Human resource performance salary accounting management method and system based on artificial intelligence

Abstract

The invention relates to the technical field of performance prediction, in particular to a human resource performance salary accounting management method and system based on artificial intelligence, and the method comprises the steps: obtaining factor data related to employee productivity, comprising factors such as a management structure, personalized post adaptation, in-service training, reward management, man-machine engineering design, a working motivation, working satisfaction, personal skills, working guarantee, organization culture and economic stability; input variables are converted into membership values of corresponding fuzzy sets through fuzzification processing, and a triangular membership function is used for representation; establishing a knowledge base, storing fuzzy rules and membership function parameters for productivity evaluation, setting the fuzzy rules according to expert experience, and associating input and output variables; fuzzy inference operation is carried out through an inference engine, a hierarchical fuzzy inference architecture is adopted, input variables are divided into multiple groups, the groups are respectively input to parallel sub-modules for processing, and an intermediate result is transmitted to a high-level module for comprehensive inference; and performing defuzzification processing on an output fuzzy result by using a center-of-mass method to obtain a clear evaluation result of employee productivity. According to the method, the evaluation accuracy and reliability are improved through hierarchical reasoning and defuzzification.

Description

Technical Field

[0001] The present invention relates to the technical field of performance prediction, and particularly to a human resource performance salary calculation and management method and system based on artificial intelligence. Background Art

[0002] Employee performance calculation is an important part of organizational human resource management, which is directly related to the overall operation efficiency and competitiveness of the organization. Traditional performance calculation methods are mostly based on quantitative statistics, scoring, or subjective evaluation, and have problems in dealing with the complexity and uncertainty of human behavior. In addition, in the existing technologies, the method based on fuzzy reasoning, although able to handle partial fuzziness, often shows the phenomenon of a sharp increase in the number of rules (rule explosion) when facing multiple input variables, resulting in high system complexity, low operation efficiency, and being difficult to maintain and expand. Therefore, there is an urgent need for an employee performance evaluation method that can effectively process complex multi-factor inputs, avoid rule explosion, and improve evaluation accuracy and interpretability. Summary of the Invention

[0003] In view of the above technical problems, the present invention provides a human resource performance salary calculation and management method and system based on artificial intelligence to solve the problems that the existing performance evaluation methods are insufficient in dealing with the complexity and uncertainty of human behavior, and the existing fuzzy recommendation methods have poor processing effects in dealing with multi-factor inputs.

[0004] Other features and advantages of the present invention will become apparent through the following detailed description, or be learned in part through the practice of the present invention.

[0005] According to one aspect of the present invention, a human resource performance salary calculation and management method based on artificial intelligence is proposed. The method includes:

[0006] Obtaining factor data related to employee productivity, where the factor data includes management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability;

[0007] Performing fuzzification processing on the factor data as input variables, and mapping each input variable to a membership degree value of a corresponding fuzzy set; the fuzzy sets of the factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively;

[0008] Build a knowledge base to pre-store the fuzzy rules and membership function parameters for employee productivity evaluation; set a number of fuzzy rules according to expert experience to associate the fuzzy linguistic values of the input variables with the fuzzy linguistic values of the output variables; the output variable is the evaluation result of employee productivity, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high" and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1) and (0.75, 1, 1) respectively;

[0009] Use an inference engine to perform fuzzy inference operations on the fuzzy rules, that is, substitute the fuzzified input variables into the fuzzy rule set to perform fuzzy logic inference to obtain the fuzzy output set corresponding to the output variable. Among them, a hierarchical fuzzy inference architecture is adopted during inference, and multiple input variables are divided into several groups and input into parallel fuzzy inference sub-modules respectively to obtain the corresponding intermediate output results; then several of the intermediate output results are used as the input of a higher-level fuzzy inference sub-module to continue fuzzy inference;

[0010] Perform defuzzification calculation on the fuzzy output set of the output variable, and use the centroid method to obtain the clear value of the employee productivity evaluation result.

[0011] Further, the hierarchical fuzzy inference architecture includes six fuzzy inference sub-modules: the first fuzzy inference sub-module takes three soft factors, namely management structure, personality-job fit and on-the-job training, as inputs and outputs the first intermediate result; the second fuzzy inference sub-module takes two hard factors, namely reward management and ergonomic design, as inputs and outputs the second intermediate result; the third fuzzy inference sub-module takes the first intermediate result and the second intermediate result as inputs and outputs the structural factor evaluation result; the fourth fuzzy inference sub-module takes three behavioral factors, namely work enthusiasm, job satisfaction and personal skills, as inputs and outputs the behavioral factor evaluation result; the fifth fuzzy inference sub-module takes three environmental factors, namely job security, organizational culture and economic stability, as inputs and outputs the environmental factor evaluation result; the sixth fuzzy inference sub-module takes the structural factor evaluation result, the behavioral factor evaluation result and the environmental factor evaluation result as inputs and outputs the final result of the employee productivity evaluation.

