Recently, the 2021 World Artificial Intelligence Conference "Data Power + Tool Power" to build a new digital base innovation forum was successfully held. The speakers gathered with more than 500 users from different industries to discuss the construction of a new digital base. At the meeting, Liu Chen, CEO of Yushufang and deputy secretary-general of Tsinghua Big Data Industry Federation, shared a speech with the theme of "Empowering Digital Docks and Realizing Digital Asset Management Capabilities Jump". This is a shorthand compilation, everything is subject to on-site information.
【Speech Abstract】:How to improve data asset management from the level of initial practice to a more advanced stage, and achieve a leap in capabilities? Liu Chen, CEO of Yushufang, proposed for the first time the new four concepts of data asset management, including value, synergy, lean and intelligence. He believes that from the previous systematic asset management to more agile asset management, data asset management can quickly achieve results!
The topic I'm sharing is the leap in data asset management capabilities. We explore today's topics from three perspectives, the first is how to understand the data base; the second is how to look at the current situation and challenges of data asset management practices; and the third is how to further achieve the ability transition of data asset management in enterprises and government agencies by combining the creation of data bases.
01
Spanning more than 30 years, what is the connotation of the data base
The data base is not just a data platform, but also a source-side system
Data base usually refers to the data platform in the sense of the main meaning, around 1980, there are many data technologies in the number of warehouses, in 2011 proposed data lake, last year's latest concept proposed is called "lake warehouse integration".
From the perspective of data asset management, data governance and data distribution, data does not only stay in the data platform. In the practice of many enterprises now, the source-end system will coexist with the data platform for a long time, for example, the original on-premises deployment is now going to the cloud, but the source-side system will not disappear after the cloud, or it will become a continuous data provider of the data platform. From the perspective of data management, if the data of the source system and the data platform cannot be managed well, the foundation of the data is not solid. From the perspective of using data, if the source data is not reliable, the data platform must also be unreliable. Therefore, from the perspective of data asset management and data governance, we must lay the foundation of all the data of the enterprise.
We do a lot of data application, mining, analysis is to get the value of data, through the construction of the data base, platform, with the ability to mine, analysis, but if the data assets are not solid, data is missing, digital transformation will be hindered, we have to lay the entire enterprise data base.
The current state and challenges of data asset management
The theory of data asset management, whether at home or internationally, has been developed for many years and is now relatively mature. The international theoretical model DMBOK, a guide to the data management body of knowledge, is divided into 10 to 11 areas, including data architecture, data model, data security, data warehouse, etc.
The national standard DCMM - data management capability maturity evaluation model released in China in 18 years relies on international data theory to divide data governance into eight fields such as data strategy and data governance. For example, we must carry out data-related underlying design, data strategic planning, data-related organization and construction, and data application, while data security, data quality, data standards, and the entire life cycle of data also need to be managed. In the DCMM model, through 8 dimensions, 28 sub-domains, and more than 400 evaluation indicators, the complete data capabilities of a unit will be evaluated.
These theories are relatively mature at the national and international levels, and many companies are practicing them. But in the process of doing data governance practice, our country in 2005 to 2010 mainly in the field of banks and operators to do data governance, between 2010 and 2015 banks went relatively fast, other industries began to initial practice, after 2015 all industries began to do data governance, because big data rises, everyone began to pay attention to data governance, data asset management. In 2018 and 2019, data came to the forefront, and various industries attached more and more importance to and practiced data asset management.
02
Years of practice, full of confusion: construction is more than enough, and the results are insufficient
Many enterprises have done the top-level planning of data governance, and may ask large consulting companies to do tens of millions of consulting and planning projects, but they find that they do not know how to land, and the top-level design and landing practice are disconnected. Senior leaders originally attached great importance to data governance work, but found that the effect of doing it for half a day was not too obvious, and over time it would stop.
Business units will think that data asset management, data governance is still more technical, they find it is not clear how to participate in these work, the real value of data asset catalogs to business units is not particularly obvious, so the participation of business units is also very low.
The digital team will feel that we have also built an organization, also released the system, and also done a lot of consulting and technical work, data models, data standards have also done a lot, but because the leadership is not very recognized, the participation of business departments is very low, the business value is not obvious, so the digital team is very confused, what to do to make the data asset management work effective?
