A method for building an enterprise data asset model based on data middle platform

Abstract

The present invention discloses a method for constructing an enterprise data asset model based on a data middle platform, including the following steps: business combing; demand analysis; business design; data demand combing and tracing; data collection and extraction; analysis layer data access technology route design; multi-dimensional and multi-layer data tracing; analysis model design; analysis model construction; analysis model training and evaluation; analysis model optimization; operating environment construction; application deployment and verification; configuration implementation. The present invention improves the enterprise platform function, studies and establishes a power grid benefit analysis model, and connects the links of people, money, and materials.

Description

Technical Field

[0001] The present invention relates to the technical field of market model construction, and in particular to a method for constructing an enterprise data asset model based on a data middle platform. Background Art

[0002] Current status of financial data middle platform construction: In recent years, the company's finance department has shown an accelerating trend in the construction of big data analysis. In 2019, a financial data asset platform was built, including data analysis applications for accounting, budget management, asset management, engineering financial management, electricity price management, and fund management. Starting from 2020, based on the overall deployment of State Grid Corporation on the construction direction of the data middle platform, the company's finance department has made efforts in data analysis applications based on the data middle platform to build a "smart shared financial platform" with six modules including key indicator dashboards, intelligent analysis centers, report management centers, professional management centers, business service centers, and grassroots service centers.

[0003] In order to implement the State Grid Corporation's work to improve quality and efficiency, assist the scenario application of the smart shared financial platform, and better ensure the practical application from the data level, it is necessary to carry out the construction of enterprise data models. Summary of the invention

[0004] Based on the technical problems existing in the background technology, the present invention proposes a method for constructing an enterprise data asset model based on a data middle platform.

[0005] The present invention proposes a method for constructing an enterprise data asset model based on a data middle platform, comprising the following steps:

[0006] S1 Business review: Investigate the data characteristics, data quality, storage methods, management methods, and business logic of the enterprise's auxiliary business decision-making indicator model, and analyze and study the survey results;

[0007] S2 Demand Analysis: Based on the research, accurately understand the specific requirements of the business needs and transform the business needs into a complete demand definition, that is, to refine the result indicators to the smallest unit;

[0008] S3 Business Design: Based on the results of project demand analysis, design the indicator factor composition to assist business decision-making, determine the specific business content that the project needs to achieve, and form a design document;

[0009] S4 Data demand sorting and tracing: sort out the specific data requirements of the indicator model that assists the enterprise in business decision-making, determine the data source, and determine the data format and screening conditions for master data extraction;

[0010] S5 Data Collection and Extraction: Collect data from internal and external business systems or offline, extract required data, export enterprise master data list, export factor data related to indicator models that assist business decision-making, and analyze and process them;

[0011] S6 Analysis Layer Data Access Technology Route Design: According to the requirements of the indicator model architecture for supporting business decision-making, complete the data access technology route, including the design of the historical stock data extraction technology route and the incremental data extraction technology route;

[0012] S7 multi-dimensional and multi-layer data traceability: decompose the relevant factors involved in the indicator model of the enterprise's auxiliary business decision-making, clarify the traceability logic, write a traceability demand application form, and clarify the granularity, time period, data acquisition logic, unit and other related information of the factors;

[0013] S8 Analysis Model Design: According to the needs of enterprise management, construct an indicator calculation model for assisting business decision-making, sort out the association relationship based on the data middle platform system table, design a wide table structure, and store the calculation results of the indicator model for assisting business decision-making;

[0014] S9 analysis model construction: Based on business content, business rules and data requirements, it is developed in the data center to establish an indicator model for assisting business decision-making. The calculation results are stored in the wide table of the data center.

[0015] S10 Analysis Model Training and Evaluation: Determine the unknown parameters in the indicator model that assists business decision-making through existing data;

[0016] S11 Analysis Model Optimization: By changing the parameter configuration of the indicator model that assists business decision-making, the model is optimized to improve the extraction and calculation efficiency;

[0017] S12 operating environment construction: complete the system environment and big data resource planning and installation configuration required for the operation of the indicator model for supporting business decision-making;

[0018] S13 Application deployment and verification: Deploy the indicator model installation and deployment work for enterprise auxiliary business decision-making, and complete the corresponding security, performance, function and stability verification work;

[0019] S14 configuration implementation: Deploy the indicator model that assists the enterprise in making operational decisions to the data center.