[0012] Further, 126 fuzzy rules are pre-stored in the knowledge base to cover the inferences of all fuzzy combinations of the input variables.

[0013] Further, the value of the factor data is determined based on the questionnaire scoring results. The scores of the factor data of each employee are averaged and normalized to the interval [0, 1] as the initial clear value of the input variable.

[0014] Furthermore, in the fuzzy inference process, for fuzzy rules with multiple premise conditions, the Min operator is used to calculate the intersection of the membership degrees of each premise condition to determine the satisfaction degree of the fuzzy rule, and the satisfaction degree is used as the truncation membership degree value of the output fuzzy set of the fuzzy rule; the Max operator is used to perform an aggregation operation on the output fuzzy sets of all the fuzzy rules to obtain the comprehensive fuzzy output result of the output variable.

[0015] Furthermore, the final crisp result of the employee productivity evaluation corresponds to a percentage value of the employee productivity level.

[0016] Furthermore, in the hierarchical fuzzy inference architecture, the intermediate output results obtained by each level of fuzzy inference sub-module are re-fuzzified before being used as the input of the next level of fuzzy inference sub-module.

[0017] According to another aspect of the present invention, there is provided an artificial intelligence-based human resource performance salary accounting and management system, comprising:

[0018] A collection module for acquiring factor data related to employee productivity, the factor data including management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability;

[0019] A fuzzy calculation module is used to fuzzify the factor data as input variables and map each input variable to the membership degree value of the corresponding fuzzy set. The fuzzy sets of the factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively. A knowledge base is established to pre-store the fuzzy rules and membership function parameters for employee productivity evaluation. According to expert experience, a number of fuzzy rules are set to associate the fuzzy linguistic values of the input variables with the fuzzy linguistic values of the output variables. The output variable is the evaluation result of employee productivity, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high", and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1), and (0.75, 1, 1) respectively. The inference engine is used to perform fuzzy inference operations on the fuzzy rules, that is, substituting the fuzzified input variables into the fuzzy rule set to perform fuzzy logic inference to obtain the fuzzy output set corresponding to the output variable. Among them, in the inference, a hierarchical fuzzy inference architecture is adopted, dividing multiple input variables into several groups and respectively inputting them into parallel fuzzy inference sub-modules to obtain the corresponding intermediate output results. Then, several of the intermediate output results are used as the input of a higher-level fuzzy inference sub-module to continue fuzzy inference. Defuzzification calculation is performed on the fuzzy output set of the output variable, and the centroid method is used to obtain the clear value of the employee productivity evaluation result.

[0020] The technical solution of the present invention has the following beneficial effects:

[0021] By incorporating various input factors such as management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability into the model, the evaluation of employee productivity is made more comprehensive.

[0022] The triangular membership function is used to define the fuzzy sets of the input and output variables, and the centroid method is used for defuzzification, ensuring the intuitiveness, interpretability of the system inference process, and the smoothness of the output results.

[0023] By designing a hierarchical fuzzy inference architecture and dividing the input variables into multiple subsystems for hierarchical inference, the number of fuzzy rules is effectively reduced (from 59049 rules that need to be defined in traditional fuzzy inference to 126 rules), significantly reducing the computational complexity and avoiding the rule explosion phenomenon.

[0024] The modular-designed fuzzy inference sub-module is convenient for system expansion and maintenance, improving the scalability and flexibility of the evaluation model.

[0025] It can simulate the expert evaluation process in a way that is closer to human cognitive logic, improving the accuracy and rationality of productivity evaluation results, and is applicable to human resource management and decision-making support in different organizational environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flowchart of a method for calculating and managing human resource performance compensation based on artificial intelligence in an embodiment of this specification;

[0027] Figure 2 It is a specific schematic diagram of a triangular membership function in an embodiment of this specification;

[0028] Figure 3 It is a structural block diagram of a system for calculating and managing human resource performance compensation based on artificial intelligence in an embodiment of this specification. DETAILED IMPLEMENTATION MANNER

[0029] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art. The features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the embodiments of the present invention. However, those skilled in the art will realize that one or more of the specific details may be omitted, or other methods, components, devices, steps, etc. may be employed. In other cases, well-known technical solutions are not shown or described in detail to avoid obscuring the various aspects of the present invention.

