Intelligent enterprise operation risk decision-making method and system based on multi-source heterogeneous data fusion

Abstract

The invention belongs to the technical field of data fusion and intelligent decision-making, and particularly discloses an enterprise operation risk intelligent decision-making method and system based on multi-source heterogeneous data fusion, and the method comprises the steps: crawling multi-source heterogeneous data of each department of an enterprise, constructing a unified data network, comprehensively reflecting the operation condition of the enterprise, and carrying out the intelligent decision-making of the enterprise operation risk. Potential risks in enterprise operation can be identified more accurately, and management comprehensiveness of enterprise operation data is realized. A dynamic resource allocation mechanism is constructed by taking continuous operation risk factors of all departments of an enterprise as key references, and multi-source heterogeneous data are automatically processed and analyzed, so that real-time monitoring and early warning of the operation risk of the enterprise are realized, resource allocation of the enterprise is optimized, and meanwhile, the enterprise can be ensured to quickly adapt to market changes; by establishing a plurality of operation progress systems of each business optimization department and analyzing the optimization expectation, targeted business optimization suggestions can be provided for each department of an enterprise.

Description

Technical Field

[0001] The present invention belongs to the technical field of data fusion and intelligent decision-making, and relates to an intelligent decision-making method and system for enterprise operation risk based on multi-source heterogeneous data fusion. Background Art

[0002] In the process of enterprise operation, accurately evaluating and managing operation risks is crucial for ensuring the sustainable and stable development of enterprises. However, with the expansion of enterprise scale and the diversification of business, the operation data generated by each department of the enterprise is increasingly huge and complex, covering multiple fields such as finance, market, supply chain, and policy, forming a multi-source heterogeneous data environment. These data not only have a wide range of sources but also diverse formats, bringing huge challenges to data integration, analysis, and utilization.

[0003] Existing methods have obvious drawbacks in dealing with enterprise operation risk assessment, including difficulties in data integration, low analysis efficiency, and inability to adapt to dynamic data changes. These problems limit the enterprise's ability to respond quickly and make effective decisions in a complex and changing market environment, resulting in the enterprise's inability to make flexible change responses in the face of operation risks due to the inability to adapt to the dynamic changes of data, and thus unable to provide timely and accurate risk early warnings and decision support for the enterprise. Summary of the Invention

[0004] To overcome the above defects of the prior art and achieve the above object, the present invention proposes the following technical solutions: The intelligent decision-making method for enterprise operation risk based on multi-source heterogeneous data fusion provided in the first aspect of the present invention includes the following: S1. Crawl multi-source data related to the operation of each department of the enterprise within a preset period, generate a multi-source heterogeneous data network for enterprise operation, and export the continuous operation risk factors of each department of the enterprise.

[0005] S2. Obtain the corresponding resource application data of each department of the enterprise, and determine the resource allocation label of the enterprise in the current application period with reference to the continuous operation risk factors of each department of the enterprise. The resource allocation label includes a regular label, a restricted label, and a surplus label.

[0006] Wherein, the resource application data includes the application amount, application purpose, and application period.

[0007] S3. Under the regular label, distribute resources according to the application amount of each department; under the restricted label and the surplus label, set department approval rules adapted to the resource allocation label, and accordingly divide each application permission department and each business optimization department of the enterprise.

[0008] S4. Based on the resource application data of each application permission department, identify the adaptability of the declared resources of each application permission department, and accordingly set the resource allocation mechanism for different application permission departments.

[0009] S5. Establish multiple operation progress systems for each business optimization department, analyze the optimization expectations of each business optimization department for the corresponding multiple operation progress systems, and generate corresponding optimization marks for each business optimization department accordingly.

[0010] The enterprise operation risk intelligent decision-making system based on multi-source heterogeneous data fusion provided by the second aspect of the present invention includes: an operation risk assessment module, which crawls multi-source data related to the operation of each department of the enterprise within a preset period, generates a multi-source heterogeneous data network for the operation of the enterprise, and exports the continuous operation risk factors of each department of the enterprise.

[0011] A resource label determination module, which obtains the corresponding resource declaration data of each department of the enterprise, and determines the resource allocation label of the enterprise in the current declaration period with reference to the continuous operation risk factors of each department of the enterprise. The resource allocation label includes a regular label, a restricted label, and a surplus label.

[0012] Among them, the resource declaration data includes the application amount, the application purpose, and the declaration period.

[0013] A department division module, under the regular label, distributes resources according to the application amount of each department; under the restricted label and the surplus label, sets department approval rules adapted to the resource allocation label, and divides each declaration permission department and each business optimization department of the enterprise accordingly.

[0014] A resource allocation module, based on the resource declaration data of each declaration permission department, identifies the adaptability of the declared resources of each declaration permission department, and sets the resource allocation mechanism for different declaration permission departments accordingly.