[0020] Preferably, for enterprises, an indicator factor system is constructed, indicators are calculated by factors, scores are assigned to indicators, and score database results are calculated.

[0021] Preferably, a data chain for assessment and evaluation calculations is established, but there are no auxiliary analysis tools for the transmission impact of data changes. It is necessary to build source system data quality monitoring tools, data access and calculation result abnormality monitoring functions, and data result auxiliary analysis tools.

[0022] Preferably, when the system performs multi-user concurrent operations, the average response time for page access shall not exceed 5 seconds; the average response time for system login shall not exceed 5 seconds; when performing simple query, add and delete services, the average response time shall not exceed 5 seconds; when performing complex comprehensive services, the average response time shall not exceed 8 seconds; when performing report statistics services, the average response time shall not exceed 60 seconds; when performing forecasting and calculation services, the average response time shall not exceed 600 seconds. The information system transaction failure rate shall not exceed 0.1%.

[0023] Preferably, the use of the data in step S5 should be error-checked and verified in the application program to ensure the correctness and integrity of the original data.

[0024] Preferably, in the data conversion process of step S4, a universal standard format is used, taking into account the format requirements of different systems and different applications. When the trusted area obtains data, the data format should be verified before use to ensure its usability.

[0025] Preferably, the step S7 carries out the work of tracing the source of the indicator factors according to the factor list sorted out by the indicator system. The source systems of the factors include ERP, financial control system, and marketing system.

[0026] Preferably, the step S8 constructs a function for monitoring the calculation results of the indicator model for assisting business decision-making, and actively issues an early warning when the calculation results of the indicator for assisting business decision-making have a significant change compared with the factor value of the previous day.

[0027] Preferably, the step S8 constructs an enterprise calculation result monitoring function, and actively issues an early warning when the indicator calculation result has a significant change compared with the previous day, or the factor value has a significant change compared with the previous day.

[0028] Preferably, an auxiliary analysis tool for enterprise data results is constructed to realize the analysis of the impact of factors on indicators and indicators on scores in the enterprise calculation results.

[0029] In the present invention, the enterprise data asset model construction method based on the data middle platform improves the enterprise platform function, studies and establishes a power grid benefit analysis model, connects the links of people, finance and materials, and accurately transmits benefit indicators such as electricity sales revenue, labor costs, and operation and maintenance costs to the smart shared financial platform, thereby assisting in improving the management awareness of all levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Deployment topology diagram for the enterprise. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0032] Reference Figure 1 , a method for constructing an enterprise data asset model based on a data middle platform, comprising the following steps:

[0033] S1 Business review: Investigate the data characteristics, data quality, storage methods, management methods, and business logic of the indicator model that assists business decision-making, and analyze and study the survey results;

[0034] S2 Demand Analysis: Based on the research, accurately understand the specific requirements of business needs and transform business needs into complete demand definitions, that is, to refine the revenue and cost quantitative indicators to the smallest unit;

[0035] S3 business design: Based on the results of project demand analysis, construct the power grid indicator factor composition, determine the specific business content that the project needs to achieve, and form a design document;

[0036] S4 Data demand sorting and tracing: sort out the specific data requirements of the indicator model that assists the enterprise in business decision-making, determine the data source, and determine the data format and screening conditions for master data extraction;

[0037] S5 Data Collection and Extraction: Collect data from internal and external business systems or offline, extract required data, export enterprise master data list, export factor data related to indicator models that assist business decision-making, and analyze and process them;

[0038] S6 Analysis Layer Data Access Technology Route Design: According to the requirements of the indicator model architecture for supporting business decision-making, complete the data access technology route, including the design of the historical stock data extraction technology route and the incremental data extraction technology route;

[0039] S7 multi-dimensional and multi-layer data traceability: decompose the relevant factors involved in the indicator model of the enterprise's auxiliary business decision-making, clarify the traceability logic, write a traceability demand application form, and clarify the granularity, time period, data acquisition logic, unit and other related information of the factors;

[0040] S8 Analysis Model Design: According to the needs of enterprise management, an indicator analysis and calculation model for assisting business decision-making is constructed. The correlation relationship is sorted out according to the data middle platform system table, and a wide table structure is designed to store the calculation results of the indicator model for assisting business decision-making;

[0041] S9 analysis model construction: Based on business content, business rules and data requirements, it is developed in the data center to establish an indicator model for assisting business decision-making. The calculation results are stored in the wide table of the data center.

