Enterprise multi-dimensional business index analysis data resource optimization method, medium and device

Abstract

This invention relates to a method, medium, and device for optimizing data resources for multi-dimensional business indicator analysis in enterprises. It collects multi-dimensional business indicator analysis requests from enterprise business systems, performs pattern recognition and feature extraction on these requests to generate business entity features, inputs these features into a resource allocation model to output priority scores and resource demand prediction results, calculates the optimal resource allocation scheme based on the priority scores and resource demand prediction results through an optimization decision model, and automatically schedules resources according to the optimal resource allocation scheme to generate business decision instructions. Finally, it outputs business decision instructions and a data resource management report. By employing a set similarity algorithm to automatically identify and quantify the resource reuse potential between different requests, it significantly improves the efficiency of enterprise data resource utilization and reduces redundant computational consumption, making resource allocation more rational and efficient; it also reduces manual intervention and improves the efficiency and response speed of enterprise data resource management.

Description

Technical Field

[0001] This invention relates to the field of business management, specifically to a method, medium, and equipment for optimizing data resources for multi-dimensional business indicator analysis of enterprises. Background Technology

[0002] In existing technologies, data query acceleration solutions for multi-dimensional business indicator analysis of enterprises mainly have the following two implementation methods:

[0003] Current Status 1: Database materialized view technology (such as Apache Doris). This technology defines materialized views through SQL, periodically or in real-time calculates and stores pre-aggregated data, and supports direct querying or transparent use of materialized views through query rewriting to accelerate queries. However, this solution has the following limitations: as a general database-level acceleration method, it is insufficient when facing complex enterprise-level indicator management systems, especially when dealing with complex business scenarios such as dimensional rhetoric, specific statistical period requirements, and composite indicator calculations. It requires a large number of external supplementary capabilities to achieve platform-based general acceleration, resulting in a low degree of completeness and platformization of the overall solution.

[0004] Current Status 2: Layered Materialization Acceleration Solution (e.g., Aloudata). This solution employs a multi-layered materialized view structure (typically including a wide table layer, a regular aggregation layer, an inter-row offset calculation layer, a complex metric layer, and a result materialization layer), maximizing the efficiency of computational resource reuse through layered design. While this solution can theoretically improve query performance, it suffers from significant implementation and operational complexity: the layered architecture makes system implementation difficult, query matching logic complex, checking dependencies between layers challenging, and there is a lack of effective means for creating and evaluating materialized views. Especially when facing large-scale enterprise analysis scenarios with tens of thousands of metrics and millions of dimension combinations, accurately identifying worthwhile dimension / metric combinations for materialization and establishing an efficient matching mechanism becomes the main technical obstacle to the practical implementation of this solution.

[0005] In summary, existing technical solutions cannot achieve an ideal balance between platformization, implementation complexity, operation and maintenance costs, and analytical performance, and cannot meet the urgent needs of modern enterprises for efficient, intelligent, and easy-to-use data resource optimization and management for multi-dimensional business indicator analysis. Summary of the Invention

[0006] In view of the above problems, the present invention provides a method, medium and device for optimizing data resources for multi-dimensional business indicator analysis of enterprises.

[0007] To achieve the above objectives, in a first aspect, the present invention provides a method for optimizing data resources for multi-dimensional business indicator analysis of enterprises, comprising:

[0008] Collect multi-dimensional business indicator analysis requests from enterprise business systems. These requests include dimension combination patterns, indicator calculation logic, query frequency characteristics, and response time requirements.

[0009] The system performs pattern recognition and feature extraction on multi-dimensional business indicator analysis requests to generate business entity features for resource optimization decisions. The feature extraction is configured to automatically identify and quantify the resource reuse potential between different requests using a set similarity algorithm.

[0010] The characteristics of business entities are input into the resource allocation model that integrates heat analysis and performance evaluation, and the priority score of each business entity and the corresponding resource demand prediction results are output.

[0011] Based on priority scoring and resource demand forecasting results, combined with predefined business value weights and computational cost constraints, the optimal resource allocation scheme for each business entity is calculated through an optimized decision-making model. The optimal resource allocation scheme includes the pre-calculation scope, update frequency, and storage resource configuration.

[0012] Based on the optimal resource allocation scheme and real-time resource capacity constraints, automated resource scheduling is performed to generate business decision instructions, including business view creation, business cache adjustment strategies, or resource reclamation.

[0013] Output business decision instructions and data resource management reports with benefit assessments and optimization suggestions.

[0014] In some embodiments, pattern recognition and feature extraction are performed on multi-dimensional business indicator analysis requests to generate business entity features for resource optimization decision-making, including:

[0015] The query patterns of multi-dimensional business indicator analysis requests are categorized to obtain either exploratory analysis requests or deterministic analysis requests.

[0016] If it is an exploratory analysis request, it will be marked as a non-optimized processing type, and the exploratory analysis request will be executed directly using real-time computing resources. Its query pattern characteristics and resource consumption will be recorded.

[0017] If it is a deterministic analysis request, extract the new dimension elements and new indicator elements contained in the deterministic analysis request to generate new business data entities;

[0018] A set similarity algorithm is used to calculate the resource overlap between newly added business data entities and existing business data entities;

[0019] When the resource overlap exceeds the preset threshold, the newly added dimension elements and newly added indicator elements will be merged into the matched existing business data entity and recorded as the business data entity to be processed.

[0020] Based on the analysis of historical request logs, the access popularity distribution and response performance indicators of each business entity in the pending business data entity are analyzed, and the historical access sequence, popularity score parameters and performance evaluation parameters of each business entity are generated.

[0021] Based on the preset business value weight configuration rules, and combined with the popularity scoring parameters and performance evaluation parameters, priority weight coefficients are assigned to the business entities of the business data entities to be processed.

[0022] Based on the dimension-indicator combination relationship, resource usage characteristics, popularity scoring parameters, performance evaluation parameters, and priority weight coefficients of the business data entities to be processed, business entity characteristics are generated.

[0023] In some embodiments, when the resource overlap does not exceed a preset threshold, the method includes:

[0024] Treat newly added business data entities as independent business data entities;

[0025] Based on a preset initial resource allocation strategy, basic computing and storage resources are allocated to independent business data entities.

[0026] Initiate access monitoring and performance tracing mechanisms for independent business data entities, record their dimensional element access patterns and indicator calculation load characteristics, and generate resource usage characteristics;

[0027] Additionally, based on the resource usage patterns of similar historical business data entities, predict the resource demand trends of independent business data entities and generate resource demand prediction parameters.

[0028] The dimension-metric combination relationship, resource usage characteristics, and resource demand prediction parameters of independent business data entities are incorporated into the feature set of existing business entities, while new historical request logs are generated.

[0029] In some embodiments, business entity characteristics are input into a resource allocation model that integrates heat analysis and performance evaluation, and the model outputs priority scores for each business entity and corresponding resource demand prediction results, including:

[0030] Based on the popularity score parameters and historical access sequences in the characteristics of business entities, a time series prediction algorithm is used to infer the future access popularity trend of each business entity and generate a popularity prediction value.

