Performance assessment method and device and readable storage medium

Abstract

The invention provides a performance appraisal method and device and a readable storage medium, and relates to the technical field of performance appraisal, and the method comprises the steps: collecting multi-dimensional performance data of employees; the collected performance data of the multiple dimensions are preprocessed; and inputting the preprocessed multi-dimensional performance data into a pre-trained retrieval enhancement generation (RAG) large model, retrieving related historical cases by using the RAG large model, and generating a performance assessment and evaluation report of the employee in combination with the preprocessed multi-dimensional performance data, the performance appraisal evaluation report comprises basic information of the employees, performance scores and rankings, and next improvement suggestions. According to the method, the device and the readable storage medium, the problems that a traditional performance assessment method is high in evaluation subjectivity, insufficient in data collection and processing, imperfect in feedback mechanism and low in intelligent degree can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of performance appraisal, and in particular to a performance appraisal method, device and readable storage medium. Background Art

[0002] Performance appraisal is an important part of corporate management, which is directly related to employees' salary, promotion and career development. However, the traditional performance appraisal method has the following shortcomings:

[0003] (1) Evaluation is highly subjective: The existing performance appraisal system often relies on the subjective judgment of managers and lacks objective and unified evaluation standards. This leads to large differences in the evaluation of the same employee by different managers, affecting the fairness and accuracy of the evaluation results. Highly subjective evaluations may cause employees to question the evaluation results, which in turn affects their work enthusiasm and teamwork atmosphere.

[0004] (2) Inadequate data collection and processing: The existing performance appraisal system has deficiencies in data collection and processing, and may not be able to fully and accurately reflect the work performance of employees. For example, the system may only rely on limited self-evaluation, peer evaluation or superior evaluation, while ignoring other important data sources. Inadequate data collection and processing may lead to appraisal results that deviate from reality and fail to provide targeted guidance for employees' career development.

[0005] (3) Imperfect feedback mechanism: The existing performance appraisal system is insufficient in providing feedback and improvement suggestions. Employees may not be able to understand their performance in a timely and accurate manner, nor can they obtain targeted improvement suggestions. An imperfect feedback mechanism may result in employees being unable to effectively improve their performance, which in turn affects the performance of the entire team.

[0006] (4) Low intelligence: The existing performance appraisal system has a low level of intelligence and is unable to fully utilize advanced technologies such as big data and artificial intelligence for efficient and accurate appraisals. The low level of intelligence may lead to low appraisal efficiency and the inability to provide employees with valuable feedback and improvement suggestions in a timely manner. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a performance appraisal method, device and readable storage medium in view of the above-mentioned deficiencies in the prior art, so as to solve the problems of strong subjectivity in evaluation, insufficient data collection and processing, imperfect feedback mechanism and low intelligence level in traditional performance appraisal methods.

[0008] In a first aspect, the present invention provides a performance appraisal method, the method comprising:

[0009] Collect multi-dimensional performance data of employees;

[0010] Preprocessing the collected multi-dimensional performance data;

[0011] The preprocessed multi-dimensional performance data is input into the pre-trained retrieval enhancement generation RAG big model, and the RAG big model is used to retrieve relevant historical cases and combine the preprocessed multi-dimensional performance data to generate the employee's performance appraisal evaluation report, wherein the performance appraisal evaluation report includes the employee's basic information, performance score and ranking, and the next improvement suggestions.

[0012] Furthermore, the collection of multi-dimensional performance data of employees specifically includes:

[0013] Collecting multi-dimensional performance data of employees from multiple business systems through an application programming interface (API) interface and / or an extraction, transformation and loading (ETL) interactive method, wherein the multiple business systems include a human resources system, a points system, and a subdivision unit system;

[0014] Collect scanned files or image data of employees' historical paper performance data by uploading them as attachments.

[0015] Furthermore, the multi-dimensional performance data of employees is collected from multiple business systems through the application programming interface API interface and / or the extraction, transformation and loading (ETL) interaction mode, specifically including:

[0016] For each of the multiple business systems, if the business system supports API interface to obtain data, the multi-dimensional performance data of employees is collected through the pre-configured API interface address, request method, return format and request cycle; if the business system supports ETL interaction, the multi-dimensional performance data of employees is collected through the pre-configured data source, parsing rules and execution cycle.

[0017] Furthermore, the multi-dimensional performance data includes at least three of the following: attendance, overtime, night shift, honor, punishment, completion of work tasks, workload, work quality, teamwork ability, and innovation ability;

[0018] The performance appraisal evaluation report also includes at least one of the following: key performance indicators KPI, work goal setting GS, employee quality and ability evaluation, work highlights of this month, and work deficiencies of this month.

[0019] Furthermore, before the pre-processed multi-dimensional performance data is input into the pre-trained retrieval enhancement generation RAG large model, the method further includes:

[0020] Configuring the relevant parameters of the RAG large model, the relevant parameters include the number of overlaps when cutting documents, the maximum number of document recalls, the minimum score of document recalls, and the creativity / randomness of responses;

[0021] The structure of the performance appraisal evaluation report is configured to determine all modules included, the title of each module and the order in which each module is arranged in the performance appraisal evaluation report.

