An employee service management method and system for human resources intelligent terminal

Abstract

The present invention discloses an employee service management method and system for a human resources intelligent terminal, which relates to the field of computer technology, including collecting employee static data and employee dynamic data through a human resources intelligent terminal, and generating an employee data set; generating an employee portrait containing employee characteristic information according to the employee data set; triggering a service linkage mechanism to call corresponding resources to generate personalized service content according to predicted service needs, and sending the personalized service content to the employee's human resources intelligent terminal; collecting employee feedback data on pushed personalized services, analyzing the feedback results, and performing parameter optimization on the matching logic of a dynamic rule engine; updating employee portraits and service rules according to the optimized dynamic rule engine; the present invention dynamically generates a rule candidate set through a dynamic rule engine, and selects the optimal rule in combination with reinforcement learning, so as to respond to changes in employee needs in real time and ensure the personalization and timeliness of service content.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an employee service management method and system for a human resources intelligent terminal. Background Art

[0002] With the continuous development of digital and intelligent management of enterprises, human resource management has gradually shifted from traditional manual processes to automated systems based on big data analysis and intelligent algorithms. In the existing technology, many companies have begun to use employee management platforms, human resource smart terminals or other information tools to collect employees' personal information, attendance records and some behavioral data, so as to achieve basic employee information management and service demand response. However, these systems usually only provide static employee data management functions and lack the ability to predict and respond to employees' dynamic needs in real time. Some systems try to push demand through simple rule engines, but the rules are relatively fixed and cannot meet the diverse service needs of employees in a dynamic environment.

[0003] However, existing technologies still have significant deficiencies in dealing with the complexity and dynamic changes of employee needs. On the one hand, the prediction of employee service needs relies on manual static rules, resulting in a low degree of personalization of service content, making it difficult to accurately match employees' actual needs; on the other hand, the lack of effective feedback mechanisms and dynamic optimization capabilities makes it difficult for the system to adapt to the changing trends of employee needs in long-term operation, and the service effect gradually declines. Based on this, there is an urgent need for a new method that can dynamically generate rules and optimize the service response process to improve the intelligence level of employee service management and user experience. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an employee service management method for a human resources intelligent terminal to solve the problem in the prior art that it is difficult to accurately match the actual needs of employees and difficult to adapt to the changing trend of employee needs.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an employee service management method for a human resources intelligent terminal, which includes collecting employee static data and employee dynamic data through a human resources intelligent terminal and generating an employee data set;

[0008] Generate employee profiles containing employee characteristic information based on employee data sets;

[0009] Based on the employee profile, the dynamic rule engine matches the rule candidate set that the employee currently corresponds to, and predicts the employee's service needs based on the matched rule candidate set;

[0010] Based on the predicted service needs, the service linkage mechanism is triggered to call corresponding resources to generate personalized service content, and the personalized service content is sent to the employee's human resources smart terminal;

[0011] Collect employees’ feedback on the personalized services they push, analyze the feedback results, and optimize the matching logic of the dynamic rule engine;

[0012] Update employee portraits and service rules based on the optimized dynamic rule engine.

[0013] As a preferred solution of the employee service management method for the human resources intelligent terminal of the present invention, wherein: the employee static data and employee dynamic data are collected through the human resources intelligent terminal, and the specific steps are as follows:

[0014] Deploy human resources intelligent terminals as data collection carriers;

[0015] Collect employee static data in real time through human resources intelligent terminals and clean the employee static data; employee static data includes: personal information and job information;

[0016] The cleaned employee static data is stored in a relational database to form an initial employee data set;

[0017] Use human resources intelligent terminals to collect employee dynamic data in real time; employee dynamic data includes clock-in records, heart rate, and workstation usage status;

[0018] The employee dynamic data is transmitted to the backend server through the MQTT protocol and stored in the time series database to form a time series dynamic data stream;

[0019] Utilizing the employee initial data set and time series dynamic data stream, the employee static data and employee dynamic data are associated based on the employee number field, and the integrity and timeliness of the data are verified through timestamp matching to generate an employee data set.

[0020] As a preferred solution of the employee service management method for a human resources intelligent terminal of the present invention, wherein: the employee portrait containing employee characteristic information is generated according to the employee data set, and the specific steps are as follows:

[0021] Extract employee static data associated with the employee number field from the employee dataset to generate basic information features of the employee;

[0022] Analyze employee dynamic data associated with the employee number field from the employee data set and calculate employee behavior pattern characteristics;

[0023] Statistical indicators are extracted from employees' historical records, and employee static data and employee dynamic data associated with the employee number field in the employee dataset are combined to generate employee portraits containing employee characteristic information using machine learning models.

