Research On The Application Of Personalized Recommendation Algorithm In E-commerce Under Big Data Environment

With the continuous expansion of e-commerce scale, personalized recommendation technology has been widely used. However, traditional recommendation system has been unable to meet the needs of the current data processing. Having a good ability for big data processing is the basic requirement of new personalized recommendation system. In addition, traditional recommendation system are often focusing on the tangible goods recommendation, and paying less attention for e-commerce logistics service recommendation. This paper researched the information personalized recommendation service in e-commerce environment, combining with the application background of big data:First, aiming at the shortcomings of the traditional algorithm, sitting about from the incremental updating of recommendation model, this paper put forward the concept of user-deviation, and based on which to build the user-deviation matrix model with incremental updating. At the same time, by combining with Hadoop cloud computing platform, this paper realized the model building by distributed. Inner incremental updating of the model and external technical support of the big data processing, the two complement each other, makes the model has high efficiency on big data processing.Second, using the user-deviation matrix as the recommendation model, this paper put forward IU-UserCF and UDB-Slope One recommendation algorithms. These two kinds of algorithms based on incremental updating recommendation model have good extensibility, not only can effectively process big data, but also have very high accuracy.Finally, the proposed algorithm is applied to actual system. Using the e-commerce logistics service as recommend object and IU-UserCF as recommendation algorithm, this paper constructed the personalized recommendation system that is oriented toward e-commerce logistics service. The e-commerce logistics service recommendation system is an exploration for the practical application of the recommendation algorithm under big data. It enriches the application scenario of the personalized recommendation technology.