[0022] Furthermore, after using the RAG big model to retrieve relevant historical cases and combining the pre-processed multi-dimensional performance data to generate the employee's performance appraisal report, the method further includes:

[0023] Sending the performance appraisal report to the employee and the management personnel corresponding to the employee;

[0024] A user interface is provided for the employee to view the performance appraisal report, and an overall performance analysis view is provided for the management staff.

[0025] Furthermore, the method further comprises:

[0026] By utilizing the RAG model, the assessment standards and weight distribution in the performance assessment system are dynamically adjusted according to the performance changes of employees and the development needs of the enterprise.

[0027] In a second aspect, the present invention provides a performance evaluation device, the device comprising:

[0028] Performance data collection module, used to collect multi-dimensional performance data of employees;

[0029] A preprocessing module, connected to the performance data collection module, for preprocessing the collected multi-dimensional performance data;

[0030] An evaluation report generation module is connected to the preprocessing module, and is used to input the preprocessed multi-dimensional performance data into a pre-trained retrieval enhancement generation RAG large model, use the RAG large model to retrieve relevant historical cases and combine the preprocessed multi-dimensional performance data to generate the employee's performance appraisal evaluation report, wherein the performance appraisal evaluation report includes the employee's basic information, performance score and ranking, and suggestions for the next step of improvement.

[0031] In a third aspect, the present invention provides a performance appraisal device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the performance appraisal method described in the first aspect.

[0032] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the performance appraisal method described in the first aspect is implemented.

[0033] The performance appraisal method, device and readable storage medium provided by the present invention. First, the multi-dimensional performance data of employees are collected; then the collected multi-dimensional performance data is preprocessed; then the preprocessed multi-dimensional performance data is input into the pre-trained retrieval enhancement generation RAG large model, and the RAG large model is used to retrieve relevant historical cases and generate the performance appraisal evaluation report of the employee in combination with the preprocessed multi-dimensional performance data, wherein the performance appraisal evaluation report includes the basic information of the employee, performance score and ranking, and the next improvement suggestions. The present invention conducts comprehensive evaluation through the RAG large model, reduces the influence of human factors, and improves the objectivity of the evaluation. By automatically generating a performance appraisal evaluation report, specific performance evaluation and personalized improvement suggestions are provided to employees, thereby improving the pertinence and effectiveness of employee performance improvement and promoting the personal growth and career development of employees. At the same time, by collecting multi-dimensional performance data of employees, multiple dimensions and factors are comprehensively considered to ensure the comprehensiveness and accuracy of performance appraisal evaluation. The problems of strong subjectivity of evaluation, insufficient data collection and processing, imperfect feedback mechanism and low intelligence level in traditional performance appraisal methods are solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a flow chart of a performance appraisal method according to Embodiment 1 of the present invention;

[0035] Figure 2 This is a schematic diagram of the structure of a performance evaluation device according to Embodiment 2 of the present invention;

[0036] Figure 3 This is a schematic diagram of the structure of a performance appraisal device according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0038] It should be understood that the specific embodiments and drawings described herein are only used to explain the present invention rather than to limit the present invention.

[0039] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0040] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0041] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0042] It can be understood that the terms "first", "second", etc. in the embodiments of the present invention are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0043] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0044] It is understood that the flowcharts and block diagrams of the present invention illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present invention. Each box in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0045] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0046] Embodiment 1:

[0047] This embodiment provides a performance appraisal method, such as Figure 1 As shown, the method includes:

[0048] Step S101: Collect multi-dimensional performance data of employees.

[0049] In this embodiment, the multi-dimensional performance data includes at least three of the following: attendance, overtime, night shift, honor, punishment, work task completion, workload, work quality, teamwork ability, and innovation ability.

[0050] Optionally, the collection of multi-dimensional performance data of employees specifically includes:

[0051] Collecting multi-dimensional performance data of employees from multiple business systems through an API (Application Programming Interface) interface and / or an ETL (Extract-Transform-Load) interaction method, wherein the multiple business systems include a human resources system, a points system, and a subdivision unit system;

[0052] Collect scanned files or image data of employees' historical paper performance data by uploading them as attachments.

[0053] It should be noted that the performance appraisal method provided in the embodiment of the present invention is applied to a performance appraisal system, which is connected to various business systems within the company to collect multi-dimensional performance data of employees. In order to improve efficiency, the collected multi-dimensional performance data of employees can be sorted and stored first. When conducting performance appraisal, the performance data in the current cycle can be counted in real time. The two are superimposed to serve as the final multi-dimensional performance data of the employees.

[0054] Optionally, the collecting of multi-dimensional performance data of employees from multiple business systems through an application programming interface (API) interface and / or an extraction, transformation and loading (ETL) interactive mode specifically includes:

[0055] For each of the multiple business systems, if the business system supports API interface to obtain data, the multi-dimensional performance data of employees is collected through the pre-configured API interface address, request method, return format and request cycle; if the business system supports ETL interaction, the multi-dimensional performance data of employees is collected through the pre-configured data source, parsing rules and execution cycle.

