Operational talent portrait assisted decision-making method and system

Abstract

This application relates to a method and system for assisting decision-making through talent profiling. The method includes the following steps: obtaining the performance records of target employees across multiple evaluation dimensions; calculating the current dimension scores of the target employee in each evaluation dimension during the current evaluation period based on the performance records; calculating the objective weights of each evaluation dimension using the entropy weight method based on the current dimension scores; and calculating the overall completion rate of the target employee based on the objective weights and the current dimension scores, thereby generating talent profile information. Through multi-dimensional hierarchical quantification and objective weighting using the entropy weight method, accurate profiling and scientific evaluation of nuclear power operation personnel are achieved.

Description

Technical Field

[0001] This invention relates to the fields of nuclear power plant operation management and artificial intelligence technology, and more specifically, to a method and system for assisting decision-making by profiling operational personnel. Background Technology

[0002] In the nuclear power industry, operations personnel are the core force ensuring the safe and stable operation of nuclear power units. Their professional competence and management efficiency are directly related to the company's development and energy security. Currently, nuclear power companies' operations positions are known for their high configuration requirements, clear talent development paths, and sound training systems, forming a complete closed-loop training system covering pre-job training, on-the-job improvement, retraining, and skills record management.

[0003] However, under the wave of refined management and digital transformation, traditional management models have gradually exposed deep-seated contradictions: personnel allocation across shift groups, departments, and companies lacks scientific data support, making it difficult to achieve the optimal combination of personnel characteristics, experience, and proficiency, and easily leading to the loss of historical information; 16 subdivided operational positions face high-frequency authorization updates (nearly 100 person-times / year), but there is a lack of a long-term mechanism for tracking personnel's knowledge and skills weaknesses; the shift system leads to fragmented time for operational personnel, seriously restricting team collaboration, learning, and experience sharing; employee information management suffers from the problems of being "scattered, chaotic, and isolated," with massive amounts of data scattered across multiple independent systems such as SAP, UE, and license management, and the data value has not been fully explored, resulting in repetitive work consuming a large amount of human resources; departments and shift groups rely on offline Excel spreadsheets to record key data, which is delayed in updates and difficult to share, making it difficult to meet dynamic management needs.

[0004] The limitations of traditional talent selection, training, utilization, and retention models in terms of evaluation methods, talent inventory, and training characteristics have become a bottleneck restricting the high-quality development of nuclear power enterprises. Against this backdrop, the construction of an intelligent operational talent profiling system is urgently needed. This system, empowered by digital technology, enables precise talent profiling, in-depth diagnosis, and targeted training, providing a new solution for talent management in nuclear power enterprises and helping the industry achieve leapfrog development in its digital transformation. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method and system for assisting decision-making by profiling talent profiles, addressing the problems existing in the prior art.

[0006] The technical solution adopted by this application to solve its technical problem is: constructing and operating a talent profile-assisted decision-making method, including the following steps: Step S1: Obtain the performance records of the target employee across multiple evaluation dimensions; Step S2: Based on the assessment records, calculate the current dimension scores of the target employee for each evaluation dimension in the current assessment period; Step S3: Based on the current dimension score, calculate the objective weight of each evaluation dimension using the entropy weight method, and calculate the overall completion rate of the target employee based on the objective weight and the current dimension score to generate talent profile information.

[0007] Furthermore, the assessment record includes multiple assessment items, each of which includes a raw score. and corresponding weights Step S2 includes: For each evaluation dimension, the raw scores of all assessment items belonging to that evaluation dimension are calculated. With the corresponding weights Perform a weighted summation to obtain the current score of this evaluation dimension in the current assessment period.

[0008] Furthermore, the operational talent profiling-assisted decision-making method also includes: Obtain the historical scores of the target employee for the same evaluation dimension across multiple historical assessment cycles; Based on the historical dimension scores, a regression model is trained to generate a score prediction model; Based on the score prediction model, the predicted dimensional score for the next assessment cycle under this evaluation dimension is predicted.

