CN-122019494-A - Enterprise data asset management method, enterprise data asset management device, computer equipment and storage medium

CN122019494ACN 122019494 ACN122019494 ACN 122019494ACN-122019494-A

Abstract

The application relates to the technical field of data asset management, and discloses an enterprise data asset management method, device, computer equipment and storage medium, wherein the method responds to full life cycle risk management of data assets generated by enterprises to carry out management optimization, and comprises the steps of constructing a management model taking real-time data assets as input and target management strategies as output; inputting the data assets into a treatment model for treatment analysis, outputting an original treatment strategy, synchronously performing toughness analysis, outputting the original toughness strategy, and adopting an attention mechanism to fuse the original treatment strategy and the original toughness strategy in the treatment model, and outputting a target treatment strategy. The apparatus, computer device and storage medium correspond to the method. The application realizes the suitability of the original treatment strategy, the effectiveness of the original toughness strategy and the effective improvement of the quality of the target strategy, thereby realizing the optimization of the early response rate of risks and the integrity of disaster recovery coverage.

Inventors

Zhou Panyi
ZHOU XIONGWEI
LI MINGRUI
LI ZHIJUN
LI HANMING
DAI JIALIN
HUANG ZICHUN

Assignees

中南大学

Dates

Publication Date: 20260512
Application Date: 20251203

Claims (10)

1. An enterprise data asset governance method, characterized by optimizing in response to full lifecycle risk management of data assets produced by an enterprise, the optimizing outputting a target governance policy employed for the data assets based on a governance analysis and a toughness analysis, the method comprising: acquiring historical data assets and corresponding historical treatment strategies thereof, and constructing a treatment model based on the historical data assets and the corresponding historical treatment strategies, wherein the treatment model takes real-time data assets as input and target treatment strategies as output; Inputting the data asset into the governance model for governance analysis and outputting an original governance strategy; And in the treatment model, adopting an attention mechanism to fuse the original treatment strategy and the original toughness strategy, and outputting the target treatment strategy.
2. The enterprise data asset governance method of claim 1, wherein the construction of the governance model comprises: Converting unstructured historical data assets into structured asset characteristics, the asset characteristics being used to characterize information required for the abatement analysis and the toughness analysis, and the asset characteristics including at least an associated entity and a timestamp; Converting an unstructured historical governance policy into structured policy features, the policy features comprising governance features corresponding to the governance analysis and toughness features corresponding to the toughness analysis, and the governance features comprising at least a conventional governance policy, the toughness features comprising at least a risk response policy; Recombining the asset characteristics, the governance characteristics and the toughness characteristics based on a time sequence, and constructing a governance strategy mapping relation based on a recombination result; And storing the treatment strategy mapping relation into a preset neural network model for training to obtain the treatment model.
3. The enterprise data asset governance method of claim 2, wherein the construction of governance policy mappings comprises: Based on the historical data assets and the corresponding historical treatment strategies, obtaining the asset characteristics, the treatment characteristics and the characteristic distribution conditions of the toughness characteristics of the historical data assets and the corresponding historical treatment strategies under the same time sequence; Extracting features according to the feature distribution condition to obtain a strategy time feature and a strategy space feature, wherein the strategy time feature is used for representing the corresponding relation of an asset feature, a treatment feature and/or a toughness feature in time, and the strategy space feature is used for representing the corresponding relation of the asset feature, the treatment feature and/or the toughness feature in space; and verifying the strategy time characteristics and the strategy space characteristics by combining with a preset accuracy threshold, and obtaining the treatment strategy mapping relation after the verification is passed.
4. The enterprise data asset governance method of claim 2, wherein the output of the raw governance policy comprises: Based on the asset characteristics converted by the real-time data asset, matching the preset treatment characteristics in the treatment model, calling the execution parameters corresponding to the conventional treatment strategy, and generating and outputting the original treatment strategy by combining the business attribution and the quality rating of the data asset, wherein the original treatment strategy at least comprises measures corresponding to data cleaning, standardized processing, storage grading and authority management and control.
5. The enterprise data asset governance method of claim 2, wherein the outputting of the raw toughness policy comprises: And carrying out risk scene recognition based on the associated entity and the timestamp of the real-time data asset, matching a risk response strategy in the toughness characteristic based on the recognition result, and generating and outputting an original toughness strategy by combining the security level and the redundancy requirement of the data asset, wherein the original toughness strategy at least comprises measures corresponding to a fault recovery flow, a backup scheme, a risk early warning threshold value and a fault tolerance adaptation rule.
6. The enterprise data asset governance method of claim 3, wherein the construction of the governance model further comprises: Based on a long-time sequence of the full life cycle of the data asset, extracting a plurality of groups of strategy time features and strategy space features which pass verification in the treatment strategy mapping relation to form a space-time feature sequence set; a space-time attention optimizing unit is additionally arranged in the neural network model, the space-time feature sequence set is used as a training sample of the space-time attention optimizing unit, and a weight adjusting rule of a time dimension and a space dimension is preset; training the space-time attention optimizing unit through the space-time feature sequence set, so that the space-time attention optimizing unit determines an early response weight based on strategy time features of a long-time sequence and a range expansion weight based on strategy space features; And associating the trained space-time attention optimizing unit with the attention mechanism of the governance model, so that the attention mechanism calls the space-time attention optimizing unit to perform early response weight and range expansion weight.
7. The enterprise data asset remediation method of claim 6, wherein the outputting of the target remediation strategy comprises: Acquiring asset characteristics corresponding to real-time data assets, and matching the strategy time characteristics and the strategy space characteristics associated in the treatment strategy mapping relation by combining a long-time sequence of the full life cycle of the data assets; The matched strategy time characteristics and the strategy space characteristics are input into a space-time attention optimizing unit of the optimized treatment model, and the early response weight and the range expansion weight are respectively determined according to a preset weight adjusting rule; substituting the advanced response weight and the range expansion weight into a fusion operation process of an attention mechanism, and adjusting the fusion proportion of the original treatment strategy and the original toughness strategy, wherein the pre-judgment adaptation of the strategy execution time is adjusted based on the advanced response weight, and the adaptation expansion of the range is executed based on the range expansion weight adjustment strategy; and fusing the original treatment strategy and the original toughness strategy according to the regulated fusion proportion, and generating and outputting the target treatment strategy.
8. An enterprise data asset governance apparatus applying the enterprise data asset governance method of any of claims 1 to 7, the apparatus comprising: the model construction module is used for acquiring the historical data assets and the corresponding historical treatment strategies thereof, so as to construct a treatment model, wherein the treatment model takes the real-time data assets as input and takes the target treatment strategies as output; The original output module is used for inputting the data asset into the governance model for governance analysis and outputting an original governance strategy; and the fusion output module is used for fusing the original treatment strategy and the original toughness strategy in the treatment model by adopting an attention mechanism and outputting the target treatment strategy.
9. An enterprise data asset remediation computer device comprising at least one processor, at least one memory and a data bus; The processor and the memory complete communication with each other through the data bus; the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the enterprise data asset management method of any of claims 1-7.
10. A storage medium having stored thereon a computer program which when executed by a processor implements the enterprise data asset management method of any one of claims 1 to 7.