[0015] A business optimization module, which establishes multiple operation progress systems for each business optimization department, analyzes the optimization expectations of each business optimization department for the corresponding multiple operation progress systems, and generates corresponding optimization marks for each business optimization department accordingly.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By crawling the multi-source heterogeneous data of each department of the enterprise and constructing a unified data network, the present invention comprehensively reflects the operation status of the enterprise, can more accurately identify potential risks in enterprise operation, and realizes the comprehensive integration and efficient management of the operation data of each department of the enterprise.

[0017] (2) With the continuous operation risk factors of each department of the enterprise as the key reference, the present invention constructs a dynamic resource allocation mechanism. By automatically processing and analyzing multi-source heterogeneous data, it realizes real-time monitoring and early warning of enterprise operation risks, ensures that decision-making suggestions can be quickly generated when potential risks are identified, not only optimizes the resource allocation of the enterprise, but also significantly improves the decision-making efficiency and response speed of the enterprise management layer, ensuring that the enterprise can quickly adapt to market changes and make scientific and reasonable decisions.

[0018] (3) By establishing multiple operation progress systems for each business optimization department and analyzing their optimization expectations, the present invention can provide targeted business optimization suggestions for enterprises, helping enterprises identify business bottlenecks, optimize business processes, improve operation efficiency, and achieve continuous improvement and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0020] Figure 1 It is a schematic flowchart of the implementation steps of the method of the present invention.

[0021] Figure 2 It is a schematic diagram of the connection of system modules of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0023] Embodiment 1 Please refer to Figure 1 As shown, the intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion proposed by the present invention includes the following: S1. Crawl multi-source data related to the operation of each department of the enterprise within a preset period, generate a multi-source heterogeneous data network for enterprise operation, and export the continuous operation risk factors of each department of the enterprise.

[0024] In a preferred implementation manner, the generation of the multi-source heterogeneous data network for enterprise operation includes: storing the collected multi-source data in corresponding databases according to the data types respectively, and each database is efficiently managed using a relational database or a non-relational database. The data types include but are not limited to finance, market, supply chain, and policy. The relational database is, for example, MySQL, and the non-relational database is, for example, MongoDB.

[0025] The multi-source data includes but is not limited to financial data (such as balance sheets, cash flows), market data (such as competitor prices, market shares), supply chain data (such as inventory turnover rates, supplier performance rates), and policy data (such as industry regulatory documents, tax incentives).

[0026] Taking each type of database as a node, an independent transmission channel is constructed for each type of database. For example, financial data corresponds to the fund flow channel, and market data corresponds to the competition situation channel. Data is exchanged and shared in real time through data interfaces, and heterogeneous data in each type of database corresponding to each department is converted into a unified format through a normalization processing method, and accordingly, multi-source data parameters corresponding to each department are generated.

[0027] Specifically, the normalization processing method is an existing technology. For example, natural language processing (NLP) technology and machine learning are used to clean, transform, and standardize the heterogeneous data in each type of database. The Min-Max normalization or Z-score standardization method is adopted to convert the data into a unified format, generating multi-source data parameters corresponding to each department to ensure the consistency and comparability of the data.

[0028] Obtain the multi-source data parameters related to the operations of each department of the enterprise during the historical reporting period and import them into the enterprise comprehensive evaluation data pool.

[0029] Establish a project application feature library and a project contribution feature library in the enterprise comprehensive evaluation data pool. Match the multi-source data parameters related to operations with the features in these two feature libraries to generate a project application feature set and a project contribution feature set, and accordingly, obtain the multi-source heterogeneous data network of the enterprise operation.

[0030] Among them, the project application feature set and the project contribution feature set respectively contain multi-dimensional data elements used to characterize the project application and project contribution situations of each department during the corresponding historical reporting period.

[0031] The project application feature set includes, but is not limited to, the input quantity of material resources and the loss quantity of equipment resources corresponding to construction projects, the supply quantity of materials and the investment amount in marketing promotion corresponding to R & D projects. The project contribution features include, but are not limited to, the construction completion task rate and cost savings rate corresponding to construction projects, and the revenue target achievement rate corresponding to R & D projects.

[0032] In a further preferred implementation manner, the export of the continuous operation risk factors of each department of the enterprise includes: based on the multi-source heterogeneous data network of the enterprise operation, obtaining the project application feature set and the project contribution feature set of each department in a number of historical reporting periods.

[0033] Taking the time period as the horizontal axis and the values of each data element in the feature set as the vertical axis, based on the time series, draw the corresponding historical operation risk trend charts of the project application feature set and the project contribution feature set of each department in a number of historical reporting periods. The historical operation risk trend charts are used to visually display the changes in the project application and project contribution features of each department in different historical reporting periods.