[0056] In this embodiment, the collection of performance data includes the following three methods:

[0057] (a) API interface: If the external system (i.e., business system) supports API interface to obtain data, the address, request method, return format, and request cycle of the interface are configured in the system. The system will then request the data back according to the configured rules and complete the storage parsing.

[0058] (b) ETL interaction: If the external system supports ETL interaction, we can configure the data source (such as the points system data source, including URL, driver, user name, password and other connection parameters), parsing rules, and execution cycle, and the system will complete the automatic data interaction according to our configuration.

[0059] (c) Historical paper performance data (i.e. attachment upload method): Scan the historical paper performance data into PDF or images, call the system attachment upload function, and complete the storage of historical paper data.

[0060] Through the above three methods, multi-channel data collection can be achieved, laying the foundation for subsequent applications.

[0061] Step S102: pre-processing the collected multi-dimensional performance data.

[0062] In this embodiment, the preprocessing step includes performing data cleaning, removing outliers, reasonably filling in missing data, and converting all data into a unified format for processing by the RAG large model.

[0063] Step S103: input the pre-processed multi-dimensional performance data into a pre-trained RAG (Retrieval-Augmented Generation) large model, use the RAG large model to retrieve relevant historical cases and combine the pre-processed multi-dimensional performance data to generate a performance appraisal report for the employee, wherein the performance appraisal report includes the employee's basic information, performance score and ranking, and suggestions for the next step of improvement.

[0064] In this embodiment, after receiving the pre-processed multi-dimensional performance data, the RAG large model uses a deep learning algorithm to conduct a comprehensive evaluation of the employee's performance. The model generates new evaluation content for each employee by retrieving relevant historical cases and analyzing the correlation and weight between data, and conducts a comprehensive evaluation of the employee's performance to obtain a performance appraisal report. During the evaluation process, the model will comprehensively consider multiple dimensions and factors to ensure the comprehensiveness and objectivity of the evaluation.

[0065] It should be noted that with the rapid development of artificial intelligence technology, especially the continuous progress of natural language processing (NLP) and deep learning technology, the RAG model, as an emerging natural language processing technology, has achieved remarkable results in tasks such as question answering, summarization, and translation. By applying the RAG model to the field of performance appraisal, it can solve the problems existing in traditional performance appraisal methods and improve the accuracy and objectivity of the appraisal.

[0066] Optionally, the performance appraisal report also includes at least one of the following: KPI (Key Performance Indicators), GS (Goal Setting), employee quality and ability evaluation, work highlights of this month, and work deficiencies of this month.

[0067] In this embodiment, in order to make the report content detailed and accurate and able to fully reflect the employee's performance, the performance appraisal evaluation report can include the employee's basic information, KPI, GS work tasks, employee quality and ability evaluation, this month's work highlights, this month's work deficiencies, next improvement suggestions, performance scores and rankings and other information.

[0068] For example, a performance appraisal report might look like this:

[0069] Report date: XX / XX / 2024

[0070] Employee name: Zhang San

[0071] Department: Marketing Department

[0072] Position: Marketing Specialist

[0073] 1. KPI (Key Performance Indicator)

[0074] Activity planning and execution effect: Zhang San planned and executed 3 marketing activities this month, with a total of 5,000 participants, a 20% increase from last month. According to the questionnaire survey after the activity, user satisfaction reached 90%.

[0075] Increased brand awareness: Through Zhang San’s promotional efforts, the brand received 10,000 social media mentions this month, a 30% increase from the previous month, of which 80% were positive mentions.

[0076] 2. GS (Work Goal Setting)

[0077] Event planning goals: Plan and execute at least 2 marketing promotion activities per month, with no less than 1,500 participants in each activity and user satisfaction not less than 85%.

[0078] Brand awareness enhancement goal: Increase brand mentions on social media by at least 20% each month through promotional activities, with positive mentions accounting for at least 75%.

[0079] 3. Evaluation of employee quality and ability

[0080] Creative planning ability: Zhang San has shown strong creative ability in event planning and is able to design novel promotion plans that meet market needs. This month, he proposed the "XX Theme Month" event plan, which successfully attracted a large number of users to participate and increased brand awareness.

[0081] Teamwork ability: Zhang San has good communication and collaboration skills in the team, and can effectively cooperate with colleagues to complete tasks together. In the "XX Theme Month" event, he took the initiative to coordinate resources from all parties to ensure the smooth progress of the event.

[0082] 4. Performance scores and rankings

[0083] Based on the RAG model, a comprehensive analysis of Zhang San's performance in the past assessment cycle was conducted. His total performance score was 85 points, ranking 3 / 20 in the department, which is an upper-middle level.

[0084] 5. Advantages and disadvantages analysis

[0085] This month's work highlights:

[0086] Strong creative planning ability: Zhang San has demonstrated outstanding creative ability in the planning and execution of marketing activities. He has proposed novel and effective promotion plans many times, effectively enhancing brand awareness and user engagement.

[0087] Good teamwork: In team projects, Zhang San can actively communicate and collaborate with colleagues to solve problems together, making important contributions to the achievement of team goals.

