Jill dyche of primary data management, Evan levy six-level model

Mdm is actually a very important basic information for modern enterprises to carry out daily operations. Almost all production and operation activities of enterprises are inseparable from the support of primary data. Therefore, primary data is one of the most important enterprise information.

However, there is no reference model similar to the SOA reference architecture for MDM management, especially the construction of MDM systems, enterprises or CIOs who are preparing to build MDM or conduct MDM management are often confused. Recently, looking at relevant information, we found that the master data hierarchies of Jill dyche and Levy, this is an important reference.

Based on the complexity of the implementation of primary data management, we can refer to Jill dyche and Evan Levy's view that primary data management can be divided into six levels, reflecting the primary data management (MDM) from low to high) different Maturity. Below we will briefly introduce these six layers:
Level 0: no master data management (MDM) is implemented)
In the case of level 0, there is no data sharing between each application of the enterprise, and no data definition element exists in the entire enterprise. For example, a company sells many products and processes the production and sales of these products by multiple independent systems. Each system processes product data and has its own list of independent products, product Data is not shared between systems. At level 0, each independent application is responsible for managing and maintaining its own key data (such as the product list and customer information), and each system does not share this information, the data is not connected.
Level 1: provides a list
Whether large or small, list management is a common method. Within the company, a logical or physical list is maintained manually. When heterogeneous systems and users need some data, they can obtain this list. Maintenance of this list, including data addition, deletion, update, and conflict handling, is handled by the staff of each department through a series of discussions and meetings. Business Rules are used to reflect the consistency of value. When business rules change or a similar situation occurs, such highly manual management processes are prone to errors. Because list management is manually managed, the quality of list maintenance depends on who participates in the change management process. If someone is absent, list maintenance will be affected.
Level 2: equivalent access (through the interface, each system is directly interconnected with the master data host)
Compared with MDM Level 1, Mdm Level 2 introduces (automatic) management of primary data. The establishment of data standards defines the access and sharing of detailed data stored in the central repository, providing strict support for sharing and using data between systems. Central repository is usually called "master data host )". This knowledge base can be a database or an application system that supports data access and sharing online.

Creating, reading, updating, and deleting (crud) are typical programming terms for processing basic functions. Even in MDM, crud processing is a basic function. If your database only supports crud processing, it does not mean that you have implemented MDM. Mdm Level 2 introduces peer-based access, which means that one application can call another application to update or refresh the required data. When the crud processing rule definition is complete, Mdm Level 2 requires the customer or "equivalent" application to format the request (and data) so as to be consistent with the MDM knowledge base. The MDM Knowledge Base provides centralized data storage and supply (provisioning ). At this stage, rule management, data quality, and change management must be customized and built as additional functions within the enterprise.
Level 3: centralized bus Processing
Compared with MDM Level 2, Mdm Level 3 breaks the organizational boundaries of independent applications, use data standards accepted by various systems to establish and maintain master data in a unified manner (data stored on the master data host of MDM Level 2 is stored separately by each system, ).

Centralized processing means building a universal, target-based platform for MDM. Most companies find that MDM is challenging their existing IT architecture: they have too many independent platforms to process primary data. Mdm Level 3 centrally accesses and controls data used across different applications and systems. This greatly reduces the complexity of application data access, greatly simplifies the management of data-oriented rules, and enables MDM to have more functions and features than a distributed environment. Data of business owners is facing the challenge of consistency. Data exists in different places, and the meaning of the data is different. The data rules vary with systems. Centralized MDM processing-using a public platform as a bus (hub)-describe a consensus and integrate theme domain data from multiple systems, this means that the use of centralized and standardized methods to convert heterogeneous operation data will be integrated no matter what it looks like in the source system. At MDM Level 3, the company manages the content of the topic domain in a centralized manner. This means that the application system, as a consumer or master data, has a consensus that data is the image of the topic data content, breaking the organizational boundaries of each independent application. Mdm Level 3 supports the existence of distributed primary reference data.
Level 4: business rules and policy support
Once data is integrated from multiple data sources, the topic domain view goes beyond the individual application and is displayed as an enterprise view. You will get a single version of the facts. When a single version of the facts is available, the inevitable response from business directors and executives is often: "prove it ". Mdm Level 4 ensures that the primary data reflects a company's business rules and processes and confirms their correctness. Mdm Level 4 introduces primary data to support rules and performs integrity checks on the MDM bus and other external systems. Because most companies are relatively complex, the rules and policies that affect business data access and operations are also relatively complex. It is impractical to assume that any single system can include and manage various types of rules related to the primary reference data. Therefore, if an MDM bus really intends to provide data accuracy within the enterprise, the support for workflow and process integration is essential.
Level 5: enterprise data concentration
At MDM level 5, bus and related master data are integrated into independent applications. There is no obvious separation between primary data and application data. They are integrated. When the primary data record details are modified, the related data elements of all applications are updated. This means that all consumption applications and source systems access the same data instance. This is essentially a closed-loop MDM: all application systems are integrated through a centrally managed primary data. At this level, all systems seem to be the same version of the truth. The operating application system and MDM content are synchronized. Therefore, when a change occurs, the operating application system is updated. In those familiar MDM Architecture styles, the persistent bus architecture, when a bus updates all the operating application systems, will reflect this change, forming a direct operational view of the change. In the registration environment, when data is updated, the bus connects to related system application transaction updates through the web service. Therefore, Mdm Level 5 provides an integrated, synchronous architecture. When a system with permissions updates a data value, all the systems in the company will reflect the change. After the system updates the data value, do not select update for the corresponding value in other systems: Mdm will make this update transparent.

From MDM Level 4 to MDM Level 5 means that MDM functionality is not specially designed or encoded in an application. This also means that primary data transmission and supply do not require dedicated development or support from the source system. All applications clearly know that they do not own or control master data. They only use data to support their own functions and processes. Because of the MDM bus and supported IT infrastructure, all applications can access the primary reference data. After completing MDM level 5, a company will connect all their applications-both operational and analytical-all access to primary data is transparent. For example, when a customer updates her status-do not register the System of the change-data changes will be broadcast to all application platforms (so consistent ). Mdm Level 5 implements the data concept as a service. Mdm Level 5 ensures a consistent enterprise image of the primary data topic domain. Defining "customer" is actually the same as accepting changes to the customer's primary data business rules for other applications. Mdm Level 5 removes the last obstacle to primary data: Unified adoption of data definitions, authorized use, and change propagation.