[0034] Perform a normal distribution fitting on the characteristic data in the historical operation risk trend chart to obtain the corresponding means and standard deviations of the project application characteristic sets and project contribution characteristic sets for each department.

[0035] Obtain the confidence levels selected by each department in the cases of project application and project contribution respectively, and determine the corresponding confidence level coefficients through the standard normal distribution table. The following relationship exists between the confidence level coefficient and the confidence level: In a normal distribution, for a given confidence level, there is a specific confidence level coefficient such that the probability that the data falls within the interval μ ± kσ is that confidence level, where μ represents the mean and σ represents the standard deviation.

[0036] Specifically, the confidence level represents the proportion of the confidence intervals containing the true values of the population parameters in multiple data, which is defined by the enterprise. Common confidence level selections are 90%, 95%, 99%, etc. In the scenario of enterprise operation risk assessment, the confidence level reflects the credibility of the estimated fluctuation range of the characteristic data of each department's project application and project contribution.

[0037] When selecting a 90% confidence level, the confidence level coefficient takes 1.645, corresponding to the z value with a cumulative probability of 0.95 in the standard normal distribution. The z value is the critical value obtained from the standard normal distribution table according to the confidence level. The z value corresponds to "the multiple of the standard deviation corresponding to the middle probability". At this time, the probability that the data in the historical operation risk trend chart falls within the interval μ ± 1.645σ is 90%, with 5% in each of the two tails, and the cumulative probability is 0.95; when selecting a 95% confidence level, the confidence level coefficient takes 1.96, corresponding to the z value with a cumulative probability of 0.975 in the standard normal distribution. At this time, the probability that the data in the historical operation risk trend chart falls within the interval μ ± 1.96σ is 95%, with 2.5% in each of the two tails, and the cumulative probability is 0.975; when selecting a 99% confidence level, the confidence level coefficient takes 2.576, corresponding to the z value with a cumulative probability of 0.995 in the standard normal distribution. At this time, the probability that the data in the historical operation risk trend chart falls within the interval μ ± 2.576σ is 99%, with 0.5% in each of the two tails, and the cumulative probability is 0.995.

[0038] Calculate the upper and lower limits of the corresponding confidence intervals for the project application characteristic sets and project contribution characteristic sets of each department according to the mean, standard deviation, and confidence level coefficient, that is, the upper limit is μ + kσ, and the lower limit is μ - kσ. The confidence interval is used to define the reasonable fluctuation range of different characteristic set data at a given confidence level.

[0039] Statistically analyze the proportion of the number of features that fall within the corresponding confidence intervals in the project application feature set and the project contribution feature set of each department, and use this proportion as the persistence index of the project application and project contribution of the historical operations of each department within the confidence interval. 、 Integrate and generate the continuous operation risk factors of each department of the enterprise. Among them, represents the number of each department. The larger the persistence index of the project application within the confidence interval, the greater the possibility that the change in the project application characteristics of this department exceeds the normal fluctuation range, and the higher the continuous operation risk factor; the larger the persistence index of the project contribution within the confidence interval, the more stable the change in the project contribution characteristics of this department within the normal fluctuation range, and the lower the continuous operation risk factor. Furthermore, the continuous operation risk factors of each department of the enterprise are derived for the enterprise to use in operation risk assessment and management decision-making.

[0040] Through crawling the multi-source heterogeneous data of each department of the enterprise and constructing a unified data network, the present invention comprehensively reflects the operation status of the enterprise, can more accurately identify potential risks in enterprise operation, and realizes the comprehensive integration and efficient management of the operation data of each department of the enterprise.

[0041] S2. Obtain the corresponding resource declaration data of each department of the enterprise, and determine the resource allocation label of the enterprise in the current declaration cycle with reference to the continuous operation risk factors of each department of the enterprise. The resource allocation label includes a regular label, a restricted label, and a surplus label.

[0042] Among them, the resource declaration data includes the application amount, the application purpose, and the declaration cycle.

[0043] In a preferred implementation manner, determining the resource allocation label of the enterprise in the current declaration cycle includes: defining that at least one department's continuous operation risk factor exceeds the preset operation risk threshold factor as a first-level restriction condition.

[0044] Obtain the total resource allocation amount of the enterprise in the current declaration cycle, and statistically analyze the corresponding comprehensive application amounts of each department of the enterprise, and then define that the total resource allocation amount is lower than the comprehensive application amount as a second-level restriction condition.

[0045] When the enterprise has at least one of the first-level restriction condition or the second-level restriction condition, determine that the resource allocation label is a restricted label.

[0046] Otherwise, when the total resource allocation amount exceeds the comprehensive application amount, determine that the resource allocation label is a surplus label; when the total resource allocation amount is equal to the comprehensive application amount, determine that the resource allocation label is a regular label.