[0030] In addition, the accompanying drawings are only schematic illustrations of the present invention. The same reference numerals in the drawings denote the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0031] The present invention provides a method for calculating and managing human resource performance compensation based on artificial intelligence for a product. Referring to Figure 1As shown, it is a flow chart of a method for calculating and managing human resource performance compensation based on artificial intelligence for a product provided by an embodiment of the present invention. The method can be applied to electronic devices such as personal computers and servers. The method can be executed by a device, which can be implemented by software and / or hardware. The method can specifically include the following steps S101 to S102:

[0032] In step S101, factor data related to employee productivity is obtained, and the factor data includes management structure, individual job adaptation, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture and economic stability.

[0033] Among them, management structure refers to the division of responsibilities, allocation of rights and responsibilities, and information flow between various management levels within an organization. A good management structure can improve work efficiency and coordination. Personality and job adaptation refers to the degree of fit between employees' personality characteristics and their job requirements. High adaptability helps employees adapt to the job environment more quickly and perform at their best. On-the-job training refers to the training programs provided by the organization for employees to continuously improve their professional skills, update their knowledge, and exchange experiences, which can promote the improvement of employees' professional quality and increase their work efficiency. Reward management refers to the use of various incentive mechanisms such as salary, bonuses, benefits, and promotion opportunities to stimulate employees' enthusiasm and work motivation, thereby improving their work output. Ergonomic design refers to whether the design of equipment, tools, and systems in the work environment conforms to ergonomic principles to reduce employee fatigue, improve comfort and work efficiency. Work motivation refers to the comprehensive performance of employees' internal and external motivations in the process of completing work tasks, which directly affects their work involvement and continuity. Job satisfaction refers to the comprehensive subjective satisfaction of employees with their own job positions, responsibilities, work environment, and organizational management. High satisfaction is usually positively correlated with high productivity. Personal skills refer to the comprehensive qualities of employees such as communication skills, problem-solving skills, collaboration skills, and innovation skills in their professional activities, which have a direct effect on improving individual and organizational performance. Job security refers to the degree of employees' perception of current career stability and future employment prospects. Good job security can reduce employee turnover and enhance work engagement. Organizational culture refers to the common values, beliefs, behavioral norms, and work atmosphere formed within the organization, which has a profound impact on employee behavior patterns, sense of belonging, and productivity. Economic stability refers to the stability of the macroeconomic environment, including the employment market situation, salary levels, inflation rates, etc. The stability of the economic environment is directly related to employees' professional confidence and consumption capacity, thereby indirectly affecting their work performance. By collecting data on the above eleven factors through questionnaires, we can systematically portray the work status and potential productivity performance of individual employees in the organization in a multi-dimensional manner, laying the foundation for subsequent comprehensive evaluation based on fuzzy reasoning.

[0034] In step S102, the factor data is subjected to fuzzification as an input variable, and each input variable is mapped to a membership degree value of a corresponding fuzzy set; the fuzzy sets of the factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively.

[0035] Among them, fuzzification is the initial step. In this step, input data is collected and its degree of belonging to the corresponding fuzzy set is determined through a membership function. The membership function is represented by a curve that maps each input point to a membership degree value ranging from 0 to 1. There can be various types of membership functions in the present invention, such as triangular, trapezoidal, piecewise, Gaussian, bell-shaped, etc. Exemplarily, for the fuzzy inference system proposed in the present invention, a triangular membership function is adopted. The purpose of this choice is to facilitate the evaluation questions in the questionnaire, because there is no form suitable for triangular fuzzy numbers specified on the ordinal scale to handle the uncertainty of interval boundaries. Triangular fuzzy numbers are derived from three-valued judgments, which include the minimum possible value 'a', the most likely value 'b', and the maximum possible value 'c'. Triangular fuzzy numbers can be expressed as M = (a, b, c), as shown in the following formula and Figure 2 shown:

[0036]

[0037] The linguistic terms assigned to the fuzzy sets for each input evaluation are "poor", "medium", and "good", and the corresponding parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1).

[0038] In step S103, a knowledge base is established, and fuzzy rules and membership function parameters for employee productivity evaluation are pre-stored; several fuzzy rules are set according to expert experience to associate the fuzzy linguistic values of input variables with the fuzzy linguistic values of output variables; the output variable is the evaluation result of employee productivity, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high", and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1), and (0.75, 1, 1) respectively.

