Financial fund performance evaluation data management system and method

Abstract

The invention discloses a financial fund performance evaluation data management system and method, and belongs to the technical field of data processing, and the system comprises a joint data collection module which is used for building a source data set; the intelligent distribution module is used for configuring dimension weights; the index construction module is used for constructing performance evaluation indexes; the scoring module is used for establishing a scoring analysis result; and the management module is used for carrying out financial fund performance evaluation management. The technical problem that the performance evaluation result lacks objectivity and accuracy due to the fact that data sources are scattered and evaluation dimensions are not comprehensive in existing financial fund performance evaluation is solved, multiple data sources are integrated, comprehensive multi-dimensional evaluation is achieved, and the performance evaluation efficiency is improved. And the objectivity and accuracy of performance evaluation are improved through intelligent distribution and analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a financial funds performance evaluation data management system and method. Background Art

[0002] In modern fiscal management, the rational allocation and use of fiscal funds are of great significance to the sustainable development of the social economy. However, the existing fiscal fund performance evaluation methods face many challenges, which limit the optimal effect of fiscal funds in the process of use. First, the dispersion of data sources is one of the main problems in the current performance evaluation of fiscal funds. The use of fiscal funds involves a variety of data sources, which are often stored in different systems. There is a lack of unified data management and integration mechanisms, which makes data integration and analysis difficult. Secondly, the evaluation dimensions of fiscal fund performance in the existing evaluation system are not comprehensive enough. They usually only focus on the input-output relationship, ignore key dimensions such as results and benefits, and cannot fully reflect the effect of the use of fiscal funds. In addition, the setting of the evaluation criteria is relatively fixed, and lacks the ability to dynamically adjust according to specific usage scenarios, which makes it difficult for the evaluation results to adapt to different fiscal fund usage scenarios, affecting the accuracy and scientificity of the evaluation. Summary of the invention

[0003] This application provides a financial funds performance evaluation data management system and method, which aims to solve the technical problems in the existing financial funds performance evaluation, such as scattered data sources and incomplete evaluation dimensions, which lead to the lack of objectivity and accuracy of performance evaluation results, and to achieve the technical effect of integrating multiple data sources, realizing comprehensive multi-dimensional evaluation, and improving the objectivity and accuracy of performance evaluation through intelligent allocation and analysis.

[0004] In view of the above problems, this application provides a financial funds performance evaluation data management system and method.

[0005] The first aspect disclosed in the present application provides a financial funds performance evaluation data management system, the system including: a joint data acquisition module, used to access the data source, perform data acquisition, and establish a source data set, the data source including a project management source, a financial source, a questionnaire storage source, and a sensor data source; an intelligent allocation module, used to obtain the use scenarios of financial funds, and configure dimension weights based on the use scenarios, the dimensions corresponding to the dimension weights include input dimension, output dimension, result dimension, benefit dimension, and impact dimension; an indicator construction module, used to divide the indicators of each dimension respectively through the indicator allocation network based on the use scenarios, establish the indicator division results, and use the indicator division results and dimension weights to construct performance evaluation indicators; a scoring module, used to extract the data features of the source data set by feature engineering after data cleaning, and use the performance evaluation indicators to score and analyze the data feature extraction results to establish the scoring analysis results; a management module, used to store the source data set and the scoring analysis results, and perform financial funds performance evaluation management.

[0006] Another aspect disclosed in the present application provides a method for managing data for performance evaluation of fiscal funds, the method comprising: accessing a data source, executing data collection, and establishing a source data set, wherein the data source comprises a project management source, a financial source, a questionnaire storage source, and a sensor data source; obtaining usage scenarios of fiscal funds, and configuring dimension weights based on the usage scenarios, wherein the dimensions corresponding to the dimension weights comprise input dimension, output dimension, result dimension, benefit dimension, and impact dimension; based on the usage scenarios, dividing indicators of each dimension respectively through an indicator allocation network, establishing indicator division results, and constructing performance evaluation indicators using the indicator division results and dimension weights; after data cleaning the source data set, extracting data features of the source data set using feature engineering, and performing scoring analysis on the data feature extraction results using performance evaluation indicators, and establishing scoring analysis results; storing the source data set and the scoring analysis results, and conducting performance evaluation management of fiscal funds.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The joint data collection module is used to access the data source, perform data collection, and establish the source data set, and the data source includes the project management source, the financial source, the questionnaire storage source, and the sensor data source; the intelligent allocation module obtains the use scenario of the financial funds, and configures the dimension weight based on the use scenario, and the dimension corresponding to the dimension weight includes the input dimension, the output dimension, the result dimension, the benefit dimension, and the impact dimension; the indicator construction module divides the indicators of each dimension through the indicator allocation network based on the use scenario, establishes the indicator division result, and constructs the performance evaluation indicator using the indicator division result and the dimension weight; the scoring module cleans the source data set, extracts the data features of the source data set by feature engineering, and uses the performance evaluation indicator to score and analyze the data feature extraction results to establish the scoring analysis result; the management module stores the source data set and the scoring analysis results to manage the performance evaluation of financial funds. It solves the technical problems of the scattered data sources and incomplete evaluation dimensions in the existing financial fund performance evaluation, which leads to the lack of objectivity and accuracy of the performance evaluation results, and achieves the technical effect of integrating multiple data sources, realizing comprehensive multi-dimensional evaluation, and improving the objectivity and accuracy of performance evaluation through intelligent allocation and analysis.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A structural schematic diagram of a financial fund performance evaluation data management system is provided for the embodiment of the present application;

