Research And Implementation Of Digital Integrated Service Platform For Small And Micro Businesses

In recent years,with the rapid development of intelligent mobile terminals and the fierce competition on the Internet,how to combine social networking with e-commerce has become one of the current research hotspots.As an important part of social e-commerce,small and micro businesses use less professional business tools,low efficiency of daily marketing and commodity promotion,and lack of rigid commercial demand.At the same time,a large number of data generated in the marketing process lack of scientific management system.In view of this phenomenon,this paper aims to solve the pain point for small and micro businesses,design and implement a digital integrated service platform.The main work of this paper is as follows1.In this paper,we build a virtual container based on the virtualapplication framework to make the third-party app run in the container,so as to fully monitor and control it,and use the auxiliary function components provided by Android system to realize RPA process automation.This scheme is more compatible than the cross process operation memory scheme,and has lower development cost for new third-party apps and their internal updates.2.For the business data of small and micro businesses,this paper builds a complete data production,storage and operation process system.Firstly,the embedded points are embedded in the specific events of the Android end of the service platform,and the collected data is transmitted online through the Kafka message queue.The offline platform uses flume for data collection.The data platform is built based on Hadoop distributed cluster,the storage medium of data warehouse is hive database,and the daily data summary calculation is based on spark computing framework.3.Based on the big data platform,this paper designs and constructs the user portrait model of small and micro businesses and their consumers.The data stored in the platform can be divided into two categories:desensitized user attribute data and behavior data.The analysis goal of user portrait model is to label each user with a specific weight to show that users have interest,preference and demand for the content.In this paper,based on the TF-IDF weight calculation and subject words,we build a business classification model of small and micro businesses.At the same time,we build a high-quality consumer screening model based on K-means clustering algorithm,genetic algorithm and SVM model.It is verified that the model has high accuracy and has certain application significance.