Performance evaluation optimization system and method based on AI

Abstract

The invention discloses an AI-based performance evaluation optimization system and method, relates to the technical field of human resource management, and solves the problems that the actual performance of an employee cannot meet a target requirement, the employee performance is determined to be poor and the performance of the employee cannot meet the target requirement due to neglect of factors influencing employee performance change in a certain time range in the prior art. The accuracy of the performance evaluation method is low; and the sense of belonging of the employees to the company is influenced. An employee performance portrait is generated according to work data, and an employee theoretical performance index is calculated based on the portrait; a state influence coefficient is generated according to the physical and mental data, an employee performance index is generated according to the state influence coefficient, an employee theoretical performance index of the employee in the current stage is calculated through historical multi-source data of the employee, and the employee theoretical performance index is adjusted according to the physical and mental data of the employee in the time period to obtain the employee performance index. The accuracy of employee performance evaluation is improved, humanistic care of a company is promoted, and the satisfaction degree of employees is improved.

Description

Technical Field

[0001] The present application belongs to the field of human resource management technology, and specifically is a performance evaluation optimization system and method based on AI. Background Art

[0002] Performance evaluation, also known as performance appraisal, performance evaluation, and employee assessment, is a formal employee evaluation system and an important basic work in human resource development and management. It aims to evaluate and measure employees' work behavior and work results in their positions through scientific methods and principles. The results of performance evaluation can directly affect the vital interests of many employees, such as salary adjustment, bonus distribution, and job promotion. At the same time, it can also provide multi-faceted support for the company.

[0003] Traditional performance evaluation systems often rely on manual questionnaires, subjective evaluations by superiors, or simple KPI assessments. These methods are highly subjective, have incomplete data collection, long evaluation cycles, and lagging feedback mechanisms. They are difficult to accurately reflect employees' true work performance and potential, and affect employee incentives and the overall effectiveness of the organization. With the development of automation, existing technologies use automation to conduct performance evaluations, which improves evaluation efficiency. However, when evaluating performance, factors that affect changes in employee performance within a certain time frame are ignored, resulting in employees' actual performance failing to meet target requirements and identifying employees as having poor performance. This results in low accuracy in performance evaluation methods and affects employees' sense of belonging to the company. Therefore, the performance evaluation system still needs further improvement. Summary of the invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes an AI-based performance evaluation optimization system and method to solve the technical problem that the prior art ignores the factors that affect the changes in employee performance within a certain time range, resulting in the actual performance of employees failing to meet the target requirements, determining that the employee's performance is poor, resulting in low accuracy of the performance evaluation method and affecting the employee's sense of belonging to the company.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides an AI-based performance evaluation and optimization system, comprising: a data collection module, a data analysis module, an early warning module and a database;

[0006] The data acquisition module is used to acquire employee data through data acquisition equipment; the employee data includes employee ID, physical and mental data, and work data;

[0007] The data analysis module: generates an employee performance portrait based on the work data corresponding to the employee ID; calculates the employee theoretical performance index based on the employee performance portrait; generates a state influence coefficient based on the physical and mental data corresponding to the employee ID; generates an employee performance index based on the state influence coefficient and the employee theoretical performance index; generates an alarm signal based on the employee performance index and the state influence coefficient;

[0008] The early warning module: makes prompts according to the alarm signal and contacts the management personnel.

[0009] Through the above steps, this application uses the employees' past diversified data sources to calculate their current theoretical performance index, and then further integrates the employees' physical and mental health data over a period of time, and makes detailed adjustments to the employees' theoretical performance index, thereby deriving a more accurate employee performance index. This not only significantly improves the accuracy of employee performance evaluation, but also fully reflects the company's deep care for employees, and effectively enhances employees' satisfaction and sense of belonging.

[0010] Furthermore, generating an employee performance profile based on the work data corresponding to the employee ID includes:

[0011] Acquire some historical work data; the historical work data includes work performance, work attitude and professional skill scores;

[0012] Define the dimensions of work performance, work attitude, and professional skills;

[0013] Calculate the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value according to the work performance, work attitude and professional skill scores;

[0014] Construct an employee performance profile matrix based on employee ID and its corresponding dimensional values;

[0015] Generate employee performance profiles based on the employee performance profile matrix.