[0011] Figure 2 A flow chart of a method for managing financial funds performance evaluation data is provided for an embodiment of the present application.

[0012] Explanation of the accompanying drawings: joint data collection module 1, intelligent allocation module 2, indicator construction module 3, scoring module 4, management module 5. DETAILED DESCRIPTION

[0013] This application provides a financial funds performance evaluation data management system and method to solve the technical problems in the existing financial funds performance evaluation, such as scattered data sources and incomplete evaluation dimensions, which lead to the lack of objectivity and accuracy of performance evaluation results. It achieves the technical effect of integrating multiple data sources, realizing comprehensive multi-dimensional evaluation, and improving the objectivity and accuracy of performance evaluation through intelligent allocation and analysis.

[0014] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0015] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a financial fund performance evaluation data management system, the system comprising:

[0016] The joint data collection module 1 is used to access the data source, perform data collection, and establish a source data set. The data source includes a project management source, a financial source, a questionnaire storage source, and a sensor data source.

[0017] Specifically, the system terminal collects and integrates data by accessing multiple data sources to build a complete source data set. These data sources cover key information in different fields, such as management data used to track project progress and resource usage in project management systems, financial data used to record fund flows and financial statements in financial systems, questionnaire data storage systems used to collect feedback and opinions, and sensor networks used to monitor and collect environmental and process data in real time. By integrating data from these multiple sources, comprehensive and real-time financial information can be obtained, providing a basis for subsequent data analysis and performance evaluation.

[0018] Intelligent allocation module 2 is used to obtain the usage scenarios of financial funds and configure dimension weights based on the usage scenarios. The dimensions corresponding to the dimension weights include input dimension, output dimension, result dimension, benefit dimension, and impact dimension.

[0019] Specifically, the system knows to first obtain the specific usage scenarios of fiscal funds. This step is to gain an in-depth understanding of the specific application of funds, such as in infrastructure, public services, education or scientific research. According to these usage scenarios, the weights of various evaluation dimensions are specifically configured through weight proportion calculation and information entropy calculation. These dimensions include input dimension, output dimension, result dimension, benefit dimension, impact dimension, etc., which are used to comprehensively measure the performance of funds. Among them, the input dimension is used to measure the input amount and resource usage of fiscal funds. The output dimension is used to evaluate the specific results or outputs produced after the use of funds. The result dimension is used to analyze the actual impact of the use of funds on the completion of target tasks. The benefit dimension is used to evaluate the economic and social benefits brought about by the input and output of funds. The impact dimension is used to measure the long-term impact of the use of funds on the wider society, economy or environment. Through this weight configuration based on usage scenarios, the importance of different dimensions in the overall performance evaluation can be dynamically adjusted, so that the evaluation is more in line with the actual situation.

[0020] Furthermore, the acquisition of the usage scenarios of the fiscal funds and the configuration of the dimension weights based on the usage scenarios include:

[0021] The weights of each dimension in the historical data are called according to the usage scenario, where each dimension weight is marked with a trust value. After the weights of each dimension are standardized, the weight proportions are reconstructed based on the trust value mark and the standardized processing results, as follows: in, represents the weight ratio of the i-th sample in the j-th dimension after smoothing, α is the smoothing coefficient, T j Represents the trust value of the jth dimension, Y ij Represents the weight value of the i-th sample in the j-th dimension after standardization, m is the total number of samples, and i represents any sample; after calculating the information entropy based on the reconstructed weight ratio, the dimension weight is configured through the information entropy.

