Data Governance Series 1: data governance framework [interpretation and analysis]

One 、 What is data governance ？

Wikipedia ： Data governance is important to ensure the accuracy of data 、 Moderate sharing and protection are essential . An effective data governance plan will improve decision-making 、 Cut costs 、 Reduce risks and improve safety compliance , Give value back to the business , And ultimately reflected in increasing revenue and profits .

The author thinks ： All the businesses that are going on to improve data quality 、 Technology and management activities fall into the category of data governance . The purpose of data governance is to control data resources effectively , Control the data , In order to improve the quality of data and the ability to realize data .

Two 、 Why data governance is needed ？

In our country , The informatization development and construction level of various industries are not balanced , Some industries are just starting . however , Whether it's the financial industry 、 Communication industry 、 Real estate industry 、 Traditional manufacturing and agriculture , The development of its informatization basically follows “ Nolan model ”. The author thinks that enterprise informatization has roughly experienced the initial chimney system construction 、 There are three major stages in the construction of integrated system in the middle stage and data management system in the later stage , It can be said that it is a process of construction before governance .

1、 The level of data quality is uneven

In modern times ,“ Data capitalization ” This concept has been understood and accepted by most people . Whether it's a business 、 Government or other organizations , More and more attention is paid to the management of data assets . However , Data is not an asset , That is to say, not all data are data assets , There's also junk data in the data . What we need to govern is the data assets that can create value for the enterprise , Not all the data .

2、 Data exchange and sharing are difficult

At the initial stage of enterprise informatization construction, there is a lack of overall informatization planning , Most of the system construction is single architecture system or package software driven by business department , Data scattered in these architectures is not unified 、 Inconsistent development languages 、 Database diversity in the system , There's even a lot of data stored on employees' personal computers , As a result, one by one “ Information Island ”. these “ Isolated island ” There is no effective connection between them , Data cannot be interconnected , You can't communicate meaningfully according to the user's instructions , The value of data cannot be fully realized . Only Unicom data , Eliminate these “ Information Island ”, In order to achieve data-driven business 、 Data driven management , To really release the value of data .

3、 Lack of effective management mechanism

at present , Many enterprises have realized the importance of data , And try to control the data flow through the business flow of the production system , However, due to the lack of effective management mechanism and some human factors , In the process of data flow , There are data maintenance errors 、 Duplicate data 、 Data inconsistency 、 Incomplete data , This leads to a lot of junk data . Unclear data property rights , Confusion of management responsibilities , Unclear management and use process , It's an important factor that causes data quality problems .

4、 There are data security risks

2018 year 3 In the Facebook 5000 Ten thousand users' information is leaked and abused , Affected by the incident ,Facebook Share prices fell sharply that day 7%, Market capitalization 360 More than $ , The Cambridge analysis company, which embezzled the data, was also established in the same year 5 Monthly stop of operation , And apply for bankruptcy . This kind of data security incident , It happens more frequently in our country , I remember clearly ：2011 year , Hackers have made it public on the Internet CSDN User database for , the height is 600 More than ten thousand plaintext registered email accounts and passwords have been exposed and leaked ;2016 year , SF employees should steal a large amount of customer information and be taken to court ;2017 year , JD employees steal users' personal information 50 Billion bars , And through various ways in the network black market . In recent years , With the development of big data , There are so many data security incidents like this . Data asset management , It is developing from traditional decentralized manual management to computer centralized management , People pay more and more attention to the security of data .

3、 ... and 、DMBOK Data governance framework for

DMBOK It's by the data management association （DAMA） Professional books on data management , a copy DAMA Data management dictionary . It has certain guidance for the construction of enterprise data governance system . notes ：DAMA Is the abbreviation of Data Management Association , Is a non-profit association of global data management and business professionals , Committed to the research and practice of data management .

DMBOK Divide data management into the following 10 Two functional areas ：

Data control ： Planning at the level of data management and usage 、 Supervise and control .

Data architecture management ： Define the blueprint for data asset management .

Data development ： Data analysis 、 Design 、 The implementation of 、 test 、 Deploy 、 Maintenance, etc .

Data operation management ： Provide technical support from data acquisition to data cleaning .

Data security management ： Ensure privacy 、 Confidentiality and appropriate access rights, etc .

Data quality management ： Definition 、 Monitoring and improving data quality .

Reference data and master data management ： Managing golden versions and copies of data .

Data warehouse and business intelligence management ： Achieve reporting and Analysis .

File and content management ： Manage data outside the database

Metadata management ： Integration of metadata 、 Control and provide metadata .

Four 、 Understanding and interpretation of data governance framework

DMBOK It gives framework suggestions for enterprise data governance , But any guiding framework document is not omnipotent . Different industries 、 Enterprises of different nature 、 Different degrees of informatization 、 Different corporate cultures , Its data governance scheme must be adapted to local conditions , Tailored . We often say ： There is no best solution, only a more suitable one . When enterprises implement data governance , Full analysis and evaluation should be done , Don't follow the wind blindly , Avoiding data governance has little effect , It's a waste of investment .