[0042] S10 Analysis Model Training and Evaluation: Determine the unknown parameters in the indicator model that assists business decision-making through existing data;

[0043] S11 Analysis Model Optimization: By changing the parameter configuration of the indicator model that assists business decision-making, the model is optimized to improve the extraction and calculation efficiency;

[0044] S12 operating environment construction: complete the system environment and big data resource planning and installation configuration required for the operation of the indicator model for supporting business decision-making;

[0045] S13 Application deployment and verification: Deploy the indicator model installation and deployment work for enterprise auxiliary business decision-making, and complete the corresponding security, performance, function and stability verification work;

[0046] S14 configuration implementation: Deploy the indicator model that assists the enterprise in making operational decisions to the data center.

[0047] In the present invention, an indicator factor system of indicators for assisting business decision-making is constructed for an enterprise, indicators are calculated by factors, scores are assigned to indicators, and score library results are calculated.

[0048] In the present invention, a data chain for assessment and evaluation calculations is established, but there is no auxiliary analysis tool for the transmission impact of data changes. It is necessary to build source system data quality monitoring tools, data access and calculation result abnormality monitoring functions, and data result auxiliary analysis tools.

[0049] In the present invention, when the system performs multi-user concurrent operations, the average response time for page access shall not exceed 5S; the average response time for system login shall not exceed 5S; when executing simple query, add and delete services, the average response time shall not exceed 5S; when executing complex comprehensive services, the average response time shall not exceed 8S; when executing report statistics services, the average response time shall not exceed 60S, and when executing forecasting and calculation services, the average response time shall not exceed 600S. The information system transaction failure rate shall not exceed 0.1%.

[0050] In the present invention, the use of the data in step S5 should be error-checked and verified in the application program to ensure the correctness and integrity of the original data.

[0051] In the present invention, the data in step S4 is converted in a universal standard format, taking into account the format requirements of different systems and different applications. When the trusted area obtains data, the format of the data should be verified before use to ensure its availability.

[0052] In the present invention, the step S7 carries out the work of tracing the source of the indicator factors according to the factor list sorted out by the indicator system. The source systems of the factors include ERP, financial control system, and marketing system.

[0053] In the present invention, the step S8 constructs a function for monitoring the calculation results of the indicator model for assisting business decision-making of the enterprise, and actively issues an early warning when the calculation result has a significant change compared with the factor value of the previous day.

[0054] In the present invention, the step S8 constructs an enterprise calculation result monitoring function, and actively issues an early warning when the indicator calculation result of the indicator system has a large change compared with the previous day, or the factor value has a large change compared with the previous day.

[0055] In the present invention, an auxiliary analysis tool for enterprise data results is constructed to realize the analysis of the influence of factors on indicators and indicators on scores in enterprise calculation results.

[0056] The present invention: constructs a data model, connects power grid assets, personnel, projects, and cost centers, forms an enterprise grid topology, designs the storage table structure of the master database, sorts out the source system of the master data and the table association relationship, and realizes the refinement of the quantitative indicators of business decision-making indicators to the smallest unit.