[0016] Furthermore, the calculation of the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value according to the work performance, work attitude and professional skill scores includes:

[0017] Obtain the work performance, work attitude and professional skill scores corresponding to the employee ID; the work performance includes the task completion rate RWL, the product defect rate CQL and the output efficiency CL; the work attitude includes the work enthusiasm level GJD and the execution level GZD; the professional skill score includes several test scores KF corresponding to several majors;

[0018] The work performance dimension value GYW is calculated by the formula GYW = α1 × RWL + α2 × (1-CQL) + α3 × CL;

[0019] The work attitude dimension value GTW is calculated by the formula GTW = α4 × GJD + α5 × GZD;

[0020] By formula Calculate the professional skill dimension value ZJW; where α1, α2 and α3 are weight coefficients, α1, α2 and α3∈(0,1), and α1+α2+α3=1; α4 and α5 are weight coefficients, α4 and α5∈(0,1), and α4+α5=1; max{} represents the maximum value.

[0021] Furthermore, the calculation of the employee theoretical performance index based on the employee performance portrait includes:

[0022] Obtain the work performance dimension value GYW, work attitude dimension value GTW and professional skills dimension value ZJW in the employee performance portrait;

[0023] By formula Calculate the employee theoretical performance index YLJZ; where β1, β2 and β3 are weight coefficients, which are calculated based on the position data corresponding to the employee ID; g is the proportional coefficient, g∈(0,1); DY, DT and DZ represent unit performance, unit attitude level and unit score, respectively.

[0024] Through the above steps, this application considers multiple aspects of employee data, not just performance, thus expanding the scope of employee performance considerations, making employee performance evaluation more comprehensive and providing more accurate data for the development of employees and the company.

[0025] Furthermore, the weight coefficient is calculated according to the position data corresponding to the employee ID, including:

[0026] Obtaining the position data corresponding to the employee ID; the position data includes the position name and position level;

[0027] Inputting the job title and job grade into the portrait emphasis grade estimation model to obtain several emphasis grades CD of the employee performance portrait; the portrait emphasis grade estimation model is constructed by an artificial intelligence model; the several emphasis grades include the work performance emphasis grade GYCD, the work attitude emphasis grade GTCD and the professional skills emphasis grade ZJCD;

[0028] By formula Calculate the weight coefficient β1 corresponding to work performance;

[0029] By formula Calculate the weight coefficient β2 corresponding to work attitude;

[0030] By formula Calculate the weight coefficient β3 corresponding to professional skills.

[0031] Furthermore, the portrait-focused level estimation model is constructed through an artificial intelligence model, including:

[0032] Obtain several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles;

[0033] Divide several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles into training data, verification data, and test data; perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set;

[0034] Select an AI model as the base model;

[0035] Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0036] By verifying the pre-trained model on the test set, we finally obtained a portrait emphasis level estimation model whose input is job title and job level, and whose output is several emphasis levels corresponding to the employee performance profile.

[0037] Furthermore, the generating of the state influence coefficient according to the physical and mental data corresponding to the employee ID includes:

[0038] Acquire a number of physical and mental data corresponding to the employee ID within a time period T; the physical and mental data include physical data and mental data; the physical data include body temperature TW and heart rate XL;

[0039] Get the normal range of body temperature and heart rate;

[0040] Inputting the mental data into a mental assessment model to obtain a mental state grade JZD; the mental assessment model is constructed by an artificial intelligence model;

[0041] Determine whether the body temperature is within the normal range;

[0042] Yes, set the body temperature influence value TWY = 0;

[0043] No, the absolute value of the difference between the body temperature and the boundary of the normal body temperature range is calculated as the body temperature impact value TYZ;

[0044] Determine whether the heart rate is within the normal range;

[0045] Yes, let the heart rate influence value XLY = 0;

[0046] No, the heart rate impact value XLY is calculated by calculating the absolute value of the difference between the heart rate and the heart rate close to the boundary of the normal heart rate range;

[0047] By formula Calculate the state influence coefficient ZYX; where β4 and β5 are weight coefficients, β4 and β5∈(0,1); γ1 and γ2 are weight coefficients, γ1 and γ2∈(0,1); k represents the number of the physical and mental data obtained within the time period T; DTW represents unit body temperature, DXL represents unit heart rate, and DZD represents unit state level.