[0022] In a preferred embodiment, the system terminal calls the stored weights of each dimension from the historical data of each data source according to the specific usage scenario. These weight data carry a trust value identifier to indicate the credibility of the weight data. This step ensures that the data can be reasonably adjusted according to historical records and trust when the weight is configured. In order to ensure that the weights of different dimensions are calculated on the same scale, the system terminal standardizes the weights of each dimension, that is, the difference between each dimension weight and the minimum value of each dimension weight in the collected original dimensional weights is calculated by the ratio of the maximum value of each dimension weight and the minimum value of each dimension weight. Through this standardization process, the weights of all dimensions are adjusted to the same magnitude, thereby obtaining a standardized processing result, which is convenient for subsequent weight ratio calculations. After the standardization process, the system terminal reconstructs the weight ratio of each dimension through the weight ratio calculation formula based on the trust value identifier and the standardized result. The weight ratio calculation formula is as follows: in, Characterizes the weight proportion of the i-th sample in the j-th dimension after smoothing. α is the smoothing coefficient, which is used to control the trade-off between historical trust value and uniform distribution in the weight proportion. If the application scenario is small and stable, a larger α can be used. If the scenario changes greatly and is uncertain, a smaller α can be selected to enhance the role of uniform distribution. For example, in long-term projects with government financial funds (urban subway construction, highway maintenance, etc.), such projects often have many years of stable historical data, and the funding usage pattern and effect expectations are relatively stable. Therefore, α can be 0.8, so that the evaluation relies more on reliable historical trust values ​​to allocate weights and ensure consistency in performance evaluation. In pilot projects implemented for the first time (green energy subsidy programs, new technology research and development subsidies, etc.), the historical data of these projects are few or not reliable enough, and there is great uncertainty in the scenario. In this case, α can be 0.2, so that the evaluation mainly relies on uniform distribution to increase the flexibility of weight allocation and avoid over-reliance on limited or unstable historical data. T j Represents the trust value of the jth dimension. ij Represents the weight value of the i-th sample in the j-th dimension after standardization. m is the total number of samples. i represents any sample. After the weight ratio reconstruction is completed, the system terminal will configure the dimension weight based on the calculated weight ratio through the information entropy calculation formula and the dimension weight configuration formula. Through the above steps, the dimension weights in the fiscal fund use scenario can be dynamically configured according to the dimension weights of historical data, standardization processing results, trust value identification and information entropy, thereby improving the accuracy and objectivity of performance evaluation.

[0023] Furthermore, after calculating the information entropy based on the reconstructed weight proportion, configuring the dimension weights based on the information entropy includes:

[0024] The information entropy is calculated by the following formula: E j Characterizes the information entropy of the jth dimension, k is a normalization constant, Configure dimension weights based on information entropy: w j is the weight of the jth dimension, and n is the total number of dimensions.

[0025] In a feasible implementation, after reconstructing the weight proportion, the system terminal uses information entropy calculation to further analyze the distribution of each dimension. The calculation formula of information entropy is as follows: Among them, E j The information entropy of the jth dimension is used to measure the uncertainty of the data in each dimension. A higher information entropy indicates that the data distribution in this dimension is more uniform, and vice versa. k is a normalization constant used to ensure that the value of information entropy is within a reasonable range and is defined as After the information entropy calculation is completed, the weights of each dimension are configured based on the calculation results of the information entropy. The configuration process involves the dimension weight configuration formula, which is as follows: w j is the weight of the jth dimension, and n is the total number of dimensions. This formula is used to configure the dimension weights based on the information entropy results. j When w is smaller (i.e., the uncertainty is lower), j The larger the value, the greater the weight of this dimension. j When w is larger, j The value is small, reducing its influence in the overall weight. Through the above steps, the weights of different dimensions can be effectively configured according to information entropy to improve the accuracy and rationality of weight allocation of different dimensions in the performance evaluation of fiscal funds.

[0026] The indicator construction module 3 is used to divide the indicators of each dimension respectively through the indicator allocation network based on the usage scenario, establish the indicator division results, and use the indicator division results and dimension weights to construct performance evaluation indicators.

