Optimization Method and Platform for Multidimensional Intelligent Evaluation of Employees Based on Artificial Intelligence

Abstract

The present application provides an optimized method and platform for multi-dimensional intelligent evaluation of employees based on artificial intelligence, which relates to the technical field of human resource management. By collecting and analyzing the functional information and relevant data of employees within a preset collection window, a first-level defect constraint is established, and the evaluation data is further analyzed to create a second-level defect constraint; after integrating the first and second defect constraints, a virtual test is executed; a joint index evaluation network is established and the virtual test results and window data sets are read to generate a first evaluation result; the self-cognition evaluation of employees and the first evaluation result are evaluated and certified to establish a second evaluation result, and the first and second evaluation results are integrated to generate a comprehensive evaluation result. The present application solves the technical problem that the existing technology leads to one-sided evaluation results due to the lack of multi-dimensional, dynamic and objective integration capabilities in the evaluation system, and improves the comprehensiveness and accuracy of employee evaluation.

Description

Technical Field

[0001] The present application relates to the technical field of human resource management, and specifically to an artificial intelligence-based employee multi-dimensional intelligent assessment optimization method and platform. Background Art

[0002] In modern enterprise management, employee evaluation is an important part of human resource management, and plays a key role in employee career development, incentive mechanisms, and talent allocation of organizations.

[0003] Existing employee evaluation methods usually rely on supervisor evaluation, performance appraisal, 360-degree feedback and other methods. These methods are mainly based on fixed evaluation standards and indicator systems to conduct regular evaluations of employees' work performance. Although these evaluation methods can reflect employees' daily work performance to a certain extent, due to the single source of evaluation data, they cannot fully cover the various abilities of employees, resulting in overly one-sided evaluation results. At the same time, the objectivity of the evaluation is difficult to guarantee. Simple performance evaluation may be affected by the difficulty of the task, and supervisor evaluation is easily interfered by factors such as personal preferences, which cannot fully reflect the true ability of employees. In addition, in a rapidly changing business environment, employees' abilities and role requirements are also constantly changing, and traditional evaluation methods are difficult to adapt to these changes in real time. They often ignore the individual differences and development potential of employees, resulting in evaluation results that cannot accurately reflect the true abilities and potential of employees. Summary of the invention

[0004] This application provides an artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method and platform, which solves the technical problem that the existing technology lacks the ability to integrate multiple dimensions, dynamism and objectivity in the evaluation system, resulting in one-sided evaluation results and failure to fully reflect the true ability and potential of employees, and achieves the technical effect of improving the comprehensiveness and accuracy of employee evaluation.

[0005] In view of the above problems, on the one hand, the present application provides an employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence, and the method includes: collecting employee data within a preset collection window to establish a window data set; obtaining employee function information of the employee, performing coverage defect analysis based on the employee function information and the window data set, and establishing a first defect constraint; performing evaluation data coverage analysis of the window data set using a general evaluation indicator to establish a second defect constraint; after integrating the first defect constraint and the second defect constraint, establishing a virtual test plan and executing employee testing to establish a virtual test result; obtaining the employee's historical database, establishing a joint indicator evaluation network based on the historical database and employee planning, and generating a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network; reading the employee's self-cognition evaluation, evaluating and authenticating the self-cognition evaluation and the first evaluation result, establishing a second evaluation result, and integrating the first evaluation result and the second evaluation result to generate a comprehensive evaluation result.

[0006] On the other hand, the present application also provides an employee multi-dimensional intelligent evaluation optimization platform based on artificial intelligence, and the platform includes: an employee data collection module, which is used to collect employee data within a preset collection window and establish a window data set; a function coverage analysis module, which is used to obtain employee function information of employees, perform coverage defect analysis based on the employee function information and the window data set, and establish a first defect constraint; an evaluation index coverage analysis module, which is used to perform evaluation data coverage analysis of general evaluation indicators on the window data set to establish a second defect constraint; a virtual testing module, which is used to integrate the first defect constraint and the second defect constraint, establish a virtual testing plan and execute employee testing, and establish a virtual test result; a joint evaluation module, which is used to obtain the employee's historical database, establish a joint indicator evaluation network based on the historical database and employee planning, and generate a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network; a comprehensive evaluation module, which is used to read the employee's self-cognition evaluation, evaluate and authenticate the self-cognition evaluation and the first evaluation result, establish a second evaluation result, and integrate the first evaluation result and the second evaluation result to generate a comprehensive evaluation result.

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

[0008] By collecting employee data within the preset collection window and establishing a window data set, real-time and dynamic employee performance data is provided for subsequent evaluation, ensuring the timeliness and relevance of the evaluation data. Obtain employee functional information, identify possible functional coverage deficiencies or defects by analyzing the differences between employee functions and actual performance, establish the first defect constraint, and provide a basis for subsequent optimization. Perform coverage analysis of general evaluation indicators on the window data set, further identify possible evaluation indicator defects by analyzing the coverage of general evaluation indicators, establish the second defect constraint, and ensure the comprehensiveness and completeness of the evaluation system. After integrating the first defect constraint and the second defect constraint, establish a virtual test plan and perform employee testing, verify the effectiveness of the defect constraint through virtual testing, and provide employees with a simulated evaluation environment to further optimize the evaluation results. Obtain the employee's historical database, establish a joint indicator evaluation network in combination with employee planning, and generate a more comprehensive and objective first evaluation result by integrating historical data and planning goals and combining virtual test results. Read the employees' self-perception evaluation, evaluate and certify the self-perception evaluation and the first evaluation results, and establish the second evaluation results. By introducing the employees' subjective evaluation, the evaluation results can be closer to the employees' actual situation, balance the internal (employee self-perception) and external (evaluation based on data and indicators) perspectives, reduce the one-sidedness of a single evaluation, and finally integrate the comprehensive evaluation results to more comprehensively and objectively reflect the employees' true abilities and potentials, and improve the comprehensiveness and objectivity of the evaluation.