[0047] S3. Under the regular label, allocate resources according to the application quotas of each department; under the restricted label and the surplus label, set the department approval rules adapted to the resource allocation label, and accordingly divide each application permission department and each business optimization department of the enterprise.

[0048] In a preferred implementation manner, the division of each application permission department and each business optimization department of the enterprise includes: setting a first-level risk threshold value corresponding to the restricted label and a second-level risk threshold value corresponding to the surplus label, and the first-level risk threshold value is higher than the second-level risk threshold value.

[0049] Under the restricted label, compare the continuous operation risk factors of each department with the first-level risk threshold value, and determine that the departments with continuous operation risk factors lower than the first-level risk threshold value are business optimization departments, and the rest are application permission departments.

[0050] Under the surplus label, compare the continuous operation risk factors of each department with the second-level risk threshold value, and determine that the departments with continuous operation risk factors lower than the second-level risk threshold value are business optimization departments, and the rest are application permission departments.

[0051] Statistically obtain each application permission department and each business optimization department of the enterprise.

[0052] S4. Based on the resource application data of each application permission department, identify the suitability of the applied resources of each application permission department, and accordingly set the resource allocation mechanism for different application permission departments.

[0053] In a preferred implementation manner, the setting of the resource allocation mechanism for different application permission departments includes: extracting the keywords of the application purposes in the corresponding resource application data of each application permission department through text mining technology, and matching them with the keyword library related to the core business of the corresponding department, and accordingly determining the suitability of the applied resources of each application permission department, and the suitability of the applied resources includes suitability and unsuitability.

[0054] The application purposes include but are not limited to "purchase of a certain equipment for a new production line", "research and development of new energy vehicle battery technology", and "optimization of the raw material supply chain".

[0055] Specifically, the determination process of the suitability of the applied resources is as follows: compare the keywords in the corresponding application purposes of each application permission department with the keyword library related to the core business of the department. If a keyword in the application purpose fails to be retrieved in the keyword library related to the core business, for example, a keyword in the application purpose is the name of a certain equipment and it fails to be retrieved in the keyword library related to the core business, then it is determined that the suitability of the applied resources of the application permission department is unsuitable, otherwise it is suitable.

[0056] For the non - compliant declaration and approval departments, based on the multi - source data related to the operation of the declaration and approval departments and the resource declaration data, analyze the rationality index of their resource declarations, and generate the rationality index of resource declarations for each declaration and approval department accordingly.

[0057] Taking the rationality index of resource declarations of each declaration and approval department as a reference, calibrate the application amount of each declaration and approval department to obtain the final calculated amount of the declaration and approval department.

[0058] Specifically, the process for determining the final calculated amount is as follows: Compare the rationality index of resource declarations of each declaration and approval department with the preset reference rationality index threshold of resource declarations. If the rationality index of resource declarations of a certain declaration and approval department is lower than the preset reference rationality index threshold of resource declarations, then multiply the rationality index of resource declarations of this declaration and approval department by the application amount of the corresponding declaration and approval department to obtain the final calculated amount of this declaration and approval department; otherwise, take the declared amount as the final calculated amount of this declaration and approval department.

[0059] For the compliant declaration and approval departments, take the declared amount as the final calculated amount.

[0060] Statistical the final calculated amount of each declaration and approval department to obtain the comprehensive declared amount. If it does not exceed the total resource allocation amount, conduct resource accounting according to the regular issued link, otherwise introduce the enterprise backup resource link to the regular issued link, and accordingly allocate resources to different declaration and approval departments.

[0061] Specifically, the regular issued link is to issue resources according to the final calculated amount of each declaration and approval department; the introduction of the enterprise backup resource link to the regular issued link means applying for enterprise backup resources on the basis of the regular issued link to supplement the corresponding resource allocation of each declaration and approval department.

[0062] In a further preferred implementation manner, for the non - compliant declaration and approval departments, based on the multi - source data related to the operation of the declaration and approval departments and the resource declaration data, analyze the rationality index of their resource declarations, including: Count the number of keywords with retrieval failures in the corresponding declared uses of the non - compliant declaration and approval departments, and take the ratio of it to the preset reference number of keywords. The reciprocal of this ratio is the business correlation degree index between the corresponding resource declaration uses and the core business of the non - compliant declaration and approval departments. The larger the business correlation degree index, the stronger the correlation between the resource declaration and the business, and the higher the rationality of its resource declaration.