[0027] Specifically, the system terminal divides each performance dimension into separate indicators through a preset indicator allocation network according to the acquired usage scenario. The indicator allocation network is a structured method for classifying and allocating evaluation indicators, which can reasonably allocate and subdivide indicators for different dimensions according to scenario requirements. Each dimension, such as input, output, result, benefit, impact, etc., is carefully allocated indicators through the indicator classification layer and indicator reconstruction layer of the network to ensure that key evaluation indicators are covered in their respective fields. After completing the indicator division of each dimension, the system terminal will generate an indicator division result, in which each dimension has its corresponding evaluation indicator set. Subsequently, these division results and the previously configured dimension weights are used to comprehensively construct a comprehensive performance evaluation indicator, that is, according to the weight of each dimension, the specific indicators in the dimension are weighted, and the weighted results of each dimension are summarized. The performance evaluation indicator enables each indicator to be appropriately weighted according to its importance in the overall evaluation, thereby forming a more objective and comprehensive performance evaluation framework. The purpose of this process is to ensure that the evaluation can not only refine the specific indicators of different dimensions, but also reasonably allocate weights on a multi-dimensional basis, so that the final performance evaluation is more in line with the actual scenario requirements and has higher accuracy.

[0028] Furthermore, based on the usage scenario, the indicator division of each dimension is performed through the indicator allocation network, and the indicator division result is established, including:

[0029] The indicator allocation network is initialized with the usage scenario, and the indicator grading layer of the initialized indicator allocation network is called to perform graded fitting on each dimension to establish an indicator grading result, wherein the graded fitting of each dimension includes at least 2-level indicator fitting; the indicator reconstruction layer of the initialized indicator allocation network is called to perform weighted reconstruction of the indicator's own weight and the grading weight based on the indicator grading result, and establish the indicator division result with the weighted reconstruction result.

[0030] In a preferred embodiment, the system terminal constructs an indicator allocation network including an input layer, an indicator classification layer, an indicator reconstruction layer, and an output layer according to the indicators to be evaluated. The input layer is used to receive the usage scenario data and the initial indicator information, the classification layer is responsible for subdividing and fitting the indicators of each dimension, the reconstruction layer is used for weighted reconstruction of the comprehensive weight and the classification weight, and the output layer is used to output the established indicator division results. Multiple nodes are set in each level, each node corresponds to a specific indicator or indicator category, and each node has a corresponding indicator weight, which is determined based on the historical data of the indicator, and is used to indicate the importance of each indicator in its dimension. In order to construct an indicator allocation network, the system terminal will obtain historical scenario data of historical usage scenarios, including historical scenario descriptions, historical weight data, historical trust values, and historical indicator sets. Between the layers of the network, the initial weights and bias parameters are defined using a random initialization method. In the indicator classification layer, a neural network is used to design an algorithm for hierarchical fitting of each dimension, and each dimension has at least two levels of indicator division, such as a primary indicator for a large category (such as an input dimension) and a secondary indicator for a more specific sub-item (such as personnel expenses and equipment costs). In the reconstruction layer, a weighted reconstruction algorithm is defined, and a weighted synthesis is performed by combining the weight of the indicator itself and the grading weight. The reconstruction layer is used to adjust and redistribute weights to ensure that the weights of indicators at all levels are reasonably distributed. Subsequently, a historical scene data is randomly extracted from the historical scene data as the first historical scene data, and the first historical scene data is input into the network, and then passed to the indicator grading layer and reconstruction layer through the input layer. The output value of each historical scene data is calculated according to the network structure and the current weight, and the predicted value of the indicator division result is obtained. The weighted cross entropy loss is then used to measure the error between the predicted output and the historical indicator set, and the loss value is calculated. After that, the back propagation algorithm is used to calculate the gradient of the loss function for each parameter in the network. According to the calculated gradient, the optimization algorithm (such as gradient descent, Adam, etc.) is used to update the network parameters to reduce the loss value. Repeat the process of forward propagation, loss calculation and parameter update, and iterate the training until the loss value converges or reaches the preset stop condition (such as the number of training rounds or loss threshold). After the training is completed, the system terminal matches the parameters of the current indicator allocation network with the usage scenario, and stores them in the indicator allocation network data table for subsequent direct calls. For other historical scenario data, the system terminal performs the same operation and stores the parameters of the indicator allocation network determined each time together with the corresponding usage scenario into the indicator allocation network data table.