[0088] Strong learning ability: Faced with market changes and new technologies, Zhang San shows a strong willingness to learn and a fast learning ability. He is able to quickly master new tools and methods and improve work efficiency.

[0089] Not enough work this month:

[0090] Data analysis capabilities need to be improved: Although Zhang San performed well in planning and execution, he was slightly lacking in data analysis and lacked in-depth analysis and insight into market trend forecasting and effect evaluation.

[0091] Time management needs to be further optimized: Although Zhang San performed well in event planning and execution overall and achieved the goals set for this month, he occasionally did not allocate time in a detailed manner in the initial planning stage of some urgent or complex projects, resulting in additional adjustments during the subsequent execution process. In order to further improve work efficiency and cope with possible emergencies in the future, Zhang San needs to continue to optimize his time management methods to ensure that work can be carried out efficiently and orderly under any circumstances.

[0092] 6. Suggestions for next improvement

[0093] Strengthen data analysis skills training: It is recommended that Zhang San participate in data analysis-related training courses to improve his sensitivity and analytical ability to market data so as to more accurately predict market trends and evaluate promotion effects.

[0094] Optimize time management methods: Recommend Zhang San to learn and practice efficient time management methods, such as the four quadrants rule, the Pomodoro technique, etc., to improve work efficiency and ensure that tasks are completed on time.

[0095] Enhance cross-departmental communication capabilities: Encourage Zhang San to strengthen communication and collaboration with other departments (such as sales department, product department) in order to better understand market demand and product characteristics, and thus plan more targeted promotion plans.

[0096] 7. Summary

[0097] Zhang San has shown strong creative planning ability and good teamwork spirit in the past assessment cycle, but he also has deficiencies in data analysis ability and time management. Through targeted training and improvement, I believe Zhang San will be able to further improve his comprehensive ability and make greater contributions to the development of the company.

[0098] Optionally, before inputting the pre-processed multi-dimensional performance data into a pre-trained retrieval enhancement generation RAG large model, the method further comprises:

[0099] Configuring the relevant parameters of the RAG large model, the relevant parameters include the number of overlaps when cutting documents, the maximum number of document recalls, the minimum score of document recalls, and the creativity / randomness of responses;

[0100] The structure of the performance appraisal evaluation report is configured to determine all modules included, the title of each module and the order in which each module is arranged in the performance appraisal evaluation report.

[0101] In this embodiment, before the RAG big model is used, the relevant parameters of the RAG big model and the structure of the performance appraisal evaluation report can be configured first, so that the RAG big model can retrieve relevant historical cases and generate the employee's performance appraisal evaluation report in combination with the preprocessed multi-dimensional performance data based on the configured relevant parameters and the structure of the performance appraisal evaluation report.

[0102] Overlap number when splitting a document: refers to the number of overlapping words between adjacent paragraphs when splitting a long document into short paragraphs for model processing.

[0103] Maximum document recall: This parameter limits the maximum number of documents that the model returns when retrieving relevant historical cases.

[0104] Minimum document recall score: This parameter sets the minimum threshold that a document score must reach for the model to retrieve relevant historical cases.

[0105] Creativity / Randomness in Responses: This parameter controls the degree of creativity and randomness of the model when generating performance appraisal evaluation reports.

[0106] The structure of the performance appraisal report: that is, all modules contained in the report (for example, it may include a basic information module, a performance score and ranking module, an improvement suggestion module, etc.), the title of each module and the order in which each module is arranged in the performance appraisal report, etc.

[0107] Optionally, after using the RAG large model to retrieve relevant historical cases and combining the pre-processed multi-dimensional performance data to generate the employee's performance appraisal report, the method further includes:

[0108] Sending the performance appraisal report to the employee and the management personnel corresponding to the employee;

[0109] A user interface is provided for the employee to view the performance appraisal report, and an overall performance analysis view is provided for the management staff.

[0110] In this embodiment, after the performance appraisal report is generated, the performance appraisal report can be fed back to employees and management in a timely manner. At the same time, a user interface for viewing the performance appraisal report is provided for the employees, and an overall performance analysis view is provided for the management personnel. The provided user interface supports downloading, printing and online discussion functions to promote feedback and communication.

[0111] In an optional embodiment, the performance appraisal system can first import the account information of all employees within the company, and then collect and pre-process the multi-dimensional performance data of each employee based on the account information of each employee, and then call the question-and-answer interface of the RAG big model to input the pre-processed multi-dimensional performance data of each employee into the RAG big model, and output the performance appraisal evaluation report of each employee through the RAG big model. Then, the performance appraisal system displays it through the user interface and sends the corresponding performance appraisal evaluation report to the employee's corresponding terminal device.

[0112] Optionally, the method further comprises:

[0113] By utilizing the RAG model, the assessment standards and weight distribution in the performance assessment system are dynamically adjusted according to the performance changes of employees and the development needs of the enterprise.