[0009] Furthermore, the operational talent profiling-assisted decision-making method also includes: The predicted dimension scores are compared and analyzed with the current dimension scores to generate development trend assessment information for the target employee under the evaluation dimensions.

[0010] Further, in step S3, calculating the objective weights of each evaluation dimension using the entropy weight method based on the current dimension score includes: Obtain the current scores of multiple employees in the same position for each evaluation dimension in the current assessment period, and construct a dimension score matrix; For each evaluation dimension column of the dimensional score matrix, a range normalization transformation is performed to map the current dimensional score of all employees under that evaluation dimension to the [0,1] interval, generating a normalized matrix; For each evaluation dimension column of the normalized matrix, calculate the percentage value of each employee under that evaluation dimension and generate a percentage matrix; Based on the proportion matrix, the entropy value of each evaluation dimension is calculated, and the entropy value is used to quantify the dispersion of employee scores under that evaluation dimension. The degree of difference is calculated based on the entropy value of each evaluation dimension, and the degree of difference is equal to 1 minus the entropy value; The objective weight of an evaluation dimension is obtained by dividing the difference of each evaluation dimension by the sum of the differences of all evaluation dimensions.

[0011] Furthermore, the range normalization transformation adopts the following formula: in, Let i be the current score of the i-th employee in the j-th evaluation dimension. , These are the minimum and maximum scores for all employees under this evaluation dimension, respectively. Not equal to ; The formula for calculating the proportion is: in, The total number of employees in the same position. To prevent zero correction items.

[0012] Furthermore, the formula for calculating the entropy value is as follows: Among them, when When = 0, it is agreed that =0, The total number of employees in the same position; The formula for calculating the degree of difference is: Furthermore, when all employees have the same current score for a certain evaluation dimension, the degree of difference in that evaluation dimension is considered. Set it directly to 0, and it will not participate in subsequent weight allocation.

[0013] Furthermore, the formula for calculating the objective weight is as follows: in, The total number of evaluation dimensions; The objective weights of all evaluation dimensions satisfy... ; Furthermore, when the sum of the differences across all evaluation dimensions When =0, all evaluation dimensions are equally weighted, that is... .

[0014] Further, in step S3, calculating the overall completion rate of each employee based on the objective weights and the current dimensional scores includes: Based on the range normalization transformation, the current dimension score of each employee is mapped to the interval [0,1] to obtain the normalized dimension score vector of the employee. The employee's overall completion score is obtained by weighting and summing the employee's normalized dimensional score vector with the aforementioned objective weights. in, Let be the overall completion rate of the i-th employee, with a value ranging from [0,1]. For the employee Normalized scores under each evaluation dimension Let j be the objective weight of the j-th evaluation dimension. The total number of evaluation dimensions; The employee's overall completion rate is obtained by weighting and summing the employee's normalized dimensional score vector with the objective weights.

[0015] This application also provides a talent profiling-assisted decision-making system, including a processor and a memory storing a computer program, wherein the processor, when executing the computer program, implements the steps of the talent profiling-assisted decision-making method described above.

[0016] The method for assisting decision-making by profiling operational personnel according to the present invention has the following beneficial effects: it includes: obtaining the performance records of target employees under multiple evaluation dimensions; calculating the current dimension scores of target employees in each evaluation dimension under the current evaluation period based on the performance records; calculating the objective weights of each evaluation dimension using the entropy weight method based on the current dimension scores; and calculating the overall completion rate of target employees based on the objective weights and the current dimension scores, thereby generating personnel profile information. Through multi-dimensional hierarchical quantification and objective weighting using the entropy weight method, accurate profiling and scientific evaluation of nuclear power operation personnel are achieved. Attached Figure Description

[0017] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart illustrating the logic of a method for using talent profiling to support decision-making. Figure 2 This is a diagram illustrating the backend configuration process; Figure 3 This is a diagram of the front-end application process; Figure 4 This is a flowchart of the data interface call process. Detailed Implementation

[0018] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the invention are now described in detail with reference to the accompanying drawings. In the following description, specific details such as particular structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known devices, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0019] like Figure 1 As shown, Figure 1 This is a flowchart illustrating the logic of a method for using talent profiling to support decision-making.