[0031] When using the indicator allocation network, the system terminal first initializes the indicator allocation network according to the acquired usage scenario, that is, matches the network parameters suitable for the scenario from the indicator allocation network data table. This process ensures that the network can reasonably allocate and divide indicators according to specific scenarios. After the network is initialized, the system terminal inputs the scenario data of the usage scenario into the indicator allocation network. The input layer of the indicator allocation network receives these data and passes them to the indicator classification layer. The indicator allocation network performs hierarchical fitting for each dimension according to the learned mapping relationship. At least two levels of indicator fitting are performed for each dimension to refine and classify the evaluation indicators at different levels. After the indicator classification fitting, the indicator classification results for each dimension are generated, which include the indicators at each level under each dimension and their hierarchical relationships. After obtaining the classification results, the initialized indicator reconstruction layer is called. This layer performs weighted reconstruction of the indicators at each level according to the established classification results. Weighted reconstruction includes comprehensive calculation of the weight of each indicator and the weight of the level to which it belongs. In the weighted reconstruction process, the comprehensive weight of each indicator is calculated to ensure that the weight distribution of the indicators at each level is reasonable. This step involves the combination of the indicator's own weight and the hierarchical weight to ensure that the reconstructed weight reflects the importance of indicators at each level. Finally, the indicator division result is generated using the weighted reconstruction result and output through the output layer. This indicator division result includes the specific indicators and their weight configuration under each dimension, which is used for subsequent performance evaluation analysis to ensure the comprehensiveness of the evaluation and the rationality of the weight.

[0032] Scoring module 4 is used to extract data features of the source data set by using feature engineering after data cleaning of the source data set, and to score and analyze the data feature extraction results by using performance evaluation indicators to establish scoring analysis results;

[0033] Specifically, when processing the source data set, the system terminal first performs data cleaning to ensure the quality and consistency of the data. Data cleaning includes steps such as removing missing values, correcting outliers, and unifying data formats, with the goal of eliminating noise and errors in the data and making the data more reliable. Subsequently, feature engineering is applied to extract useful data features from the source data set. This step uses principal component analysis, feature fusion, and other methods to identify and extract information that can represent the key features of the data, thereby enhancing the expressiveness and interpretability of the data. After the feature extraction is completed, the constructed performance evaluation indicators are used to analyze and score the extraction results. During the scoring analysis process, the system terminal matches the extracted data features with the performance evaluation indicators. Each performance evaluation indicator corresponds to one or more data features for scoring calculation. Then, according to the importance of different dimensions, the weights of the performance evaluation indicators are applied to the relevant data features. The weight is used to adjust the contribution of each feature in the final score. For example, a dimension has a higher weight, which means that it accounts for a larger proportion in the score. After that, the score corresponding to each feature is calculated, that is, the indicator weight corresponding to the feature is multiplied by the value of the feature, and all calculation results are summarized to generate a comprehensive score on each dimension. After calculating the scores of all features, the calculated scores are normalized by the extreme value normalization method to ensure that the scores are within a unified range for easy comparison. Then, the scores of each dimension are weighted and summed to obtain the final comprehensive score. This final comprehensive score and the scores of each dimension are stored to establish the score analysis results. This process takes into account the importance of each dimension to ensure that the final score result can accurately reflect the overall performance of the sample.

[0034] Furthermore, after the source data set is cleaned, feature engineering is used to extract data features of the source data set, including:

[0035] The principal component analysis tool is used to extract the principal components of the source data set after data cleaning, and the principal component extraction result is established; the principal component extraction result is fused with homologous features to establish a first fusion result; the first fusion result is fused with multi-source data to establish a second fusion result; and data features are extracted according to the second fusion result.

[0036] In a preferred embodiment, after the source data set is cleaned, the system terminal uses the principal component analysis (PCA) tool to perform dimensionality reduction processing on the cleaned data. PCA projects high-dimensional data into low-dimensional space through linear transformation, and extracts the principal component with the largest explained variance in the data. The extracted principal components are used to construct the principal component extraction results, which are important representatives of the information in the data set. In the principal component extraction results, data features with the same source are selected, and the homologous features are fused by weighted average or other combination methods to form a new feature set, that is, the first fusion result including homologous features. Subsequently, the first fusion result is fused with data features from other sources by weighted average or splicing, such as different sensor data, questionnaire data or financial data. After the fusion is completed, a comprehensive feature set containing multi-source information is formed, that is, the second fusion result. After that, the features in the second fusion result are extracted, that is, the correlation between the features in the second fusion result and the performance target of the fiscal funds is calculated by using the Pearson correlation coefficient or the Spearman correlation coefficient, and the calculated correlation coefficient is compared with the correlation threshold (such as 0.5), and the data features greater than or equal to the correlation threshold are extracted. Through the above steps, principal component analysis and multi-level fusion technology can be used to extract high-value comprehensive data features from different sources and different levels to ensure the representativeness of the data and the accuracy of the analysis.

[0037] Management module 5 is used to store source data sets and scoring analysis results, and to conduct financial fund performance evaluation management.

