4. The core MIS construction of enterprises in MIS construction takes a lot of human and labor, and what is it going to get? This is the problem that the user is most concerned. Generally speaking, it is desirable to get as much as possible with advanced hardware devices and as many functions as possible. In fact, as the American scholar Martin is pointed out when data-centered: The data processing process of the enterprise is variable, but its basic data is stable. We use the categorical organization to the data platform in the database. The fundamental task of MIS construction is to organize data platforms under manual mode into a unified data platform. Because MIS hardware devices are replaced due to the upgrade of the device, the functions developed will also be replaced due to changes in management; and the data platform can be upgraded as the database is upgraded. As long as the basic data does not change, it can have a long time. Use it. Therefore, the data platform is the core of the construction of the system. Systems with data platforms have the following features: systematic. Based on the unified data platform, various summary and analysis results can be obtained by the system, thus reducing processing links, achieving management of management work specifications and saving human resources. In this way, the data processing work of each department is connected to MIS to form a systematic processing. rationality. The data in the enterprise has different categories, and the data is classified according to these categories. Some departmental subsystems are independent, and the amount of data is also large, which belongs to different categories, which constitute the main structure of the data platform. The data platform organized by this slogable class has a reasonable structure. In addition, the accumulation and improvement of the data platform is arranged, so on the basis of the unified data platform, the system can guarantee the system a unified and reasonable whole. Adaptability. In actual management, a large number of data processing does not require complex mathematical calculations, as long as there is a detailed and reasonable data, with existing development tools, it is still more convenient. On the other hand, the various functions developed are directly obtained by the data platform, which is independent and functional changes only involve independent procedures, which makes the system's functionality and changes, can adapt to actual Working needs. In order to form an MIS with a data platform as the core, the following method can be adopted: the method of dividing the management. For larger systems, the original system can be decomposed into multiple small systems according to the category of the data, which arrays their development order, thereby converting the problem of establishing a unified data platform to establish multiple categories of small data platforms. The problem. By developing various subsystems, various types of data platforms are established, and finally, it will become a unified data platform. Gradual method. Identify a large amount of data must pass the user and have a process. To this end, in the process of developing each subsystem, it is necessary to invest in the trial operation as soon as possible, according to the various problems proposed by the user, to attract users by meeting the needs of users; and in the process of repeated modification, truly Understand the properties of the user's intent and data, on this basis can organize data to achieve the purpose of establishing a data platform. Direct abstract method. The general development habits are to describe the original system with a data flow diagram and developed on the basis of these materials. In fact, this is a system structure that simulates the original working mode and does not form a system structure that is the core of the data platform. The organizational data only needs to grasp the source of data, the category, and the frequency of the frequency, and directly describe the data platform directly from the structure of the data platform, the inlet of the data, and the frequency of data. After organizing data, you can directly develop programs to get a variety of results.