[0020] The technical solution adopted by this application to solve its technical problem is: constructing and operating a talent profile-assisted decision-making method, including the following steps: Step S1: Obtain the performance records of the target employee across multiple evaluation dimensions; In this step, it should be noted that the assessment records include operation logs, job standard procedures, on-site working condition data, and historical samples of the same job. After data cleaning, data alignment, and data verification, the above data forms a dataset of computable indicators.

[0021] Step S2: Based on the assessment records, calculate the current dimension scores of the target employee for each evaluation dimension in the current assessment period; In this step, it should be noted that the dimensional score is obtained by weighted summation of the scores of secondary indicators belonging to the same evaluation dimension. First, the raw score is calculated according to the quantitative calculation formula of each secondary indicator. Then, the raw score is normalized. Finally, the normalized scores of the secondary indicators belonging to the same evaluation dimension are weighted and summed with their corresponding weights to obtain the current dimensional score of that evaluation dimension.

[0022] Step S3: Based on the current dimension scores, calculate the objective weights of each evaluation dimension using the entropy weight method, and calculate the overall completion rate of the target employee based on the objective weights and the current dimension scores to generate talent profile information.

[0023] In this step, it should be noted that the entropy weight method is an objective weighting method that determines the weight by analyzing the dispersion of employee scores under each evaluation dimension: if the difference in employee scores under a certain dimension is greater, then the entropy value of that dimension is smaller, the difference is greater, and the objective weight is higher; conversely, if the scores tend to be consistent, then the weight is lower.

[0024] In one embodiment, basic employee information, training records, authorization information, and operation logs are synchronized from the SAP system, the UE training system, the license management system, and the operation log system via API interfaces. Data synchronization is performed via scheduled tasks, with a full synchronization executed every morning at midnight, and a real-time interface ensuring that incremental data latency is less than 2 hours. The synchronized data enters the data platform, where data cleaning, alignment, and verification are performed sequentially.

[0025] Data cleaning includes: unifying the date format to "YYYY-MM-DD", mapping enumerated values ​​(such as gender and job level) to the system's standard dictionary, generating temporary numbers for missing employee IDs according to rules and logging them, and truncating out-of-bounds operation durations according to thresholds and marking them as abnormal.

[0026] Data alignment includes: using the employee ID as the global primary identifier, linking the "employee name, department, and position" in the SAP system with the "training record and certificate number" in the training system to the same employee profile master table.

[0027] Data validation includes verifying whether employee IDs are empty, whether departments exist, whether job positions match, and whether training records are complete. If data anomalies are detected, the system automatically triggers an alarm and records it in the validation log for manual review. After the above data governance, a standardized dataset is generated.

[0028] It should be noted that multiple evaluation dimensions include at least one of attitude, learning, skills, improvement, experience, and management. The secondary indicators under a single evaluation dimension include at least one of accuracy, timeliness, stability, and violation penalties. Under a single secondary indicator, there may be one or more measurable operational data items.

[0029] In one embodiment, the original scores of the secondary indicators include: secondary indicators under the skill dimension include: equipment operation proficiency (belonging to the accuracy rate indicator); secondary indicators under the experience dimension include: inspection path accuracy (belonging to the timeliness indicator) and anomaly identification sensitivity (belonging to the accuracy rate indicator); secondary indicators under the safety dimension include: emergency procedure standardization (belonging to the violation penalty indicator).

[0030] The quantitative calculation formula for the accuracy of the inspection path is as follows: in, This is the actual inspection path point sequence. This is a standard inspection path point sequence. This represents the edit distance between the actual path and the standard path. The quantitative calculation formula for the proficiency in operating the equipment is as follows: in, For correct operating procedures, This represents the total number of operation steps. The quantitative calculation formula for the standardization of the emergency response process is as follows: in, The step completion rate, For the order accuracy, η is the time hit rate, η1, η2, η3 are the corresponding weight coefficients and satisfy η1+η2+η3=1, Viol is the number of violations, and λv is the violation penalty coefficient.