[0057] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for constructing an enterprise data asset model based on a data middle platform, characterized in that: The steps include: S1 Business review: Investigate the business status of the enterprise's business indicator model in terms of data characteristics, data quality, storage methods, management methods, and business logic, and analyze and study the survey results; S2 Demand Analysis: Based on the research, accurately understand the specific requirements of business needs and transform business needs into complete demand definitions, that is, to refine auxiliary decision-making indicators to the smallest unit; S3 Business Design: Based on the results of project demand analysis, design the indicator factor composition to assist business decision-making, determine the specific business content that the project needs to achieve, and form a design document; S4 Data demand sorting and tracing: sort out the specific data requirements of the enterprise's auxiliary business decision-making indicator model, determine the data source, and determine the data format and screening conditions for master data extraction; S5 Data Collection and Extraction: Collect data from internal and external business systems or offline, extract required data, export enterprise master data list, export factor data related to auxiliary business decision indicator model, and analyze and process; S6 Analysis Layer Data Access Technology Route Design: According to the enterprise auxiliary business decision-making indicator model architecture requirements, complete the data access technology route, including the historical stock data extraction technology route design and incremental data extraction technology route design; S7 multi-dimensional and multi-layer data traceability: decompose the relevant factors involved in the enterprise auxiliary business decision-making indicator model, clarify the traceability logic, write the traceability demand application form, and clarify the factor granularity, time period, data acquisition logic, and unit-related information; S8 analysis model design: According to the needs of enterprise management, build an enterprise auxiliary business decision-making indicator calculation model, sort out the association relationship based on the data middle platform system table, design a wide table structure, and store the calculation results of the auxiliary business decision-making indicator model; S9 analysis model construction: Based on business content, business rules and data requirements, it is developed in the data center to establish an enterprise auxiliary business decision-making indicator model, and the calculation results are stored in the wide table of the data center; S10 Analysis Model Training and Evaluation: Determine the unknown parameters in the enterprise's auxiliary business decision-making indicator model through existing data; S11 Analysis Model Optimization: By changing the parameter configuration of the enterprise's auxiliary business decision-making indicator model, the model is optimized to improve the extraction and calculation efficiency; S12 operating environment construction: complete the planning, installation and configuration of the system environment and big data resources required for the operation of the enterprise's auxiliary business decision-making indicator model; S13 Application deployment and verification: Deploy the enterprise auxiliary business decision-making indicator model installation and deployment work, and complete the corresponding security, performance, function and stability verification work; S14 configuration implementation: deploy the enterprise auxiliary business decision-making indicator model to the data center; For enterprises, we build an indicator factor system for substations and account managers, calculate indicators based on factors, assign points to indicators, and calculate the score database results; Build a data chain for assessment and evaluation calculations, but there is no auxiliary analysis tool for the transmission impact of data changes. It is necessary to build source system data quality monitoring tools, data access and calculation result abnormality monitoring functions, and data result auxiliary analysis tools; When the system performs multi-user concurrent operations, the average response time for page access shall not exceed 5S; the average response time for system login shall not exceed 5S; when executing simple query, addition and deletion services, the average response time shall not exceed 5S; when executing complex comprehensive services, the average response time shall not exceed 8S; when executing report statistics services, the average response time shall not exceed 60S, and the average response time for executing forecasting and calculation services shall not exceed 600S. The information system transaction failure rate shall not exceed 0.1%.

2. According to the method for constructing an enterprise data asset model based on a data middle platform according to claim 1, it is characterized in that: The use of the data in step S5 should be error-checked and verified in the application program to ensure the correctness and integrity of the original data.

3. According to the method for constructing an enterprise data asset model based on a data middle platform according to claim 1, it is characterized in that: In the data conversion process of step S4, a universal standard format is used, taking into account the format requirements of different systems and different applications. When the trusted area obtains data, the format of the data should be verified before use to ensure its availability.

4. According to the method for constructing an enterprise data asset model based on a data middle platform according to claim 1, it is characterized in that: The step S7 carries out the work of tracing the indicator factors according to the factor list sorted out by the indicator system, and the source systems of the factors include ERP, PMS, financial control system, and marketing system.

5. The method for constructing an enterprise data asset model based on a data middle platform according to claim 1, characterized in that: The step S8 constructs a function for monitoring the calculation results of the enterprise auxiliary business decision-making indicator model, and actively issues an early warning when the calculation result has a large change compared with the value of the previous day.

6. The method for constructing an enterprise data asset model based on a data middle platform according to claim 1, characterized in that: Build auxiliary analysis tools for enterprise data results to realize analysis of the impact of factors on indicators and indicators on scores in enterprise calculation results.

Images