[0031] Extract performance evaluation parameters from the characteristics of business entities and combine them with real-time resource capacity constraints. Use the resource bottleneck identification model to determine the performance bottleneck points of each business entity under the current and predicted load, and obtain performance analysis results.

[0032] The heat prediction value and performance analysis results are input into the multi-objective optimization decision model. The multi-objective optimization decision model analyzes the business value weight and computational cost constraints at the same time, and outputs the priority score of each business entity.

[0033] Based on priority scores, a quantitative relationship between priority scores and specific resource requirements is established. Based on the quantitative relationship, resource requirement prediction results are generated for each business entity. The resource requirement prediction results include computational resource quotas, storage space configurations, and data update frequency requirements.

[0034] Feedback calibration is performed on the resource demand forecast results, and the final priority scores and resource demand forecast results for each business entity are output.

[0035] In some embodiments, a resource bottleneck identification model is used to determine the performance bottleneck points of each business entity under current and predicted loads, and performance analysis results are obtained, including:

[0036] Collect query response time, computing resource utilization and data throughput indicators of each business entity in the historical period, and construct a performance baseline model. The performance baseline model is a reference benchmark constructed by collecting historical performance data of each business entity in the normal operation period.

[0037] Based on the performance evaluation parameters in the characteristics of business entities, simulate the growth trend of resource demand of each business entity under the predicted load.

[0038] By comparing and analyzing the simulated resource demand growth trend with real-time resource capacity constraints, key business entities at risk of resource contention can be identified.

[0039] For the identified key business entities, root cause analysis algorithms are used to locate their performance bottleneck types, including compute-intensive bottlenecks, storage-intensive bottlenecks, or network transmission-intensive bottlenecks.

[0040] Based on different performance bottleneck types, corresponding bottleneck severity scores and bottleneck resolution priority rankings are generated;

[0041] The output includes performance analysis results that include the type of performance bottleneck, severity score, and resolution priority ranking.

[0042] In some embodiments, based on priority scoring and resource demand forecasting, combined with predefined business value weights and computational cost constraints, an optimal resource allocation scheme for each business entity is calculated through an optimization decision model, including:

[0043] A multi-objective optimization model is constructed with the objective functions of maximizing business value and minimizing computational cost. Business value is obtained by multiplying the priority score by the business value weight, and computational cost is represented by multiplying the resource demand forecast by the unit resource cost.

[0044] The total resource capacity, the minimum resource guarantee for a single business entity, and the maximum response time are used as constraints for the multi-objective optimization model.

[0045] The Pareto optimality algorithm is used to solve the multi-objective optimization model to obtain the Pareto solution set for resource allocation of each business entity;

[0046] Based on preset business strategy preferences, the optimal compromise solution is selected from the Pareto solution set of resource allocation to determine the computing resource quota, storage space allocation and data update frequency of each business entity, which is the optimal resource allocation scheme.

[0047] In some embodiments, selecting the optimal compromise solution from the Pareto solution set for resource allocation based on preset business strategy preferences includes:

[0048] Establish a business strategy preference evaluation matrix, which includes weight configurations for three dimensions: maximizing resource utilization, maximizing business value, and optimizing cost control.

[0049] For each solution in the Pareto solution set of resource allocation, a multi-dimensional benefit evaluation is performed, and the comprehensive score of each solution in three dimensions—resource utilization, business value realization, and cost control—is calculated.

[0050] The comprehensive scores of the three dimensions are fused according to the weight configuration in the business strategy preference evaluation matrix by weighted summation to obtain the comprehensive preference score of each solution;

[0051] Based on the ranking results of the comprehensive preference scores, the solution with the highest score is selected as the optimal compromise solution and output.

[0052] In some embodiments, automated resource scheduling is performed based on the optimal resource allocation scheme and real-time resource capacity constraints, generating business decision instructions including business view creation, business cache adjustment strategies, or resource reclamation, including:

[0053] The computing resource quota in the optimal resource allocation scheme is matched with the real-time available computing resource capacity to identify resource gaps or redundancy and obtain the resource matching analysis results.

[0054] Based on the resource matching analysis results, a resource scheduling strategy engine is used to generate specific resource operation instructions, including:

[0055] When a resource gap is identified in a high-priority business entity, a business view creation instruction is generated.

[0056] When resource utilization is detected to be lower than the preset utilization threshold, a business cache adjustment strategy instruction is generated.

[0057] When resource redundancy is identified and the business value score is lower than the elimination threshold, a resource reclamation instruction is generated.

[0058] Based on the data update frequency requirements of business entities and real-time data changes, formulate differentiated data update strategies and associate these strategies with corresponding resource operation instructions.

[0059] The execution priority of each instruction is assessed based on business value weight, resource urgency, and scope of impact.

[0060] Output business decision instructions with execution priority markers. These instructions include business view creation instructions, business cache adjustment strategy instructions, and resource reclamation instructions, along with the expected benefit assessment and execution constraints for each instruction.

[0061] In a second aspect, the present invention also provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the method described in the first aspect.

[0062] In a third aspect, the present invention also provides an electronic device including a memory and a processor, the memory being used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method described in the first aspect.

[0063] Unlike existing technologies, the above technical solution collects multi-dimensional business indicator analysis requests from enterprise business systems, performs pattern recognition and feature extraction on these requests to generate business entity features, inputs these features into a resource allocation model to output priority scores and resource demand prediction results, calculates the optimal resource allocation scheme through an optimization decision model based on the priority scores and resource demand prediction results, and automatically schedules resources according to the optimal resource allocation scheme to generate business decision instructions, ultimately outputting business decision instructions and a data resource management report. This technical solution significantly improves enterprise data resource utilization efficiency and reduces redundant computational consumption by automatically identifying and quantifying the resource reuse potential between different requests using a set similarity algorithm; it achieves accurate resource demand prediction based on a resource allocation model that integrates heat analysis and performance evaluation, making resource allocation more rational and efficient; the optimization decision model combining business value weights and computational cost constraints effectively controls computational costs while ensuring business analysis needs are met; and the automated resource scheduling mechanism generates business decision instructions, greatly reducing manual intervention and improving enterprise data resource management efficiency and response speed.

[0064] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0065] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on this application.

[0066] In the accompanying drawings of the instruction manual:

[0067] Figure 1 This is a schematic diagram illustrating steps S101 to S106 as described in the specific implementation method.

[0068] Figure 2 This is a schematic diagram illustrating steps S201 to S208 as described in the specific implementation method;

[0069] Figure 3 This is a schematic diagram illustrating steps S301 to S305 as described in the specific implementation method.

[0070] Figure 4 This is a schematic diagram illustrating steps S401 to S405 as described in the specific implementation method.

[0071] Figure 5 The following is a schematic diagram illustrating steps S501 to S506 as described in the specific implementation method. Detailed Implementation

[0072] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0073] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0074] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0075] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0076] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.