The author thinks that enterprise data governance should consider the following elements ：

1、 The object of data governance

Everyone is talking about data governance , But what data needs to be governed ？ We say data governance is not all data governance , It's about governance of enterprise data assets . that , The problem is coming. , What are data assets ？ And how to identify data assets ？

Wikipedia definition ： Data assets belong to the digital property of ordinary individuals and enterprises , Data assets are an extension of intangible assets , No physical form . Its essence is that data, as an economic resource, participates in the economic activities of enterprises , Reduce and eliminate the risks in the economic activities of enterprises , Provide a reasonable basis for the management control and scientific decision-making of the enterprise , It is expected to bring economic benefits to the enterprise .

The author thinks , Data assets do not have physical form , But it must be a virtual form of physical mapping in the network world . For enterprises , people 、 equipment 、 product 、 materiel 、 software system 、 database 、 And any kind of data that involves using files as carriers , All belong to the data assets of the enterprise .

Although we have defined data assets , But the data governance focus of different industries is also different . Data governance needs to understand the needs of the industry 、 Enterprise demands , In different industries 、 Different enterprises should have different differentiation schemes . When enterprises implement data governance , The first step is to identify and define data assets , Define the object and scope of data governance , Do a good job in the top-level design of data governance ！

2、 The opportunity of data governance

In recent years, I visited some enterprises due to my work , The economic situation is different 、 The characteristics of the industry are different 、 The degree of informatization is different 、 Data governance is not the same .

The first kind of enterprises ： The economy is strong , Informatization started earlier , The degree of informatization is relatively high , Such as ：XX Bank 、 The state grid , They have formed a systematic data governance system .

The second kind of enterprises ： Have a certain economic strength 、 The degree of informatization is relatively good , But the early information blind suggestions , Bought a bunch of software packages , Built a bunch of systems , Although the system is used more or less , But the effect is not good , Talking about data governance , The customer felt headache ： What data does an enterprise have ？ Where is all this data distributed ？ How to start data governance ？

The third kind of enterprise ： Economic strength is relatively weak , There are also enterprises that have just started informatization , Most of these enterprises still rely on paper or offline business models , Some enterprises use financial software or ERP System , Data is stored in personal computers or production systems , There is basically no data governance . Most of the small and medium-sized private manufacturing enterprises in China are at this level .

How to choose the time of enterprise data governance ？ It's data before governance , It's better to build a good data governance system before building an application system ？ For the above different types of enterprises , The timing of its data governance choice and the design of its system construction can never be generalized .

For the first kind of enterprises , There is a relatively perfect data governance system , What's more, we need to strengthen data security 、 Data applications 、 Data innovation , Steadily improve data management 、 Data application and data liquidity ;

For the second kind of enterprise monomer architecture system, there are many , Information island serious , There must be multiple sources of data 、 repeat 、 Inconsistencies, etc , Its data governance is imminent ;

For the third kind of enterprises , In the wave of digitization , Although informatization is weak , But if we have a good data base , It's not enterprise reform and innovation , Realization “ Overtaking in curve ” The best time .

3、 Who will implement it 、 Who will lead

Enterprises often have such a misunderstanding , Many people think that data governance is a matter of the information sector and has nothing to do with the business sector . Previously, we said that data governance is the governance of enterprise data assets , Since it's an asset , We must confirm the right . The production of enterprise data assets 、 Use should have clear responsibility department , Obviously, the production and ownership Department of data assets should be the business department , At most, the information department is just a data asset custody department . The author has also repeatedly stressed the data problem of enterprises ,80% It's about business and management ,20% It's a technical problem .

therefore , Enterprise data governance should be led by senior leaders , The business department is responsible for , The information department implements , The participation of all members of the enterprise . All employees of enterprises should cultivate data thinking and data awareness , Of course, it's a long-term process , It's also not easy , It needs to accumulate and precipitate bit by bit , And continue to integrate into the corporate culture .

4、 The content of data governance

Data governance is long term 、 Complex engineering , It involves the organizational system 、 Standard system 、 Process system 、 Technical system and evaluation system , Contains data standards 、 Data quality 、 Master data 、 Metadata 、 Data security and other aspects . Because of the nature of the business 、 Business characteristics 、 Different management models , It is necessary to establish a data governance framework that meets the current situation and needs of enterprises , To guide the development of enterprise data governance .

The following is my personal understanding of the data governance framework , I hope the experts in the industry can correct the deficiencies , Looking forward to communicating with you ：

Organization system ： The implementation of data governance projects is by no means a departmental matter , It cannot be solved in a single Department of the enterprise . It needs to be considered from the whole organization , Establish a professional data management organization system , Confirm the right of data assets , Clarify the corresponding governance systems and standards , Cultivate data governance awareness throughout the organization . This needs to be IT Collaborate with business units , And must work together consistently , To improve the reliability and quality of data , So as to support key business and management decisions , And ensure compliance with regulations .