[0048] Furthermore, the mental assessment model is constructed through an artificial intelligence model, including:

[0049] Obtaining certain historical mental data and their corresponding mental state levels;

[0050] Dividing a number of historical mental data and their corresponding mental state levels into training data, verification data, and test data; performing data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set;

[0051] Select an AI model as the base model;

[0052] Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0053] By verifying the pre-trained model on the test set, we finally obtain a mental assessment model whose input is mental data and output is the mental state level.

[0054] Furthermore, generating the employee performance index according to the state influence coefficient and the employee theoretical performance index includes:

[0055] Obtain the state influence coefficient ZYX and the employee theoretical performance index YLJZ;

[0056] By the formula YJZ = YLJZ × (1-ZYX γ )Calculate the employee performance index YJZ; where γ is the exponential coefficient, γ>1.

[0057] Furthermore, generating an alarm signal according to the employee performance index and the status impact coefficient includes:

[0058] Obtain the actual performance index and status influence coefficient of the employee and the employee performance index within the time period T;

[0059] The calculation method of the employee's actual performance index is the same as the calculation method of the employee's theoretical performance index; and the data in the calculation method all use the corresponding data within the time period T;

[0060] Determine whether the state influence coefficient is greater than the influence threshold; if yes, generate an employee poor health alarm signal; if no, determine whether the state influence coefficient is greater than D times the influence threshold; if yes, generate an employee poor health warning signal; if no, do nothing; where D is the proportional coefficient, D∈(0,1);

[0061] Determine whether the employee's actual performance index is greater than the employee's performance index; if yes, do nothing; if no, generate an employee performance unsatisfactory warning signal; when the employee's actual performance index is less than the employee's performance index for N consecutive times, generate an employee resignation prompt signal; where N is a positive integer.

[0062] Another aspect of the present invention provides an AI-based performance evaluation optimization method, comprising:

[0063] S0: Get employee data;

[0064] S1: Generate employee performance profiles based on work data corresponding to employee IDs; calculate employee theoretical performance indexes based on employee performance profiles;

[0065] S2: Generate a state influence coefficient based on the physical and mental data corresponding to the employee ID; generate an employee performance index based on the state influence coefficient and the employee theoretical performance index;

[0066] S3: Generate an alarm signal based on the employee performance index and the status impact coefficient; make a prompt based on the alarm signal and contact the management personnel.

[0067] Compared with the prior art, the beneficial effects of this application are:

[0068] 1. This application generates an employee performance portrait according to the work data corresponding to the employee ID; calculates the employee theoretical performance index according to the employee performance portrait; generates a state influence coefficient according to the physical and mental data corresponding to the employee ID; generates an employee performance index according to the state influence coefficient and the employee theoretical performance index; generates an alarm signal according to the employee performance index and the state influence coefficient, calculates the employee theoretical performance index at the current stage through the employee's historical multi-source data, and adjusts the employee theoretical performance index according to the employee's physical and mental data within a time period to obtain the employee performance index, thereby improving the accuracy of employee performance evaluation, promoting the company's humanistic care, and improving employee satisfaction.

[0069] 2. This application generates adaptive weights for employee position data, assigns different weights to various parameters in the employee performance portraits of different employees, achieves flexible adjustment of evaluation standards, avoids the disconnection between static evaluation standards and actual work, and improves the timeliness and relevance of performance evaluation results.

