Research And Implementation Of APP Behavior Analysis Based On Android Log

With the "Internet + Agriculture" strategy put forward,all kinds of agricultural e-commerce platform have been rapid development.People in the e-commerce platform shopping process,usually left a large number of shopping behavior patterns reflect the behavior of shopping information.Applying the modern information technology of big data and artificial intelligence to dig out the behavior data of users' shopping habits and tendencies,it is of great significance to build a complete user portrait and achieve accurate and personalized recommendation.At present,the analysis of user purchase behavior mainly concentrates on the background server-side log and can not obtain all the purchasing behavior information of the user on the client side.This research is aimed at the "one village one product online shopping mall" Android mobile client,elaborated from the mobile client development,front-end collection of logs to the user purchase behavior analysis of the entire process.The main research work is as follows:1)research android development process common technical problems,including android development framework and various components to use.Through the project needs analysis and program design,the main completion of the user module,product recommendation display module,shopping cart module,order module and the purchase of the development and implementation of modules.2)research log acquisition technology.The main indicators of user purchase behavior include user browsing,favorites,adding to the shopping cart and submitting orders.3)Analyze the common user analysis of user behavior analysis.According to the analysis of user purchase behavior of this subject,Logistic regression algorithm and GBDT algorithm are respectively used,and the idea of GBDT + LR algorithm fusion is applied to the analysis of user purchase behavior.This article presents a complete process from the Android client development,log acquisition to user behavior analysis.The project adopted in the program can be applied in similar situations to promote the use of innovative research results.