[0038] Specifically, after obtaining the scoring analysis results, the system terminal stores the source data set and the scoring analysis results. This process ensures that all data and analysis results are stored securely and systematically for subsequent review, tracking and analysis. The stored source data set retains the complete, cleaned original information and extracted key features, which is convenient for backtracking or further processing when needed. The scoring analysis results include detailed scores and comprehensive scores, reflecting the contribution of different dimensions and features in performance evaluation. By storing these data and analysis results, the system terminal realizes the continuous management and optimization of fiscal funds performance, provides a reliable data foundation for relevant departments, and supports them in making reasonable adjustments in future decision-making and fund allocation, thereby improving the efficiency of fund use.

[0039] Furthermore, the system also includes:

[0040] The permission management module 5 is used to configure data access rights for user roles and establish user verification, which includes linkage verification of identity and location. When a user applies for data access, the linkage verification of identity and location is called to identify the user. If the identification result is a pass result, a permission instruction is generated; the data management module 5 is used to call the data display within the permission according to the permission instruction, and permit the user to perform data management within the permission.

[0041] In a preferred embodiment, the financial fund performance evaluation data management system has a rights management module 5 and a data management module 5 in addition to the joint data collection module 1, the intelligent allocation module 2, the indicator construction module 3, the scoring module 4, and the management module 5. In the rights management module 5, the system terminal configures corresponding data access rights for different user roles to ensure that only authorized users can access specific data content. In order to enhance data security, a user verification mechanism is established, including linkage verification of identity and location. When a user applies for data access, the system terminal will call this verification mechanism to verify the user's identity and location information. The linkage verification performs identity identification by comparing the user's current location with a preset trusted location (such as an office location or a registered device location). If the verification result confirms that the user meets the identity and location requirements, an access permission instruction will be generated to allow the user to access the data. This mechanism not only ensures the security of the data, but also provides users with a more rigorous access control process to effectively prevent unauthorized access. In the data management module 5, once the access permission instruction is generated, the user can call the data within his authority for display and operation. The system terminal loads the relevant data according to the permission instruction to ensure that the user can only view and manage the data content that meets his role authority. This means that users can perform management operations such as data query, editing, updating, or other permission-related operations, but cannot access or change other data beyond their authorized scope. In this way, not only the data usage needs of users are met, but also the security and integrity of data are maintained through strict permission control. This process ensures that users can effectively perform data management tasks within the scope of their permissions, while preventing unauthorized access and modification.

[0042] Furthermore, the system also includes:

[0043] The feedback module is used to receive user feedback and establish a feedback optimization data set; the intervention impact module is used to generate real-time optimization response data based on the feedback optimization data set and the preset intervention index, and optimize the indicator construction module 3 through the real-time optimization response data.

[0044] In a preferred embodiment, the financial fund performance evaluation data management system also includes a feedback module and an intervention impact module. In the feedback module, the system receives user feedback information during operation, whether it is the user's opinions on the data usage experience, function suggestions or problems found. After these feedbacks are collected and sorted by the system terminal, a feedback optimization data set is created. This data set contains the user's specific feedback content, timestamp, user role information, etc., as the basis for subsequent analysis and optimization. In the intervention impact module, after receiving user feedback and establishing a feedback optimization data set, the system terminal will analyze it in combination with a preset intervention index. The intervention index is a parameter used to measure the priority and depth of the system response optimization needs, which is determined by weighted summation of user role weights, severity and feedback frequency. The system terminal generates real-time optimization response data based on these feedbacks and intervention indexes. This data contains specific measures required for optimization (such as adjusting the hierarchical layer algorithm of the indicator allocation network, reallocating the weights of indicators, etc.), priority arrangements, etc. After generating the real-time optimization response data, the system terminal adjusts the parameters and logic of the indicator construction module 3 according to the optimization suggestions therein, for example, changing the weights in the indicator allocation network, updating the indicator grading method, deleting invalid indicators, etc. Through this optimization, the system terminal can more accurately configure and allocate indicators in different usage scenarios, making them more in line with the actual needs in user feedback and improving flexibility and accuracy in performance evaluation.

[0045] In summary, the financial fund performance evaluation data management system provided by the embodiment of the present application has the following technical effects:

[0046] The joint data collection module 1 is used to access the data source, perform data collection, and establish a source data set, wherein the data source includes a project management source, a financial source, a questionnaire storage source, and a sensor data source; the intelligent allocation module 2 is used to obtain the use scenario of fiscal funds, and configure the dimension weight based on the use scenario, wherein the dimension corresponding to the dimension weight includes the input dimension, the output dimension, the result dimension, the benefit dimension, and the impact dimension; the indicator construction module 3 is used to divide the indicators of each dimension through the indicator allocation network based on the use scenario, establish the indicator division result, and construct the performance evaluation indicator using the indicator division result and the dimension weight; the scoring module 4 is used to extract the data features of the source data set by feature engineering after data cleaning of the source data set, and perform scoring analysis on the data feature extraction result by using the performance evaluation indicator to establish the scoring analysis result; the management module 5 is used to store the source data set and the scoring analysis result, and perform the performance evaluation management of fiscal funds. Through the above steps, the technical problems of the scattered data sources and incomplete evaluation dimensions in the existing fiscal funds performance evaluation, which lead to the lack of objectivity and accuracy of the performance evaluation results, are solved, and the technical effects of integrating multiple data sources, realizing comprehensive multi-dimensional evaluation, and improving the objectivity and accuracy of performance evaluation through intelligent allocation and analysis are achieved.

[0047] Embodiment 2, based on the same inventive concept as a financial fund performance evaluation data management system in the above embodiment, Figure 2 As shown, the embodiment of the present application provides a method for managing financial fund performance evaluation data, the method comprising:

[0048] Access the data source, perform data collection, and establish a source data set, wherein the data source includes a project management source, a financial source, a questionnaire storage source, and a sensor data source; obtain the usage scenarios of fiscal funds, and configure dimension weights based on the usage scenarios, wherein the dimensions corresponding to the dimension weights include input dimension, output dimension, result dimension, benefit dimension, and impact dimension; based on the usage scenarios, perform indicator division of each dimension through an indicator allocation network, establish indicator division results, and use the indicator division results and dimension weights to construct performance evaluation indicators; after data cleaning of the source data set, use feature engineering to extract data features of the source data set, and use performance evaluation indicators to perform scoring analysis on the data feature extraction results to establish scoring analysis results; store the source data set and scoring analysis results, and conduct performance evaluation management of fiscal funds.

[0049] Furthermore, the intelligent allocation module 2 includes:

[0050] The weights of each dimension in the historical data are called according to the usage scenario, where each dimension weight is marked with a trust value. After the weights of each dimension are standardized, the weight proportions are reconstructed based on the trust value mark and the standardized processing results, as follows: in, represents the weight ratio of the i-th sample in the j-th dimension after smoothing, α is the smoothing coefficient, T j Represents the trust value of the jth dimension, Y ij Represents the weight value of the i-th sample in the j-th dimension after standardization, m is the total number of samples, and i represents any sample; after calculating the information entropy based on the reconstructed weight ratio, the dimension weight is configured through the information entropy.

[0051] Furthermore, the intelligent allocation module 2 includes:

[0052] The information entropy is calculated by the following formula: E j Characterizes the information entropy of the jth dimension, k is a normalization constant, Configure dimension weights based on information entropy: w j is the weight of the jth dimension, and n is the total number of dimensions.

[0053] Furthermore, the indicator construction module 3 includes:

[0054] The indicator allocation network is initialized with the usage scenario, and the indicator grading layer of the initialized indicator allocation network is called to perform graded fitting on each dimension to establish an indicator grading result, wherein the graded fitting of each dimension includes at least 2-level indicator fitting; the indicator reconstruction layer of the initialized indicator allocation network is called to perform weighted reconstruction of the indicator's own weight and the grading weight based on the indicator grading result, and establish the indicator division result with the weighted reconstruction result.

[0055] Furthermore, scoring module 4 includes:

[0056] The principal component analysis tool is used to extract the principal components of the source data set after data cleaning, and the principal component extraction result is established; the principal component extraction result is fused with homologous features to establish a first fusion result; the first fusion result is fused with multi-source data to establish a second fusion result; and data features are extracted according to the second fusion result.

[0057] Furthermore, the management module 5 includes:

[0058] Configure data access rights for user roles and establish user verification, which includes linkage verification of identity and location. When a user applies for data access, the linkage verification of identity and location is called to identify the user. If the identification result is a pass result, a permission instruction is generated; the data within the permission is called for display according to the permission instruction, and the user is allowed to perform data management within the permission.

[0059] Furthermore, the management module 5 includes:

[0060] Receive user feedback and establish a feedback optimization data set; generate real-time optimization response data based on the feedback optimization data set and the preset intervention index, and optimize the indicator construction module 3 through the real-time optimization response data.

[0061] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0062] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to the selection of multiple elements individually or in whole. Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application intends to include these modifications and variations.