[0024] As a preferred solution of the employee service management method for a human resources intelligent terminal of the present invention, wherein: based on the employee portrait, the rule candidate set currently corresponding to the employee is matched by a dynamic rule engine, and the service needs of the employee are predicted according to the matched rule candidate set. The specific steps are as follows:

[0025] Standardize and embed the data in the employee portrait into time series to generate the feature matrix X;

[0026] The feature matrix X is input into the generator of the adversarial network for dynamic rule engine matching to generate the rule candidate set corresponding to the employee. The expression is:

[0027]

[0028] Among them, R'(t) represents the rule candidate set dynamically generated at time point t, T represents the length of the time window, is the integral symbol, D(G(X,W g )) is the output of the discriminator model, D represents the discriminator function whose input is the generation rule, G(X,W g ) is the generator function, W g represents the generator weight parameter, λ is the regularization coefficient, n is the total number of rules in the historical rule set, i is the index of the rule, P i (R) is the frequency of occurrence of the ith rule R in the historical records, and dt represents a small increment in the time dimension;

[0029] The generated rule candidate set R'(t) is screened and optimized through reinforcement learning, and decisions are generated to predict service demand;

[0030] According to the decision of predicting service demand, select the rule that maximizes the total benefit from the candidate rule set R'(t);

[0031] According to the definition of the largest rule, the corresponding service demand type is extracted as the employee's service demand.

[0032] As a preferred solution of the employee service management method for human resource intelligent terminal described in the present invention, wherein: according to the predicted service demand, the service linkage mechanism is triggered to call the corresponding resources to generate personalized service content, and the personalized service content is sent to the employee's human resource intelligent terminal. The specific steps are as follows:

[0033] Based on the predicted service requirements, the corresponding resources in the resource mapping table are matched and verified to see if they pass. Specific verification indicators include the running status of the resources, external API availability, and permission configuration.

[0034] After the verification is passed, the resources that have passed the verification are marked as available resources;

[0035] Based on the matched available resources, combined with the employee static data and employee dynamic data in the employee data set, resource call request parameters are generated, and resource calls are initiated with the adapted interface protocol according to the resource call request parameters;

[0036] Call the interface of the matching resource, parse the response data returned by the resource to extract the service data, and organize the service data into structured data;

[0037] Generate personalized service content based on the structured data returned by the resources and the type of service requirements;

[0038] Choose a push method based on employees’ notification channel preferences;

[0039] Through the push interface, personalized service content is sent to the employee's human resources smart terminal according to the selected push method.

[0040] As a preferred solution of the employee service management method for human resources intelligent terminal of the present invention, wherein: the feedback data of employees on the pushed personalized services are collected, the feedback results are analyzed, and the matching logic of the dynamic rule engine is optimized. The specific steps are as follows:

[0041] The feedback data in the personalized service content is collected and pushed in real time through the interface, and the employee number, service ID and timestamp are attached. After being transmitted to the backend server, it is stored in the document database;

[0042] Clean the feedback data stored in the document database, verify the data's association integrity, and generate a standardized structured feedback data set;

[0043] Aggregate and analyze the structured feedback dataset by service type, time, and employee characteristics, and calculate the average rating, reception rate, and sentiment score of different service types;

[0044] Identify the performance differences of rule matching through the average ratings, reception rates, and sentiment scores of different service types, and then evaluate the fit between push content and employee needs;

[0045] Optimize the matching logic of the dynamic rule engine based on the degree of fit.

[0046] As a preferred solution of the employee service management method for human resource intelligent terminal of the present invention, wherein: the employee portrait and service rules are updated according to the optimized dynamic rule engine, and the specific steps are as follows:

[0047] According to the optimized dynamic rule engine, the rules are associated with employee portraits through employee numbers to form a corresponding relationship between rules and portraits;

[0048] Based on the correspondence between rules and portraits, combined with dynamic employee data collected in real time, the behavioral regularity characteristics in employee portraits are updated;

[0049] Based on the updated employee portrait, extract the static feature fields involved in rule matching, compare them one by one with the static features in the employee portrait, and improve the employee portrait;

[0050] Adjust rule weights based on improved employee profiles, generate new rules based on service feedback data, and update service rules;

[0051] Use the updated employee portraits and service rules to verify the matching effect of the rule output and check the compatibility of the portraits and rules.