[0009] In summary, this application achieves a comprehensive assessment of employee capabilities by collecting and analyzing data from different dimensions and combining employee self-perception with historical performance, significantly improving the comprehensiveness and accuracy of employee evaluations, and providing more scientific and effective support for employee development and corporate management.

[0010] 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

[0011] Figure 1 A flowchart of an artificial intelligence-based employee multi-dimensional intelligent assessment optimization method provided in an embodiment of the present application.

[0012] Figure 2 A schematic diagram of the process of establishing a joint indicator evaluation network in the artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method provided in an embodiment of the present application.

[0013] Figure 3A schematic diagram of the process of generating a first evaluation result in the artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method provided in an embodiment of the present application.

[0014] Figure 4 A schematic diagram of the structure of an artificial intelligence-based employee multi-dimensional intelligent assessment optimization platform provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: employee data collection module 10 , function coverage analysis module 20 , evaluation index coverage analysis module 30 , virtual testing module 40 , joint evaluation module 50 , comprehensive evaluation module 60 . DETAILED DESCRIPTION

[0016] The embodiments of the present application provide an artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method and platform, which solves the technical problem in the prior art that the evaluation system lacks multi-dimensional, dynamic and objective integration capabilities, resulting in one-sided evaluation results and failure to fully reflect the true abilities and potentials of employees, thereby achieving the technical effect of improving the comprehensiveness and accuracy of employee evaluation.

[0017] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides an employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence, and the method includes:

[0018] Step S1: Collect employee data within a preset collection window and establish a window data set.

[0019] Specifically, the preset collection window is a pre-set time range or data collection range. For example, it can be set to one month or one quarter, and data on employees is collected during this time period, including task completion, work quality, communication records, etc. For example, the internal office system of the enterprise, such as the OA system (office automation system), can be used to collect employee attendance data, task completion status, etc.; or the employee's contribution data in the project can be collected through special project management software. The collected data is sorted and stored in a database or data warehouse to form a window data set.

[0020] By collecting data within the preset collection window, employees' work data can be collected in a timely and comprehensive manner, providing real-time and dynamic data support for subsequent evaluations.

[0021] Step S2: Acquire the employee function information of the employee, perform coverage defect analysis based on the employee function information and the window data set, and establish a first defect constraint.

[0022] Specifically, the employee's employee function information is obtained from the company's job description or the job specification document of the human resources department, that is, the job requirements of the employee's position, including the skills required for the position, work tasks, and work standards to be achieved. For example, for a software engineer position, the function information may include proficiency in a certain programming language, completing the development tasks of the software module on time, etc. Then the employee function information is compared and analyzed with the window data set to find out the gap or uncovered part between the employee's actual data (window data set) and the employee function information. For example, the employee function information includes code review, but there is no corresponding data in the window data set. Through coverage defect analysis, the data that has not been fully collected is found, so as to define the first defect constraint. This first defect constraint is a constraint condition or rule derived from defect analysis, which is used to ensure that there are no omissions or errors in the data, or to point out the incompleteness of the data. For example, the function information of employee A includes code writing, code review, and demand analysis. In the window data set, it is found that employee A has data on code writing and demand analysis, but no data on code review. Therefore, the first defect constraint is established, requiring the addition of evaluation indicators for code review in the evaluation system.

[0023] By covering defect analysis and establishing the first defect constraint, the evaluation system can be optimized to ensure that the evaluation system can fully cover all key functions of employees and improve the comprehensiveness and accuracy of the evaluation.

[0024] Step S3: performing evaluation data coverage analysis of a common evaluation index on the window data set to establish a second defect constraint.

[0025] Specifically, the general evaluation indicators are a set of universal evaluation standards used to measure employees' work performance, ability, contribution, etc. For example, work efficiency, communication skills, task completion quality, etc. Interactively obtain the general evaluation indicators of the company or organization where the employee is located, evaluate and analyze the data in the window data set, and find out where the evaluation indicators are insufficiently covered. For example, some indicators are not reflected or insufficiently covered in the collected data. Based on the results of the evaluation data coverage analysis, a second defect constraint is established to point out the lack of these indicators in the window data set. Exemplarily, the company's general evaluation indicators include work performance, teamwork, and innovation ability. In the window data set, it was found that there was data on work performance and innovation ability, but no data on teamwork. Therefore, a second defect constraint is established to require the addition of evaluation indicators for teamwork to the evaluation system.

[0026] Through the evaluation data coverage analysis and the establishment of the second defect constraint, the evaluation system is further optimized to ensure that the evaluation system can comprehensively cover all important evaluation dimensions, avoid omissions in certain aspects of the evaluation, and thus improve the comprehensiveness and accuracy of the evaluation.

