Big Data-Spark-based Machine Learning-Intelligent Customer System Project Practice

Big Data-Spark-based Machine Learning-Intelligent Customer System Project Actual Combat

Network Disk Address: https://pan.baidu.com/s/19-U_XwIUZUENNgmz6VysMQ Password: j288

Alternate Address (Tencent Micro Cloud): https://share. weiyun.com/90d4d9196a9d86c74aabe8df78cdc624 Password: uq2q3w


Project background:
1. Nowadays, many platform customer relationship management systems use traditional CRM, all information is manually entered, and users can only find and mark upstream and downstream enterprises or information manually. .

2. The arrival of the era of big data, artificial intelligence, cloud computing, advanced analysis, and the replacement of manual parts with machines. From the trend of 2016, machine learning and artificial intelligence (AI) will completely change CRM in the next few years. .

Therefore, we have transformed the CRM, which mainly includes the following parts:
1. Enterprise information and commodity information come from the Internet, and machine learning is used to automatically analyze and classify.
2. The commodity information entered by the user and the commodity information newly crawled from the Internet are all classified by the model calculated by machine learning.
3. Machine learning automatically calculates the upstream and downstream of enterprise and supply and demand information.
4. Machine learning automatically optimizes the computing model at regular intervals.


The basic idea of ​​the whole project is how to crawl a large amount of data into Hbase through crawlers, and then use ETL tools to initially transform and filter the data to mongodb, extract mongodb data for cleaning, calculate the model and put it in hdfs. Subsequent incoming data is used to calculate the type of data through the model. The project system mainly includes front-end + back-end + machine learning. The front-end uses React Native, Native, the back-end uses Dubbo+Spring+java, and the machine learning is implemented by Spark. The machine learning-spark code of this project runs on mesos.