[0114] It should be noted that traditional performance appraisal systems often lack the ability to make dynamic adjustments based on changes in employee performance. Once the appraisal criteria or methods are determined, it is difficult to flexibly adjust them according to actual conditions. The lack of a dynamic adjustment mechanism may lead to a large deviation between the appraisal results and the actual performance of employees, and fail to reflect the growth and progress of employees in a timely manner. The present invention utilizes the RAG large model to intelligently and dynamically adjust key parameters such as appraisal criteria and weights according to changes in employee performance and the development needs of the enterprise, thereby achieving the flexibility and adaptability of the performance appraisal system and ensuring that the appraisal criteria are always highly consistent with the actual performance of employees and the development needs of the enterprise.

[0115] In a specific embodiment, the performance appraisal method is applied to a performance appraisal system, which is mainly composed of the following modules:

[0116] (1) Data collection module: responsible for collecting employee performance data, including but not limited to attendance, overtime, night shift (if applicable), historical evaluation information, score ranking information, honors, penalties, work task completion, workload, work quality, teamwork ability, innovation ability and other dimensions. This data can be obtained through the company's internal human resources system, score system, small unit system, ERP (Enterprise Resource Planning), CRM (Customer Relationship Management) and other business systems, or collected through employee self-evaluation and peer evaluation to ensure the comprehensiveness and accuracy of the performance appraisal report.

[0117] (2) Preprocessing module: cleans and standardizes the collected data to ensure data quality and facilitate subsequent analysis.

[0118] (3) RAG big model analysis module: This module is the core of the system. It uses the pre-trained RAG big model to conduct in-depth analysis of pre-processed performance data. The RAG big model integrates machine learning algorithms and natural language processing technology, and can comprehensively consider performance data from multiple dimensions to generate performance appraisal reports.

[0119] Among them, the RAG big model can comprehensively evaluate employee performance based on the collected data by retrieving relevant historical cases (including employees' own historical performance appraisal cases) and generating new evaluation content. At the same time, the model can also generate personalized improvement suggestions based on the evaluation results.

[0120] (4) Evaluation report generation module: Based on the analysis results of the RAG model, the module automatically generates employee performance appraisal reports, including basic employee information, KPIs (Key Performance Indicators), GS (Goal Setting) work tasks, employee quality and ability evaluation, highlights of this month's work, deficiencies in this month's work, suggestions for next improvement, performance scores and rankings, and other information.

[0121] Among them, the next improvement suggestions can be based on the preset improvement strategy library combined with the individual characteristics of employees, historical performance trends and industry best practices, and put forward targeted improvement suggestions from the aspects of skill improvement, project participation, career planning, etc. The strategy library can include the weights of the three contents of KPI, GS work tasks, and employee quality and ability evaluation, as well as additional plus and minus points (for example, extra points for special contributions and minus points for major mistakes).

[0122] (5) Feedback module: Provide timely feedback of performance appraisal reports to employees and management so that they can understand the performance status in a timely manner and take corresponding improvement measures.

[0123] (6) User interface module: provides employees with an interface for viewing personal performance appraisal reports, and provides managers with an overall performance analysis view to facilitate the formulation of team management strategies.

[0124] (7) Dynamic adjustment and optimization module: The weight distribution of different indicators has a great impact on the final result. The RAG model can dynamically adjust the assessment criteria, weights and other parameters according to the performance changes of employees and the development needs of the enterprise. At the same time, this module can also continuously optimize the RAG model to improve the accuracy and reliability of the evaluation.

[0125] Based on the above performance appraisal system, the corresponding performance appraisal methods include:

[0126] Step 1: The data collection module regularly collects performance data from the company's internal business systems (such as human resources 2.0, points system, small unit system, etc.) and employee self-evaluation, mutual evaluation and other channels through API interface, ETL interaction or offline manual sorting.

[0127] Step 2: To ensure the accuracy and consistency of the data, the preprocessing module performs data cleaning, removes outliers, reasonably fills in missing data, and converts all data into a unified format for processing by the RAG large model.

[0128] Step 3: The RAG large model analysis module receives the pre-processed data and uses deep learning algorithms to comprehensively evaluate the employee's performance. The model generates new evaluation content for each employee by retrieving relevant historical cases and analyzing the correlation and weight between data, and comprehensively evaluates the employee's performance. During the evaluation process, the model will comprehensively consider multiple dimensions and factors to ensure the comprehensiveness and objectivity of the evaluation.

[0129] Step 4: The evaluation report generation module automatically generates a performance appraisal report based on the output results of the RAG large model, which includes the employee's basic information, KPI, GS work tasks, employee quality and ability evaluation, this month's work highlights, this month's work deficiencies, next improvement suggestions, performance scores and rankings, etc., so that the report content is detailed and accurate and can fully reflect the employee's performance.

[0130] Step 5: The feedback module will provide timely feedback of performance appraisal evaluation reports to employees and management.

[0131] Step 6: The user interface module displays the performance appraisal report to employees and managers, supports downloading, printing and online discussion functions, and promotes feedback and communication.

[0132] Step 7: The dynamic adjustment and optimization module dynamically adjusts the assessment criteria, weights and other parameters based on the feedback results and actual needs, and continuously optimizes the RAG large model.

[0133] The performance appraisal method has the following specific features:

[0134] a) Customization of RAG big model: Customized training of RAG big model is carried out according to the specific needs of performance appraisal to optimize its capabilities in multi-dimensional data analysis and comprehensive evaluation generation.

