Design And Implementation Of Sales Lead Recommendation System Based On Big Data

In recent years,traditional B2 B companies have struggled to face the new form of "Internet +",but the degree of digitization between enterprises is quite different.For the sales department that is vital to the life of the company,it also faces the problems of extremely fierce competition in the industry,increasing difficulty in developing new customers,and rising costs of acquiring new customers.Aiming at the above problems,based on big data technology,combined with enterprise-specific data,on the premise of ensuring system compatibility,design and implement a sales lead recommendation system with ease of use,high responsiveness and high accuracy.The sales lead recommendation system is based on the BI big data platform and is mainly divided into two parts.One part is responsible for mass data storage and recommendation calculation,and the other part is responsible for system data display and operation processing.For the first part,use Scrapy crawler technology to collect vertical industry-specific data.Use Hadoop's HDFS distributed file system to store massive data including operation data,business data,crawler data,and other data.Combine technology such as Flume,Sqoop,and Hive to finish the ETL operation.And the sales leads are obtained through the similarity calculation of Spark SQL and Spark MLlib combined with Jieba word segmentation.Finally,the tasks are placed on the Azkaban task scheduler for updating.For the second part,use Spring Boot and Mybatis framework technology to write and logically process the back-end part.Use relational databases such as My SQL and Postgre SQL to store user information,customer information,and clue information.Use Redis non-relational databases to upgrade special processing speed.Combine with the front-end pages implemented by the Vue open source framework for data display,page interaction and operation processing.Due to the task scheduling strategy and the addition of customers,the calculation of recommendation clues is not timely,so Elasticsearch is used to retrieve the data to facilitate the temporary recommendation of new customers.The recommendation system is being tested internally.After verification,sales leads are accurate,which solves the problem of sales staff's difficulty in finding customers.It really breaks the traditional B2 B industry's situation of searching for potential customers through traditional manual methods,and liberates the staff.