybrid recommendation module To provide users with recommendation services.Among them,the recommendation based on demographics can solve the cold start problem of users,the collaborative filtering recommendation can make recommendations without the content attributes of items,and the model-based recommendation algorithm can provide users with real-time recommendation services.Based on the above algorithm,the Spark distributed computing engine is used to implement the algorithm recommended in this article,which solves the problem of server performance degradation under massive data.MangoDB is used as the database to store data,Redis is used as the cache database,Flume and Kafka process real-time data,and finally achieve E-commerce recommendation system based on big data platform was introduced and the system was demonstrated.