[0031] When a single secondary indicator includes multiple measurable operational data items, the secondary indicator value... Normalize: in, For personnel During the assessment period Next The raw scores of each secondary indicator (the raw scores are derived from multiple measurable operational data items using a quantitative calculation formula). and These are the minimum and maximum values ​​of the j-th secondary indicator in the sample of the same job position, respectively. To prevent a zero correction term (taking the smallest positive number to prevent the denominator from being zero), the weights are calculated using a combination of subjective and objective methods, with subjective weighting. Obtained from AHP(xx), objective weights Obtained by the entropy weight method, and then merged into: in, For the first The final weighting of each secondary indicator The fusion coefficient has a range of values. This is used to adjust the ratio of subjective weight to objective weight; This indicates that the sum of the weights of all secondary indicators is 1.

[0032] Dimensional Score: in, For personnel Under the assessment period t, the first Dimensional scores of each primary dimension Indicates the first The second-level indicator belongs to the first One primary dimension, For the first The weights of each secondary indicator, For the first The normalized scores of each secondary indicator.

[0033] in, The comprehensive score (out of 100) of personnel u during the assessment period t. For the first The weights of each first-level dimension (satisfying) =1, For the first The dimensional score of each primary dimension.

[0034] Furthermore, the assessment record includes multiple assessment items, each of which includes a raw score. and corresponding weights Step S2 includes: for each evaluation dimension, obtaining the raw scores of all assessment items belonging to that evaluation dimension. With corresponding weights Perform a weighted summation to obtain the current score of this evaluation dimension in the current assessment period.

[0035] Specifically, assuming the "experience dimension" includes two assessment items: inspection path accuracy and anomaly detection sensitivity, the raw score for inspection path accuracy... =72 points, weight =0.6, the raw score for anomaly detection sensitivity. 66.7, Weight =0.4, then the current dimension score of the experience dimension is... =0.6×72+0.4×66.7=43.2+26.68=69.88 points.

[0036] Furthermore, the talent profiling-based decision-making method also includes: obtaining the historical dimension scores of the target employee for the same evaluation dimension under multiple historical assessment cycles; training a regression model based on the historical dimension scores to generate a score prediction model; and predicting the predicted dimension score for the next assessment cycle under the same evaluation dimension based on the score prediction model.

[0037] In one embodiment, for employee Zhang San's "skill dimension," the historical dimension scores for the past six assessment cycles are obtained: [75.2, 78.5, 80.1, 82.3, 81.7, 84.2]. A linear regression model is used for training to obtain the prediction model. , where t is the assessment cycle number. Based on this model, the predicted dimension score for the 7th assessment cycle is: 1.8×7+74.5=12.6+74.5=87.1 points.

[0038] In another embodiment, a multilayer perceptron-based method for predicting dimensional scores is used. This prediction model employs a multilayer perceptron classifier / regressor from the Weka machine learning library. During the model building phase, for a specific evaluation dimension, the system extracts the employee's historical evaluation records from the main employee evaluation table (assess_exam_user_main) and selects valid historical evaluation sequences. As training data, among which Let the time of the r-th assessment be mapped to a time feature. (That is, the assessment year multiplied by 12 plus the assessment month). This is the total score for this dimension in this assessment (retrieved from the main table entity, corresponding to one of the six dimensions: attitude, learning, skills, improvement, experience, and management). The system will remove records corresponding to invalid assessment numbers from the training set, and will use time characteristics... Total score of dimensions Two-dimensional instances are constructed for training. During training, WekaInstances objects are created, with attributes including the numerical assessment time (assessTime) and dimensional score (score). The class index is set to the column containing the score to use regression mode. The hidden layer structure is set to 3 layers (HiddenLayers=3), and the training iteration time is 500 times (TrainingTime=500). The buildClassifier is called to complete the model training. In the prediction phase, for the next assessment time of the target... The system first calculates the corresponding time features. ,Will( Substitute ,0) into the trained model to obtain the predicted score for that dimension. Furthermore, predicted values ​​should be limited to the historical minimum and maximum scores to avoid anomalous extrapolation. For model evaluation, it is recommended to report at least the mean absolute error on the reserved validation set or rolling backtesting. ; and root mean square error If the business uses interval constraints (such as applying upper and lower bounds relative to the historical mean to the predicted value), the proportion falling within the constraint interval can be added as an availability indicator. Through the above multilayer perceptron prediction method, the system can scientifically predict the capability development of employees in future assessment cycles based on their historical dimensional score trends, providing forward-looking decision support for talent profiling.