[0135] b) Personalized improvement suggestions: Based on employees’ personal performance data and combined with organizational development strategies, highly personalized improvement suggestions are generated to enhance employee satisfaction and career development motivation.

[0136] c) Dynamic adjustment mechanism: The system can continuously optimize the parameters of the RAG model based on feedback and new performance data to ensure the accuracy and timeliness of the evaluation system.

[0137] Through the performance appraisal method provided by the embodiment of the present invention, enterprises can achieve a comprehensive and objective evaluation of employee performance, effectively improve performance management efficiency, and promote employee personal growth and the improvement of the overall team efficiency. At the same time, the flexibility and scalability of the system also provide the possibility of introducing more performance indicators and optimizing the evaluation model in the future.

[0138] In a specific example, a company wants to develop a performance appraisal system to solve the problems of traditional performance appraisal methods, such as strong subjectivity, insufficient data collection and processing, imperfect feedback mechanism, and low intelligence. This leads to inaccurate appraisal results and makes it difficult to truly reflect the actual performance level of employees. Therefore, the development of a new system aims to improve the overall level of human resource management in the company, promote the personal growth and career development of employees, and thus create greater value for the company. The following are the specific implementation steps:

[0139] 1. Log in to the performance management system and import the account information of all employees in the company.

[0140] 2. Configure data collection rules:

[0141] (a) API interface: If the external system (i.e., business system) supports API interface to obtain data, the address, request method, return format, and request cycle of the interface are configured in the system. The system will then request the data back according to the configured rules and complete the storage parsing.

[0142] (b) ETL interaction: If the external system supports ETL interaction, we can configure the data source (such as the points system data source, including URL, driver, user name, password and other connection parameters), parsing rules, and execution cycle, and the system will complete the automatic data interaction according to our configuration.

[0143] (c) Historical paper performance data (i.e. attachment upload method): Scan the historical paper performance data into PDF or images, call the system attachment upload function, and complete the storage of historical paper data.

[0144] Through the above three methods, multi-channel data collection can be achieved, laying the foundation for subsequent applications.

[0145] 3. Configure the RAG large model:

[0146] (a) Select the trained RAG large model.

[0147] (b) De-duplication and clean the historical data that has been stored in the database and then perform vector processing.

[0148] (c) Configure the amount of overlap when cutting documents, the maximum number of document recalls, the minimum document recall score, and the creativity / randomness of responses.

[0149] (d) Configure the structure of the performance appraisal report, i.e. the title and arrangement order of each module.

[0150] 4. Feed the processed data to the RAG large model.

[0151] 5. Call the RAG question-and-answer interface to output the performance appraisal report.

[0152] 6. Push the performance appraisal report to the person being evaluated and the leader, and query it through the web page.

[0153] 7. The person being evaluated fills out an improvement plan and feeds it into the RAG model as an input factor for the next evaluation cycle, thus realizing a rolling dynamic evaluation feedback mechanism.

[0154] In another specific embodiment, the performance evaluation method may include the following steps:

[0155] S1. Regularly collect performance data from the company's internal business systems (such as HR 2.0, points system, small unit system, etc.) and employee self-evaluation and mutual evaluation channels through ETL tools, API interfaces or offline manual sorting.

[0156] S2. Perform data cleaning on the collected performance data, remove outliers, fill in missing data reasonably, and convert all data into a unified format for processing by the RAG large model.

[0157] S3. Based on the pre-processed performance data, a deep learning algorithm is used to comprehensively evaluate the performance of employees. The model generates a comprehensive performance score for each employee by analyzing the correlation and weight between data.

[0158] S4. Based on the output results of the RAG big model, a performance appraisal evaluation report is automatically generated, including the employee's basic information, KPI, GS work tasks, employee quality and ability evaluation, this month's work highlights, this month's work deficiencies, next improvement suggestions, rankings and other information. Among them, the next improvement suggestions are targeted improvement suggestions from the aspects of skill improvement, project participation, career planning, etc., based on the employee's personal characteristics, historical performance trends and industry best practices.

[0159] S5. Display performance appraisal reports to employees and managers, support downloading, printing and online discussion functions, and promote feedback and communication.

[0160] It should be noted that the performance appraisal method provided by the present invention realizes accurate appraisal and comprehensive evaluation of employee performance by deeply integrating the advanced Retrieval-Augmented Generation (RAG) large model technology, and puts forward targeted improvement suggestions, thereby comprehensively improving the human resource management efficiency of the enterprise and the personal development motivation of employees.

[0161] Specifically, the present invention has the following beneficial effects:

[0162] a) Improve the accuracy of performance appraisal: Through the retrieval function of the RAG large model, the system can automatically locate historical data related to current employee performance from massive data, and combine it with real-time performance data to generate a more objective and comprehensive performance analysis report. This not only reduces the subjectivity of human evaluation, but also ensures the accuracy and fairness of the appraisal results.

[0163] b) Enhanced data-driven decision support: This invention makes full use of the data processing and analysis capabilities of the RAG large model to explore the potential laws and trends in performance data and provide scientific decision-making basis for enterprise management. Through data-driven decision-making, enterprises can more effectively formulate and adjust human resource policies and optimize resource allocation.