[0070] 3. This application takes into account the physical and mental data of employees during working hours, adjusts the theoretical performance index of employees according to their physical and mental state, and obtains an employee performance index related to the physical and mental data of employees, thereby improving the accuracy of performance evaluation and enhancing employees' sense of belonging to the company. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0072] Figure 1 This is a schematic diagram of the principle of an AI-based performance evaluation and optimization system for this application;

[0073] Figure 2 A flow chart for generating an alarm signal for this application;

[0074] Figure 3 This is a flow chart of an AI-based performance evaluation optimization method for this application. DETAILED DESCRIPTION

[0075] The technical solution of the present application will be described clearly and completely in conjunction with the embodiments below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0076] See also Figure 1 , the first aspect of the present application provides an AI-based performance evaluation and optimization system, including: a data acquisition module, a data analysis module, an early warning module and a database;

[0077] Data collection module: obtain employee data through data collection equipment; employee data includes employee ID, physical and mental data, and work data; data collection equipment includes various sensors, etc.;

[0078] Data analysis module: Generate employee performance portraits based on work data corresponding to employee IDs. Employee performance portraits are data that describe employees in all dimensions that affect performance; calculate employee theoretical performance indexes based on employee performance portraits. Employee theoretical performance indexes refer to the performance index that employees should theoretically achieve; generate state influence coefficients based on physical and mental data corresponding to employee IDs. State influence coefficients refer to the coefficient of the degree of state of employees affected by physical and mental data; generate employee performance indexes based on state influence coefficients and employee theoretical performance indexes. Employee performance indexes refer to the performance index that employees should actually achieve; generate alarm signals based on employee performance indexes and state influence coefficients;

[0079] Early warning module: Prompts are given based on alarm signals and management personnel are contacted; alarm signals include alarm signals for employees’ poor health, warning signals for employees’ poor physical condition, and warning signals for employees’ unsatisfactory performance.

[0080] In this embodiment, the employee performance profile is generated according to the work data corresponding to the employee ID, including:

[0081] Obtain some historical work data; historical work data includes work performance, work attitude and professional skills scores;

[0082] Define the dimensions of work performance, work attitude, and professional skills;

[0083] Calculate the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value according to the work performance, work attitude and professional skill scores;

[0084] Construct an employee performance profile matrix based on employee ID and its corresponding dimensional values;

[0085] Generate employee performance profiles based on the employee performance profile matrix.

[0086] In this embodiment, the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value are calculated according to the work performance, work attitude and professional skill scores, including:

[0087] Get the work performance, work attitude and professional skill scores corresponding to the employee ID; work performance includes task completion rate RWL, product defect rate CQL and output efficiency CL; work attitude includes work enthusiasm level GJD and execution level GZD; professional skill scores include several test scores KF corresponding to several majors;

[0088] The work performance dimension value GYW is calculated by the formula GYW = α1 × RWL + α2 × (1-CQL) + α3 × CL; as the task completion rate and output efficiency increase and the product defect rate decreases, it means that the employee's work performance is better, so the work performance dimension value will increase accordingly;

[0089] The work attitude dimension value GTW is calculated by the formula GTW = α4 × GJD + α5 × GZD; the more positive the employees are at work and the higher their execution ability is, the better their work attitude is, so the work attitude dimension value increases accordingly;

[0090] By formula Calculate the professional skill dimension value ZJW; the higher the test scores corresponding to several majors of an employee, the higher the professional level of the employee, and therefore the professional skill dimension value will increase accordingly; among them, α1, α2 and α3 are weight coefficients, α1, α2 and α3∈(0,1), and α1+α2+α3=1, and the specific value is set according to experience; α4 and α5 are weight coefficients, α4 and α5∈(0,1), and α4+α5=1, and the specific value is set according to experience; max{} means taking the maximum value.

[0091] Calculating the employee theoretical performance index according to the employee performance portrait in this embodiment includes:

[0092] Obtain the work performance dimension value GYW, work attitude dimension value GTW and professional skills dimension value ZJW in the employee performance portrait;

[0093] By formula Calculate the employee's theoretical performance index YLJZ; where β1, β2 and β3 are weight coefficients, and the weight coefficients are calculated based on the position data corresponding to the employee ID; g is the proportional coefficient, g∈(0,1), and the specific value is set according to experience. The setting of g is to make the employee's theoretical performance index YLJZ∈(0,1); DY, DT and DZ represent unit performance, unit attitude level and unit score respectively, and the specific values ​​are set according to experience; employee performance evaluation cannot be considered simply from the perspective of performance, but also from other perspectives; with the increase of work performance dimension value, work attitude dimension value and professional skills dimension value, the employee's theoretical performance index will also increase accordingly.