[0063] Derive the project application feature set within the current declaration cycle from the multi - source data related to the operation of the non - compliant declaration and approval departments, and obtain the industry - average application feature set. Calculate the proportion of the difference value of the corresponding features of the two in the industry average value to obtain the deviation rate of each feature, and then accumulate to obtain the comprehensive deviation rate of the two. and map it to an inverse relationship with the resource application rationality index such as where is used to avoid abnormal calculation of the formula when the value is 0. The smaller the comprehensive deviation rate, the larger the resource application rationality index, indicating that the declared resource quantity is more in line with the actual demand, and thus the higher the rationality.

[0064] Obtain the application quotas of the misfitting declared permission departments in several historical declaration cycles, and examine the consistency index of the change trend in the time series of their historical application quotas with those of the declared permission departments.

[0065] Specifically, adopt the time series analysis method to calculate the growth difference of the corresponding application quotas in several historical declaration cycles compared with those in the previous declaration cycle, and count the proportion of the number of declaration cycles with the growth difference within the preset range. This proportion is the historical declaration trend consistency index. Among them, the higher the proportion, the better the historical declaration trend consistency and the higher the resource declaration rationality.

[0066] According to the influencing degrees of the above-mentioned business relevance index, resource demand rationality index, and change trend consistency index on the resource declaration rationality, different weights are assigned to them respectively. For example, through empirical evaluation, the influencing degrees of each index on the resource declaration rationality are determined as After multiplying each index by its corresponding weight and summing them up, an initial evaluation value is obtained.

[0067] The initial evaluation value is converted into a resource declaration rationality index within the range of 0 - 1 through normalization, specifically: (initial evaluation value - minimum possible initial evaluation value) / (maximum possible initial evaluation value - minimum possible initial evaluation value), where the minimum possible initial evaluation value is the result calculated by taking the minimum value of each index and according to the weights, and the maximum possible initial evaluation value is the result calculated by taking the maximum value of each index and according to the weights.

[0068] The present invention takes the continuous operation risk factors of each department of the enterprise as the key reference, constructs a dynamic resource allocation mechanism, and through automated processing and analysis of multi-source heterogeneous data, realizes real-time monitoring and early warning of the enterprise operation risks, ensures that decision-making suggestions can be quickly generated when potential risks are identified, not only optimizes the resource allocation of the enterprise, but also significantly improves the decision-making efficiency and response speed of the enterprise management layer, and ensures that the enterprise can quickly adapt to market changes and make scientific and reasonable decisions.

[0069] S5. Establish multiple operation progress systems for each business optimization department, analyze the optimization expectations of the corresponding multiple operation progress systems of each business optimization department, and generate corresponding optimization marks for each business optimization department accordingly.

[0070] In a preferred embodiment, analyzing the optimization expectations of each business optimization department for the corresponding multiple operation progress systems includes: formulating the key milestone time plans for each stage of the operation projects corresponding to the business optimization department, and tracking in real time the deviation between the actual progress and the planned progress of the projects, and generating a time progress system accordingly. The operation projects include, but are not limited to, application R & D projects and construction projects. Example: For an Internet company planning to launch a new social application, the key milestone times for each stage include, but are not limited to, the project start time, the prototype design completion time, the development completion time, the testing completion time, and the online time.

[0071] Defining the multi-dimensional resource input plans such as human, material, and financial resources required for the operation projects corresponding to the business optimization department, recording the actual usage data of project resources, tracking the matching degree between the resource usage data and the resource input plan, and generating a resource progress system accordingly. Among them, the actual usage data of project resources includes the actual input amount of resources, the idle resource amount, and the total project output.

[0072] Setting the quality standards and acceptance indicators for the operation projects in different stages, regularly evaluating and detecting the project quality of the operation projects corresponding to the business optimization department, recording the quality compliance situation and the quality improvement progress, and generating a quality progress system accordingly.

[0073] Based on the time progress system, obtain the deviation rate of the actual completion duration and the planned completion duration of the project, that is, (actual completion duration - planned completion duration) / planned completion duration × 100%; count the on-time achievement rate of each stage of key milestones, that is, on-time achievement rate = the number of key milestones achieved on time / total number of key milestones × 100%; integrate and evaluate the time efficiency optimization expectations of the corresponding business optimization department, that is, time efficiency optimization expectation = deviation rate × (1 - on-time achievement rate), and this expectation is used to measure the optimization space and expected effect of the project in the time dimension.

[0074] Based on the resource progress system, calculate the input ratio of the actual resource input and the planned resource input, that is, input ratio = actual input amount of resources / planned input amount of resources × 100%; calculate the resource idle rate, that is, resource idle rate = idle resource amount / total amount of resources × 100%; calculate the resource consumption per unit output, that is, resource consumption per unit output = actual input amount of resources / total project output, and accumulate the input ratio, resource idle rate, and resource consumption per unit output to obtain the resource utilization optimization expectations of the corresponding business optimization department, and this expectation reflects the optimization potential and expected goals of the project in the resource utilization dimension.

