Design And Implementation Of Small And Medium Customer Growth Clues Access System

Nowadays,the Internet industry is booming,and rapid development will inevitably be accompanied by fierce competition.From a large background,Baidu has been losing money due to the shrinking stock market and the loss of profits.Aside from the data,there are several issues that need attention: first,the entire advertising industry is in a downturn,and all advertising-related companies are inevitably affected;second,Baidu has invested too much money in areas such as AI,in which at present,it is still difficult to monetize;thirdly,in recent years,Baidu has encountered many word-of-mouth crises in succession,triggering a large number of users to turn back,and causing Baidu to lose its competitiveness.In order to help the company get out of the trough,we put forward a solution for the clue access of small and medium-sized customers' growth in view of the problem of the loss of small and medium-sized customers.The system described in this thesis is the implementation of the clue access scheme.It mainly includes the clue collection subsystem and clue layering subsystem.The clue collection subsystem mainly completes the collection,storage,sorting and distribution of clue data,and provides data sources for other modules,which is the basis of the entire system management data;The clue layering subsystem is to use a certain strategy to label all clues in the library in the form of related aggregation and classification.The collection part of this system uses Flask framework,the layered part uses Spring MVC framework,the verification code service is developed based on tensorflow,and the storage tool uses Mysql search engine for Elastic Search.The system first sets the rules for thread crawling through the basic setting module of the thread collection subsystem,and then provides the necessary technical services for the thread crawling module and the thread analysis module through the public service module.The thread crawling is accomplished through the crawler technology.Then clue analysis converts web page information into clue data.With the source of clue data,the clue layering subsystem can use these clues for classification management and optimization,so that clues can be used more effectively.Through the implementation of effective reminders and layered mechanisms,we can accelerate the follow-up of clues,improve the conversion rate of clues,improve the quality of clues,and realize the healthy growth of small and medium-sized customers.The author participated in the development of the entire thread access system and its external systems,and independently completed the development of the thread collection subsystem and the thread layering subsystem,including basic rule setting for clue collection,public services,clue capture and analysis,data cleaning of clue layers,rule setting,layer calculation and data presentation.The application effect of the completed work is to establish and perfect the process of customer management and incubation for the client-to-clue system for the clue-layer system,and lay a foundation for the process of clue-to-cash flow recovery.Help the company to follow up on potential small and medium-sized customers and existing customers,defeat competitors,and win customer trust.For the clue collection subsystem,ICP filing clues are very important clue resources.At present,resource centers or sales collect ICP clues by themselves,there are problems of high cost,high security risks,and poor timeliness.This system collects ICP clues through the Pangu side and stores them in real time to improve efficiency and reduce costs...