[0052] In a second aspect, the present invention provides an employee service management system for a human resources intelligent terminal, including a data set generation module, a portrait generation module, a demand prediction module, a service push module, a feedback collection module and a rule update module;

[0053] The data set generation module is used to collect employee static data and employee dynamic data through the human resources intelligent terminal and generate an employee data set;

[0054] The portrait generation module is used to generate an employee portrait containing employee characteristic information based on the employee data set;

[0055] The demand prediction module is used to match the rule candidate set currently corresponding to the employee through the dynamic rule engine based on the employee portrait, and predict the employee's service demand according to the matched rule candidate set;

[0056] The service push module is used to generate personalized service content and send the personalized service content to the employee's human resources smart terminal;

[0057] The feedback collection module is used to collect employees' feedback data on the pushed personalized services, analyze the feedback results, and optimize the matching logic of the dynamic rule engine;

[0058] The rule updating module is used to update employee portraits and service rules according to the optimized dynamic rule engine.

[0059] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the employee service management method for a human resources intelligent terminal as described in the first aspect of the present invention is implemented.

[0060] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the employee service management method for a human resources intelligent terminal as described in the first aspect of the present invention is implemented.

[0061] The beneficial effects of the present invention are as follows: the present invention significantly improves the intelligence level and accuracy of employee service management by combining employee portraits, a dynamic rule engine and a service feedback optimization mechanism; first, employee portraits are generated by integrating static data and dynamic data, which can comprehensively reflect the basic information, behavioral patterns and predictive characteristics of employees, and provide data support for the accurate prediction of service needs; secondly, a dynamic rule engine is used to dynamically generate a rule candidate set, and the optimal rules are selected in combination with reinforcement learning, which can respond to changes in employee needs in real time and ensure the personalization and timeliness of service content; in addition, a closed-loop feedback mechanism is introduced, and the rule engine and portrait data are dynamically optimized through the analysis of employee feedback data on services, which can continuously learn and adapt to the changing trends of employee needs; it has significant advantages in service accuracy, response efficiency and long-term adaptability, thereby providing a more efficient and intelligent solution for the human resource management of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0063] Figure 1 This is a flow chart of the employee service management method for a human resources intelligent terminal in Example 1.

[0064] Figure 2 This is a module diagram of the system for employee service management of human resources intelligent terminals in Example 1. DETAILED DESCRIPTION

[0065] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0066] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0067] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0068] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides an employee service management method for a human resources intelligent terminal, comprising the following steps:

[0069] S1. Collect employee static data and employee dynamic data through human resources intelligent terminals and generate employee data sets.

[0070] Furthermore, deploying human resource intelligent terminals as data collection carriers;

[0071] Among them, human resources intelligent terminals refer to devices that can collect, process and transmit data, including hardware devices (such as workstation sensors, smart bracelets) and software terminals (such as mobile applications, enterprise WeChat applets); data collection carriers, that is, hardware and software tools used to collect static and dynamic data of employees, as the main source of data input.

[0072] It should be noted that by deploying human resources intelligent terminals, the ability to collect employee data from multiple dimensions can be achieved, which not only ensures the comprehensiveness of the collection (covering static and dynamic data), but also ensures the real-time nature of the collection (continuity of dynamic data). These data provide a high-precision data foundation for subsequent data analysis and service optimization.

[0073] Employee static data is collected in real time through human resources intelligent terminals, and data quality verification is used to clean employee static data; employee static data includes: personal information and job information.

[0074] It should be noted that this step realizes the efficient collection of employee static data through human resources intelligent terminals, and improves the quality (accuracy and completeness) of data through standardized verification and data cleaning, laying a solid foundation for subsequent storage and integration. Cleaned static data can avoid correlation problems caused by data errors and ensure overall reliability.

[0075] The cleaned static data of employees is stored in a relational database to form the initial data set of employees. Among them, "relational database" is a database management system based on the relational model (such as MySQL, PostgreSQL), which stores data through a table structure and supports efficient query and data management; the initial data set of employees is a structured data set stored in the database after normalization, which contains the static personal information of employees and is the basis for data integration and analysis.

[0076] It should be noted that storing cleaned employee static data in a relational database can achieve structured management, and data can be quickly accessed and operated through SQL queries. At the same time, the employee initial data set provides a static information foundation for the subsequent association and integration of dynamic data, and supports batch analysis and modeling needs.

[0077] Utilize HR smart terminals to collect employee dynamic data in real time; employee dynamic data includes clock-in records, heart rate, and workstation usage status. Dynamic data refers to employee behavior and status data that changes over time, such as clock-in time, heart rate (bpm), workstation usage time, etc. Real-time collection: Through HR smart terminals, dynamic data is collected and transmitted at a higher frequency (such as seconds or minutes).

[0078] Preferably, the real-time collection of dynamic data can reflect the behavioral patterns and physiological status of employees, and provide a real-time basis for employee health management and work efficiency evaluation. For example, through the abnormal detection of heart rate data, employees' health problems can be discovered and responded to in a timely manner.