[0077] Without further limitations, the use of terms such as “comprising,” “including,” “having,” or other similar open-ended expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0078] As understood in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0079] The processor described in the embodiments of this application can be implemented by hardware, firmware, software, or a combination thereof. It can be a circuit, one or more of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, or a microprocessor. It also includes other physical, biological, or chemical structures that can implement the same or equivalent functions as the processors listed above, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some or all of the steps in the computer program or method involved in the various embodiments of this application, or any combination of the steps mentioned therein.

[0080] The computer program involved in the embodiments can be stored in a computer device readable storage medium, which includes, but is not limited to, disks, magnetic tapes, magnetic cards, floppy disks, flash memory, optical disks, optical cards, read-only memory (ROM), random access memory (RAM), erasable programmable ROM (EPROM), and electrically erasable programmable ROM (EEPROM), etc., and also includes other biological, physical, or chemical structures that can achieve the same or equivalent functions as the storage media listed above, such as DNA, RNA, proteins, and other units with information storage capabilities. In specific embodiments, the storage medium involved can be one of the above-mentioned media types, or a combination of the above-mentioned media types. In different embodiments, the computer program involved in the embodiments can be centrally stored in a single medium, or distributed and stored in multiple media. The memory containing the computer device readable storage medium can be non-volatile memory or random access memory. These computer device readable storage media can be built into the device, or can be connected to the device involved in the embodiments as an external device or part of an external device. In some embodiments, the memory having a computer device readable storage medium is deployed locally; in other embodiments, the memory may be deployed remotely from the processor, for example, as a network-attached memory accessed via RF circuitry or an external port and a communication network, wherein the communication network may be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), or a suitable combination thereof, as long as computer device access to the memory is enabled. Furthermore, the computer program involved in the embodiments may be stored in plaintext / ciphertext form, or it may be designed as training data, integrated and recombined through model training and implicitly stored in the parameter states of a deep neural network or other machine learning model.

[0081] Please see Figure 1 In a first aspect, this embodiment provides a method for optimizing data resources for multi-dimensional business indicator analysis of enterprises, including:

[0082] S101. Collect multi-dimensional business indicator analysis requests from enterprise business systems. Multi-dimensional business indicator analysis requests include dimension combination patterns, indicator calculation logic, query frequency characteristics, and response time requirements.

[0083] S102. Perform pattern recognition and feature extraction on multi-dimensional business indicator analysis requests to generate business entity features for resource optimization decisions. Feature extraction is configured to automatically identify and quantify the resource reuse potential between different requests using a set similarity algorithm.

[0084] S103. Input the characteristics of the business entities into the resource allocation model that integrates heat analysis and performance evaluation, and output the priority score of each business entity and the corresponding resource demand prediction results.

[0085] S104. Based on priority scoring and resource demand prediction results, combined with predefined business value weights and computational cost constraints, the optimal resource allocation scheme for each business entity is calculated through an optimized decision model. The optimal resource allocation scheme includes the pre-calculation scope, update frequency, and storage resource configuration.

[0086] S105. Perform automated resource scheduling based on the optimal resource allocation scheme and real-time resource capacity constraints, and generate business decision instructions including business view creation, business cache adjustment strategy or resource reclamation.

[0087] S106. Output business decision instructions and data resource management reports with benefit assessments and optimization suggestions.

[0088] In step S101, a multi-dimensional business indicator analysis request refers to a data query request initiated by the enterprise's business system that includes specific analytical requirements. The dimension combination mode represents the business dimensions involved in the analysis and their combination methods; the indicator calculation logic defines the business indicators to be calculated and their calculation formulas; the query frequency characteristic reflects the historical frequency pattern of this analysis request; and the response time requirement specifies the business-level limitation on the query completion time. This information is collected through the enterprise's existing query log monitoring system and business metadata management system, ensuring that resource optimization decisions are based on complete business requirement characteristics.

[0089] In step S102, pattern recognition and feature extraction refer to the process of automatically classifying and mining the collected analysis requests. Business entity features refer to a standardized set of features that, after integration, represent a class of similar analysis needs. A set similarity algorithm is used to calculate the degree of similarity between different analysis requests in terms of dimensional and indicator combinations, thereby identifying request groups that can share computing resources. The potential for resource reuse is quantified by the shareable computing power among similar requests. By replacing manual classification with automated algorithms, the efficiency and accuracy of resource reuse identification are significantly improved.

[0090] In step S103, the resource allocation model integrating heat analysis and performance evaluation is an intelligent decision-making model that comprehensively considers business access hotspots and system performance indicators. Heat analysis predicts future access trends based on historical query frequency characteristics, while performance evaluation combines actual system operating indicators (such as query latency and resource utilization) to determine the effectiveness of the current resource allocation. Priority scoring is used to rank the urgency of resource allocation for different business entities, and the resource demand prediction results quantify the computing, storage, and update resources required by each business entity.

[0091] In step S104, the optimization decision model is a multi-objective constrained optimization model, where the business value weight is determined by the business importance, and the computational cost constraint is set based on the enterprise's IT resource budget. The pre-computation scope defines which data needs to be pre-computed and stored, the update frequency determines the data refresh cycle, and the storage resource configuration specifies the storage medium and capacity allocation. By solving for the resource allocation scheme that maximizes business value under cost constraints, a balance between resource investment and business returns is achieved.

[0092] In step S105, automated resource scheduling refers to the process of automatically executing resource allocation operations according to the optimal resource allocation scheme, where real-time resource capacity constraints reflect the currently available computing and storage resources of the system. Business view creation refers to generating a materialized view based on the pre-calculated scope; business cache adjustment strategy refers to optimizing cache allocation based on access patterns; and resource reclamation refers to releasing idle or low-value resources. By replacing manual intervention with automated operations, the timely and accurate execution of the resource allocation scheme is ensured.

[0093] In step S106, the data resource management report is a comprehensive resource optimization effect evaluation report. The benefit evaluation quantifies the performance improvement and cost savings brought about by resource optimization, the optimization suggestions provide further improvement directions, and the resource optimization results are presented to managers in a visual way.

[0094] This embodiment achieves efficient utilization of enterprise data resources through fully automated management of the entire process from business request collection to resource scheduling and execution. Intelligent feature extraction based on a set similarity algorithm ensures the accuracy of resource reuse; a resource allocation model integrating multi-dimensional evaluation indicators provides a scientific basis for decision-making; an automated resource scheduling mechanism guarantees the timely implementation of the solution; and the final data resource management report provides guidance for continuous optimization. It is easy to implement within the enterprise's existing technical architecture and can effectively improve data resource utilization efficiency and reduce operating costs.

[0095] Please see Figure 2 In some embodiments, pattern recognition and feature extraction are performed on multi-dimensional business indicator analysis requests to generate business entity features for resource optimization decisions, including:

[0096] S201. Classify the query patterns of multi-dimensional business indicator analysis requests to obtain either exploratory analysis requests or deterministic analysis requests;

[0097] S202. If it is an exploratory analysis request, mark the exploratory analysis request as a non-optimized processing type, use real-time computing resources to directly execute the exploratory analysis request, and record its query pattern characteristics and resource consumption.