[0039] Among them, building a knowledge base is an important part of constructing an employee productivity evaluation model. Specifically, the knowledge base contains fuzzy rules and membership function parameters for evaluating employee productivity. The construction of the knowledge base is based on expert experience. According to expert experience and the understanding of various relevant factors of employee productivity, a number of fuzzy rules are set. The form of each rule is "if... then...", and these rules are used to describe the fuzzy relationship between input variables and output variables. The fuzzy linguistic values of input variables (such as "poor", "medium", "good") and the fuzzy linguistic values of output variables (such as "very low", "low", "high", "very high") are related through these rules. For the output variable of employee productivity evaluation, five fuzzy linguistic values are defined: "very low", "low", "medium", "high", and "very high". These fuzzy linguistic values use triangular membership functions to describe their membership degrees.

[0040] In step S104, a fuzzy inference operation is performed on the fuzzy rules by using an inference engine, that is, the fuzzified input variables are substituted into the fuzzy rule set to perform fuzzy logic inference, and a fuzzy output set corresponding to the output variable is obtained. Among them, when inferring, a hierarchical fuzzy inference architecture is adopted, multiple input variables are divided into several groups, and are respectively input into parallel fuzzy inference sub-modules to obtain corresponding intermediate output results; then several of the intermediate output results are used as the input of a higher-level fuzzy inference sub-module to continue fuzzy inference.

[0041] Among them, the core task of the inference engine is to substitute the fuzzified input variables into the preset fuzzy rule set, perform fuzzy logic inference, and thus obtain the fuzzy output set of the output variables. The key to this process is fuzzy inference, that is, mapping the fuzzy input data to the corresponding fuzzy output through the rule system. During the inference process, in order to cope with the complexity of multiple input variables and rules, the system adopts a hierarchical fuzzy inference architecture. This architecture divides multiple input variables into several groups according to a certain logic and inputs them into parallel fuzzy inference sub-modules for inference respectively. These parallel sub-modules can independently process different groups of inputs and thus obtain their respective intermediate output results. After obtaining the intermediate output results of each sub-module, these intermediate results will be input into a higher-level fuzzy inference sub-module for comprehensive processing and further inference at a higher level. Through this layer-by-layer inference method, information can be integrated at multiple levels, thus obtaining the final employee productivity evaluation result. The specific steps of the fuzzy inference operation are as follows: input fuzzified data, substitute the input data obtained and fuzzified in step S101 into the fuzzy rule set for inference; sub-module processing, divide the input data into different groups according to the hierarchical architecture and input them into each parallel fuzzy inference sub-module; intermediate output results, each sub-module respectively obtains intermediate output results according to its rule set, and these results are the fuzzy output set; layer-by-layer inference, pass the intermediate output results of multiple sub-modules layer by layer to higher-level fuzzy inference sub-modules for inference at a higher level, and finally obtain a comprehensive result. Based on the hierarchical structure of this embodiment, the problem of rule explosion can be effectively avoided, and at the same time, the inference efficiency is improved, making the entire system more efficient and scalable. Through this layer-by-layer inference and grouped processing method, the inference engine can efficiently process complex multi-dimensional inputs, thus greatly reducing the computational amount in the inference process of the employee productivity evaluation system, while still maintaining high accuracy and rationality.

[0042] In step S105, defuzzification calculation is performed on the fuzzy output set of the output variable, and the centroid method is used to obtain the crisp value of the employee productivity evaluation result.

[0043] In the last step of the fuzzy inference process, defuzzification is used to convert the output fuzzy set into a definite numerical value as the final result of the employee productivity evaluation. The present invention adopts the centroid of area (COA) method as the defuzzification technique. The COA method performs defuzzification based on the centroid position of the fuzzy set. Specifically, the COA method calculates the center of gravity (i.e., the "center" of the fuzzy set) of the fuzzy set, and determines this center position by performing a weighted average of the distribution of the fuzzy set along the x-axis. The centroid of area method can effectively solve the problems of sharp increase in fuzzy rules and computational amount (rule explosion) existing in traditional fuzzy inference, and it provides a simple way to obtain the crisp value of the fuzzy output.

[0044] Specifically, when defuzzifying, the COA method is used to calculate the centroid of the fuzzy set, and the formula is as follows:

[0045]

[0046] Where x i is the value of the i-th point in the fuzzy set, and μ i (x i ) is the membership degree value of this point, and n is the number of points in the fuzzy set. Exemplarily, taking the triangular membership function as an example, when the fuzzy output set is "low" or "high", etc., its centroid value is calculated through the above formula to obtain a clear output value, that is, the evaluation result of employee productivity.