[0027] Step S4: After integrating the first defect constraint and the second defect constraint, a virtual test plan is established and employee testing is performed to establish a virtual test result.

[0028] Specifically, the first defect constraint and the second defect constraint are integrated into the evaluation system to form a comprehensive evaluation framework. Based on the comprehensive evaluation framework, a virtual test plan is designed to evaluate the performance of employees in specific functions and evaluation indicators. Employees are tested according to the virtual test plan, and the virtual test results are recorded. These results can be used as supplementary data for evaluating employee capabilities in order to form a more comprehensive evaluation result. Exemplarily, the first defect constraint requires the addition of evaluation indicators for code review, and the second defect constraint requires the addition of evaluation indicators for teamwork. Based on these constraints, a virtual test plan is established, including code review tests and teamwork tests. Employee A is tested according to the virtual test plan using the company's internal simulation test platform or a dedicated online testing tool, and the test results are recorded. The test results are stored in a MySQL database to form virtual test results.

[0029] By integrating the first defect constraint and the second defect constraint, establishing a virtual test plan and executing employee testing, it is possible to supplement and enrich employee evaluation information and further improve the accuracy and comprehensiveness of the evaluation.

[0030] Step S5: Acquire the employee's historical database, establish a joint indicator evaluation network according to the historical database and employee planning, and generate a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network.

[0031] Specifically, the historical database of employees is obtained from the company's database. The historical database stores relevant data on employees' past work performance, including previous performance data, project participation, etc. At the same time, according to the career plan formulated by the company for employees or the employees' own development plan, including promotion goals, training needs, career paths and other information. Then, a joint indicator evaluation network is constructed using machine learning algorithms or data mining technology. This joint indicator evaluation network is an intelligent algorithm model that uses multiple indicators and data sources to comprehensively evaluate employees. It generates the first evaluation result by reading the data of employees in virtual tests and window data sets and performing comprehensive analysis.

[0032] By establishing a joint indicator evaluation network and combining historical data with employee planning, we can comprehensively consider employees' historical performance and future plans in the evaluation process, so that the evaluation results not only reflect the employees' past performance, but also predict their future career development potential, thereby improving the comprehensiveness and scientific nature of the evaluation.

[0033] Step S6: Read the employee's self-evaluation, evaluate and authenticate the self-evaluation and the first evaluation result, establish a second evaluation result, and integrate the first evaluation result and the second evaluation result to generate a comprehensive evaluation result.

[0034] Specifically, the employee's self-evaluation, that is, the employee's self-evaluation of his or her work performance, is read through the employee self-evaluation questionnaire or online evaluation system. Then the self-evaluation is compared and verified with the first evaluation result, and the consistency or difference between the two is analyzed to establish the second evaluation result. Finally, the first evaluation result and the second evaluation result are integrated, for example, by using the weighted average method to generate a comprehensive evaluation result. By combining the employee's subjective evaluation with the objective evaluation based on data and algorithms, the one-sidedness of a single evaluation is avoided, making the final comprehensive evaluation result more comprehensive, objective and fair.

[0035] Further, such as Figure 2 As shown, in step S5, a historical database of employees is obtained, and a joint indicator evaluation network is established based on the historical database and employee planning, including:

[0036] Step S51: extracting employee characteristics from the historical database, matching parallel employees using the employee characteristics, and configuring calibration evaluation indicators based on the parallel employees.

[0037] Step S52: Analyze the employee plan and establish a personal plan and a target development plan.

[0038] Step S53: reconstructing the calibration evaluation index weights based on the personal plan and the target development plan, and establishing a joint index evaluation network.

[0039] Specifically, various characteristic data of employees are extracted from the historical database, such as work efficiency, problem-solving ability, teamwork ability, communication skills, etc., and other employees with similar characteristics to the target employee are found as parallel employees through similarity matching algorithms (such as cosine similarity and Euclidean distance) using the extracted employee characteristics. Based on the performance of the parallel employee, the evaluation indicators applicable to the target employee are determined, and an initial weight is assigned to each evaluation indicator. The initial weight can be determined based on the performance and importance of the parallel employee. By mining employee characteristics and finding parallel employees to configure evaluation indicators, the evaluation indicators are more in line with the actual situation of the employees, have stronger pertinence, and can more accurately reflect the performance of employees in specific areas related to themselves.

[0040] Analyze the documents or data of employee planning to extract personal planning and target development planning. For data in text form, text analysis methods in natural language processing technology can be used. For example, use lexical analysis tools to decompose text into meaningful words and phrases to identify content related to career development and business goals. For plans in the form of tables or structured data, data processing software (such as Excel) can be used for analysis. Among them, personal planning is the career development plan of individual employees, such as promotion goals, skill improvement, etc.; target development planning is the development goal set by the enterprise for employees based on their own development needs and employee positions, such as required skills, knowledge improvement, and functional expansion. By analyzing the employee's career plan, the employee's career development direction can be accurately identified, providing a personalized basis for the subsequent evaluation system, so that the evaluation results are consistent with the development needs of employees and enterprises.

