Analysis And Research Of Spark E-commerce User Behavior Data

As the times progress and develop,data becomes more and more important to us.The arrival of big data has innovated the technology of computers and other industries,and has brought us into the era of big data.Due to the rapid development of e-commerce,many e-commerce platforms now use big data technology or cloud computing for data management^[1].Currently,the most used big data frameworks are Hadoop and Spark.By analyzing user behavior data obtained from e-commerce platforms,they can guess and recommend users'favorite products to meet user needs.Therefore,understanding user behavior is a necessary condition for the development of the e-commerce industry.This article is based on the analysis of big data.The data is based on the real data provided by the operators of the relevant platforms.These data are processed,mined,and analyzed to obtain corresponding results.By analyzing the behavior of users and using a combination of means clustering algorithm,Naive Bayes method,decision tree algorithm and other methods to integrate and classify these data,the e-commerce platform can predict users'favorite products based on these classified data.Provide targeted products to users and save each other's time.The main research work of this paper is as follows:?1?This article starts with the pre-processing of e-commerce user behavior data,user behavior characteristic data mining,and user behavior analysis.It classifies user data and analyzes the data separately by classification to obtain the results.?2?There will be some redundant data in the process of data processing.First,we must remove and separate these data to avoid wasting time in the analysis process,and then perform related operations on the data and judge the user's basic information by classification,such as:Whether it is a new user of the platform or whether there are historical orders on the platform.The framework used in this article is the Spark framework.It introduces the basic concepts of Spark and comparison with other frameworks.The application of such frameworks is also a major part of this article.Highlights.?3?This article also uses a combination of some algorithms,such as:clustering algorithms,decision tree algorithms,and naive Bayes methods based on mathematical classification and statistical algorithms to process the data,and calculate the weighted results of features and ordinary data through calculation.Compare and classify.?4?Finally,through the construction of the virtual machine and the construction of the Spark framework environment,the user behavior data of the e-commerce platform is analyzed to obtain the corresponding user purchase information data,and analyze the purchase intention of users based on these data.