Research And Realization Of Supermarket Customer Purchase Behavior Analysis Based On Big Data

With the advent of the Internet era, the supermarket industry generally began to use the database system to collect and store sales information,these data often contain huge commercial value. Digging out valuable information from a large amount of data can help supermarket marketers formulate effective marketing strategies to increase the operating profit of supermarkets.With the increase of data size, data mining and analysis is facing a severe test,selecting an effective big data processing platform becomes critical. Spark is an efficient and reliable big data processing framework. Compared with the general data processing framework, Spark has a higher processing performance. Based on this, in this paper, Spark technology is used to analyze the purchase behavior of supermarket customers.In this paper, we first build a Spark distributed cluster environment based on YARN resource manager and the corresponding application development environment, and introduce two classical algorithms in data mining: Clustering Algorithm and Association Rules Algorithm, and analyzing implementation of K-means Clustering Algorithm and FP-Growth Association Rules Algorithm on Spark Platform. Then, the user data of the supermarket is processed and analyzed, and the potential features of the users are extracted by the analysis result. The user is clustered by using the K-means algorithm based on Spark platform. Through the analysis of the experimental results, the reliability of the Spark platform is verified by the experiment, and the running efficiency of the Spark platform is better than that of the Hadoop platform. At last, we process and analyze the commodity data and transaction information of the supermarket, and use the FP-Growth algorithm based on Spark platform to carry on the Association Rules mining of the supermarket commodity. Through the analysis of the experimental results, we verify the reliability of the Association Rules, according to the mining association rules we can provide supermarket operators with some suggestions.