[0094] Through the above steps, this embodiment considers multiple aspects of employee data, not just performance, thus expanding the scope of employee performance considerations, making employee performance evaluation more comprehensive and providing more accurate data for the development of employees and the company.

[0095] The weight coefficient in this embodiment is calculated based on the position data corresponding to the employee ID, including:

[0096] Get the position data corresponding to the employee ID; the position data includes the position name and position level;

[0097] Inputting the job title and job grade into the portrait emphasis grade estimation model to obtain several emphasis grades CD of the employee performance portrait; the portrait emphasis grade estimation model is constructed through an artificial intelligence model; the several emphasis grades include the work performance emphasis grade GYCD, the work attitude emphasis grade GTCD and the professional skills emphasis grade ZJCD;

[0098] By formula Calculate the weight coefficient β1 corresponding to work performance;

[0099] By formula Calculate the weight coefficient β2 corresponding to work attitude;

[0100] By formula The weight coefficient β3 corresponding to the professional skill is calculated; the weight coefficients β1, β2 and β3 in this embodiment are obtained after normalizing several emphasis levels.

[0101] Through the above steps, this embodiment implements adaptive weight allocation for each employee's performance calculation based on the employee's position data, intelligently assigns appropriate and personalized weights to each parameter in each employee's performance portrait, and can flexibly adjust the evaluation criteria to ensure that it fits the actual work scenario, effectively avoiding the problem of static evaluation criteria being out of touch with reality, which not only improves the timeliness of performance evaluation results, but also provides the company with more accurate and reliable employee performance feedback.

[0102] The portrait-focused level estimation model in this embodiment is constructed through an artificial intelligence model, including:

[0103] Obtain several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles;

[0104] Divide several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles into training data, verification data, and test data; perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set; the ratio between the training set, the test set, and the verification set is 7:2:1;

[0105] Select an artificial intelligence model as the basic model; the artificial intelligence model includes a BP neural network model, etc.;

[0106] Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0107] By verifying the pre-trained model on the test set, we finally obtained a portrait emphasis level estimation model whose input is job title and job level, and whose output is several emphasis levels corresponding to the employee performance profile.

[0108] In this embodiment, the state influence coefficient is generated according to the physical and mental data corresponding to the employee ID, including:

[0109] Acquire several physical and mental data corresponding to the employee ID within the time period T; the physical and mental data include physical data and mental data; the physical data include body temperature TW and heart rate XL; the time period T is set according to experience, and in this embodiment, the time period T is set to 1 month; the mental data refers to the data for evaluating the mental state of the employee, including questionnaires, etc.;

[0110] Obtain the normal range of body temperature and the normal range of heart rate; the normal range of body temperature and the normal range of heart rate are set according to experience. In this embodiment, the normal range of body temperature is set to 36° C. to 37.2° C., and the normal range of heart rate is set to 60 beats / minute to 120 beats / minute;

[0111] The mental data is input into the mental assessment model to obtain the mental state level JZD; the mental assessment model is constructed through an artificial intelligence model;

[0112] Determine whether the body temperature is within the normal range;

[0113] Yes, set the body temperature influence value TWY = 0;

[0114] No, the absolute value of the difference between the body temperature and the boundary of the normal body temperature range is calculated as the body temperature impact value TYZ;

[0115] Determine whether the heart rate is within the normal range;

[0116] Yes, let the heart rate influence value XLY = 0;

[0117] No, the heart rate impact value XLY is calculated by calculating the absolute value of the difference between the heart rate and the heart rate close to the boundary of the normal heart rate range;