[0098] S203. If it is a deterministic analysis request, extract the new dimension elements and new indicator elements contained in the deterministic analysis request and generate new business data entities.

[0099] S204. Use a set similarity algorithm to calculate the resource overlap between newly added business data entities and existing business data entities;

[0100] S205. When the resource overlap exceeds the preset threshold, the newly added dimension elements and newly added indicator elements will be merged into the matched existing business data entity and recorded as the business data entity to be processed.

[0101] S206. Based on historical request logs, analyze the access popularity distribution and response performance indicators of each business entity in the business data entity to be processed, and generate historical access sequence, popularity score parameters and performance evaluation parameters for each business entity.

[0102] S207. Based on the preset business value weight configuration rules, and combined with the popularity scoring parameters and performance evaluation parameters, assign priority weight coefficients to the business entities of the business data entities to be processed.

[0103] S208. Based on the dimension-indicator combination relationship, resource usage characteristics, popularity scoring parameters, performance evaluation parameters, and priority weight coefficients of the business data entities to be processed, generate business entity characteristics.

[0104] In step S201, query pattern classification refers to dividing the analysis request into different processing types based on its business characteristics. Among them, exploratory analysis request represents a free exploratory query without preset dimensions and indicator ranges, while deterministic analysis request refers to a fixed pattern query with clear dimension combinations and indicator calculation logic.

[0105] In step S202, the non-optimized processing type indicates that this type of request does not participate in the resource optimization process and is directly allocated real-time computing resources for execution. The query pattern feature record includes the dimensional combination tendency and computational complexity, and the resource consumption record records the actual amount of computing and storage resources used to ensure the real-time requirements of exploratory queries.

[0106] In step S203, the newly added dimension elements and newly added indicator elements refer to the dimension attributes and indicator calculation requirements in the request that are not covered by the existing business data entities. The newly added business data entities are temporary data entities formed by integrating these elements, ensuring that each deterministic analysis request can be normalized into a standard data processing unit.

[0107] In step S204, the resource overlap is calculated using a set similarity algorithm to quantify the degree of overlap between newly added and existing business data entities in terms of dimension and indicator combinations. Preferably, the Jaccard similarity coefficient is used to calculate the similarity between the dimension set and the indicator set, and then a weighted average is used to obtain a comprehensive overlap score.

[0108] In step S205, the preset threshold is set based on historical optimization results, typically between 50% and 70%. When the overlap exceeds the threshold, the newly added element is merged into the most similar existing entity to form a business data entity to be processed. This effectively reduces data entity redundancy and improves resource reuse efficiency.

[0109] In step S206, the historical access sequence records the access patterns of entities in different time periods. The popularity score parameter is calculated based on the access frequency and recent access trends. The performance evaluation parameters include average response time and resource utilization efficiency indicators.

[0110] In step S207, the business value weight configuration rule is set according to the importance of the business department and the business impact. The priority weight coefficient is obtained by weighted fusion heat score and performance evaluation parameters, and is used to identify the priority level of resource allocation for business entities.

[0111] In step S208, the business entity feature is a composite feature vector formed by integrating all the aforementioned parameters. The dimension-index combination relationship defines the data scope of the entity, the resource usage feature describes the operational characteristics of the entity, and the popularity, performance, and weight parameters provide quantitative evaluation indicators.

[0112] This embodiment achieves intelligent identification and standardized processing of analysis requests through a refined request classification and feature extraction mechanism. It differentiates between exploratory and deterministic requests to ensure efficient resource optimization; employs a set similarity algorithm to effectively identify reusable resources and reduce redundancy; and constructs a complete feature profile of business entities based on multi-dimensional parameters.

[0113] Please see Figure 3 In some embodiments, when the resource overlap does not exceed a preset threshold, the method includes:

[0114] S301. Treat the newly added business data entity as an independent business data entity;

[0115] S302. Allocate basic computing and storage resources to independent business data entities based on the preset initial resource configuration strategy;

[0116] S303. Start the access monitoring and performance tracking mechanism for independent business data entities, record their dimension element access patterns and indicator calculation load characteristics, and generate resource usage characteristics.

[0117] S304, and based on the resource usage patterns of similar historical business data entities, predict the resource demand trends of independent business data entities and generate resource demand prediction parameters.

[0118] S305. Incorporate the dimension-indicator combination relationship, resource usage characteristics, and resource demand prediction parameters of independent business data entities into the existing business entity feature set, and generate new historical request logs.

[0119] In step S301, an independent business data entity refers to a new business data unit that has low similarity to existing business data entities and requires independent resource allocation. This determination is based on the comparison results of the aforementioned resource overlap and preset threshold, ensuring that each unique analysis requirement can obtain an independent resource management unit.

[0120] In step S302, the initial resource configuration strategy refers to the rules for allocating basic resources to newly created independent business data entities, including initial computing resource quotas and storage space allocation. Preferably, the basic resource configuration can refer to the historical average resource usage of similar business entities, or adopt the default resource configuration template specified by the enterprise, to ensure that the new entity can start and run normally.

[0121] In step S303, the access monitoring and performance tracking mechanism refers to a system that continuously monitors the operating status of independent business data entities. The dimension element access pattern records the access frequency and combination method of each dimension, and the indicator calculation load characteristic records the CPU, memory and other resource consumption during the indicator calculation process.

[0122] In step S304, resource demand trend prediction is based on the growth patterns of historically similar business data entities, and a time series forecasting algorithm is used to analyze resource usage growth patterns. Resource demand prediction parameters include quantitative indicators such as expected increase in computational resources and storage space expansion needs, which are used to guide future resource expansion decisions.

[0123] In step S305, the newly generated business entity characteristics and historical request logs will be integrated into the existing management system for unified management. The dimension-indicator combination relationship defines the business scope of the business entity, resource usage characteristics provide actual operational data, and resource demand forecasting parameters support forward-looking resource planning. This information together improves the enterprise's data resource management map.

[0124] This embodiment provides a complete solution for creating and managing independent entities to address new business needs with low resource overlap. By establishing independent monitoring and prediction mechanisms, it ensures that each business data entity receives a resource management strategy that matches its characteristics, guaranteeing the normal operation of new businesses and achieving comprehensive coverage and dynamic adaptation of resource management.

[0125] Please see Figure 4 In some embodiments, business entity characteristics are input into a resource allocation model that integrates heat analysis and performance evaluation, and the model outputs priority scores for each business entity and corresponding resource demand prediction results, including:

[0126] S401. Based on the popularity rating parameters and historical access sequences in the characteristics of business entities, a time series prediction algorithm is used to deduce the future access popularity trend of each business entity and generate a popularity prediction value.