[0041] Based on the personal plan and target development plan obtained in the previous steps, the weights of the established calibrated evaluation indicators are reconstructed. For example, if the employee's personal plan focuses on developing technical capabilities, and the target development plan is to improve the innovation level of the project, then in the evaluation indicators, the weights of relevant indicators such as technical capability improvement indicators and innovation capability indicators will be increased accordingly. The weight reconstruction process can be achieved by establishing a mathematical model, such as the analytic hierarchy process (AHP). By constructing a hierarchical model, taking the personal plan and target development plan as the criterion layer, and the calibrated evaluation indicators as the solution layer, the relative weights of each evaluation indicator are calculated, and a joint indicator evaluation network is established. Through weight reconstruction, the weights of the evaluation indicators are more reasonable, and the evaluation system can be dynamically adjusted according to the development needs of individual employees and enterprises, ensuring that the evaluation results can better reflect the performance of employees in meeting their own and enterprise development goals, and improving the scientificity and effectiveness of the evaluation.

[0042] Furthermore, the joint index evaluation network is as follows:

[0043] ;in, Characterize employees at time The combined indicator score of The total number of indicator dimensions, Representing any dimension, Characterization The weight of the dimension indicator, Characterization The score of the dimension indicator, Characterization The score of the dimension indicator, Characterizes the time decay factor, Characterization Dimensional indicators and The interaction coefficients of the dimension indicators.

[0044] Specifically, the above joint index evaluation network is mainly composed of two parts. The first part , mainly considering the contribution of each individual dimension after considering the time factor; the second part The influence of the relationship between different dimensions on the overall evaluation of employees is considered. The two parts are added together to obtain the joint indicator score of the employee at time t. In this joint indicator evaluation network, the introduction of the time decay factor enables the evaluation results to be dynamically adjusted over time, which can reflect that recent performance has a greater impact on the current evaluation than long-term performance, and is more in line with the actual situation. The consideration of the interaction coefficient between indicators can reflect the mutual influence relationship between different evaluation dimensions and more accurately measure the comprehensive performance of employees in a complex working environment.

[0045] This joint indicator evaluation network takes into account multiple factors, making employee evaluation more comprehensive, dynamic and reasonable, and better reflecting the comprehensive performance of employees under the influence of different time and multi-dimensional interactions. Further, step S51 includes:

[0046] Step S511: establishing a key ratio, performing feature normalization processing on the employee characteristics, performing principal component screening of the normalized data according to the key ratio, and establishing a main employee characteristic.

[0047] Step S512: after reconstructing the important proportions of the main employee characteristics, perform employee similarity matching of internal employees and establish similarity matching results.

[0048] Step S513: Obtain the matching degree of the similar matching results, use the matching degree as the first filtering condition, obtain the matching number of the similar matching results, use the matching number as the second filtering condition, use the first filtering condition and the second filtering condition to perform internal matching filtering, and establish an internal filtering set, wherein the internal filtering set is configured with an internal trust identifier.

[0049] Step S514: configuring external matching constraints with the internal trust identifier, performing similar matching of external employees based on the main employee characteristics after reconstructing the important proportions, and performing similar screening using the external matching constraints to establish an external screening set.

[0050] Step S515: establishing parallel employees according to the internal filter set and the external filter set.

[0051] Specifically, a key ratio is set in advance, which is the ratio of the number of more important key features in employee characteristics. For example, the key ratio is set to 0.8, which means that the principal components screened out are expected to cover 80% of employee characteristic information. Use maximum-minimum normalization or Z-score normalization to normalize employee characteristics, and convert feature data of different dimensions and ranges into the same standard range. Use principal component analysis (PCA) to reduce the dimension of normalized data, that is, extract the most representative principal features from multiple employee characteristics according to the key ratio and record them as the main employee characteristics. Through feature normalization and principal component analysis, redundant data is effectively removed, thereby reducing the complexity of the data, highlighting the key information that best represents employee characteristics, and improving the efficiency and accuracy of subsequent matching and analysis.

[0052] Since some redundant features have been removed, the weights of the selected main employee features need to be readjusted and allocated. For example, a total of 8 employee features are extracted from the historical database, and the sum of the weights of the 8 features is 1. Through principal component analysis, 5 main employee features are determined, and the weights are reconstructed according to the original weight ratio of the 5 main employee features so that the sum of the weights of the 5 main employee features is 1. After reconstructing the important proportions of the main employee features, use a similarity matching algorithm (such as cosine similarity and Euclidean distance) to calculate the similarity between the target employee and other internal employees, record similar matching employees and their similarities, and form a similar matching result.

[0053] Extract the matching degree (i.e., the similarity between employees in the similar matching results) and the matching quantity (i.e., the number of employees similar to the target employee in the similar matching results). Use the matching degree as the first screening condition, for example, set a minimum matching degree threshold, such as 0.8, which means that only internal employees with a matching degree greater than or equal to 0.8 can be selected. Use the matching quantity as the second screening condition, for example, set the minimum matching quantity to 5, which means that there must be at least 5 internal employees that meet the matching condition. Use these two screening conditions to perform internal matching screening and filter out internal employees that meet the conditions. The set of internal employees after screening is the internal screening set, and configure the internal trust degree mark for the employees in the internal screening set. This internal trust degree mark is used to indicate the credibility or reliability of the internal employees after the internal matching screening. The trust degree mark can be set according to the matching degree, for example, the higher the matching degree, the higher the trust degree mark. By setting the two screening conditions of matching degree and matching quantity, internal employees with high similarity to the target employee and sufficient number can be more accurately screened from the similar matching results. The internal trust degree mark helps to further distinguish the reliability of different employees in the internal screening set, providing a more accurate reference for subsequent operations.

