Shulou(Shulou.com)06/03 Report--

Recently handed over the previous big data project, to make a summary of the previous project content. It can also be regarded as combing the structure of the project, and it can also be regarded as a summary of the early stage, laying a foundation for later study.

Clean up data

For traditional industries, it is generally a gimmick to say that they want to engage in big data, because the amount of data before is not very large, so it is basically based on some statistical analysis. At this stage, your understanding of the data is particularly important! The knowledge involved here includes data cleaning and related ETL technology. In other words, you have to do data analysis, where the data is very important, when you do not know the location of your data, your analysis will have nothing to talk about. And there must be a lot of problems in the original data. At this point, your cleaning process is to gain an in-depth understanding of the original data. It is also particularly important to say that a good data analyst must be a good business person. Because only when you know more about the data, you can better complete and replace it. To put it more colloquially, you have to convert the raw data into data that PC can read.

There is also a 4:3:3 rule that your raw data should be trained from the three dimensions of testing, training, and verification to form a cycle so that your data can be more successful in the end. When your data is stored, it is also important to use structured or unstructured data. It also determines the speed of your later reading!

Analysis data

This step is to be done in conjunction with the business, how much do you understand the business. Combined with business needs to analyze data, rather than simply understand the data, different industries and different types of work have different understanding of the same data. In contrast, business people need to have a deeper understanding of the data. How do you analyze your data and understand the special values in it. How to find the target data you require is particularly important.

Analyzing the data also has a bearing on the success or failure of your project. This personal feeling is also an important area for product managers to grasp. First of all, as a product manager, you can't know all the industries very well, in this case, you are bound to be able to understand the value of the data to the maximum. In this step, you need to communicate deeply with the business staff to ensure that you have a detailed understanding of the data before you can stand out in the next step.

Algorithm selection

Some people say that this point involves research and development, as a product manager does not need to focus on. But from a personal point of view, this is equally important. Because the improper selection of your initial algorithm will lead to errors in the later results. In other words, you have to choose the basic things as soon as you come up.

In the aspect of algorithm selection, the personal feeling is to combine the business to implement. First of all, it is necessary to figure out what indicators are the main focus of the business side. The parameters related to this index are those, and how these parameters affect these indicators. As for the accuracy of the algorithm, this can be continuously improved by refining the granularity of the data. Different codes have different resource scheduling for the system, and if your understanding of the algorithm determines the speed of your final product response to the maximum!

Demand analysis

Some people say that this piece is the most important. Why don't you put it in the first part, but in the last part. Because deeply feel, in the traditional industry, the needs of users are not clear, or not so clear. Or maybe the user's needs can be guided. For a long time, individuals have divided the needs of users into four categories: strong demand, weak demand, real demand and false demand.

Sometimes, it is necessary to distinguish these needs. The product manager is required to have a background in the relevant industry. Because different industries, different companies have different needs for people. How to mine the needs of users and transform these requirements into products that can be realized on the ground. This is very demanding for the product manager.

Departmental communication

Big data product, I will divide it into three lines, one is the product, one is the business, the other is the research and development. This involves communication between departments. Many users of the business need to go through the product people to feedback to the R & D, and the R & D also needs the product people to put their work into the actual project.

Big data, to the above. The leadership may not understand what big data can do. This requires the product staff to make it clear to the leadership in a popular language. For cooperative manufacturers, there must be correct guidance in order to let the other side see the possibility of cooperation. So as to provide power for the development of the project.

Big data project, from the point of view of a product manager to participate in this project, only to find that what he has learned is so little in practical application. The desire of the traditional industry for big data is no longer just based on the concept, but the real landing, the real auxiliary business to create value. In this respect, the demand for a product manager will only become higher and higher.

I am glad that the opening topic of the paper passed smoothly some time ago. Big data still has a long way to go. Let's go and cherish it.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.