[0127] S402. Extract performance evaluation parameters from the characteristics of business entities, and combine them with real-time resource capacity constraints. Use the resource bottleneck identification model to determine the performance bottleneck points of each business entity under the current and predicted load, and obtain the performance analysis results.

[0128] S403. Input the heat prediction value and performance analysis results into the multi-objective optimization decision model. The multi-objective optimization decision model analyzes the business value weight and computational cost constraints at the same time, and outputs the priority score of each business entity.

[0129] S404. Based on the priority score, establish a quantitative relationship between the priority score and the specific resource requirements. Generate resource requirement prediction results for each business entity based on the quantitative relationship. The resource requirement prediction results include the calculation resource quota, storage space configuration and data update frequency requirements.

[0130] S405. Feedback calibration of resource demand forecast results, and output the final priority score and resource demand forecast results for each business entity.

[0131] In step S401, the time series prediction algorithm uses mature prediction models such as ARIMA or LSTM to extrapolate the trend of access volume changes within a specific future time period based on the time pattern characteristics in historical access sequences and the periodicity in the popularity scoring parameters. The generated popularity prediction value quantifies the expected access pressure of each business entity in the future, providing a time-dimensional reference for resource pre-allocation.

[0132] In step S402, the resource bottleneck identification model analyzes performance evaluation parameters such as response time and throughput, and combines them with the availability of resources such as CPU, memory, and storage in real-time resource capacity constraints to identify the performance bottleneck points of each business entity under the current load. It also predicts the type of bottleneck (compute-intensive, storage-intensive, or I / O-intensive) that may occur under the future load corresponding to the heat prediction value obtained in step S401. The performance analysis results clarify the urgency and type of resource requirements of each entity.

[0133] In step S403, the multi-objective optimization decision model aims to maximize business value and minimize computational cost. Business value is represented by the product of business value weight and predicted popularity value, while computational cost is represented by the product of resource demand and unit resource cost. By solving for the Pareto optimal solution set, and under the premise of satisfying real-time resource capacity constraints, a priority score that comprehensively considers business importance and cost-effectiveness is output.

[0134] In step S404, the resource demand mapping algorithm establishes a quantitative relationship between priority scores and specific resource demands, mapping the priority scores to computing resource quotas (such as the number of CPU cores and memory size), storage space configurations (such as storage capacity and storage type), and data update frequency requirements (such as real-time updates, hourly updates, or daily updates). Preferably, a regression model or lookup table based on historical data is used to achieve an accurate mapping from scores to resource demands.

[0135] In step S405, the feedback calibration mechanism dynamically adjusts the parameter settings of the prediction model by comparing the deviation between historical prediction values ​​and actual resource usage. During calibration, key considerations are given to influencing factors such as seasonality and abnormal fluctuations in business activities to ensure that the final output priority score and resource demand prediction results are highly consistent with actual business needs.

[0136] This embodiment achieves a precise mapping from business characteristics to resource requirements through multi-stage analysis, prediction, and optimization decision-making. Heat analysis based on time-series forecasting provides forward-looking insights into resource demand; resource bottleneck identification ensures the targeted and effective allocation of resources; a multi-objective optimization model balances business value and cost control; and a feedback calibration mechanism guarantees the accuracy and adaptability of the prediction results.

[0137] Please see Figure 5 In some embodiments, a resource bottleneck identification model is used to determine the performance bottleneck points of each business entity under current and predicted loads, and performance analysis results are obtained, including:

[0138] S501. Collect query response time, computing resource utilization and data throughput indicators of each business entity in the historical period, and construct a performance baseline model. The performance baseline model is a reference benchmark constructed by collecting historical performance data of each business entity in the normal operation period.

[0139] S502. Based on the performance evaluation parameters in the characteristics of business entities, simulate the growth trend of resource demand of each business entity under the predicted load.

[0140] S503. Compare and analyze the simulated resource demand growth trend with the real-time resource capacity constraints to identify key business entities that are at risk of resource contention.

[0141] S504. For the identified key business entities, use root cause analysis algorithms to locate their performance bottleneck types, including compute-intensive bottlenecks, storage-intensive bottlenecks, or network transmission-intensive bottlenecks.

[0142] S505. Generate corresponding bottleneck severity scores and bottleneck resolution priority rankings based on different performance bottleneck types.

[0143] S506. Outputs performance analysis results including performance bottleneck type, severity score, and resolution priority ranking.

[0144] In step S501, the performance baseline model is a reference benchmark constructed by collecting historical performance data of each business entity during its normal operating cycle. The query response time reflects the system processing efficiency, the computing resource utilization rate represents the CPU and memory utilization rate, and the data throughput index measures the I / O processing capability. Together, they constitute the performance profile of the business entity and provide a comparison benchmark for bottleneck identification.

[0145] In step S502, the resource demand growth trend simulation is based on the load characteristic indicators in the performance evaluation parameters. It uses prediction methods such as linear regression or exponential smoothing to deduce the changes in the demand for computing, storage and network resources of each business entity under the predicted load, so as to ensure the rationality of the prediction results.

[0146] In step S503, resource contention risk identification is achieved by comparing simulated demand with real-time resource capacity constraints. When the predicted resource demand exceeds the optimal threshold of available system resources, the business entity is determined to have resource contention risk. Key business entities refer to those entities that both have resource contention risk and possess high business value weight.

[0147] In step S504, the root cause analysis algorithm uses machine learning methods such as decision trees or random forests to locate bottleneck types by analyzing the correlation between performance indicators: computationally intensive bottlenecks are characterized by a high correlation between CPU utilization and query response time; storage-intensive bottlenecks show a strong correlation between disk I / O and response time; and network transmission-intensive bottlenecks are characterized by a positive correlation between network bandwidth utilization and data transmission latency.

[0148] In step S505, the bottleneck severity score is calculated based on the resource gap ratio and the scope of business impact, with a value ranging from 0 to 100. The resolution priority ranking comprehensively considers the severity score, business value weight, and resolution cost, using a weighted scoring method to determine the processing order.

[0149] In step S506, the performance analysis results are output in a structured data format, including bottleneck type classification, severity quantification score, and processing priority recommendations.

[0150] This embodiment achieves precise location and prioritization of performance issues through a systematic bottleneck identification and analysis method. A performance baseline based on historical data provides a scientific comparison standard; multi-dimensional simulation analysis ensures comprehensive risk identification; root cause analysis algorithms enable accurate classification of bottleneck types; and quantitative scoring and prioritization provide clear decision-making basis for resource optimization. The entire solution helps enterprises quickly locate and resolve performance bottlenecks, improving system resource utilization efficiency.

[0151] In some embodiments, based on priority scoring and resource demand forecasting, combined with predefined business value weights and computational cost constraints, an optimal resource allocation scheme for each business entity is calculated through an optimization decision model, including:

[0152] A multi-objective optimization model is constructed with the objective functions of maximizing business value and minimizing computational cost. Business value is obtained by multiplying the priority score by the business value weight, and computational cost is represented by multiplying the resource demand forecast by the unit resource cost.