[0054] Configure external matching constraints with internal trust identifiers, such as external matching thresholds or external matching numbers. For example, if the internal trust identifier is "high", the matching degree of external employees is required to be no less than 0.7. Then, similar matching of external employees is performed based on the main employee characteristics after reconstructing the important proportions. For example, similarity matching algorithms (such as cosine similarity and Euclidean distance) are used to calculate the similarity between the target employee and the external employee in each main employee characteristic dimension, and the similarity of each dimension is weighted and summed according to the reconstructed main employee characteristic weights to determine the final external matching degree. These external employee data can be obtained from external talent pools or employee databases of partner companies. Use external matching constraints to perform similarity screening, screen out external employees who meet the conditions, and form an external screening set. The setting of external matching constraints makes external employee similarity matching more targeted, avoids meaningless large-scale search and matching, improves matching efficiency, and the established external screening set provides an external reference for finding parallel employees.

[0055] The employees in the internal screening set and the external screening set are merged to form a parallel employee set. The combination of internal and external employee resources to determine parallel employees can more comprehensively consider similar employees from different sources, provide a richer and more comprehensive reference for configuring calibration evaluation indicators, and make the evaluation indicators more reasonable and scientific.

[0056] Further, such as Figure 3 As shown, in step S5, after reading the virtual test result and the window data set with the joint index evaluation network, a first evaluation result is generated, including:

[0057] Step S54: the virtual test results and the window data set are labeled and then fused to establish a fused data set.

[0058] Step S55: extracting interactive behaviors from the fused data set and establishing an interactive feature set.

[0059] Step S56: Acquire workload data that is temporally associated with the interaction feature set in the fused data set, and establish interaction-workload association.

[0060] Step S57: extracting emotion data from the fused data set, establishing an emotion feature set, and establishing an emotion-workload association based on the emotion feature set.

[0061] Step S58: establishing additional evaluation results through interaction-workload association and emotion-workload association, and adding the additional evaluation results to the first evaluation results.

[0062] Specifically, specific identifiers are added to the virtual test results and the window dataset to distinguish the source or type of the data, and then the two are fused together to form a new dataset, namely, the fused dataset.

[0063] Use data processing tools (such as Python's Pandas library) to clean and preprocess the fused data set, and use natural language processing (NLP) technology or graph theory methods to extract interaction behavior data between employees, such as communication frequency, number of cooperative projects, etc. For example, extract communication frequency by analyzing email content, and extract the number of cooperative projects through the project management system. Organize the extracted interaction behavior data into an interaction feature set and store it in the database. By extracting interaction behavior data, it can reflect the employee's ability in teamwork and communication, providing a more comprehensive perspective for subsequent evaluation.

[0064] In the fused data set, according to the time information of the interaction behavior in the interaction feature set, the corresponding workload data is searched. For example, if the interaction feature set indicates that employees have frequent communication and collaboration (interaction behavior) during a project, then the workload data such as the amount of work tasks and working hours completed by employees during this project are searched. The matched data is sorted and analyzed, and the interaction-workload association is established by matching timestamps or time intervals, so that the workload of employees during the interaction process can be deeply understood, which helps to evaluate the work efficiency and contribution of employees in the process of team collaboration, so that the evaluation results can more comprehensively reflect the actual work performance of employees.

[0065] The fused data set is mined for emotional data, such as employee satisfaction feedback at work, emotional tendencies in emails, etc. For example, the fused data set contains data such as employee questionnaire results and emotional expressions in work diaries, and natural language processing technology is used for emotional analysis. For example, the emotional tendency in the sentence is judged using the emotional dictionary and machine learning algorithm, and the features related to the emotion are extracted to form an emotional feature set. Then, the emotion-workload association is established based on the emotional feature set and workload data. For example, the trend of workload changes under different emotional states is analyzed through statistical analysis methods. By extracting emotional data and establishing the emotion-workload association, the impact of employees' emotional states on workload is considered, and the work ability and efficiency of employees under different emotions can be more comprehensively evaluated.

[0066] According to the interaction-workload correlation and emotion-workload correlation, the additional evaluation results are obtained by analysis and calculation. For example, if it is found that the workload of employees is significantly improved under positive emotions, and the work efficiency is higher in a good interaction process, then the corresponding scores or evaluation levels can be given as additional evaluation results according to these correlations. Then the additional evaluation results are added to the first evaluation results, which can be done by weighted addition or direct merging. By adding additional evaluation results, the first evaluation results are further improved, making the employee evaluation more comprehensive and detailed, and more accurately reflecting the comprehensive performance of employees under the influence of different factors.

[0067] Further, step S6 includes:

[0068] Step S61: calling the dimension evaluation results in the first evaluation results, and performing dimension-by-dimension comparison based on the dimension evaluation results and the self-cognition evaluation to establish dimension comparison results.

[0069] Step S62: input the dimension comparison result into the cognitive bias channel, and use the output cognitive bias as the second evaluation result.

