Data operation is not difficult. The key is have you found the "G" point?

On January 15th, the first in-depth training for Alibaba Cloud big data partners was held in Beijing. Our company was invited to participate. I and two R&D students came to Alibaba Wangjing Park again.

> The content of the first part of the training, the introduction and application of Shujia

In addition to introductory content, there are still "dry stuff", which is the practice of MaxCompute. MaxCompute was originally called ODPS, www.aliyun.com/product/odps, and is the foundation of Ali's entire big data solution.


Come up, tell us what data analysis is about, say we are here to learn big data, why do we need to study data analysis?
Because: big data is not for blowing! There


are many benefits that can be obtained through data analysis, but the most important thing is to obtain the ability to operate data!
Why do data operations?
Because: Baidu, Alibaba, Tencent, Apple, Google, Amazon's operations are digitized one by one, and one game is making money!


Let's talk about data warehouses again. Why do we talk about this? Don't we want data-based operations? How to operate without data, if there is too much data, it will naturally need to be stored in the warehouse!
The field of data warehousing has been developing for more than 30 years. Big data is not subverting, but inheriting and expanding related domain knowledge.
The analysis and processing process of "big" data and "small" data is basically the same.


Have you heard of star schemas? So have you heard of the paradigm model?
Anyway, in data analysis application scenarios, data duplication is not much concerned, and data is usually organized in a star model; online business application scenarios need to avoid duplication of data storage as much as possible, and a paradigm model is usually used for data storage.
A star schema is a data modeling method that organizes data into dimensions and fact tables. Data organized in a star schema is more conducive to understanding and analysis.
This modeling method has been used for data warehouses for 30 years, and big data warehouses will continue to be used.
Click to read the full text: http://click.aliyun.com/m/9753/