[0039] Furthermore, the talent profiling-based decision-making method also includes: comparing and analyzing the predicted dimension scores with the current dimension scores to generate assessment information on the development trend of the target employee under the evaluation dimensions.

[0040] Specifically, if Zhang San's current skill dimension score is 84.2 points and the predicted score for the next period is 87.1 points, then the development trend is assessed as "showing an upward trend, with an expected increase of 2.9 points." If the predicted score is lower than the current score, then the assessment is "there is a risk of decline, and close attention is recommended."

[0041] Further, in step S3, the objective weight of each evaluation dimension is calculated using the entropy weight method based on the current dimension scores, including: obtaining the current dimension scores of multiple employees in the same position for each evaluation dimension under the current assessment period, and constructing a dimension score matrix; performing range normalization transformation on each evaluation dimension column of the dimension score matrix to map the current dimension scores of all employees under that evaluation dimension to the [0,1] interval, generating a normalized matrix; calculating the proportion of each employee under that evaluation dimension for each evaluation dimension column of the normalized matrix, generating a proportion matrix; calculating the entropy value of each evaluation dimension based on the proportion matrix, the entropy value is used to quantify the dispersion of employee scores under that evaluation dimension; calculating the degree of difference based on the entropy value of each evaluation dimension, the degree of difference is equal to 1 minus the entropy value; dividing the degree of difference of each evaluation dimension by the sum of the degree of difference of all evaluation dimensions to obtain the objective weight of that evaluation dimension.

[0042] Specifically, assuming there are 5 employees in a nuclear reactor inspection position, the score matrix for the three dimensions of skills, experience, and safety is as follows: First, range normalization is performed on each dimension column. Taking the empirical dimension column [70, 75, 72, 78, 74] as an example, the minimum value is 70 and the maximum value is 78. After normalization, it becomes [0, 0.625, 0.25, 1.0, 0.5]. Then, the proportion matrix is ​​calculated; finally, the entropy value of each dimension is calculated. Finally, the degree of difference is calculated. The objective weights are then obtained by normalization.

[0043] Furthermore, the range normalization transformation is performed using the following formula: in, Let i be the current score of the i-th employee in the j-th evaluation dimension. , These are the minimum and maximum scores for all employees under this evaluation dimension, respectively. Not equal to ; The formula for calculating the percentage is: in, The total number of employees in the same position. To prevent zero correction items.

[0044] Furthermore, the formula for calculating entropy is: Among them, when When = 0, it is agreed that =0, The total number of employees in the same position; The formula for calculating the degree of difference is: Furthermore, when all employees have the same current score for a certain evaluation dimension, the degree of difference in that evaluation dimension is considered. Set it directly to 0, and it will not participate in subsequent weight allocation.

[0045] Furthermore, the formula for calculating objective weights is: in, To evaluate the total number of dimensions, The difference in the j-th evaluation dimension; The objective weights of all evaluation dimensions satisfy... ; Furthermore, when the sum of the differences across all evaluation dimensions When =0, all evaluation dimensions are equally weighted, that is... .

[0046] Further, in step S3, calculating the overall completion score for each employee based on the objective weights and the current dimensional scores includes: mapping each employee's current dimensional score to the [0,1] interval using range normalization transformation to obtain the employee's normalized dimensional score vector; and weighting the employee's normalized dimensional score vector with the objective weights to obtain the employee's overall completion score. in, Let be the overall completion rate of the i-th employee, with a value ranging from [0,1]. For the employee Normalized scores under each evaluation dimension Let j be the objective weight of the j-th evaluation dimension. To evaluate the total number of dimensions, the employee's normalized dimension score vector is weighted and summed with the objective weights to obtain the employee's overall completion rate.