[0075] Based on the quality progress system, the quality compliance rate of the project is statistically calculated, that is, the quality compliance rate = the number of quality-compliant projects / the total number of projects × 100%; the quality defect repair rate is statistically calculated, that is, the quality defect repair rate = the number of repaired quality defects / the total number of discovered quality defects × 100%; the quality improvement progress parameter is recorded, that is, the quality improvement progress parameter = the average growth value of quality indicators at different stages / the preset reference growth value. The quality compliance rate of the project, the quality defect repair rate, and the quality improvement progress parameter are accumulated to obtain the corresponding quality improvement and optimization expectations of the business optimization department. This expectation reflects the optimization direction and expected effectiveness of the project in the quality dimension.

[0076] According to the corresponding acquisition methods of the time efficiency optimization expectation, resource utilization optimization expectation, and quality improvement optimization expectation, the optimization expectations of each business optimization department for the corresponding multiple operation progress systems are statistically obtained, providing data support and reference basis for business optimization decisions.

[0077] In a further preferred implementation manner, the generation of the corresponding optimization marks for each business optimization department includes: based on the optimization expectations of each business optimization department for the corresponding multiple operation progress systems, determining the optimization requirement levels of each business optimization department in the three dimensions of time, resources, and quality. The optimization requirement levels are divided into high level, medium level, and low level. Then, the progress systems at different levels in each business optimization department are marked for optimization, providing a data basis for the enterprise's strategic planning and business expansion.

[0078] Exemplarily, the marking rules are set as follows: For each business optimization department, an independent optimization marking system is established for the three dimensions of time, resources, and quality respectively. The marking adopts the combination form of "department identifier - dimension identifier - level identifier", where the department identifier is the unique code or name abbreviation of the business optimization department; the dimension identifiers are represented by "T" for the time dimension, "R" for the resource dimension, and "Q" for the quality dimension respectively; the level identifiers are represented by "H" for the high level, "M" for the medium level, and "L" for the low level respectively.

[0079] After determining the optimization requirement level of a certain business optimization department in a certain dimension, the corresponding optimization mark is generated according to the above marking rules. For example, if the code of a certain business optimization department is "BOD001" and its optimization requirement level in the time dimension is high level, then the optimization mark for the time dimension progress system of this department is "BOD001-T-H".

[0080] Regularly (such as monthly or quarterly), according to the latest optimization expectations of the operation progress systems of each business optimization department, re-evaluate their optimization requirement levels in the three dimensions of time, resources, and quality, and update the corresponding optimization marks. The updated marks should cover the original marks, and information such as the update time and update reason should be recorded to ensure the accuracy and traceability of the marks.

[0081] By establishing multiple operation progress systems for each business optimization department and analyzing their optimization expectations, the present invention can provide targeted business optimization suggestions for enterprises, helping enterprises identify business bottlenecks, optimize business processes, improve operation efficiency, and achieve continuous improvement and sustainable development.

[0082] Embodiment 2 Please refer to Figure 2 As shown, based on Embodiment 1, the enterprise operation risk intelligent decision-making system based on multi-source heterogeneous data fusion provided by the second aspect of the present invention includes: an operation risk assessment module, a resource label determination module, a department division module, a resource allocation module, and a business optimization module.

[0083] The operation risk assessment module, the resource label determination module, the department division module, the resource allocation module, and the business optimization module are connected in sequence.

[0084] The operation risk assessment module crawls multi-source data related to the operation of each department of the enterprise within a preset period, generates a multi-source heterogeneous data network for the operation of the enterprise, and exports the continuous operation risk factors of each department of the enterprise.

[0085] The resource label determination module obtains the corresponding resource declaration data of each department of the enterprise, and determines the resource allocation label of the enterprise in the current declaration period with reference to the continuous operation risk factors of each department of the enterprise. The resource allocation label includes a regular label, a restriction label, and a surplus label.

[0086] Among them, the resource declaration data includes the application amount, the application purpose, and the declaration period.

[0087] The department division module distributes resources according to the application amount of each department under the regular label; under the restriction label and the surplus label, it sets department approval rules adapted to the resource allocation label, and accordingly divides each declaration permission department and each business optimization department of the enterprise.

[0088] The resource allocation module identifies the adaptability of the declared resources of each declaration permission department based on the resource declaration data of each declaration permission department, and accordingly sets the resource allocation mechanism for different declaration permission departments.

[0089] The business optimization module establishes multiple operation progress systems for each business optimization department, analyzes the optimization expectations of the corresponding multiple operation progress systems of each business optimization department, and accordingly generates corresponding optimization marks for each business optimization department.