Claims

1. A financial fund performance evaluation data management system, characterized in that: The system comprises: A joint data collection module is used to access data sources, perform data collection, and establish source data sets. The data sources include project management sources, financial sources, questionnaire storage sources, and sensor data sources. An intelligent allocation module is used to obtain the usage scenarios of financial funds and configure dimension weights based on the usage scenarios. The dimensions corresponding to the dimension weights include input dimension, output dimension, result dimension, benefit dimension, and impact dimension. An indicator construction module, which is used to divide the indicators of each dimension respectively through the indicator allocation network based on the usage scenario, establish the indicator division results, and construct the performance evaluation indicators using the indicator division results and dimension weights; The scoring module is used to extract data features of the source data set by using feature engineering after data cleaning of the source data set, and to perform scoring analysis on the data feature extraction results by using performance evaluation indicators to establish scoring analysis results; The management module is used to store source data sets and scoring analysis results, and to conduct financial fund performance evaluation management.

2. A financial fund performance evaluation data management system as claimed in claim 1, characterized in that: The obtaining of usage scenarios of financial funds and configuring dimension weights based on the usage scenarios also includes: The weights of each dimension in the historical data are called according to the usage scenario, where each dimension weight is marked with a trust value; After standardizing the weights of each dimension, the weight ratio is reconstructed based on the trust value identification and standardization results, as follows: in, represents the weight ratio of the i-th sample in the j-th dimension after smoothing, α is the smoothing coefficient, T j Represents the trust value of the jth dimension, Y ij represents the weight value of the i-th sample in the j-th dimension after standardization, m is the total number of samples, and i represents any sample; After calculating the information entropy based on the reconstructed weight ratio, the dimension weight is configured through the information entropy.

3. A financial fund performance evaluation data management system as claimed in claim 2, characterized in that: After calculating the information entropy based on the reconstructed weight proportion, configuring the dimension weights based on the information entropy also includes: The information entropy is calculated by the following formula: E j Characterizes the information entropy of the jth dimension, k is a normalization constant, Configure dimension weights based on information entropy: w j is the weight of the jth dimension, and n is the total number of dimensions.

4. A financial fund performance evaluation data management system as claimed in claim 1, characterized in that: Based on the usage scenario, the indicator allocation network is used to divide the indicators of each dimension respectively, and the indicator division result is established, which also includes: Initialize the indicator allocation network according to the usage scenario, call the indicator grading layer of the initialized indicator allocation network, perform graded fitting on each dimension, and establish an indicator grading result, wherein the graded fitting of each dimension includes at least 2-level indicator fitting; The indicator reconstruction layer of the initialized indicator allocation network is called to perform weighted reconstruction of the indicator's own weight and the grading weight based on the indicator grading result, and the indicator division result is established with the weighted reconstruction result.

5. A financial fund performance evaluation data management system as claimed in claim 1, characterized in that: After the source data set is cleaned, the data features of the source data set are extracted by feature engineering, which further includes: Use the principal component analysis tool to extract the principal components of the source data set after data cleaning and establish the principal component extraction results; Performing homologous feature fusion on the principal component extraction results to establish a first fusion result; Perform multi-source data fusion on the first fusion result to establish a second fusion result; Data features are extracted according to the second fusion result.

6. The financial fund performance evaluation data management system according to claim 1, characterized in that: The system further comprises: The permission management module is used to configure data access rights for user roles and establish user verification, which includes linkage verification of identity and location. When a user applies for data access, the linkage verification of identity and location is called to identify the user. If the identification result is a pass result, a permission instruction is generated; The data management module is used to call the data display within the authority according to the permission instruction and allow the user to perform data management within the authority.

7. A financial fund performance evaluation data management system as claimed in claim 1, characterized in that: The system further comprises: Feedback module, used to receive user feedback and establish feedback optimization data set; The intervention impact module is used to generate real-time optimization response data based on the feedback optimization data set and the preset intervention index, and optimize the indicator construction module through the real-time optimization response data.

8. A method for managing financial fund performance evaluation data, characterized in that: The method is performed by a financial fund performance evaluation data management system according to any one of claims 1 to 7, comprising: Access data sources, perform data collection, and establish source data sets. The data sources include project management sources, financial sources, questionnaire storage sources, and sensor data sources; Obtain the usage scenarios of fiscal funds, and configure dimension weights based on the usage scenarios. The dimensions corresponding to the dimension weights include input dimension, output dimension, result dimension, benefit dimension, and impact dimension; Based on the usage scenario, the indicator division of each dimension is performed through the indicator allocation network, the indicator division result is established, and the performance evaluation indicator is constructed using the indicator division result and the dimension weight; After cleaning the source data set, use feature engineering to extract the data features of the source data set, and use performance evaluation indicators to perform scoring analysis on the data feature extraction results to establish scoring analysis results; Store source data sets and scoring analysis results to conduct financial funds performance evaluation management.

Images