[0047] This method constructs an intelligent talent management hub, directly addressing the management pain points of scattered, disorganized, and isolated employee information. By building a unified data management architecture, it achieves efficient integration and orderly storage of employee information. At the data presentation level, the system innovatively adopts a visualization solution combining team suggestion prompts, dynamic bar charts, and multi-dimensional radar charts. This provides an intuitive and vivid way to deeply analyze employee data, helping managers quickly grasp key information such as talent structure and performance. Simultaneously, the platform is equipped with flexible dimension management functions. The data types required change as indicators and rules in the evaluation template are added or removed. Data collection employs a dual mode: an automatic mode, which uses configurable scheduling strategies to call external talent information interfaces, updating personnel master data after snapshot alignment and consistency verification, suitable for high-frequency, standardized master data synchronization; and a manual mode, where users, based on the system's provided import template, write assessment details and secondary dimensions into the evaluation record table through table parsing, suitable for low-frequency, manually verified, or external system-unavailable scoring detail entry. These two modes are not interchangeable and can dynamically expand data dimensions according to enterprise development needs, effectively improving the system's sustainability and scalability. Through this series of designs, the system continuously empowers enterprise talent management and strategic decision-making, promoting human resource management towards intelligence and refinement.

[0048] This application focuses on three core technologies: First, an algorithm for constructing an operational assessment index system. This focuses on the algorithm's unique mathematical modeling methods, intelligent weight allocation mechanism, and customized index generation process for nuclear power positions. Patent applications safeguard algorithmic innovation, while software copyright registration covers the algorithm's source code, design documents, and flowcharts, preventing code plagiarism and document tampering, thus ensuring technological exclusivity in the nuclear power employee operational assessment management software market. Second, multi-source operational data collection and precise integration technology. This focuses on core technical aspects such as the adaptation architecture design of data interfaces, dedicated algorithms for data cleaning and alignment, and cross-system data integration processes. Patent applications ensure a technological barrier in the field of nuclear power data acquisition and processing. Third, a visualized operational assessment and team comparative analysis model, encompassing the model's unique visualization algorithm, the mathematical model for team comparative analysis, and the scoring algorithm.

[0049] This invention has the following advantages: First, it precisely addresses industry pain points. It focuses on the deep-seated problems of talent management in the nuclear power industry, such as lack of data support for cross-departmental deployment, insufficient tracking of personnel knowledge and skills, shift work hindering collaborative learning, and fragmented and chaotic employee information management. Directly addressing the shortcomings of traditional management models, it constructs an intelligent talent profiling system, forming a complete solution from data collection and analysis to application, systematically solving multiple problems existing in traditional management models. Second, it improves management efficiency and scientific rigor. Leveraging artificial intelligence and big data analytics, it achieves in-depth mining and precise analysis of talent data, providing a scientific basis for talent management and avoiding the limitations of relying on experience and subjective judgment. Simultaneously, by integrating scattered employee information through digital warehouses and data analysis networks, it breaks down information silos, achieving centralized data management and efficient utilization, avoiding duplication of work, and reducing labor costs. Third, it supports enterprise digital transformation. In the context of digital transformation in the nuclear power industry, this invention achieves intelligent upgrades in talent management through digital empowerment, providing strong support for enterprises in their digital transformation. Furthermore, by constructing a scientifically complete talent competency model and job competency profile, it promotes the synergistic improvement of organizational effectiveness and individual capabilities, stimulating innovation. Fourth, it enables personalized and differentiated development. This invention can formulate personalized talent development plans based on the characteristics of different positions and personnel, meeting the differentiated development needs of different levels and professionals within the enterprise. Simultaneously, it provides strong support for the construction of the enterprise's operational talent pipeline through precise talent profiling and competency models, ensuring the quality and quantity of talent reserves. Fifth, it enhances the enterprise's competitiveness and sustainable development capabilities. Through scientific personnel allocation and precise skills assessment, it achieves the optimal combination of talent, improving the overall operational efficiency of the enterprise. Furthermore, the special nature of the nuclear power industry dictates that the professional competence and management effectiveness of operational personnel are directly related to energy security. This invention, by improving talent management, provides a more reliable guarantee for the safe and stable operation of nuclear power units, thereby laying a solid foundation for the sustainable development of the enterprise.