[0079] Furthermore, edge computing technology can be introduced to preliminarily process dynamic data (such as average and peak calculation) on the human resources intelligent terminal to reduce transmission pressure and improve collection efficiency. For example, when collecting heart rate, only the maximum and minimum values ​​in the interval are transmitted. This method can save bandwidth while retaining the key features of dynamic data.

[0080] The dynamic data of employees is transmitted to the backend server through the MQTT protocol and stored in the time series database to form a time series dynamic data flow; specifically, the MQTT protocol: a lightweight message transmission protocol suitable for data transmission of IoT devices, with the characteristics of low latency and low bandwidth consumption; time series database: a database specially designed for storing and querying time series data (such as InfluxDB, TimescaleDB), supporting efficient time dimension query and analysis; time series dynamic data flow: refers to the continuous transmission and storage of dynamic data in the form of time series, reflecting the changing trajectory of data over time.

[0081] Preferably, the MQTT protocol can achieve low-latency and efficient transmission between the human resources intelligent terminal and the backend server, while the time series database can store and manage high-frequency dynamic data, laying the foundation for subsequent time series analysis (such as trend prediction and anomaly detection).

[0082] Utilize the employee initial data set and time series dynamic data stream, associate employee static data and employee dynamic data based on the employee number field, verify the integrity and timeliness of the data through timestamp matching, and generate an employee data set; specifically, associate employee static data and employee dynamic data: based on the common fields of employee static data and employee dynamic data (such as employee_id), organically combine the two, verify the integrity and timeliness of the data through timestamp matching, and generate an employee data set; timestamp matching: utilize the timestamp field of dynamic data to ensure the timeliness and accuracy of the data, and eliminate data records with mismatched or duplicated times.

[0083] Preferably, employee static data provides basic information about employees, and employee dynamic data provides information about employees' behaviors and status. The employee data set generated through association and integration can achieve all-round insights. Timestamp matching ensures the real-time and consistency of data, effectively avoiding erroneous analysis results caused by data mismatch.

[0084] Furthermore, an employee profile containing employee characteristic information is generated based on the employee dataset. The specific steps are as follows:

[0085] Employee static data associated with the employee number field is extracted from the employee data set to generate basic information features of the employee. Specifically, employee static data, as the basic layer of employee portraits, provides a complete description of employee identity attributes and can quickly establish a basic feature set for employees.

[0086] Analyze employee dynamic data associated with the employee number field from the employee data set and calculate the employee's behavioral pattern characteristics.

[0087] Preferably, dynamic data reflects the behavior patterns of employees in a specific time period and can reveal their work habits and status (such as regular working hours and work pressure). Dynamic analysis can detect abnormal behavior patterns (such as excessive overtime and abnormal heart rate fluctuations), providing a basis for timely intervention.

[0088] Statistical indicators are extracted from the employee's historical records, and the employee static data and employee dynamic data associated with the employee number field in the employee data set are combined to generate an employee portrait containing employee characteristic information using a machine learning model. Among them, historical records refer to data records accumulated over a long period of time within an organization, which usually contain information related to the employee's past behavior and performance.

[0089] S2. Based on the employee portrait, the dynamic rule engine is used to match the rule candidate set that the employee currently corresponds to, and the employee's service needs are predicted based on the matched rule candidate set.

[0090] Furthermore, the data in the employee portraits are standardized and time series embedded to generate the feature matrix X.

[0091] The feature matrix X is input into the generator of the adversarial network for dynamic rule engine matching to generate the rule candidate set corresponding to the employee. The expression is:

[0092]

[0093] Among them, R'(t) represents the rule candidate set R' dynamically generated at time point t, argmax R is a mathematical operator, which means that among all possible rule sets R, the rule set that maximizes the objective function value is selected. T represents the length of the time window. is an integral symbol, indicating that the objective function is continuously summed within the time range [0, T], D(G(X,W g )) is the output of the discriminator model, which represents the evaluation result of the discriminator on the generator's generated rules (the value range is [0,1], the closer to 0, the more the rule meets the requirements), D is the discriminator function whose input is the generated rules, G(X,W g ) is the generator function, which represents the generator function based on the feature matrix X and the generator weight parameter W g , output rule, λ is the regularization coefficient, which represents the balance between the quality of control rule generation and the frequency of historical rules, n is the total number of rules in the historical rule set, i is the index of the rule, P i (R) is the frequency of occurrence of the ith rule R in the historical records, indicating the business rationality or acceptability of the rule, and dt represents a small increment in the time dimension.

[0094] Specifically, the probability that the generator's generation rule is rejected by the discriminator within the time window T is calculated by integration, thereby optimizing the generation effect of the generator.