[0047] In one embodiment, the hierarchical fuzzy inference architecture includes six fuzzy inference sub-modules: the first fuzzy inference sub-module takes three soft factors, namely management structure, personality-position adaptation, and on-the-job training, as inputs and outputs a first intermediate result; the second fuzzy inference sub-module takes two hard factors, namely reward management and ergonomic design, as inputs and outputs a second intermediate result; the third fuzzy inference sub-module takes the first intermediate result and the second intermediate result as inputs and outputs a structural factor evaluation result; the fourth fuzzy inference sub-module takes three behavioral factors, namely work enthusiasm, job satisfaction, and personal skills, as inputs and outputs a behavioral factor evaluation result; the fifth fuzzy inference sub-module takes three environmental factors, namely job security, organizational culture, and economic stability, as inputs and outputs an environmental factor evaluation result; the sixth fuzzy inference sub-module takes the structural factor evaluation result, the behavioral factor evaluation result, and the environmental factor evaluation result as inputs and outputs the final result of the employee productivity evaluation.

[0048] Among them, the hierarchical fuzzy inference architecture is to cope with the complexity of multiple input variables and solve the rule explosion problem in traditional fuzzy inference. Specifically, this architecture divides multiple input variables into different categories, assigns them to different sub-modules for parallel inference, and finally obtains the final employee productivity evaluation result through multi-level inference.

[0049] The hierarchical fuzzy inference architecture consists of six fuzzy inference sub-modules. Each sub-module is responsible for processing specific input factors and outputting intermediate results or final evaluation results. Among them:

[0050] The first fuzzy inference sub-module: Input: management structure, personality-position adaptation, on-the-job training (these three belong to soft factors); Output: the first intermediate result, which represents the evaluation of the structural factors of the employee.

[0051] The second fuzzy inference sub-module: Input: Reward management and ergonomic design (these two belong to hard factors). Output: The second intermediate result, which represents the evaluation of the employee's hard factors.

[0052] The third fuzzy inference sub-module: Input: The first intermediate result and the second intermediate result (representing the structural factor and the hard factor evaluation results respectively). Output: The structural factor evaluation result, which is a comprehensive evaluation of the employee's overall work structure and hard conditions.

[0053] The fourth fuzzy inference sub-module: Input: Work enthusiasm, job satisfaction, and personal skills (these three belong to behavioral factors). Output: The behavioral factor evaluation result, which describes the behavioral attitudes and skill levels demonstrated by the employee at work.

[0054] The fifth fuzzy inference sub-module: Input: Job security, organizational culture, and economic stability (these three belong to environmental factors). Output: The environmental factor evaluation result, reflecting the impact of the organizational environment where the employee is located on their productivity.

[0055] The sixth fuzzy inference sub-module: Input: The structural factor evaluation result, the behavioral factor evaluation result, and the environmental factor evaluation result (representing the employee's work structure, behavioral state, and environmental conditions respectively). Output: The final employee productivity evaluation result, which gives a final productivity score by comprehensively considering all factors.

[0056] The hierarchical fuzzy inference architecture avoids the explosive growth of the number of rules by reasonably grouping the input variables and sending each group of data to an independent inference sub-module for processing. In addition, this architecture can effectively improve the processing efficiency and inference accuracy of the system, and finally provide accurate employee productivity evaluation results.

[0057] In one embodiment, 126 fuzzy rules are pre-stored in the knowledge base for covering the inferences of all fuzzy combinations of the input variables.

[0058] Among them, in the knowledge base of the system, 126 fuzzy rules are pre-stored. The setting of these rules is based on the experience of experts and the understanding of various relevant factors of employee productivity. Each rule adopts the form of "if... then..." to describe the fuzzy relationship between the input variables and the output variables. The number of fuzzy rules is calculated by the following formula:

[0059] N = M V ;

[0060] Where: N is the number of fuzzy rules, M is the number of membership functions for each input variable, which is 3 in this implementation: representing "poor", "medium", and "good", and V is the number of input variables (which is 2 in this embodiment, i.e., the soft factor and the hard factor each form a group). For example, if M = 3 and V = 2, the number of rules N is 3 2 = 9 rules. If the number of input variables V = 3, the number of rules will increase sharply (increase to 27 rules). To avoid the problem of rule explosion, this embodiment limits the number of rules to 126 through a hierarchical fuzzy inference architecture, which ensures the high efficiency of system calculation.

[0061] These 126 fuzzy rules cover all possible fuzzy combinations of input variables. Specifically, the rules describe how to evaluate the employee productivity under different input conditions, ensuring that inferences can be made for each possible input. The setting of the rules not only needs to consider the logical relationships between different factors but also needs to reflect the performance of employees under various conditions according to the actual work situation.

[0062] By storing these 126 fuzzy rules in the knowledge base, the system can cover all fuzzy combinations of input variables and evaluate the employee productivity in different situations. This knowledge base provides the basis for the fuzzy inference system, ensures that the inference process can be completed efficiently, and at the same time guarantees the rationality and accuracy of the results.