[0153] The total resource capacity, the minimum resource guarantee for a single business entity, and the maximum response time are used as constraints for the multi-objective optimization model.

[0154] The Pareto optimality algorithm is used to solve the multi-objective optimization model to obtain the Pareto solution set for resource allocation of each business entity;

[0155] Based on preset business strategy preferences, the optimal compromise solution is selected from the Pareto solution set of resource allocation to determine the computing resource quota, storage space allocation and data update frequency of each business entity, which is the optimal resource allocation scheme.

[0156] In this embodiment, the multi-objective optimization model refers to a mathematical model that simultaneously optimizes two objectives: maximizing business value and minimizing computational cost. Business value is quantified by the product of priority score and business value weight, reflecting the importance of the business. Computational cost is represented by the product of resource demand forecast and unit resource cost, reflecting the economy of resource consumption. The dual-objective balance ensures that resource allocation meets business needs while controlling costs.

[0157] Total resource capacity constraints limit the upper limit of available system resources, minimum resource guarantees for single business entities ensure that each entity receives a basic resource quota, and maximum response time constraints guarantee business performance requirements. Pareto optimality algorithms employ NSGA-II or multi-objective particle swarm optimization, using evolutionary computation to obtain a set of non-dominated solutions constituting the Pareto solution set. Each solution represents a different balance between business value and computational cost.

[0158] Business strategy preferences are based on corporate strategic goals, such as prioritizing maximizing business value or minimizing costs. When selecting the optimal compromise solution from the Pareto solution set, decision-making methods such as weighted summation or ε-constraint are used. The final determined computing resource quotas, storage space allocations, and data update frequencies constitute the optimal resource allocation scheme.

[0159] This embodiment achieves a scientific balance between business value and resource costs through a multi-objective optimization method. The use of Pareto optimality ensures the diversity and superiority of the solution set; flexible adaptation to specific enterprise needs is achieved through business strategy preferences; and comprehensive consideration of constraints guarantees the feasibility of the solution implementation.

[0160] In some embodiments, selecting the optimal compromise solution from the Pareto solution set for resource allocation based on preset business strategy preferences includes:

[0161] Establish a business strategy preference evaluation matrix, which includes weight configurations for three dimensions: maximizing resource utilization, maximizing business value, and optimizing cost control.

[0162] For each solution in the Pareto solution set of resource allocation, a multi-dimensional benefit evaluation is performed, and the comprehensive score of each solution in three dimensions—resource utilization, business value realization, and cost control—is calculated.

[0163] The comprehensive scores of the three dimensions are fused according to the weight configuration in the business strategy preference evaluation matrix by weighted summation to obtain the comprehensive preference score of each solution;

[0164] Based on the ranking results of the comprehensive preference scores, the solution with the highest score is selected as the optimal compromise solution and output.

[0165] In this embodiment, the business strategy preference evaluation matrix refers to an evaluation framework used to quantify a company's resource allocation strategy preferences, which includes weighted configurations for three dimensions: maximizing resource utilization, maximizing business value, and optimizing cost control. The weights of each dimension are determined by the company's strategic objectives. For example, when prioritizing business response speed, the weight of business value can be increased, while when focusing on operational efficiency, the weight of resource utilization can be increased.

[0166] Multi-dimensional benefit assessment refers to the process of quantitatively evaluating each candidate solution in the Pareto solution set. The resource utilization dimension calculates the efficiency of the solution in utilizing system resources; the business value realization dimension assesses the degree to which the solution supports business objectives; and the cost control dimension measures the economic performance of the solution. Each dimension uses standardized indicators for a percentage-based scoring system.

[0167] The weighted summation algorithm linearly weights the scores of the three dimensions according to preset weights. The calculation formula can be: Overall Preference Score = W1 × Resource Utilization Score + W2 × Business Value Score + W3 × Cost Control Score, where W1, W2, and W3 are the weights of each dimension and their sum is 1. Based on the ranking results of the overall preference scores, the solution with the highest score is selected as the optimal compromise solution. This solution achieves the best balance across the three dimensions and best meets the overall strategic needs of the enterprise.

[0168] This embodiment achieves the scientific selection of the optimal solution that best aligns with the company's strategy from multiple Pareto optimal solutions by establishing a quantitative evaluation matrix and a weighted decision-making mechanism. The evaluation matrix comprehensively considers the multi-dimensional impact of resource allocation; the weighted summation algorithm ensures the interpretability and operability of the decision-making process; and the final selected compromise solution satisfies both business needs and the company's strategic orientation. The entire selection process is standardized and transparent, providing a reliable basis for corporate resource allocation decisions.

[0169] In some embodiments, automated resource scheduling is performed based on the optimal resource allocation scheme and real-time resource capacity constraints, generating business decision instructions including business view creation, business cache adjustment strategies, or resource reclamation, including:

[0170] The computing resource quota in the optimal resource allocation scheme is matched with the real-time available computing resource capacity to identify resource gaps or redundancy and obtain the resource matching analysis results.

[0171] Based on the resource matching analysis results, a resource scheduling strategy engine is used to generate specific resource operation instructions, including:

[0172] When a resource gap is identified in a high-priority business entity, a business view creation instruction is generated.

[0173] When resource utilization is detected to be lower than the preset utilization threshold, a business cache adjustment strategy instruction is generated.

[0174] When resource redundancy is identified and the business value score is lower than the elimination threshold, a resource reclamation instruction is generated.

[0175] Based on the data update frequency requirements of business entities and real-time data changes, formulate differentiated data update strategies and associate these strategies with corresponding resource operation instructions.

[0176] The execution priority of each instruction is assessed based on business value weight, resource urgency, and scope of impact.

[0177] Output business decision instructions with execution priority markers. These instructions include business view creation instructions, business cache adjustment strategy instructions, and resource reclamation instructions, along with the expected benefit assessment and execution constraints for each instruction.

[0178] In this embodiment, resource matching analysis refers to the process of comparing the planned resource allocation in the optimal resource allocation scheme with the real-time available resources of the system. By calculating the difference between resource demand and available resources, resource gaps or resource redundancies are accurately identified. Resource gaps indicate that resources are in short supply, while resource redundancy indicates that resources are over-allocated. The resource scheduling strategy engine is an intelligent decision-making system based on a rule engine, which automatically generates corresponding operation instructions based on the resource matching analysis results.

[0179] For resource gaps in high-priority business entities, generate business view creation instructions to improve computing efficiency;

[0180] When resource utilization is lower than a preset utilization threshold, a commercial cache adjustment strategy instruction is generated to improve resource utilization efficiency.

[0181] When resource redundancy exists and the business value score is lower than the elimination threshold, a resource reclamation instruction is generated to release idle resources.