Description

Enterprise data asset management method, enterprise data asset management device, computer equipment and storage medium Technical Field The application relates to the technical field of data asset management, in particular to an enterprise data asset management method, an enterprise data asset management device, computer equipment and a storage medium. Background The conventional enterprise data asset management method focuses on conventional management with single dimension such as data cleaning, storage grading and the like, or performs toughness measures such as data disaster recovery and the like in an isolated manner, and the management mode has two main core defects: Firstly, the treatment strategy lacks the prospective of the whole life cycle, and the potential risk is difficult to respond in advance; Secondly, the fusion degree of toughness strategies such as conventional management and data disaster recovery is low, and space-time suitability is poor, such as lag of disaster recovery execution time and incomplete coverage. Such defects can lead to significant loss of abatement when the data asset is at risk of being lost, damaged or unusable. Thus, a need exists for a new solution for enterprise data asset management. Disclosure of Invention The application aims to provide an enterprise data asset management method, an enterprise data asset management device, computer equipment and a storage medium, so as to solve the technical problem that the conventional management and data disaster recovery in the data asset management in the prior art are poor in synergy. To achieve the above objects, the present application provides an enterprise data asset governance method for governance optimization in response to full lifecycle risk management of data assets produced by an enterprise, the optimization outputting a target governance policy employed for the data assets based on a governance analysis and a toughness analysis, the method comprising: acquiring historical data assets and corresponding historical treatment strategies thereof, and constructing a treatment model based on the historical data assets and the corresponding historical treatment strategies, wherein the treatment model takes real-time data assets as input and target treatment strategies as output; Inputting the data asset into the governance model for governance analysis and outputting an original governance strategy; And in the treatment model, adopting an attention mechanism to fuse the original treatment strategy and the original toughness strategy, and outputting the target treatment strategy. Preferably, the construction of the governance model includes: Converting unstructured historical data assets into structured asset characteristics, the asset characteristics being used to characterize information required for the abatement analysis and the toughness analysis, and the asset characteristics including at least an associated entity and a timestamp; Converting an unstructured historical governance policy into structured policy features, the policy features comprising governance features corresponding to the governance analysis and toughness features corresponding to the toughness analysis, and the governance features comprising at least a conventional governance policy, the toughness features comprising at least a risk response policy; Recombining the asset characteristics, the governance characteristics and the toughness characteristics based on a time sequence, and constructing a governance strategy mapping relation based on a recombination result; And storing the treatment strategy mapping relation into a preset neural network model for training to obtain the treatment model. Preferably, the construction of the treatment strategy mapping relation includes: Based on the historical data assets and the corresponding historical treatment strategies, obtaining the asset characteristics, the treatment characteristics and the characteristic distribution conditions of the toughness characteristics of the historical data assets and the corresponding historical treatment strategies under the same time sequence; Extracting features according to the feature distribution condition to obtain a strategy time feature and a strategy space feature, wherein the strategy time feature is used for representing the corresponding relation of an asset feature, a treatment feature and/or a toughness feature in time, and the strategy space feature is used for representing the corresponding relation of the asset feature, the treatment feature and/or the toughness feature in space; and verifying the strategy time characteristics and the strategy space characteristics by combining with a preset accuracy threshold, and obtaining the treatment strategy mapping relation after the verification is passed. Preferably, the output of the original abatement strategy includes: Based on the asset characteristics converted by the real-time data asset, matching the prese