[0095] Specifically, the frequency of rules in historical records is used to regularize the generated results, thus avoiding the generation of rules that are too rare or do not conform to business logic.

[0096] The generated rule candidate set R'(t) is screened and optimized through reinforcement learning, and a decision to predict service demand is generated. The expression is:

[0097]

[0098] Among them, π(R) represents the decision to predict the service demand R, E[R f(π,x,β'(t))] is the expectation operator, R f is the reward function, β represents the balance coefficient, Q(X,R,t) is the state action value function, x π is the decision variable under the decision function π;

[0099] Reward function R f (π,X,R'(t)), the expression is:

[0100] R f (π,X,R'(t))=α·S(π,R'(t))+γ·F(X,R'(t))-δ·C(R'(t));

[0101] Among them, S(π,R'(t)) represents the matching score between the decision function and the rule, F(X,R'(t)) represents the score of the rule satisfying the employee's needs, C(R'(t)) represents the rule complexity penalty term, α represents the weight of controlling the decision function matching score S(π,R'(t)), γ represents the weight of the demand satisfaction score F(X,R'(t)), and δ represents the weight of controlling the rule complexity penalty term C(R'(t));

[0102] The state action value function Q(X,R,t) is expressed as:

[0103]

[0104] Among them, R f (π,X,R'(t+k)) represents the instantaneous benefit value of the rule candidate set R'(t) to the feature matrix X under the decision function π at the future time t+k, Tt represents the remaining time length from the current time point t to the end of the time window, k represents the time step, γ k Represents the discount factor raised to a power.

[0105] Based on the decision of predicting service demand, the rule that maximizes the total benefit is selected from the candidate rule set R'(t).

[0106] According to the definition of the largest rule, the corresponding service demand type is extracted as the employee's service demand.

[0107] It should be noted that the “rule that maximizes the total benefit” means that at a certain time point t, a rule is selected from the candidate rule set R'(t), which can maximize the immediate benefit (through the reward function R fThe rule can be used to measure the current status of employees, such as the degree to which the current needs of employees are met), and maximize the long-term benefits (measured by the state-action value function Q, such as the positive impact of the rule on the future status of employees or business goals), while comprehensively considering the execution cost and complexity of the rules, and finally obtain the rule with the highest comprehensive benefit (immediate benefit + long-term benefit - penalty term).

[0108] S3. Based on the predicted service demand, the service linkage mechanism is triggered to call corresponding resources to generate personalized service content, and the personalized service content is sent to the employee's human resources smart terminal.

[0109] Furthermore, based on the predicted service needs, the corresponding resources in the resource mapping table (such as internal ERP modules, external API interfaces, etc.) are matched and verified to see if they pass. Specific verification indicators include the operating status of the resources, external API availability, and permission configuration to ensure that the resources can be called normally.

[0110] After the verification passes, the resources that pass the verification are marked as available resources.

[0111] Based on the matched available resources, combined with the employee static data and employee dynamic data in the employee data set, resource call request parameters are generated, and resource calls are initiated with the adapted interface protocol according to the resource call request parameters.

[0112] Call the interface of the matching resource, parse the response data returned by the resource to extract the service data, and organize the service data into structured data.

[0113] Generate personalized service content based on the structured data returned by the resources and the type of service requirements.

[0114] Select the push method based on employees' notification channel preferences (such as mobile notifications, desktop messages, voice reminders, etc.).

[0115] For example: for employees who prefer instant reminders, mobile push notifications are used; for employees who prefer voice notifications, content is sent through a speech synthesis interface.

[0116] Through the push interface, personalized service content is sent to the employee's human resources smart terminal (such as mobile phone, desktop client or smart speaker) according to the selected push method.

[0117] S4. Collect employees’ feedback data on the pushed personalized services, analyze the feedback results, and optimize the parameters of the matching logic of the dynamic rule engine.

[0118] Furthermore, the feedback data in the personalized service content pushed is collected in real time through the interface, and the employee number, service ID and timestamp are attached, transmitted to the background server and stored in the document database.

[0119] Among them, the interface represents the bridge for data interaction, which is usually implemented in the form of API (application programming interface) and is used to transfer data between different services; the document database is a NoSQL database that stores data in the form of documents, usually in JSON or BSON format, and supports flexible structured and unstructured data storage.

[0120] Clean the feedback data stored in the document database, verify the data's association integrity, and generate a standardized structured feedback data set.

[0121] Preferably, through cleaning, the accuracy and completeness of the feedback data can be ensured, providing a reliable data basis for subsequent analysis and processing.