[0090] It should be noted that: for the formulas presented above, through the principle of dimensional consistency and mathematical standardization means (such as normalization processing, dimensionless parameter conversion, or unit system unification), physical quantities with different attributes can be translated into unitless standard values or superimposable parameters of the same dimension, thereby eliminating the interference of different dimensions on the operation logic, and enabling the formulas to have mathematical operation rationality and objective law adaptability while retaining the characteristics of the original data distribution. The above is only an exemplary embodiment of the present invention and should not be used to limit the scope of the present invention.

Claims

1. An intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion, characterized in that The following are included: S1. Crawl multi-source data related to the operations of each department of an enterprise within a preset period, generate a multi-source heterogeneous data network for enterprise operations, and export the continuous operation risk factors of each department of the enterprise. S2. Obtain the corresponding resource declaration data of each department of the enterprise, and determine the resource allocation labels of the enterprise in the current declaration period with reference to the continuous operation risk factors of each department of the enterprise. The resource allocation labels include normal labels, restricted labels, and surplus labels. Among them, the resource declaration data includes the application amount, application purpose, and declaration period. S3. Under the normal label, allocate resources according to the application amount of each department; under the restricted label and surplus label, set department approval rules adapted to the resource allocation label, and accordingly divide each declaration permission department and each business optimization department of the enterprise. S4. Based on the resource declaration data of each declaration permission department, identify the adaptability of the declared resources of each declaration permission department, and accordingly set the resource allocation mechanism for different declaration permission departments. S5. Establish multiple operation progress systems for each business optimization department, analyze the optimization expectations of the corresponding multiple operation progress systems of each business optimization department, and accordingly generate the corresponding optimization marks for each business optimization department.

2. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 1, wherein The generation of the multi-source heterogeneous data network for enterprise operations includes: Store the collected multi-source data in the corresponding databases according to the data type. Take each type of database as a node, and build an independent transmission channel. Convert the heterogeneous data in each type of database of each department into a unified format through a normalization processing method, and accordingly generate the multi-source data parameters of each department. Obtain the multi-source data parameters related to the operations of each department of the enterprise in the historical declaration period, and import them into the enterprise comprehensive evaluation data pool. Establish a project application feature library and a project contribution feature library in the enterprise comprehensive evaluation data pool, match the multi-source data parameters related to the operations with the features in these two feature libraries, generate a project application feature set and a project contribution feature set, and accordingly obtain the multi-source heterogeneous data network for enterprise operations.

3. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 2, wherein The export of the continuous operation risk factors of each department of the enterprise includes: Based on the multi-source heterogeneous data network for enterprise operations, obtain the project application feature set and the project contribution feature set of each department in several historical declaration periods. Draw the corresponding historical operation risk trend charts of the project application feature set and the project contribution feature set of each department in several historical declaration periods based on the time series. Perform normal distribution fitting on the feature data in the historical operation risk trend charts to obtain the corresponding mean and standard deviation of the project application feature set and the project contribution feature set of each department. Obtain the confidence levels selected by each department in the case of project application and project contribution respectively, and determine the corresponding confidence level coefficients through the standard normal distribution table. Calculate the upper and lower limits of the confidence intervals of the project application feature set and the project contribution feature set of each department according to the mean, standard deviation, and confidence level coefficients. Statistically analyze the proportion of the number of features that fall within the corresponding confidence intervals in the project application feature set and the project contribution feature set of each department, and use this proportion as the persistence index of the project application and project contribution of each department's historical operations within the confidence interval. Integrate and generate the continuous operation risk factors of each department of the enterprise, and then export them.

4. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 1, characterized in that The resource allocation labels for determining the enterprise in the current application cycle include: Define that at least one department's continuous operation risk factor exceeding the preset operation risk threshold factor is the first-level restriction condition; Obtain the total resource allocation amount of the enterprise in the current application cycle, and statistically analyze the corresponding comprehensive application amounts of each department of the enterprise. Then define that the total resource allocation amount is lower than the comprehensive application amount as the second-level restriction condition; When the enterprise has at least one of the first-level restriction conditions or the second-level restriction conditions, determine that the resource allocation label is a restricted label; Otherwise, when the total resource allocation amount exceeds the comprehensive application amount, determine that the resource allocation label is a surplus label; when the total resource allocation amount is equal to the comprehensive application amount, determine that the resource allocation label is a regular label.

5. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 1, characterized in that, The division of each application permission department and each business optimization department of the enterprise includes: setting the first-level risk threshold value corresponding to the restricted label and the second-level risk threshold value corresponding to the surplus label, and the first-level risk threshold value is higher than the second-level risk threshold value; Under the restricted label, compare the continuous operation risk factors of each department with the first-level risk threshold value, and determine that the department with a continuous operation risk factor lower than the first-level risk threshold value is a business optimization department, and the rest are application permission departments; Under the surplus label, compare the continuous operation risk factors of each department with the second-level risk threshold value, and determine that the department with a continuous operation risk factor lower than the second-level risk threshold value is a business optimization department, and the rest are application permission departments; Statistically obtain each application permission department and each business optimization department of the enterprise.

6. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 2, wherein The setting of the resource allocation mechanism for different application permission departments includes: extracting the keywords of the declared uses in the corresponding resource declaration data of each application permission department through text mining technology, and matching them with the keyword library related to the core business of the corresponding department, and accordingly determining the suitability of the declared resources of each application permission department, including suitability and unsuitability; For the application permission departments with unsuitability, analyze the resource declaration rationality index based on the multi-source data related to the operation and the resource declaration data of the department, and accordingly generate the resource declaration rationality index of each application permission department; Taking the resource declaration rationality index of each application permission department as a reference, correct the application amount of each application permission department to obtain the final accounting amount of the department; For the application permission departments with suitability, take the declared amount as the final accounting amount; Statistically analyze the final accounting amounts of each application permission department to obtain the comprehensive application amount. If it does not exceed the total resource allocation amount, conduct resource accounting according to the regular issued link, otherwise introduce the enterprise's backup resource link to the regular issued link, and accordingly allocate resources to different application permission departments.

7. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 6, characterized in that, For the inapplicable declaration and approval departments, based on the multi-source data related to the operation of the declaration and approval departments and the resource declaration data, analyze their resource declaration rationality indicators, including: Count the number of keywords that fail to be retrieved in the corresponding declared uses of the inapplicable declaration and approval departments, and calculate the ratio of this number to the preset reference keyword number. The reciprocal of this ratio is the business correlation degree indicator between the corresponding resource declaration uses and the core business of the inapplicable declaration and approval departments; Derive the project application feature set within the current declaration cycle from the multi-source data related to the operation of the inapplicable declaration and approval departments, and obtain the industry average application feature set. Based on this, calculate the comprehensive deviation rate between the two, and map it to an inverse relationship with the resource application rationality indicator; Obtain the application amounts of the inapplicable declaration and approval departments in several historical declaration cycles, and examine the consistency indicator of the change trend of the application amounts in the time series with the historical application amounts of the declaration and approval departments; Assign different weights to each indicator according to its influence degree on the resource declaration rationality, and then determine the resource declaration rationality indicator within the range of 0-1 through normalization; 8. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 1, characterized in that, The analysis of the optimization expectations of the corresponding multiple operation progress systems of each business optimization department includes: Formulate the operation planning data of the operation projects of the business optimization departments, and track the actual operation data of the projects in real time. Based on this, generate the time progress system, resource progress system, and quality progress system; Based on the time progress system, calculate the corresponding time efficiency optimization expectation of the business optimization department; Based on the resource progress system, calculate the corresponding resource utilization optimization expectation of the business optimization department; Based on the quality progress system, calculate the corresponding quality improvement optimization expectation of the business optimization department; Statistically obtain the optimization expectations of the corresponding multiple operation progress systems of each business optimization department; 9. The intelligent decision-making method for enterprise operation risks based on multi-source heterogeneous data fusion according to claim 8, characterized in that The generation of the corresponding optimization marks for each business optimization department includes: based on the optimization expectations of the corresponding multiple operation progress systems of each business optimization department, determine the optimization requirement levels of each business optimization department in the three dimensions of time, resources, and quality. The optimization requirement levels are divided into high level, medium level, and low level. Then, perform optimization marks on the progress systems at different levels in each business optimization department; 10. An intelligent decision-making system for enterprise operation risks based on multi-source heterogeneous data fusion, characterized in that, Including: The operation risk assessment module crawls the multi-source data related to the operation of each department of the enterprise within a preset cycle, generates a multi-source heterogeneous data network for the operation of the enterprise, and exports the continuous operation risk factors of each department of the enterprise; The resource label determination module obtains the corresponding resource declaration data of each department of the enterprise, and takes the continuous operation risk factors of each department of the enterprise as a reference to determine the resource allocation label of the enterprise in the current declaration cycle. The resource allocation label includes a regular label, a restricted label, and a surplus label; Among them, the resource declaration data includes the application amount, application use, and declaration cycle; The department division module distributes resources according to the application amounts of each department under the regular label; under the restricted label and the surplus label, sets the department approval rules adapted to the resource allocation label, and accordingly divides each declaration and approval department and each business optimization department of the enterprise; A resource allocation module, based on the resource declaration data of each application and approval department, identifies the suitability of the declared resources of each application and approval department, and accordingly sets up the resource allocation mechanisms for different application and approval departments; A business optimization module, establishes multiple operation progress systems for each business optimization department, analyzes the optimization expectations of each business optimization department for the corresponding multiple operation progress systems, and accordingly generates corresponding optimization marks for each business optimization department.

Images