[0118] By formula Calculate the state influence coefficient ZYX; wherein, β4 and β5 are weight coefficients, β4 and β5∈(0,1), and the specific values ​​are set according to experience; γ1 and γ2 are weight coefficients, γ1 and γ2∈(0,1), and the specific values ​​are set according to experience. In this embodiment, γ1 and γ2 are both set to 0.5; k represents the serial number of the physical and mental data obtained within the time period T; DTW represents the unit body temperature, DXL represents the unit heart rate, and DZD represents the unit state level, and the specific values ​​are set according to experience; calculate the state influence coefficient each time within the time period T and perform an average operation to obtain the state influence coefficient within the time period T; as the employee's physical data deviates more from its normal range, his physical state is not good, which will increase the state influence coefficient; as the employee's mental state level increases, it means that the employee's mental state is not good, which increases the state influence coefficient, so the state influence coefficient increases accordingly.

[0119] Through the above steps, this embodiment takes the physical and mental data of employees during working hours into consideration, adjusts the theoretical performance index of employees according to their physical and mental states, and obtains an employee performance index related to the physical and mental data of employees, thereby improving the accuracy of performance evaluation and enhancing employees' sense of belonging to the company.

[0120] The mental assessment model in this embodiment is constructed through an artificial intelligence model, including:

[0121] Obtaining certain historical mental data and their corresponding mental state levels;

[0122] Several historical mental data and their corresponding mental state levels are divided into training data, verification data and test data; data preprocessing is performed on the training data, verification data and test data to obtain training set, verification set and test set; the ratio between the training set, the test set and the verification set is 7:2:1;

[0123] Select an artificial intelligence model as the basic model; artificial intelligence includes convolutional neural network models, etc.;

[0124] Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model;

[0125] By verifying the pre-trained model on the test set, we finally obtain a mental assessment model whose input is mental data and output is the mental state level.

[0126] In this embodiment, the employee performance index is generated according to the state influence coefficient and the employee theoretical performance index, including:

[0127] Obtain the state influence coefficient ZYX and the employee theoretical performance index YLJZ;

[0128] By the formula YJZ = YLJZ × (1-ZYX γ ) calculates the employee performance index YJZ; where γ is the exponential coefficient, γ>1, and the specific value is set according to experience; the higher the employee's state influence coefficient is, the worse his physical and mental condition is during the time period T, so it is necessary to reduce his corresponding employee theoretical performance index, and thus the employee performance index is reduced accordingly.

[0129] See also Figure 2 In this embodiment, the alarm signal is generated according to the employee performance index and the status impact coefficient, including:

[0130] Obtain the actual performance index and status influence coefficient of the employee and the employee performance index within the time period T; the calculation method of the actual performance index of the employee is the same as the calculation method of the theoretical performance index of the employee; and the data in the calculation method all use the corresponding data within the time period T;

[0131] Determine whether the state influence coefficient is greater than the influence threshold, which is set based on experience; if yes, generate an employee poor health alarm signal; if no, determine whether the state influence coefficient is greater than D times the influence threshold; if yes, generate an employee poor health alarm signal; if no, do nothing; where D is a proportional coefficient, D∈(0,1), the specific value is set based on experience, and in this embodiment, D is set to 0.7;

[0132] Determine whether the employee's actual performance index is greater than the employee's performance index; if yes, do nothing; if no, generate an employee performance unsatisfactory warning signal; when the employee's actual performance index is less than the employee's performance index for N consecutive times, generate an employee dissuasion prompt signal; wherein N is a positive integer, and the specific value is set based on experience. In this embodiment, N is set to 3.

[0133] Through the above steps, this embodiment performs early warning operations on the entire process of performance evaluation. On the one hand, it judges the physical condition of the employee, and on the other hand, it judges the performance results of the employee. The early warning process is realized automatically, the time required for manual evaluation is reduced, and the efficiency of the performance evaluation system is improved.

[0134] See also Figure 3 Another aspect of the present application provides an AI-based performance evaluation optimization method, including:

[0135] S0: Get employee data;

[0136] S1: Generate employee performance profiles based on work data corresponding to employee IDs; calculate employee theoretical performance indexes based on employee performance profiles;

[0137] S2: Generate a state influence coefficient based on the physical and mental data corresponding to the employee ID; generate an employee performance index based on the state influence coefficient and the employee theoretical performance index;

[0138] S3: Generate an alarm signal based on the employee performance index and the status impact coefficient; make a prompt based on the alarm signal and contact the management personnel.