[0122] The structured feedback dataset is aggregated and analyzed by service type, time, and employee characteristics, and the average rating, reception rate, and sentiment score of different service types are calculated.

[0123] Specifically, by grouping and counting structured feedback data, we can aggregate and analyze it by service type (such as training courses, welfare recommendations, etc.), time (such as daily, weekly, monthly), and employee characteristics (such as position, department, age group). The specific method is to calculate the feedback indicators for each group of data, including: average score (summarizing employees' scores on the service and calculating the mean), acceptance rate (statistical proportion of those who choose to accept the service) and sentiment score (based on the mean of text feedback sentiment score). These statistical results can reflect the performance differences of different service types in different time periods and employee groups, and provide data support for optimizing service design and rule matching.

[0124] Furthermore, an example is given as follows: suppose an enterprise pushes three types of services (such as training courses, welfare recommendations and project allocation), and collects feedback data on a weekly basis for different employee groups (such as department managers and front-line employees); the results show that in "training courses", the average score of department managers is 4.8, the acceptance rate is 90%, and the sentiment score is 0.95, while the average score of front-line employees is 3.5, the acceptance rate is 60%, and the sentiment score is 0.7, indicating that department managers are more interested in and have higher demand for this service; in "welfare recommendations", the score of front-line employees is 4.2, the acceptance rate is 80%, and the sentiment score is 0.85, while the score of department managers is only 3.0, the acceptance rate is 40%, and the sentiment score is 0.5, indicating that welfare recommendations are more suitable for front-line employees; through such aggregate analysis, the service push strategy can be adjusted for different groups, such as strengthening the training content for department managers and optimizing the welfare recommendation plan for front-line employees.

[0125] Identify the performance differences of rule matching through the average ratings, reception rates and sentiment scores of different service types, and then evaluate the fit between the pushed content and employee needs.

[0126] Specifically, by correlating the average ratings, acceptance rates, and sentiment scores of different service types with the rule matching results, the performance differences of each rule in different dimensions are calculated. For example, for each service type, the feedback indicators of the matching rules (such as rating distribution, acceptance rate trend, and sentiment score mean) are counted and compared with the expected target values ​​to identify the shortcomings of the rules in meeting employee needs. In addition, clustering or classification algorithms can be used to divide the rules into three categories: high match, partial match, and low match, so as to accurately locate the rules that need to be optimized, and adjust the weights and logic of the rule generator based on these rules.

[0127] Furthermore, an example is given as follows: suppose three types of services are pushed to employees in different positions: training courses, welfare recommendations and project allocation; the results show that the average score of the "training courses" rule is 4.5, the acceptance rate is 85%, and the sentiment score is 0.9, indicating that the rule is highly consistent with the needs of this group; the score of the "welfare recommendation" rule is 3.2, the acceptance rate is 50%, and the sentiment score is 0.6, indicating that the rule effect is average; and the score of the "project allocation" rule is 2.8, the acceptance rate is 30%, and the sentiment score is 0.4, indicating that the rule matching effect is poor; by comparing these differences, the shortcomings of the "welfare recommendation" and "project allocation" rules can be identified, and the rule generation logic can be further adjusted (such as adding personalized parameters) to improve the accuracy of rule matching.

[0128] Optimize the matching logic of the dynamic rule engine based on the degree of fit.

[0129] Specifically, optimizing the matching logic of the dynamic rule engine means adjusting the generator weights so that it gives priority to generating rule candidate sets with high satisfaction; optimizing the discriminator classification logic to improve the accuracy of rule classification and matching; and adjusting the reward function weights to balance the rule matching score, requirement satisfaction score, and complexity penalty.

[0130] S5. Update employee portraits and service rules based on the optimized dynamic rule engine.

[0131] Furthermore, according to the optimized dynamic rule engine, the rules are associated with the employee portraits through the employee numbers, forming a corresponding relationship between the rules and the portraits.

[0132] Specifically, obtain the generated optimization rule result R(t) from the dynamic rule engine, including rule ID, matching score, rule content and other information, and record the generation timestamp; match the feature data (static features, dynamic features and predictive features) stored in the employee portrait through associated fields (such as employee_id) to form a one-to-one correspondence between the rule and the portrait; store the rule result in a temporary data table, and the fields include rule ID, employee number, rule matching score and rule content, etc., for subsequent portrait updates and rule verification steps; perform data integrity check on the rule generation result (such as checking whether key fields are missing) to ensure that the association between the rule and the portrait is accurate; write the correspondence between the rule and the portrait into the database, and pass the key fields to the next step to update the portrait features.

[0133] Based on the correspondence between rules and portraits, combined with dynamic employee data collected in real time, the behavioral pattern characteristics in the employee portraits are updated.