[0070] Specifically, the first evaluation result is the result of evaluating employees based on multiple dimensions, and the dimension evaluation result is the evaluation result of each individual dimension in the first evaluation result (such as work ability, task completion quality, etc.). Call the dimension evaluation result in the first evaluation result, and then compare each dimension evaluation result with the self-perception evaluation on the same dimension. For example, if one dimension in the first evaluation result is work efficiency, its evaluation result is "good", and the employee's self-perception evaluation is "excellent" in the dimension of work efficiency, record the difference between the two as the dimension comparison result. The dimension comparison result reflects the difference between the first evaluation result and the self-perception evaluation in each dimension. By comparing dimensions one by one, the difference between the first evaluation result and the self-perception evaluation in each dimension can be accurately found, providing a detailed data basis for the subsequent analysis of employees' cognitive bias, which helps to gain a deeper understanding of the relationship between employees' cognition of themselves and their actual evaluation.

[0071] The dimension comparison results are input into the cognitive bias channel. The cognitive bias channel is a model or algorithm used to analyze the dimension comparison results and identify the cognitive bias of employees. Exemplarily, this cognitive bias channel can be a rule-based system. For example, if the dimension comparison results show that the employee's self-perception on a certain dimension is higher than the actual evaluation, and the difference exceeds a certain threshold, it is judged as a positive cognitive bias (overestimating oneself); if the self-perception is lower than the actual evaluation and the difference exceeds the threshold, it is judged as a negative cognitive bias (underestimating oneself). The cognitive bias channel can also be a model based on machine learning. By training a large number of dimension comparison result samples, the model learns the relationship between different comparison results and cognitive biases, and then outputs cognitive biases. Finally, the output cognitive bias is used as the second evaluation result.

[0072] By establishing dimension comparison results and inputting them into the cognitive bias channel to generate the second evaluation results, it is possible to identify and adjust employees' self-cognitive biases, generate more objective and accurate evaluation results, and provide a more comprehensive and scientific basis for the final comprehensive evaluation.

[0073] Furthermore, generating a comprehensive evaluation result in the method described in the embodiment of the present application also includes:

[0074] Step S6-1: Establishing team collaboration correlation factors according to the employee function information.

[0075] Step S6-2: Obtain the evaluation data of the employee by the team members, and generate a third evaluation result according to the evaluation data and the team collaboration related factors.

[0076] Step S6-3: Integrate the first evaluation result, the second evaluation result, and the third evaluation result to generate a comprehensive evaluation result.

[0077] Specifically, based on the functional information of employees, their roles and importance in team collaboration are analyzed, and a team collaboration relevance factor is assigned to each employee. The factor can be a numerical value or a weight, reflecting the degree of contribution of the employee in team collaboration. Exemplarily, the functional information of employee A includes code writing, code review, and requirements analysis. By analyzing these functional information, it is determined that code review and requirements analysis are highly correlated with team collaboration. Therefore, a higher team collaboration relevance factor is assigned to employee A, such as 0.8. Establishing a team collaboration relevance factor through employee functional information can quantify the characteristics of employees in team collaboration from the perspective of employee functions, providing an important reference for the subsequent evaluation of team members' evaluation of employees.

[0078] Collect team members' evaluation data on employees through questionnaires, internal team evaluation systems, etc., and then process these evaluation data in combination with team collaboration related factors. For example, if the team collaboration related factors indicate that a certain employee has a greater influence in team collaboration, then when calculating the third evaluation results, the team members' evaluation weight of the employee's team collaboration may be increased accordingly. Mathematical methods such as weighted average method can be used to adjust the weight of the evaluation data according to the team collaboration related factors, and finally generate the third evaluation results. The team member evaluation data after considering the team collaboration related factors can more accurately reflect the actual performance of employees in the team, making the third evaluation results more comprehensive and objective, and can also reflect the impact of employee functions on team collaboration evaluation.

[0079] The first evaluation result, the second evaluation result and the third evaluation result are integrated. A weighted summation method can be used. For example, different weights are set according to the importance that the enterprise attaches to different evaluation results, and then the three evaluation results are added according to the weights to obtain a comprehensive evaluation result. It is also possible to directly merge related evaluation indicators to form a comprehensive evaluation indicator system to generate a comprehensive evaluation result.

[0080] By establishing team collaboration correlation factors, obtaining team member assessment data, and combining the first and second evaluation results, a comprehensive evaluation result is generated, which not only focuses on individual capabilities, but also integrates team collaboration performance and self-perception bias, thereby providing a more comprehensive, objective, and accurate evaluation result. This multi-dimensional comprehensive evaluation method greatly improves the accuracy of employee evaluation, avoids the bias caused by a single evaluation dimension, and provides a more accurate and valuable basis for enterprises to manage employees, motivate, and plan career development.

[0081] In summary, the employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence provided by the embodiment of the present application has the following technical effects:

[0082] The embodiments of the present application ensure the timeliness and comprehensiveness of the evaluation data by collecting data through preset collection windows and obtaining employee function information; identify and optimize potential defects in the evaluation system by covering defect analysis and virtual testing; further improve the accuracy and objectivity of the evaluation by establishing a joint indicator evaluation network and integrating historical data; and make the evaluation results closer to the actual situation of employees through the integration of employee self-perception evaluation, thereby enhancing employees' acceptance and recognition of the evaluation results. In addition, a team collaboration correlation factor is established based on employee function information, and the third evaluation result is generated by combining the evaluation data of team members, which further considers the performance of employees in team collaboration and makes the evaluation results more comprehensive. Overall, the embodiments of the present application significantly improve the comprehensiveness and accuracy of employee evaluations, and provide more scientific and effective support for employee development and enterprise management.