[0164] c) Provide personalized improvement suggestions: Based on the generation function of the RAG large model, the system can automatically generate personalized improvement suggestions based on the performance characteristics and needs of each employee. These suggestions not only put forward specific improvement measures for employees' shortcomings, but also encourage employees to give full play to their strengths and maximize their personal career development.

[0165] d) Improve the efficiency of human resource management: Through the automated and intelligent performance appraisal process, the present invention significantly reduces the workload of the human resource management department and improves work efficiency. At the same time, the real-time feedback mechanism of the system helps enterprises to promptly discover and solve problems in human resource management, ensuring the sustainable and healthy development of the enterprise.

[0166] The performance appraisal method provided by the embodiment of the present invention first collects multi-dimensional performance data of employees; then pre-processes the collected multi-dimensional performance data; then inputs the pre-processed multi-dimensional performance data into the pre-trained retrieval enhancement generation RAG large model, uses the RAG large model to retrieve relevant historical cases and combines the pre-processed multi-dimensional performance data to generate the performance appraisal evaluation report of the employee, wherein the performance appraisal evaluation report includes the basic information of the employee, performance score and ranking, and next improvement suggestions. The present invention uses the RAG large model for comprehensive evaluation, reduces the influence of human factors, and improves the objectivity of the evaluation. By automatically generating a performance appraisal evaluation report, specific performance evaluation and personalized improvement suggestions are provided to employees, thereby improving the pertinence and effectiveness of employee performance improvement and promoting the personal growth and career development of employees. At the same time, by collecting multi-dimensional performance data of employees, multiple dimensions and factors are comprehensively considered to ensure the comprehensiveness and accuracy of performance appraisal evaluation. The problems of strong subjectivity of evaluation, insufficient data collection and processing, imperfect feedback mechanism and low intelligence level in traditional performance appraisal methods are solved.

[0167] Embodiment 2:

[0168] like Figure 2As shown, this embodiment provides a performance evaluation device for executing the above performance evaluation method, including:

[0169] Performance data collection module 11, used to collect multi-dimensional performance data of employees;

[0170] A preprocessing module 12, connected to the performance data collection module 11, for preprocessing the collected multi-dimensional performance data;

[0171] The evaluation report generation module 13 is connected to the preprocessing module 12, and is used to input the preprocessed multi-dimensional performance data into the pre-trained retrieval enhancement generation RAG large model, use the RAG large model to retrieve relevant historical cases and combine the preprocessed multi-dimensional performance data to generate the employee's performance appraisal evaluation report, wherein the performance appraisal evaluation report includes the employee's basic information, performance score and ranking, and the next improvement suggestions.

[0172] Optionally, the performance data collection module 11 includes:

[0173] A first collection unit is used to collect multi-dimensional performance data of employees from multiple business systems through an application programming interface (API) interface and / or an extraction, transformation and loading (ETL) interaction method, wherein the multiple business systems include a human resources system, a points system, and a subdivision unit system;

[0174] The second collection unit is used to collect scanned files or image data of employees' historical paper performance data by uploading them as attachments.

[0175] Optionally, the first collecting unit is specifically configured to include:

[0176] For each of the multiple business systems, if the business system supports API interface to obtain data, the multi-dimensional performance data of employees is collected through the pre-configured API interface address, request method, return format and request cycle; if the business system supports ETL interaction, the multi-dimensional performance data of employees is collected through the pre-configured data source, parsing rules and execution cycle.

[0177] Optionally, the multi-dimensional performance data includes at least three of the following: attendance, overtime, night shift, honor, punishment, completion of work tasks, workload, work quality, teamwork ability, and innovation ability;

[0178] The performance appraisal evaluation report also includes at least one of the following: key performance indicators KPI, work goal setting GS, employee quality and ability evaluation, work highlights of this month, and work deficiencies of this month.

[0179] Optionally, the device further comprises:

[0180] A model parameter configuration module, used to configure the relevant parameters of the RAG large model, including the number of overlaps when cutting documents, the maximum number of document recalls, the minimum score of document recalls, and the creativity / randomness when responding;

[0181] The report structure configuration module is used to configure the structure of the performance appraisal evaluation report, determine all modules included, the title of each module and the arrangement order of each module in the performance appraisal evaluation report.

[0182] Optionally, the device further comprises:

[0183] An evaluation report sending module, used to send the performance appraisal evaluation report to the employee and the management personnel corresponding to the employee;

[0184] The interface providing module is used to provide the employee with a user interface for viewing the performance appraisal evaluation report and to provide the management personnel with an overall performance analysis view.

[0185] Optionally, the device further comprises:

[0186] The dynamic adjustment module is used to utilize the RAG model to dynamically adjust the assessment standards and weight distribution in the performance assessment system according to the performance changes of employees and the development needs of the enterprise.

[0187] Embodiment 3:

[0188] refer to Figure 3 This embodiment provides a performance appraisal device, including a memory 21 and a processor 22. The memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the performance appraisal method in Example 1.