[0134] Specifically, extract the employee number and related dynamic data (such as clock-in records, workstation usage data) associated with the rule matching results, and perform time series analysis on the latest dynamic data (such as calculating continuous working time and activity frequency); adjust the behavioral regularity characteristics (such as working time patterns, behavioral preferences, etc.) according to the pattern changes of dynamic data, and store the updated regularity characteristics in the dynamic feature field of the portrait; use the rule results and historical dynamic data to recalculate the predictive features of the portrait (such as future health risk predictions, job demand predictions) through machine learning models (such as regression analysis, deep learning models); integrate the updated behavioral regularity characteristics and predictive features into the employee portrait, and verify the integrity of the portrait data (such as checking whether there are missing values); store the adjusted portrait features in a document-based database for use in subsequent rule set optimization and verification.

[0135] Based on the updated employee portrait, extract the static feature fields involved in the rule matching, compare them one by one with the static features in the employee portrait, and improve the employee portrait.

[0136] Specifically, for missing features, the latest data is retrieved from the human resources intelligent terminal management platform to supplement them; for features with differences, they are directly updated to the field values ​​in the rule generation results, and the corrected data is written to the relational database.

[0137] Adjust rule weights based on improved employee portraits, generate new rules based on service feedback data, and update service rules.

[0138] Specifically, based on the improved employee portrait, first extract the static and dynamic feature data in the portrait, analyze the mapping relationship between it and the feature fields of the existing service rules, and preliminarily adjust the rule weights to make the weight values ​​more consistent with the distribution of the current portrait features; quantify the matching degree between the rule and the portrait by calculating the adaptability score of the rule. The adaptability score can be calculated by the matching degree between the importance weight of the portrait feature and the rule triggering condition. For rules with higher adaptability scores, increase the weight appropriately, and reduce the weight for rules with lower scores to ensure that the rules maintain a high correlation with the dynamic features of the portrait; at the same time, combine the feedback data of the push service, aggregate and analyze indicators such as the average score, reception rate and sentiment score of the service type, calculate the service feedback performance score of each rule, comprehensively compare the service feedback score with the adaptability score, and modify the rule weight value by dynamically adjusting the formula so that the weight can reflect the comprehensive performance of the rule in portrait matching and service feedback; for rules with poor performance in service feedback, identify potential logical defects or adaptation problems by analyzing their triggering conditions and adaptation features.

[0139] For example, if the coverage of the trigger conditions is too narrow or the feature correlation is insufficient, the trigger conditions can be expanded or the rule logic can be redefined through correlation analysis to address these problems. At the same time, based on the portrait features and feedback data, a new set of rule candidates is generated using the rule generation algorithm. The candidate rules are screened and verified, and rules with high adaptability and high feedback performance are added to the rule set. Rules with too low weights or that have not been triggered for a long time are removed to ensure the simplicity and efficiency of the rule set. Finally, the effect of the optimized rule set is verified through simulation tests or actual applications, and its performance in portrait matching, service demand prediction, and feedback performance is evaluated. The rule weights and rule logic are continuously iterated and optimized in subsequent portrait updates and service feedback data to ensure that the service system can always adapt to employee needs efficiently and accurately.

[0140] Use the updated employee portraits and service rules to verify the matching effect of the rule output and check the compatibility of the portraits and rules.

[0141] Specifically, static features, dynamic features and predictive features are extracted from the updated employee portraits and used as the input feature matrix X to verify the matching effect of the rule output; all rules in the rule set are re-matched with the feature matrix X, the matching score of each rule to the portrait is calculated, and the matching results are recorded; the matching results are statistically analyzed to calculate the average matching score, coverage (the proportion of rules that match the employee portrait) and accuracy (the proportion of rules that meet the requirements); based on the verification results, an evaluation report on the adaptability of the rules and portraits is generated to clarify the effects of the rule output under different combinations of portrait features, and to mark the rules with insufficient adaptability; the rules with insufficient adaptability are optimized and iterated, and the verification results are fed back to the rule engine as input for the next round of rule optimization.

[0142] The present embodiment also provides an employee service management system for a human resources intelligent terminal, including: a portrait generation module, a demand prediction module, a service push module, a feedback collection module and a rule update module; the portrait generation module is used to collect static data and dynamic data of employees through the human resources intelligent terminal, and generate employee portraits containing employee characteristic information; the demand prediction module is used to match the rule set currently corresponding to the employee through the dynamic rule engine based on the employee portrait, and predict the employee's service needs in combination with historical data and real-time dynamic data; the service push module is used to trigger the service linkage mechanism to call corresponding resources based on the predicted service needs, and push personalized service content to employees through the human resources intelligent terminal; the feedback collection module is used to collect employee feedback data on the push service, analyze the feedback results, and optimize the parameters of the matching logic of the dynamic rule engine; the rule update module is used to update the employee portrait and service rules according to the optimized dynamic rule engine.

