Thinking of building Data team

In the face of more and more hot http://www.aliyun.com/zixun/aggregation/14294.html "> of large data, will be in Jinshan software two years experience and later thinking simple do a summary, hope to have some usefulness."

The main problem facing the data team is the concept, build an internal wiki platform, through the relevant rules set, the concept of Intelligent learning organization to follow, there are many benefits.

First, the data-related content "defined", in the process of continuous improvement to form a team concept, everyone can participate in it, both the formulation, but also the executor, avoid communication common many non-technical problems arise.

Second, Team Running-in is a long-term process, is a mule is a horse, come out to slip. Wiki's approach is relatively fair and intelligent.

Third, the cumulative value of the superposition of ways to reduce the risk of failure.

Establishment and improvement of data-related logic database

With the help of wiki platform, the logic relation between business and data is summed up and summarized, and the knowledge of data is never too much.

Data processing Gadgets

"ID Number" "" Mobile phone Number "" IP Address Library "" Chinese words "and so on data analysis tools are complete, the important role of data preprocessing is no doubt, the processing of data noise is sometimes more practical than the data analysis model.

Direction and phased target

The direction of data analysis is often neglected, only the direction is correct, others are valuable.

Data is a long-term work that is beyond the comprehension of the average person and needs to be decomposed into executable phases.

From the failure to sum up the lesson, the person is the most essential reason, the team builds needs the reasonable combination, the individual strongest is inferior to the overall coordination.