[0050] This application also provides a talent profile-assisted decision-making system, including a processor and a memory storing a computer program. When the processor executes the computer program, it implements the steps of the talent profile-assisted decision-making method described above.

[0051] like Figure 2 , Figure 3 and Figure 4 As shown, regarding account authentication, when a user logs in, the system identifies the current user's role by querying the database, and matches the corresponding data viewing and operation permissions according to the role, ensuring that managers at different levels can only access talent information within their authorized scope, thus protecting data security.

[0052] Regarding the settings for primary and secondary dimensions: Users can customize various primary and secondary dimensions for data acquisition according to their actual management needs, enabling intuitive management of all quantifiable evaluation indicators. Furthermore, this configuration offers excellent scalability; when an enterprise adds new evaluation dimensions, the system automatically updates the data import template, allowing users to input data based on the latest template, flexibly adapting to business development.

[0053] Regarding personnel information configuration, the system standardizes and unifies scattered and disorganized personnel information, and stores employee information completely in the database. It also supports setting independent access permissions for each personnel, granting different personnel different operating permissions, and realizing refined permission management.

[0054] Users first access the front-end system and initiate a request through the "Get Menu Items" function. The front-end system then sends a "Get Menu Permissions" request to the back-end system's permission system. The permission system returns the corresponding menu permission information based on the user's role, and the front-end system displays the operable menu functions to the user accordingly. When the user "operates on menu functions" on the front-end interface, the front-end system sends a business request to the back-end system via "Access Back-end API." Before executing the business logic, the back-end system calls the "Back-end Authentication Service" to perform "API authentication," that is, to verify whether the current user has permission to perform the operation. The back-end authentication service further relies on the "Menu Permission Management" module in the "Permission System" to complete the final permission verification. The entire process realizes end-to-end permission control from front-end menu display to back-end API access, ensuring that users can only operate system functions within their authorized scope.

[0055] Specifically, the system adopts the Spring Boot 2.7.18 + MySQL 8.0 technology stack, based on a layered architecture design. The system includes: a data acquisition module, which interfaces with over 10 external systems such as SAP, UX, and license management systems via APIs to achieve automatic synchronization of multi-source heterogeneous data; a data governance module, which performs data cleaning, alignment, and validation to generate standardized datasets; an indicator calculation module, which calculates the original scores, normalized scores, and dimensional scores of secondary indicators according to configured quantitative calculation formulas; an entropy weight calculation module, which calculates the entropy value, difference, and objective weight of each dimension based on sample data from the same job position; a comprehensive completion calculation module, which calculates the comprehensive completion degree based on objective weights and normalized dimensional scores; a profile generation module, which generates talent profile information based on the calculation results, including radar charts, weakness diagnoses, and training suggestions; and a prediction module, which predicts dimensional scores for the next assessment cycle based on an MLP neural network model. The system supports concurrent access for thousands of users and has been deployed via Docker containers, possessing high availability and horizontal scalability.

[0056] It is understood that the above embodiments only illustrate preferred embodiments of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can freely combine the above technical features without departing from the concept of the present invention, and can also make several modifications and improvements, all of which fall within the protection scope of the present invention. Therefore, all equivalent transformations and modifications made with respect to the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A method for using talent profiling to assist decision-making, characterized in that, Includes the following steps: Step S1: Obtain the performance records of the target employee across multiple evaluation dimensions; Step S2: Based on the assessment records, calculate the current dimension scores of the target employee for each evaluation dimension in the current assessment period; Step S3: Based on the current dimension score, calculate the objective weight of each evaluation dimension using the entropy weight method, and calculate the overall completion rate of the target employee based on the objective weight and the current dimension score to generate talent profile information.