[0189] The memory 21 is connected to the processor 22. The memory 21 may be a flash memory, a read-only memory or other memory. The processor 22 may be a central processing unit or a single-chip microcomputer.

[0190] Embodiment 4:

[0191] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the performance appraisal method in the above-mentioned embodiment 1 is implemented.

[0192] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0193] In summary, the performance appraisal method, device and readable storage medium provided by the embodiment of the present invention first collect the multi-dimensional performance data of employees; then pre-process the collected multi-dimensional performance data; then input the pre-processed multi-dimensional performance data into the pre-trained retrieval enhancement generation RAG large model, use the RAG large model to retrieve relevant historical cases and combine the pre-processed multi-dimensional performance data to generate the performance appraisal evaluation report of the employee, wherein the performance appraisal evaluation report includes the basic information of the employee, performance score and ranking, and the next improvement suggestions. The present invention conducts comprehensive evaluation through the RAG large model, reduces the influence of human factors, and improves the objectivity of the evaluation. By automatically generating a performance appraisal evaluation report, specific performance evaluation and personalized improvement suggestions are provided to employees, thereby improving the pertinence and effectiveness of employee performance improvement and promoting the personal growth and career development of employees. At the same time, by collecting multi-dimensional performance data of employees, multiple dimensions and factors are comprehensively considered to ensure the comprehensiveness and accuracy of performance appraisal evaluation. The problems of strong subjectivity of evaluation, insufficient data collection and processing, imperfect feedback mechanism and low intelligence level in traditional performance appraisal methods are solved.

[0194] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A performance appraisal method, characterized in that: The method comprises: Collect multi-dimensional performance data of employees; Preprocessing the collected multi-dimensional performance data; The preprocessed multi-dimensional performance data is input into the pre-trained retrieval enhancement generation RAG big model, and the RAG big model is used to retrieve relevant historical cases and combine the preprocessed multi-dimensional performance data to generate the employee's performance appraisal evaluation report, wherein the performance appraisal evaluation report includes the employee's basic information, performance score and ranking, and the next improvement suggestions.

2. The method according to claim 1, characterized in that: The collection of multi-dimensional performance data of employees specifically includes: Collecting multi-dimensional performance data of employees from multiple business systems through an application programming interface (API) interface and / or an extraction, transformation and loading (ETL) interactive method, wherein the multiple business systems include a human resources system, a points system, and a subdivision unit system; Collect scanned files or image data of employees' historical paper performance data by uploading them as attachments.

3. The method according to claim 2, characterized in that The multi-dimensional performance data of employees is collected from multiple business systems through the application programming interface API interface and / or extraction, transformation and loading (ETL) interaction mode, specifically including: For each of the multiple business systems, if the business system supports API interface to obtain data, the multi-dimensional performance data of employees is collected through the pre-configured API interface address, request method, return format and request cycle; if the business system supports ETL interaction, the multi-dimensional performance data of employees is collected through the pre-configured data source, parsing rules and execution cycle.

4. The method according to claim 1, characterized in that The multi-dimensional performance data includes at least three of the following: attendance, overtime, night shift, honor, punishment, completion of work tasks, workload, work quality, teamwork ability, and innovation ability; The performance appraisal evaluation report also includes at least one of the following: key performance indicators KPI, work goal setting GS, employee quality and ability evaluation, work highlights of this month, and work deficiencies of this month.

5. The method according to claim 1, characterized in that Before inputting the pre-processed multi-dimensional performance data into the pre-trained retrieval enhancement generation RAG large model, the method further includes: Configuring the relevant parameters of the RAG large model, including the number of overlaps when cutting documents, the maximum number of document recalls, the minimum score of document recalls, and the creativity / randomness of responses; The structure of the performance appraisal evaluation report is configured to determine all modules included, the title of each module and the order in which each module is arranged in the performance appraisal evaluation report.

6. The method according to claim 1, characterized in that After using the RAG big model to retrieve relevant historical cases and combining the pre-processed multi-dimensional performance data to generate the employee's performance appraisal report, the method further includes: Sending the performance appraisal report to the employee and the management personnel corresponding to the employee; A user interface is provided for the employee to view the performance appraisal evaluation report, and an overall performance analysis view is provided for the management staff.

7. The method according to claim 1, characterized in that The method further comprises: By utilizing the RAG model, the assessment standards and weight distribution in the performance assessment system are dynamically adjusted according to the performance changes of employees and the development needs of the enterprise.

8. A performance appraisal device, characterized in that: The device comprises: Performance data collection module, used to collect multi-dimensional performance data of employees; A preprocessing module, connected to the performance data collection module, for preprocessing the collected multi-dimensional performance data; An evaluation report generation module is connected to the preprocessing module, and is used to input the preprocessed multi-dimensional performance data into a pre-trained retrieval enhancement generation RAG large model, use the RAG large model to retrieve relevant historical cases and combine the preprocessed multi-dimensional performance data to generate the employee's performance appraisal evaluation report, wherein the performance appraisal evaluation report includes the employee's basic information, performance score and ranking, and suggestions for the next step of improvement.

9. A performance appraisal device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the performance appraisal method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the performance appraisal method according to any one of claims 1 to 7 is implemented.

Images