[0139] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0140] The working principle of this application is as follows: by acquiring employee data; generating an employee performance portrait based on the work data corresponding to the employee ID; calculating the employee theoretical performance index based on the employee performance portrait; generating a state influence coefficient based on the physical and mental data corresponding to the employee ID; generating an employee performance index based on the state influence coefficient and the employee theoretical performance index; generating an alarm signal based on the employee performance index and the state influence coefficient; making a prompt based on the alarm signal and contacting the management personnel, calculating the employee theoretical performance index at the current stage through the employee's historical multi-source data, and adjusting the employee theoretical performance index based on the employee's physical and mental data within a time period to obtain the employee performance index, thereby improving the accuracy of employee performance evaluation, promoting the company's humanistic care, and improving employee satisfaction, avoiding the problem that the existing technology ignores the factors that affect employee performance changes within a certain time range, so that the actual performance of the employee does not meet the target requirements, and the employee's performance is determined to be poor, resulting in low accuracy of the performance evaluation method and affecting the employee's sense of belonging to the company.

[0141] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical method of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A performance evaluation and optimization system based on AI, characterized in that: include: Data collection module, data analysis module, early warning module and database; The data acquisition module is used to acquire employee data through data acquisition equipment; the employee data includes employee ID, physical and mental data, and work data; The data analysis module: generates an employee performance profile based on the work data corresponding to the employee ID; calculates the employee theoretical performance index based on the employee performance profile; generates a state influence coefficient based on the physical and mental data corresponding to the employee ID; Generate an employee performance index based on the status impact coefficient and the employee theoretical performance index; generate an alarm signal based on the employee performance index and the status impact coefficient; The early warning module: makes prompts according to the alarm signal and contacts the management personnel.

2. The AI-based performance evaluation and optimization system according to claim 1, characterized in that: Generating an employee performance profile based on the work data corresponding to the employee ID includes: Acquire some historical work data; the historical work data includes work performance, work attitude and professional skill scores; Define the dimensions of work performance, work attitude, and professional skills; Calculate the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value according to the work performance, work attitude and professional skill scores; Construct an employee performance profile matrix based on employee ID and its corresponding dimensional values; Generate employee performance profiles based on the employee performance profile matrix.

3. The AI-based performance evaluation and optimization system according to claim 2, characterized in that: The calculation of the corresponding work performance dimension value, work attitude dimension value and professional skill dimension value according to the work performance, work attitude and professional skill scores includes: Obtain the work performance, work attitude and professional skill scores corresponding to the employee ID; the work performance includes the task completion rate RWL, the product defect rate CQL and the output efficiency CL; the work attitude includes the work enthusiasm level GJD and the execution level GZD; the professional skill score includes several test scores KF corresponding to several majors; The work performance dimension value GYW is calculated by the formula GYW = α1 × RWL + α2 × (1-CQL) + α3 × CL; The work attitude dimension value GTW is calculated by the formula GTW = α4 × GJD + α5 × GZD; By formula Calculate the professional skill dimension value ZJW; where α1, α2 and α3 are weight coefficients, α1, α2 and α3∈(0,1), and α1+α2+α3=1; α4 and α5 are weight coefficients, α4 and α5∈(0,1), and α4+α5=1; max{} represents the maximum value.

4. The AI-based performance evaluation and optimization system according to claim 1, characterized in that: The calculation of the employee theoretical performance index based on the employee performance portrait includes: Obtain the work performance dimension value GYW, work attitude dimension value GTW and professional skills dimension value ZJW in the employee performance portrait; By formula Calculate the employee theoretical performance index YLJZ; where β1, β2 and β3 are weight coefficients, which are calculated based on the position data corresponding to the employee ID; g is the proportional coefficient, g∈(0,1); DY, DT and DZ represent unit performance, unit attitude level and unit score, respectively.