2. The method for assisting decision-making by operating talent profiling according to claim 1, characterized in that, The assessment record includes multiple assessment items, each of which includes a raw score and a corresponding weight. Step S2 includes: For each evaluation dimension, the original scores of all assessment items belonging to that evaluation dimension are weighted and summed with their corresponding weights to obtain the current dimension score for that evaluation dimension in the current assessment period.

3. The method for assisting decision-making by operating talent profiling according to claim 1, characterized in that, The operational talent profiling-assisted decision-making method also includes: Obtain the historical scores of the target employee for the same evaluation dimension across multiple historical assessment cycles; Based on the historical dimension scores, a regression model is trained to generate a score prediction model; Based on the score prediction model, the predicted dimensional score for the next assessment cycle under this evaluation dimension is predicted.

4. The method for assisting decision-making by operating talent profiling according to claim 3, characterized in that, The operational talent profiling-assisted decision-making method also includes: The predicted dimension scores are compared and analyzed with the current dimension scores to generate development trend assessment information for the target employee under the evaluation dimensions.

5. The method for assisting decision-making by operating talent profiling according to claim 1, characterized in that, In step S3, calculating the objective weights of each evaluation dimension using the entropy weight method based on the current dimension score includes: Obtain the current scores of multiple employees in the same position for each evaluation dimension in the current assessment period, and construct a dimension score matrix; For each evaluation dimension column of the dimensional score matrix, a range normalization transformation is performed to map the current dimensional score of all employees under that evaluation dimension to the [0,1] interval, generating a normalized matrix; For each evaluation dimension column of the normalized matrix, calculate the percentage value of each employee under that evaluation dimension and generate a percentage matrix; Based on the proportion matrix, the entropy value of each evaluation dimension is calculated, and the entropy value is used to quantify the dispersion of employee scores under that evaluation dimension. The degree of difference is calculated based on the entropy value of each evaluation dimension, and the degree of difference is equal to 1 minus the entropy value; The objective weight of an evaluation dimension is obtained by dividing the difference of each evaluation dimension by the sum of the differences of all evaluation dimensions.

6. The method for assisting decision-making by operating talent profiling according to claim 5, characterized in that, The range normalization transformation is performed using the following formula: in, Let i be the current score of the i-th employee in the j-th evaluation dimension. , These are the minimum and maximum scores for all employees under this evaluation dimension, respectively. Not equal to ; The formula for calculating the proportion is: in, The total number of employees in the same position. To prevent zero correction items.

7. The method for assisting decision-making by operating talent profiling according to claim 5, characterized in that, The formula for calculating the entropy value is: Among them, when When = 0, it is agreed that =0, The total number of employees in the same position; The formula for calculating the degree of difference is: Furthermore, when all employees have the same current score for a certain evaluation dimension, the degree of difference in that evaluation dimension is considered. Set it directly to 0, and it will not participate in subsequent weight allocation.

8. The method for assisting decision-making by operating talent profiling according to claim 5, characterized in that, The formula for calculating the objective weight is: in, The total number of evaluation dimensions, The difference in the j-th evaluation dimension; The objective weights of all evaluation dimensions satisfy... ; Furthermore, when the sum of the differences across all evaluation dimensions When =0, all evaluation dimensions are equally weighted, that is... .

9. The method for assisting decision-making by operating talent profiling according to claim 5, characterized in that, In step S3, calculating the overall completion rate of each employee based on the objective weights and the current dimensional scores includes: Based on the range normalization transformation, the current dimension score of each employee is mapped to the interval [0,1] to obtain the normalized dimension score vector of the employee. The employee's overall completion score is obtained by weighting and summing the employee's normalized dimensional score vector with the aforementioned objective weights. in, Let be the overall completion rate of the i-th employee, with a value ranging from [0,1]. For the employee Normalized scores under each evaluation dimension Let j be the objective weight of the j-th evaluation dimension. The total number of evaluation dimensions; The employee's overall completion rate is obtained by weighting and summing the employee's normalized dimensional score vector with the objective weights.

10. A talent profiling-based decision-making support system, comprising a processor and a memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the talent profiling-assisted decision-making method according to any one of claims 1-9.

Images