In the process of data asset management, the implementation of the process is still a lot of manual, many in the Internet field or in the data analysis, the use of these areas have a relatively mature intelligent technology, but in the field of data governance has not been well applied, people doing this work are physically and mentally exhausted. This is the four main problems that we have put together, how to solve them?
03
With the concept of the new four modernizations, promote the transition of data asset management capabilities
I put forward some solutions from my own perspective, how to improve data asset management from the level of initial practice to a more advanced stage, to achieve a leap in capabilities? Here we put forward the new four modernization concepts of data asset management for the first time.
The first is value. The main embodiment of value is that no matter what kind of data asset management work you do, the leadership cares about and the business department cares about should be your point of strength, you must carry out data asset management work around the business scenarios, business processes, and business applications that everyone cares about most, so that everything we do can help the business department improve their efficiency, and even help them reduce costs and increase efficiency.
Value-based emphasizes the need to carry out data governance work for business scenarios. A lot of work has been done, such as data accountability, but the origin should be the business scenario. In the past, the data team often did a lot of technical rule checks, and then lost to the business department, we should now start from the business scenario, in the business scenario to carry out all the data asset management work, then the participation of business personnel will be very high, because what you do will directly help them.
The second is synergy. Why put it in second place? In the past, we did data asset management is it IT team, information team has a small team or department of information governance, by these people with external service providers or product vendors to sort out data standards, data asset catalog construction, data standard construction, business department participation is very low. Moreover, these tasks are often in the headquarters of the group or the headquarters of a provincial company, and the participation of the business personnel at the grass-roots level below is also very low, or very fragmented, and the work is also very fragmented. So we need to bring the two together with the idea of synergy.
Collaboration, we believe that data asset management to achieve several collaborations, is to take the enterprise's data assets as the core, first of all, the data assets themselves these small cycles, small work collaboration, including the construction of the data middle platform, the construction process of the data base, the coordination of the operation and maintenance process. There is also the need to promote data-based collaboration between multiple businesses. What is the basis of this synergy? In fact, it should be our data accountability system, from the company's decision-making level, there is a leading group for data asset management, each business department has its own data focal department, Huawei is called data steward, and some are called data management specialists. We must pay attention to the front-line grass-roots business personnel should also participate, so there will be many special working groups in the middle, and the front-line personnel must have special working groups in order to solve specific data problems, so form a three-dimensional data accountability mechanism from horizontal to horizontal and vertical to the end, so that all parties can effectively coordinate.
The second collaboration we are talking about is that the previous data governance work was done inside the data center, which is not enough, we need to extend to the source business system. Whether it is deployed at the source or on-premises, if you only do data governance in the middle office, there are still breakpoints in many upstream business processes, and if you can't improve in the business process, there will still be problems. From the application side to the data inside the data middle office to the source-side business layer, it is necessary to coordinate.
The third is lean. Lean refers to the single point of work to further help us form small closed loops and large closed loops.
Lean talks about the work of data management of these assets, including data standards, data accountability, and the need to form its own small closed loop. From the formulation of standards to the implementation of standards to the landing of standards, a small closed loop needs to be formed. Second, we must integrate these tasks into our entire data platform construction process, whether it is building a business system or intelligence. And each stage should be integrated into the work related to asset management to achieve lean. At the same time, in the DCMM standard, emphasis is placed on quantification, and many enterprises evaluate that it belongs to the second level, in fact, after a period of practice, the quantitative indicator system can reach the fourth level.
Finally, through intelligent technology to truly improve the efficiency of data governance work. In the past, it was all about people combing and inventorying data assets, but now it is necessary to build an intelligent engine to manage the internal data assets of the enterprise based on technologies such as knowledge graphs.
Finally, let's briefly summarize that in the current digital transformation process, data asset management is no longer a traditional concept, and some new changes are needed. Our goal should be to pay more attention to the embodiment of value, and the scope should pay more attention to the global data scenario, including the ability of tools, as well as the institutional ability and organizational ability of data asset management software. Our model will change from the previous systematic asset management to more agile asset management, so that we can achieve rapid results!