[0083] Embodiment 2, as Figure 4 As shown, based on the same inventive concept as the above-mentioned embodiment 1, the embodiment of the present application provides an employee multi-dimensional intelligent evaluation optimization platform based on artificial intelligence, and the platform includes:

[0084] The employee data collection module 10 is used to collect employee data within a preset collection window and establish a window data set.

[0085] The function coverage analysis module 20 is used to obtain the employee function information of the employee, perform coverage defect analysis according to the employee function information and the window data set, and establish a first defect constraint.

[0086] The evaluation index coverage analysis module 30 is used to perform evaluation data coverage analysis of the general evaluation index on the window data set to establish a second defect constraint.

[0087] The virtual test module 40 is used to integrate the first defect constraint and the second defect constraint, establish a virtual test plan and execute employee testing to establish a virtual test result.

[0088] The joint evaluation module 50 is used to obtain the employee's historical database, establish a joint indicator evaluation network according to the historical database and the employee plan, and generate a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network.

[0089] The comprehensive evaluation module 60 is used to read the employee's self-evaluation, evaluate and authenticate the self-evaluation and the first evaluation result, establish a second evaluation result, and integrate the first evaluation result and the second evaluation result to generate a comprehensive evaluation result.

[0090] Furthermore, the joint evaluation module 50 of the embodiment of the present application is also used to perform the following steps:

[0091] Employee characteristics are extracted from the historical database, and the employee characteristics are used to match parallel employees, and calibration evaluation indicators are configured based on the parallel employees; the employee plan is analyzed to establish a personal plan and a target development plan; the calibration evaluation indicator weights are reconstructed based on the personal plan and the target development plan, and a joint indicator evaluation network is established.

[0092] Furthermore, the joint index evaluation network is as follows:

[0093] ;in, Characterize employees at time The combined indicator score of The total number of indicator dimensions, Representing any dimension, Characterization The weight of the dimension indicator, Characterization The score of the dimension indicator, Characterization The score of the dimension indicator, Characterizes the time decay factor, Characterization Dimensional indicators and The interaction coefficients of the dimension indicators.

[0094] Furthermore, the joint evaluation module 50 of the embodiment of the present application is also used to perform the following steps:

[0095] Establish key proportions, perform feature normalization processing on the employee characteristics, perform principal component screening of normalized data according to the key proportions, and establish main employee characteristics; after reconstructing the important proportions of the main employee characteristics, perform employee similarity matching of internal employees and establish similarity matching results; obtain the matching degree of the similarity matching results, use the matching degree as the first screening condition, obtain the number of matches of the similarity matching results, use the number of matches as the second screening condition, use the first screening condition and the second screening condition to perform internal matching screening, and establish an internal screening set, wherein the internal screening set is configured with an internal trust identifier; configure external matching constraints with the internal trust identifier, perform external employee similarity matching based on the main employee characteristics after reconstructing the important proportions, and use the external matching constraints to perform similarity screening to establish an external screening set; establish parallel employees according to the internal screening set and the external screening set.

[0096] Furthermore, the joint evaluation module 50 of the embodiment of the present application is also used to perform the following steps:

[0097] The virtual test results and the window data set are labeled and then fused to establish a fused data set; interactive behavior is extracted from the fused data set to establish an interactive feature set; workload data that is time-correlated with the interactive feature set in the fused data set is acquired to establish an interactive-workload association; emotion data is extracted from the fused data set to establish an emotional feature set, and an emotional-workload association is established based on the emotional feature set; additional evaluation results are established through the interactive-workload association and the emotional-workload association, and the additional evaluation results are added to the first evaluation results.

[0098] Furthermore, the comprehensive evaluation module 60 of the embodiment of the present application is also used to perform the following steps:

[0099] Call the dimension evaluation result in the first evaluation result, and compare each dimension one by one based on the dimension evaluation result and the self-cognition evaluation to establish a dimension comparison result; input the dimension comparison result into the cognitive bias channel, and use the output cognitive bias as the second evaluation result.

[0100] Furthermore, the comprehensive evaluation module 60 of the embodiment of the present application is also used to perform the following steps:

[0101] Establish a team collaboration correlation factor based on the employee function information; obtain team members' evaluation data on the employee, and generate a third evaluation result based on the evaluation data and the team collaboration correlation factor; integrate the first evaluation result, the second evaluation result, and the third evaluation result to generate a comprehensive evaluation result.

[0102] Through the above detailed description of the employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence, those skilled in the art can clearly understand the employee multi-dimensional intelligent evaluation optimization platform based on artificial intelligence in this embodiment. For the platform disclosed in Example 2, since it corresponds to the method disclosed in Example 1, it has corresponding functional modules and beneficial effects. For relevant matters, please refer to the description of the method part.