Standard system ： The standard system of data governance is multi-level , Include ： international standard 、 National standard 、 Industry standard 、 Enterprise standards, etc . The content of the enterprise data standard system shall cover ： Metadata standards 、 Master data standard 、 Refer to data standards 、 Data index standard, etc . The effectiveness of data governance , To a large extent, it depends on the rationality of data standards and the degree of unified implementation . The construction of enterprise data standard system should not only meet the current actual needs , It can also focus on the future and be in line with national and international standards .

Process system ： Data governance process system , Provide evidence-based management measures for the development of data governance 、 Specify the business process of data governance 、 Accountability system for data governance 、 Personnel roles and job responsibilities 、 The supporting environment of data governance and the promulgation of rules and regulations of data governance 、 Process, etc . Establish data production 、 The circulation 、 Use 、 file 、 Eliminate the entire lifecycle management process . Enterprises should focus on the object of data governance ： Data quality 、 Data standards 、 Master data 、 Metadata 、 Establish corresponding systems and processes for data security and other contents .

Evaluation system ： The establishment of data evaluation and assessment system is the basis for enterprises to implement and implement the relevant standards of data governance 、 The foundation of system and process . Establish a clear assessment system , In practice, a data accountability system can be established according to the specific conditions of different enterprises and the requirements of their future development , Set assessment indicators and assessment methods , And linked to personal performance . Assessment indicators include two aspects , One is the production of data 、 Evaluation and assessment indicators of management and application processes , The other is the evaluation index of data quality .

Technical system ： Data governance includes data governance tools and technologies , The overall shall include metadata management 、 Master data management 、 Data standards management 、 Data quality management and data security management .

Metadata management ： Metadata management is the management of business metadata involved in an enterprise 、 Technical Metadata 、 Manage metadata for inventory 、 Integration and management , According to science 、 Effective mechanism to manage metadata , And for developers 、 End users provide metadata services , To meet the business needs of users , Development of enterprise business system and data analysis platform 、 Support the maintenance process . With the help of change report 、 Impact analysis and other applications , Control data quality 、 Reduce the ambiguity of business terms and establish a good communication channel between business and Technology , Further improve the credibility of various data 、 Maintainability 、 Adaptability and integrability .

Data standards management ： Data standards apply to business data description 、 Information management and application system development , It can be used as the standardized definition and unified interpretation of the data involved in operation and management , It can also be used as the basis of information management , At the same time, it is also the basis for data definition during application system development . Involving national standards 、 Industry standard 、 Enterprise standards and local standards , Associate when defining metadata entities or elements . Data standards need to be continuously supplemented and improved 、 Update, optimize and accumulate , In order to better support business development and system integration .

Master data management ： Master data management is through the use of relevant processes 、 Technology and solutions , Effective management process of enterprise core data . Master data management involves all participants in master data , Such as user 、 Applications 、 Business process, etc , Create and maintain enterprise core data consistency 、 integrity 、 Relevance and correctness . Master data is widely used and shared data inside and outside the enterprise , Known as one of the enterprise data assets “ Gold data ”, Master data management is the fulcrum of enterprise digital transformation , It is the core part of enterprise data governance .

Data quality management ： Establish data quality management system , Define data quality management objectives 、 Control objects and indicators 、 Define data quality inspection rules 、 Perform data quality check , Production data quality report . Through the data quality problem processing process and related functions, the closed-loop management from discovery to processing of data quality problems is realized , So as to promote the continuous improvement of data quality .

Data security management ： At present, most people know that data security is very important , But in reality , Data security is often overlooked , Only when there are data security problems or even accidents , People realize that they have to do something for data security . Data security should run through the whole process of data governance , We should ensure that management and technology walk on two legs . In terms of Management , Establish data security management system 、 Set data security standards 、 Cultivate the data security awareness of all staff . Technically , Data security includes ： Data storage security 、 Transmission security and interface security . Of course , Safety and efficiency are always a contradiction , The stricter the data security control , The more limited the application of data . Enterprises need to be safe 、 Find a balance between efficiency .

5、 ... and 、 Data governance framework summary

Again , The implementation of data governance needs to be adapted to local conditions , No matter what kind of data governance system is established 、 What kind of data governance technology to adopt , The purpose is to achieve the goal of data governance , namely ： Through effective means of data resource control , Management and control of data , In order to improve the quality of data and the ability to realize data . Data governance system and framework , It's just a reference for enterprise data governance , Can't copy and apply , We can't govern for the sake of governance .
————————————————

Link to the original text ：https://blog.csdn.net/kuangfeng88588/article/details/89952992