5. The AI-based performance evaluation and optimization system according to claim 4, characterized in that: The weight coefficient is calculated based on the position data corresponding to the employee ID, including: Obtaining the position data corresponding to the employee ID; the position data includes the position name and position level; Inputting the job title and job grade into the portrait emphasis grade estimation model to obtain several emphasis grades CD of the employee performance portrait; the portrait emphasis grade estimation model is constructed by an artificial intelligence model; the several emphasis grades include the work performance emphasis grade GYCD, the work attitude emphasis grade GTCD and the professional skills emphasis grade ZJCD; By formula Calculate the weight coefficient β1 corresponding to work performance; By formula Calculate the weight coefficient β2 corresponding to work attitude; By formula Calculate the weight coefficient β3 corresponding to professional skills.

6. The AI-based performance evaluation and optimization system according to claim 5, characterized in that: The portrait-focused grade estimation model is constructed through an artificial intelligence model, including: Obtain several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles; Divide several historical job titles and job levels and several emphasis levels of their corresponding employee performance profiles into training data, verification data, and test data; perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set; Select an AI model as the base model; Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model; By verifying the pre-trained model on the test set, we finally obtained a portrait emphasis level estimation model whose input is job title and job level, and whose output is several emphasis levels corresponding to the employee performance profile.

7. The AI-based performance evaluation and optimization system according to claim 1, characterized in that: The generating of the state influence coefficient according to the physical and mental data corresponding to the employee ID includes: Acquire a number of physical and mental data corresponding to the employee ID within a time period T; the physical and mental data include physical data and mental data; the physical data include body temperature TW and heart rate XL; Get the normal range of body temperature and heart rate; Inputting the mental data into a mental assessment model to obtain a mental state grade JZD; the mental assessment model is constructed by an artificial intelligence model; Determine whether the body temperature is within the normal range; Yes, set the body temperature influence value TWY = 0; No, the absolute value of the difference between the body temperature and the boundary of the normal body temperature range is calculated as the body temperature impact value TYZ; Determine whether the heart rate is within the normal range; Yes, let the heart rate influence value XLY = 0; No, the heart rate impact value XLY is calculated by calculating the absolute value of the difference between the heart rate and the heart rate close to the boundary of the normal heart rate range; By formula Calculate the state influence coefficient ZYX; where β4 and β5 are weight coefficients, β4 and β5∈(0,1); γ1 and γ2 are weight coefficients, γ1 and γ2∈(0,1); k represents the number of the physical and mental data obtained within the time period T; DTW represents unit body temperature, DXL represents unit heart rate, and DZD represents unit state level.

8. The AI-based performance evaluation and optimization system according to claim 7, characterized in that: The mental assessment model is constructed through an artificial intelligence model, including: Obtaining certain historical mental data and their corresponding mental state levels; Dividing a number of historical mental data and their corresponding mental state levels into training data, verification data, and test data; performing data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set; Select an AI model as the base model; Train the basic model with the training set, and adjust the learning rate and hyperparameters on the validation set to obtain the pre-trained model; By verifying the pre-trained model on the test set, we finally obtain a mental assessment model whose input is mental data and output is the mental state level.

9. The AI-based performance evaluation and optimization system according to claim 1, characterized in that: The generating of the employee performance index according to the state influence coefficient and the employee theoretical performance index includes: Obtain the state influence coefficient ZYX and the employee theoretical performance index YLJZ; By the formula YJZ = YLJZ × (1-ZYX γ )Calculate the employee performance index YJZ; where γ is the exponential coefficient, γ>1.

10. An AI-based performance evaluation optimization method, applied to an AI-based performance evaluation optimization system according to any one of claims 1 to 9, characterized in that: include: S0: Get employee data; S1: Generate employee performance profile based on work data corresponding to employee ID; Calculate the employee's theoretical performance index based on the employee's performance portrait; S2: Generate a state influence coefficient based on the physical and mental data corresponding to the employee ID; generate an employee performance index based on the state influence coefficient and the employee theoretical performance index; S3: Generate an alarm signal based on the employee performance index and the status impact coefficient; make a prompt based on the alarm signal and contact the management personnel.

Images