[0063] In one implementation, the value of the factor data is determined based on the questionnaire scoring results. The scores of the factor data for each employee are averaged and normalized to the [0, 1] interval as the initial crisp value of the input variable.

[0064] Among them, the evaluation of employee productivity is based on multiple relevant factors, and the scores of these factors are collected through questionnaires. The data of each factor corresponds to the performance of an employee and is quantified through expert scoring or self-evaluation by the employee. For each factor affecting employee productivity (such as management structure, personality-job fit, on-the-job training, etc.), a questionnaire is designed for experts or employees to score according to their actual feelings. The scoring can use the Likert scale, for example, the scores range from 1 (very poor) to 5 (very good). After collecting the scores of all employees, for each factor, first, the scores of each employee are averaged. This step ensures that the scores of each factor are representative and can effectively reflect the general situation of the factor in the employee group. For the convenience of subsequent fuzzification processing, all scores need to be normalized, that is, each score value is converted to the [0, 1] interval. The normalization formula is:

[0065]

[0066] where x is the original score, x min and x max are the minimum and maximum values of the scores respectively.

[0067] Through questionnaire scoring, averaging, and normalization, the data for each factor of each employee is finally used as input variables in the form of clear values within the interval [0, 1]. This process ensures the standardization of the data, enabling the subsequent fuzzy inference process to proceed smoothly, while improving the unity and accuracy of data processing.

[0068] In one embodiment, during the fuzzy inference process, for fuzzy rules with multiple preconditions, the Min operator is used to calculate the intersection of the membership degrees of each precondition to determine the satisfaction degree of the fuzzy rule, and the satisfaction degree is used as the truncated membership degree value of the output fuzzy set of the fuzzy rule; the Max operator is used to perform an aggregation operation on the output fuzzy sets of all the fuzzy rules to obtain the comprehensive fuzzy output result of the output variable.

[0069] Each fuzzy rule has multiple preconditions. For example, "If the management structure is'medium' and the personality - job fit is 'good', then...". In fuzzy inference, it is necessary to combine the membership degrees of these preconditions to determine whether the rule is satisfied.

[0070] The Min operator is used to calculate the intersection of the membership degrees of the preconditions, that is, the minimum value among all the membership degrees of the preconditions is taken as the satisfaction degree of the rule. For example, if the membership degree of precondition A is 0.8 and the membership degree of condition B is 0.6, then the satisfaction degree of the rule is 0.6 (taking the minimum value). The formula is expressed as:

[0071] μ rule =min(μ A (x A ),μ B (x B ),...,μ n (x n ));

[0072] where μ A (x A ),μ B (x B ),...,μ n (x n ) are the membership degrees of each precondition.

[0073] The calculated satisfaction degree is used as the truncation membership degree value of the output fuzzy set of this fuzzy rule. That is, in the output fuzzy set, the output value corresponding to this rule uses the satisfaction degree to truncate the membership degree. This process ensures that the influence of different preconditions is reflected in the final output. Aggregate the output fuzzy sets of all fuzzy rules, and use the Max operator to process the output results of multiple rules. The Max operator will select the maximum membership degree value among the output fuzzy sets of multiple rules. Specifically, for each output variable, calculate the influence of all fuzzy rules on this output variable, and select the membership degree with the greatest influence as the final result. The formula is expressed as:

[0074] μ output =max(μ rule1 (x),μ rule2 (x),...,μ ruleN (x));

[0075] where μ ruleN (x) is the membership degree output by each rule.

[0076] That is, the Min operator is used to calculate the intersection of multiple preconditions to determine the satisfaction degree of the fuzzy rule. And, the Max operator is used to aggregate the outputs of all rules, and select the maximum membership degree value as the final fuzzy output. This method ensures that in complex multi-rule and multi-input situations, the relationship between rules can be effectively processed, and a reasonable comprehensive fuzzy output result can be obtained.

[0077] In one embodiment, the final crisp result of the employee productivity evaluation corresponds to a percentage value of the employee productivity level.

[0078] Among them, after defuzzification, the final crisp result usually presents in the form of a percentage value, that is, the productivity level of the employee is in the range of 0 to 100%. The closer to 100% indicates higher employee productivity, and closer to 0% indicates lower productivity.

[0079] In one embodiment, in the hierarchical fuzzy inference architecture, the intermediate output result obtained by each level of fuzzy inference sub-module is re-fuzzified before being used as the input of the next level of fuzzy inference sub-module.

[0080] Among them, in the hierarchical fuzzy inference architecture, the output result of each layer represents the processing result of the inference sub-module of this layer, usually a fuzzy output. To ensure the accuracy of information transfer between different levels, the intermediate output result of each level needs to be re-fuzzified.

