Whether you are a simple database system, or a complex system is even as a data warehouse, use Sybase or Oracle, and even simple use of Access or FoxPro, as long as it is a relational database, it is inevitable to design the tab. When you write code, you don't have to consider these things, because the previous design work has already done it, it seems to be the same, isn't it the second paradigm or a third paradigm? Later, I found out that I didn't easily, and the bad data structure made of the table, all the code has changed, which is probably what we have experienced by the database system. After several lessons, I explored the way I found data, and I took it out with you. (Take the library as an example, the rotten can not be rotatable).
The first step is to find some object. Database management is the data required by the company or department, and the business is related, and it is inevitably related to the object. You said that there is no matter in the personnel system. In fact, the files in the personnel system are the original form of data. It is not a file before the computer works, which is the object of management. The things in the book management system are books. According to the corresponding book number, author, publishing house, published an annual month, inventory, inventory location and other data items. These data are often the focus of data consistency and integrity consideration.
In the second step, find someone. People-oriented, people around the corresponding items must have relationships. In the book management system, there is natural book administrator and readers. The database management system itself is to facilitate man-managed data, but it often likes to record the situation of managed data, preventing unknown responsibilities, and a main line in data management. It is used to call the library administrator as an employee. The reader is called the customer, and the corresponding ID number, name, gender, photo, address, identity, etc.
The third step is to find a document. There is often a large number of documents in a specific business, such as enterprising orders, book orders, but also books, etc. At this time, you will find that people who have already found wells in front of them are used here. If the analysis is good, the data items needed here are basically available, as long as the reference is available. Because these documents are often used, the main indexing is often determined here.
Step 4, find reports and summaries. Statistics and reports are indispensable features of each database system. Its source is often the need for actual business. Some have ready-made reports. If you have, you will think that you can't make it easy for users to work to the superior department. The system is reported to summarize the following level, which is not recognized by the user's especially managers, you will wait for the rework. At this time, full communication will further improve the data items designed in the first three steps. In the book management system, we not only designed reports of commonly used collections, lending statistics, update statistics, etc., and daily business-related reports, but also made a variety of book utilization statistics, damage (lost) statistics, lending statistics, etc. The decision-making report, these designs are occasionally seen from the routine report of the administrators, which greatly meets the actual use of the system, but also not mentioned in many books.
At this point, most of the data items should be found. An additional data item is a routine business process analysis, from the object to the object or from the item, find new business relationship data, which will also be a large change in the rapid model method. There is no way, the update of the management method will inevitably bring picker changes, but basic above four data is not much change.