[0182] The differentiated data update strategy is formulated based on the data update frequency requirements and real-time data change frequency of business entities. It distinguishes between high-frequency changing data and low-frequency static data, and associates the update strategy with corresponding resource operation instructions to ensure a balance between data consistency and resource efficiency. Execution priority assessment comprehensively considers three dimensions: business value weight, resource urgency, and impact scope. A weighted scoring method is used to calculate priority scores, ensuring that critical business needs and urgent requirements receive priority resource support. The final output business decision instruction includes execution priority markings and detailed expected benefit assessments, including quantitative indicators such as expected performance improvements and cost savings estimates. It also clearly defines execution constraints such as execution time windows and resource dependencies, providing comprehensive guidance for resource scheduling operations.

[0183] This embodiment achieves precise execution and dynamic optimization of resource allocation through intelligent resource scheduling strategies and a refined instruction generation mechanism. Matching analysis based on real-time resource status ensures the accuracy of scheduling decisions; multi-condition instruction generation rules cover various resource management scenarios; a priority evaluation mechanism guarantees the resource needs of critical businesses; and complete benefit assessments and constraint descriptions provide a reliable basis for decision execution. The entire scheduling process achieves automation and intelligence in resource management.

[0184] In a second aspect, the present invention also provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the method described in the first aspect.

[0185] In a third aspect, the present invention also provides an electronic device including a memory and a processor, the memory being used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method described in the first aspect.

[0186] To facilitate understanding, the following application examples are provided:

[0187] This example proposes a single-layer materialization acceleration scheme, aiming to effectively accelerate queries when querying data in complex indicator management systems. The example includes the following parts:

[0188] 1. For user / system queries, distinguish between random queries and routine queries;

[0189] (1) Random query refers to a query in which the user has not predefined the dimension / indicator range, such as the analyst randomly selecting data;

[0190] (2) Routine queries are for queries with predefined dimensions / metric ranges. Examples include BI dashboards, metric data retrieval, and periodic reports.

[0191] (3) The invention is only for routine queries;

[0192] 2. For routine queries, the concept of a "dataset" is introduced;

[0193] (1) A defined “dataset” containing defined dimensions and metrics;

[0194] (2) When a user adds a new “dataset”, it is necessary to check and judge the similarity with the existing “dataset”. If the similarity is high (repetition > 50%), it is recommended that the user use the existing dataset and merge the newly added dimension / metric into the existing dataset; if it cannot be merged, a new “dataset” should be added.

[0195] (3) Users can query specific dimensions / indicators of the "dataset";

[0196] 3. Create a materialization acceleration table using the dataset as the sole reference;

[0197] (1) Create a materialized acceleration table for the request hotspot dataset;

[0198] ① Record request logs and periodically extract hot data sets (focus on hot dimensions / metrics);

[0199] ②Accelerate the execution of requests for frequently requested datasets;

[0200] (2) Create a materialized speed-up table for the slow query dataset;

[0201] ① Record request logs and periodically extract slow query datasets (focus on hot dimensions / metrics);

[0202] ② Create a materialized speed-up table for the slow query dataset;

[0203] (3) Create materialized acceleration tables (e.g., core reports) for key datasets;

[0204] ① Select any dataset and create a materialization acceleration table;

[0205] (4) For a single dataset, multiple materialization acceleration tables can be created based on different combinations of dimensions / indicators;

[0206] 4. Query acceleration matching and SQL rewriting;

[0207] (1) Accelerate the materialization of tables based on “dataset” + dimension / metric matching;

[0208] (2) If multiple tables are hit at the same time, select the accelerated table with the fewest partition data;

[0209] (3) SQL rewriting: Replacing the original table query with a request for a materialized table to speed up the query;

[0210] (4) No materialization acceleration was hit, continue with the original query;

[0211] 5. Accelerated physical-chemical assessment and post-treatment;

[0212] (1) Record materialized hit logs;

[0213] (2) Regularly assess the usage of physical and chemical tables and clean up low-frequency physical and chemical acceleration tables;

[0214] This example has the following advantages:

[0215] 1. Introduce the concept of "dataset" and select materialized tables for creation based on a combination of manual selection, hot requests, and slow query requests;

[0216] 2. Limiting the "dataset" during matching improves matching efficiency. For large "metric management systems" with many materialized acceleration tables, it can effectively reduce matching time.

[0217] 3. Single-layer materialization accelerates / matches, simplifying engineering implementation and facilitating the location of abnormal data links.

[0218] Unlike existing technologies, the above technical solution collects multi-dimensional business indicator analysis requests from enterprise business systems, performs pattern recognition and feature extraction on these requests to generate business entity features, inputs these features into a resource allocation model to output priority scores and resource demand prediction results, calculates the optimal resource allocation scheme through an optimization decision model based on the priority scores and resource demand prediction results, and automatically schedules resources according to the optimal resource allocation scheme to generate business decision instructions, ultimately outputting business decision instructions and a data resource management report. This technical solution significantly improves enterprise data resource utilization efficiency and reduces redundant computational consumption by automatically identifying and quantifying the resource reuse potential between different requests using a set similarity algorithm; it achieves accurate resource demand prediction based on a resource allocation model that integrates heat analysis and performance evaluation, making resource allocation more rational and efficient; the optimization decision model combining business value weights and computational cost constraints effectively controls computational costs while ensuring business analysis needs are met; and the automated resource scheduling mechanism generates business decision instructions, greatly reducing manual intervention and improving enterprise data resource management efficiency and response speed.

[0219] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A method for optimizing data resources for multi-dimensional business indicator analysis of enterprises, characterized in that, include: Collect multi-dimensional business indicator analysis requests from enterprise business systems. These multi-dimensional business indicator analysis requests include dimension combination patterns, indicator calculation logic, query frequency characteristics, and response time requirements. The multi-dimensional business indicator analysis requests are subjected to pattern recognition and feature extraction to generate business entity features for resource optimization decisions. The feature extraction is configured to automatically identify and quantify the resource reuse potential between different requests using a set similarity algorithm. The characteristics of the business entities are input into the resource allocation model that integrates heat analysis and performance evaluation, and the priority score of each business entity and the corresponding resource demand prediction results are output. Based on the priority score and resource demand prediction results, combined with the predefined business value weight and computing cost constraints, the optimal resource allocation scheme for each business entity is calculated through an optimized decision model. The optimal resource allocation scheme includes the pre-calculation scope, update frequency and storage resource configuration. Based on the optimal resource allocation scheme and real-time resource capacity constraints, automated resource scheduling is performed to generate business decision instructions, including business view creation, business cache adjustment strategies, or resource reclamation. Output the aforementioned business decision instructions and a data resource management report with benefit assessment and optimization suggestions; The multi-dimensional business indicator analysis request is subjected to pattern recognition and feature extraction to generate business entity features for resource optimization decision-making, including: The query patterns of multi-dimensional business indicator analysis requests are categorized to obtain either exploratory analysis requests or deterministic analysis requests. If it is an exploratory analysis request, then the exploratory analysis request is marked as a non-optimized processing type, the exploratory analysis request is executed directly using real-time computing resources, and its query pattern characteristics and resource consumption are recorded; If it is a deterministic analysis request, then extract the new dimension elements and new indicator elements contained in the deterministic analysis request to generate a new business data entity; A set similarity algorithm is used to calculate the resource overlap between newly added business data entities and existing business data entities; When the resource overlap exceeds a preset threshold, the newly added dimension elements and newly added indicator elements are merged into the matched existing business data entity and recorded as the business data entity to be processed. Based on the analysis of historical request logs, the access popularity distribution and response performance indicators of each business entity in the business data entity to be processed are analyzed, and the historical access sequence, popularity score parameters and performance evaluation parameters of each business entity are generated. According to the preset business value weight configuration rules, and in combination with the popularity scoring parameters and performance evaluation parameters, priority weight coefficients are assigned to the business entities of the business data entities to be processed. Based on the dimension-index combination relationship, resource usage characteristics, popularity scoring parameters, performance evaluation parameters, and priority weight coefficients of the business data entity to be processed, the characteristics of the business entity are generated.

2. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 1, characterized in that, When the resource overlap does not exceed a preset threshold, the method includes: The newly added business data entity is treated as an independent business data entity; Based on a preset initial resource allocation strategy, basic computing and storage resources are allocated to the independent business data entities. Initiate an access monitoring and performance tracing mechanism for the independent business data entity, record its dimensional element access patterns and indicator calculation load characteristics, and generate resource usage characteristics; Furthermore, based on the resource usage patterns of historically similar business data entities, the resource demand trend of the independent business data entity is predicted, and resource demand prediction parameters are generated. The dimension-index combination relationship, resource usage characteristics, and resource demand prediction parameters of the independent business data entities are incorporated into the existing business entity feature set, and new historical request logs are generated simultaneously.

3. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 1, characterized in that, The characteristics of the business entities are input into a resource allocation model that integrates heat analysis and performance evaluation. The model outputs priority scores for each business entity and corresponding resource demand prediction results, including: Based on the popularity rating parameters and historical access sequences in the characteristics of the business entities, a time series prediction algorithm is used to infer the future access popularity trend of each business entity and generate a popularity prediction value. The performance evaluation parameters are extracted from the characteristics of the business entities and combined with real-time resource capacity constraints. The performance bottleneck points of each business entity under the current and predicted load are determined by the resource bottleneck identification model to obtain the performance analysis results. The predicted heat value and performance analysis results are input into the multi-objective optimization decision model. The multi-objective optimization decision model simultaneously analyzes the business value weight and computational cost constraints, and outputs the priority score of each business entity. Based on the priority score, a quantitative relationship between the priority score and specific resource requirements is established. Based on the quantitative relationship, resource requirement prediction results are generated for each business entity. The resource requirement prediction results include the calculation of resource quotas, storage space configuration, and data update frequency requirements. The resource demand forecast results are fed back and calibrated to output the final priority scores and resource demand forecast results for each business entity.

4. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 3, characterized in that, The process involves using a resource bottleneck identification model to determine the performance bottlenecks of each business entity under current and predicted loads, and obtaining performance analysis results, including: Collect query response time, computing resource utilization and data throughput indicators of each business entity in the historical period, and construct a performance baseline model. The performance baseline model is a reference benchmark constructed by collecting historical performance data of each business entity in the normal operation period. Based on the performance evaluation parameters in the characteristics of the business entities, simulate the growth trend of resource demand of each business entity under the predicted load. By comparing and analyzing the simulated resource demand growth trend with real-time resource capacity constraints, key business entities at risk of resource contention can be identified. For the identified key business entities, root cause analysis algorithms are used to locate their performance bottleneck types, including compute-intensive bottlenecks, storage-intensive bottlenecks, or network transmission-intensive bottlenecks. Based on different performance bottleneck types, corresponding bottleneck severity scores and bottleneck resolution priority rankings are generated; The output includes performance analysis results that include the type of performance bottleneck, severity score, and resolution priority ranking.

5. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 1, characterized in that, Based on the aforementioned priority scoring and resource demand forecast, and combined with predefined business value weights and computational cost constraints, the optimal resource allocation scheme for each business entity is calculated through an optimized decision-making model, including: A multi-objective optimization model is constructed with the objectives of maximizing business value and minimizing computational cost. The business value is calculated by multiplying the priority score by the business value weight, and the computational cost is represented by multiplying the resource demand forecast by the unit resource cost. The total resource capacity, the minimum resource guarantee for a single business entity, and the maximum response time are used as constraints for the multi-objective optimization model. The Pareto optimality algorithm is used to solve the multi-objective optimization model to obtain the Pareto solution set for resource allocation of each business entity; Based on preset business strategy preferences, the optimal compromise solution is selected from the Pareto solution set of resource allocation to determine the computing resource quota, storage space allocation and data update frequency of each business entity, which is the optimal resource allocation scheme.

6. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 5, characterized in that, The selection of the optimal compromise solution from the Pareto solution set of resource allocation based on preset business strategy preferences includes: Establish a business strategy preference evaluation matrix, which includes weight configurations for three dimensions: maximizing resource utilization, maximizing business value, and optimizing cost control. For each solution in the Pareto solution set of resource allocation, a multi-dimensional benefit evaluation is performed, and the comprehensive score of each solution in three dimensions—resource utilization, business value realization, and cost control—is calculated. The comprehensive scores of the three dimensions are fused according to the weight configuration in the business strategy preference evaluation matrix by weighted summation to obtain the comprehensive preference score of each solution; Based on the ranking results of the comprehensive preference scores, the solution with the highest score is selected as the optimal compromise solution and output.

7. The method for optimizing data resources for multi-dimensional business indicator analysis of enterprises according to claim 1, characterized in that, Based on the optimal resource allocation scheme and real-time resource capacity constraints, automated resource scheduling is performed to generate business decision instructions, including business view creation, business cache adjustment strategies, or resource reclamation, including: The computing resource quota in the optimal resource allocation scheme is matched and analyzed with the real-time available computing resource capacity to identify resource gaps or redundancies and obtain the resource matching analysis results. Based on the resource matching analysis results, a resource scheduling strategy engine is used to generate specific resource operation instructions, including: When a resource gap is identified in a high-priority business entity, a business view creation instruction is generated. When resource utilization is detected to be lower than the preset utilization threshold, a business cache adjustment strategy instruction is generated. When resource redundancy is identified and the business value score is lower than the elimination threshold, a resource reclamation instruction is generated. Based on the data update frequency requirements of business entities and real-time data changes, formulate differentiated data update strategies and associate these strategies with corresponding resource operation instructions. The execution priority of each instruction is assessed based on business value weight, resource urgency, and scope of impact. Output business decision instructions with execution priority markers. These instructions include business view creation instructions, business cache adjustment strategy instructions, and resource reclamation instructions, along with the expected benefit assessment and execution constraints for each instruction.

8. A computer-readable storage medium storing computer program instructions thereon, characterized in that, The computer program instructions, when executed by a processor, implement the method as described in any one of claims 1 to 7.

9. An electronic device comprising a memory and a processor, characterized in that, The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method as described in any one of claims 1 to 7.

Images