[0081] Based on the same idea, as Figure 3 shown, a human resource performance salary calculation and management system based on artificial intelligence is provided, including:

[0082] The data collection module 201 is configured to obtain factor data related to employee productivity, where the factor data includes management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability.

[0083] The fuzzy computing module 202 is configured to perform fuzzification processing on the factor data as input variables, and map each input variable to a membership degree value of a corresponding fuzzy set; the fuzzy sets of the factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively; establish a knowledge base, and pre-store fuzzy rules and membership function parameters for employee productivity evaluation; set a number of fuzzy rules according to expert experience, and associate the fuzzy linguistic values of the input variables with the fuzzy linguistic values of the output variables; the output variable is the evaluation result of employee productivity, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high", and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1), and (0.75, 1, 1) respectively; perform fuzzy inference operations on the fuzzy rules by using an inference engine, that is, substitute the fuzzified input variables into the fuzzy rule set to perform fuzzy logic inference, and obtain a fuzzy output set corresponding to the output variable. Among them, in the inference, a hierarchical fuzzy inference architecture is adopted, multiple input variables are divided into several groups, and are respectively input into parallel fuzzy inference sub-modules to obtain corresponding intermediate output results; then, several of the intermediate output results are used as inputs of a higher-level fuzzy inference sub-module to continue fuzzy inference; perform defuzzification calculation on the fuzzy output set of the output variable, and use the centroid method to obtain a clear value of the evaluation result of employee productivity.

[0084] As can be seen from the above system, in this embodiment, by incorporating various input factors such as management structure, personalized job adaptation, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability into the model, the evaluation of employee productivity becomes more comprehensive. The fuzzy sets of input and output variables are defined using triangular membership functions, and the centroid method is used for defuzzification, ensuring the intuitiveness, interpretability of the system inference process, and the smoothness of the output results. By designing a hierarchical fuzzy inference architecture and dividing the input variables into multiple subsystems for hierarchical inference, the number of fuzzy rules is effectively reduced (from 59,049 rules that need to be defined in traditional fuzzy inference to 126 rules), significantly reducing the computational complexity and avoiding the rule explosion phenomenon. The modularized fuzzy inference sub-module is adopted, which is convenient for system expansion and maintenance, and improves the scalability and flexibility of the evaluation model. It can simulate the expert judgment process in a way that is closer to human cognitive logic, improving the accuracy and rationality of the productivity evaluation results, and is applicable to human resource management and decision support in different organizational environments.

[0085] The specific details of each module in the above system have been described in detail in the implementation manner of the method part. For the undisclosed detailed content, reference can be made to the implementation manner content of the method part, so it will not be elaborated here.

[0086] Through the description of the above embodiments, those skilled in the art can easily understand that the exemplary embodiments described here can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.) or on the network, including several instructions to enable a computing device (which can be a personal computer, server, terminal device, or network device, etc.) to execute the method according to the exemplary embodiments of the present invention.

[0087] In addition, the above drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present invention, rather than for limiting purposes. It is easy to understand that the processes shown in the above drawings do not indicate or limit the time sequence of these processes. Additionally, it is also easy to understand that these processes can be executed synchronously or asynchronously in, for example, multiple modules.

[0088] It should be noted that although several modules or units of devices for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the exemplary embodiments of the present invention, the features and functions of the two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0089] Other embodiments of the present invention will be readily apparent to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention, which follow the general principles of the invention and include known common general knowledge or conventional technical means in the technical field not disclosed in the present invention. The specification and embodiments are only to be considered as exemplary, and the true scope and spirit of the invention are pointed out by the claims.

[0090] It should be understood that the present invention is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present invention is only limited by the appended claims.

Claims

1. A human resource performance salary calculation and management method based on artificial intelligence, characterized in that The method includes: Obtaining factor data related to employee productivity, where the factor data includes management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability; Fuzzifying the factor data as input variables and mapping each input variable to a membership degree value of a corresponding fuzzy set; the fuzzy sets of the factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively; Establishing a knowledge base to pre-store fuzzy rules and membership function parameters for employee productivity assessment; setting a number of fuzzy rules according to expert experience to associate the fuzzy linguistic values of input variables with the fuzzy linguistic values of output variables; the output variable is the evaluation result of employee productivity, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high", and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1), and (0.75, 1, 1) respectively; Using an inference engine to perform fuzzy inference operations on the fuzzy rules, that is, substituting the fuzzified input variables into the fuzzy rule set to perform fuzzy logic inference to obtain a fuzzy output set corresponding to the output variable. Among them, in the inference, a hierarchical fuzzy inference architecture is adopted, dividing multiple input variables into several groups and respectively inputting them into parallel fuzzy inference sub-modules to obtain corresponding intermediate output results; then using several of the intermediate output results as the input of a higher-level fuzzy inference sub-module to continue fuzzy inference; Performing defuzzification calculation on the fuzzy output set of the output variable and obtaining a crisp value of the evaluation result of employee productivity by using the centroid method.

2. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, wherein, The hierarchical fuzzy inference architecture includes six fuzzy inference sub-modules: the first fuzzy inference sub-module takes three soft factors of management structure, personality-job fit, and on-the-job training as inputs and outputs a first intermediate result; The second fuzzy inference sub-module takes two hard factors of reward management and ergonomic design as inputs and outputs a second intermediate result; the third fuzzy inference sub-module takes the first intermediate result and the second intermediate result as inputs and outputs a structural factor evaluation result; The fourth fuzzy inference sub-module takes three behavioral factors of work motivation, job satisfaction, and personal skills as inputs and outputs a behavioral factor evaluation result; The fifth fuzzy inference sub-module takes three environmental factors of job security, organizational culture, and economic stability as inputs and outputs an environmental factor evaluation result; The sixth fuzzy inference sub-module takes the structural factor evaluation result, the behavioral factor evaluation result, and the environmental factor evaluation result as inputs and outputs the final result of the employee productivity assessment.

3. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, wherein 126 fuzzy rules are pre-stored in the knowledge base for covering the inferences of all fuzzy combinations of the input variables.

4. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, wherein The value of the said factor data is determined based on the questionnaire scoring results. The scores of the said factor data for each employee are averaged and normalized to the interval [0, 1] as the initial crisp value of the said input variable.

5. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, characterized in that, In the fuzzy inference process, for a fuzzy rule with multiple premise conditions, the Min operator is used to calculate the intersection of the membership degrees of each premise condition to determine the satisfaction degree of the said fuzzy rule, and the satisfaction degree is used as the truncation membership degree value of the output fuzzy set of the said fuzzy rule. The Max operator is used to perform an aggregation operation on the output fuzzy sets of all the said fuzzy rules to obtain the comprehensive fuzzy output result of the said output variable.

6. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, characterized in that, The final crisp result of the said employee productivity assessment corresponds to the percentage value of the employee productivity level.

7. The method for calculating and managing human resource performance compensation based on artificial intelligence according to claim 1, wherein, In the said hierarchical fuzzy inference architecture, the intermediate output result obtained by each level of fuzzy inference sub-module is re-fuzzified before being used as the input of the next level of fuzzy inference sub-module.

8. An artificial intelligence-based human resource performance salary calculation and management system, characterized in that, Including: A collection module for obtaining factor data related to employee productivity. The factor data includes management structure, personality-job fit, on-the-job training, reward management, ergonomic design, work motivation, job satisfaction, personal skills, job security, organizational culture, and economic stability. A fuzzy calculation module for fuzzifying the said factor data as input variables and mapping each input variable to the membership degree value of the corresponding fuzzy set. The fuzzy sets of the said factor data include fuzzy linguistic values "poor", "medium", and "good", and their corresponding triangular membership function parameters are (0, 0, 0.5), (0, 0.5, 1), and (0.5, 1, 1) respectively. A knowledge base is established to pre-store the fuzzy rules and membership function parameters for employee productivity assessment. A number of fuzzy rules are set according to expert experience to associate the fuzzy linguistic values of the input variables with the fuzzy linguistic values of the output variables. The output variable is the employee productivity evaluation result, and its fuzzy set includes fuzzy linguistic values "very low", "low", "medium", "high", and "very high", and the corresponding triangular membership function parameters are (0, 0, 0.25), (0, 0.25, 0.5), (0.25, 0.5, 0.75), (0.5, 0.75, 1), and (0.75, 1, 1) respectively. The inference engine is used to perform fuzzy inference operations on the said fuzzy rules, that is, substituting the fuzzified input variables into the said fuzzy rule set to perform fuzzy logic inference to obtain the fuzzy output set corresponding to the output variable. Among them, in the inference, a hierarchical fuzzy inference architecture is adopted, dividing multiple input variables into several groups and respectively inputting them into parallel fuzzy inference sub-modules to obtain the corresponding intermediate output results. Then, several of the said intermediate output results are used as the input of a higher-level fuzzy inference sub-module to continue fuzzy inference. Defuzzification calculation is performed on the fuzzy output set of the said output variable, and the centroid method is used to obtain the crisp value of the said employee productivity evaluation result.

Images