[0103] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method, characterized in that: The method comprises: Collect employee data within the preset collection window and establish a window data set; Acquire employee function information of the employee, perform coverage defect analysis according to the employee function information and the window data set, and establish a first defect constraint; Performing evaluation data coverage analysis of a general evaluation index on the window data set to establish a second defect constraint; After integrating the first defect constraint and the second defect constraint, a virtual test plan is established and employee testing is performed to establish a virtual test result; Acquire a historical database of employees, establish a joint indicator evaluation network according to the historical database and employee planning, and generate a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network; Reading the employee's self-perception evaluation, evaluating and authenticating the self-perception evaluation and the first evaluation result, establishing a second evaluation result, and integrating the first evaluation result and the second evaluation result to generate a comprehensive evaluation result; Establishing team collaboration correlation factors based on the employee function information; Acquire evaluation data of the employee by team members, and generate a third evaluation result according to the evaluation data and the team collaboration correlation factor; After integrating the first evaluation result, the second evaluation result, and the third evaluation result, a comprehensive evaluation result is generated.

2. The artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method according to claim 1, characterized in that: The acquiring of the employee's historical database and establishing a joint indicator evaluation network according to the historical database and the employee plan include: Extracting employee characteristics from the historical database, matching parallel employees using the employee characteristics, and configuring calibration evaluation indicators based on the parallel employees; Conduct planning analysis on the employee plans and establish personal plans and target development plans; The calibrated evaluation index weights are reconstructed based on the personal plan and the target development plan, and a joint index evaluation network is established.

3. The employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence as claimed in claim 2 is characterized in that: The joint index evaluation network is as follows: ； in, Characterize employees at time The combined indicator score of The total number of indicator dimensions, Representing any dimension, Characterization The weight of the dimension indicator, Characterization The score of the dimension indicator, Characterization The score of the dimension indicator, Characterizes the time decay factor, Characterization Dimensional indicators and The interaction coefficients of the dimension indicators.

4. The employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence as claimed in claim 2 is characterized in that: The matching of parallel employees by using the employee characteristics and configuring calibration evaluation indicators based on the parallel employees include: Establishing a key ratio, performing feature normalization processing on the employee characteristics, performing principal component screening of the normalized data according to the key ratio, and establishing a main employee characteristic; After reconstructing the important proportions of the main employee characteristics, performing employee similarity matching of internal employees and establishing similarity matching results; Obtaining the matching degree of the similar matching result, taking the matching degree as the first screening condition, obtaining the matching number of the similar matching result, taking the matching number as the second screening condition, performing internal matching screening by using the first screening condition and the second screening condition, and establishing an internal screening set, wherein the internal screening set is configured with an internal trust indicator; Using the internal trust identifier to configure external matching constraints, performing similar matching of external employees based on the main employee characteristics after reconstructing the important proportions, and using the external matching constraints to perform similar screening to establish an external screening set; Parallel employees are established according to the internal filter set and the external filter set.

5. The employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence as claimed in claim 1 is characterized in that: After reading the virtual test result and the window data set by the joint indicator evaluation network, generating a first evaluation result includes: The virtual test result and the window data set are marked and then fused to establish a fused data set; Extracting interactive behaviors from the fused data set to establish an interactive feature set; Acquire workload data that is temporally associated with the interaction feature set in the fused data set, and establish an interaction-workload association; Extracting emotion data from the fused data set, establishing an emotion feature set, and establishing an emotion-workload association based on the emotion feature set; An additional evaluation result is established through interaction-workload association and emotion-workload association, and the additional evaluation result is added to the first evaluation result.

6. The employee multi-dimensional intelligent evaluation optimization method based on artificial intelligence as claimed in claim 1 is characterized in that: The step of evaluating and authenticating the self-perception evaluation and the first evaluation result to establish a second evaluation result includes: Calling the dimension evaluation results in the first evaluation results, and performing dimension comparisons one by one based on the dimension evaluation results and the self-cognition evaluation to establish dimension comparison results; The dimension comparison result is input into the cognitive bias channel, and the output cognitive bias is used as the second evaluation result.

7. An AI-based employee multi-dimensional intelligent evaluation optimization platform, characterized by: The platform is used to execute the artificial intelligence-based employee multi-dimensional intelligent evaluation optimization method according to any one of claims 1 to 6, comprising: The employee data collection module is used to collect employee data within a preset collection window and establish a window data set; A coverage defect analysis module, used for obtaining employee function information of an employee, performing coverage defect analysis according to the employee function information and the window data set, and establishing a first defect constraint; A general evaluation analysis module, used to perform evaluation data coverage analysis of general evaluation indicators on the window data set to establish a second defect constraint; A virtual testing module, used to integrate the first defect constraint and the second defect constraint, establish a virtual testing plan and perform employee testing to establish a virtual testing result; A joint evaluation module, used for acquiring a historical database of employees, establishing a joint indicator evaluation network according to the historical database and employee planning, and generating a first evaluation result after reading the virtual test result and the window data set with the joint indicator evaluation network; A comprehensive evaluation module is used to read the employee's self-evaluation, evaluate and authenticate the self-evaluation and the first evaluation result, establish a second evaluation result, and integrate the first evaluation result and the second evaluation result to generate a comprehensive evaluation result; A team collaboration correlation factor is established based on the employee function information, and evaluation data of the employee by team members is obtained. A third evaluation result is generated based on the evaluation data and the team collaboration correlation factor. The first evaluation result, the second evaluation result, and the third evaluation result are integrated to generate a comprehensive evaluation result.

Images