[0143] This embodiment also provides a computer device, which is suitable for the employee service management method of a human resources intelligent terminal, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the employee service management method for a human resources intelligent terminal as proposed in the above embodiment.

[0144] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0145] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the employee service management method for a human resources intelligent terminal proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0146] In summary, the present invention significantly improves the intelligence level and accuracy of employee service management by combining employee portraits, dynamic rule engines and service feedback optimization mechanisms; first, employee portraits are generated by integrating static data and dynamic data, which can comprehensively reflect the basic information, behavioral patterns and predictive characteristics of employees, and provide data support for the accurate prediction of service needs; secondly, the dynamic rule engine dynamically generates a set of rule candidates, and combines reinforcement learning to screen out the optimal rules, which can respond to changes in employee needs in real time and ensure the personalization and timeliness of service content; in addition, a closed-loop feedback mechanism is introduced, and the rule engine and portrait data are dynamically optimized through the analysis of employee feedback data on services, which can continuously learn and adapt to the changing trends of employee needs; it has significant advantages in service accuracy, response efficiency and long-term adaptability, thereby providing a more efficient and intelligent solution for the human resource management of enterprises.

[0147] Example 2, referring to Table 1, is the second example of the present invention. To further verify the technical solution of the present invention, experimental simulation data of the employee service management method for a human resources intelligent terminal is provided.

[0148] The experiment selected 50 employees of a certain company as test samples, covering different departments and job levels, including 20 people in the production department, 15 people in the R&D department, 10 people in the administrative department and 5 people in the management level.

[0149] First, we deployed a human resources intelligent terminal, including a smart bracelet (used to collect employee dynamic data such as heart rate, workstation usage data, etc.) and a human resources intelligent terminal management platform (used to enter static data such as employee personal information, job information, etc.). Static data collection: Enter employee name, age, gender, job position, salary and other information through the human resources intelligent terminal management platform, and store it in a relational database after cleaning to form an initial employee data set. Dynamic data collection: Collect employee heart rate, workstation usage data and attendance punch-in records in real time through smart bracelets, transmit them to the background using the MQTT protocol, and store them in a time series database to form a time series dynamic data stream.

[0150] Next, static data and dynamic data are associated through employee numbers, and the timestamp mechanism is used to verify data integrity to generate an employee data set. The machine learning model is used to analyze employee behavior patterns and predictive features, and ultimately a structured employee portrait is generated.

[0151] Secondly, the generative adversarial network (GAN) is used to dynamically generate rule candidate sets, and the rule screening process is optimized through reinforcement learning to predict employees' service needs. For example, based on the characteristics of employee heart rate changes and workstation usage time in dynamic data, it is predicted that employees may need psychological counseling services or short-term vacations; based on the predicted service needs, the service linkage mechanism is triggered to call the matching service resources in the resource mapping table, such as psychological counseling appointments or welfare application forms. The service content is sent to the employee's device in the form of push notifications through the human resources smart terminal.

[0152] Secondly, we collect employees’ feedback data on push services, extract sentiment scores using natural language processing after cleaning, and aggregate and analyze the feedback data. We calculate indicators such as the average rating and reception rate of service types and optimize the parameters of the dynamic rule engine.

[0153] Finally, the implementation method of the present invention is compared with the traditional human resource intelligent terminal, wherein the traditional method only uses static data and cannot realize dynamic service prediction and personalized push.

[0154] The details are shown in Table 1 below:

[0155] Table 1 Comparative test data of employee service management methods

[0156]

[0157] Through the above test data, the following analysis conclusions can be drawn:

[0158] Regarding response time: the average response time of the new method is significantly better than that of traditional human resources intelligent terminals. For example, in the psychological counseling appointment scenario, the response time of traditional human resources intelligent terminals is 120 seconds, while the new method only takes 45 seconds, an efficiency improvement of 62.5%. This advantage is mainly due to the real-time prediction capability of the dynamic rule engine and the rapid response of the resource linkage mechanism.

[0159] Regarding the service acceptance rate: the service acceptance rate is an important indicator to measure the matching degree of employees to the pushed service content; the acceptance rate of psychological counseling appointment services of traditional human resources intelligent terminals is 60%, while the new method reaches 95%, an increase of 58.3%; this shows that the method of the present invention can more accurately predict the service needs of employees, and the service content is more in line with actual needs.

[0160] Regarding the feedback emotion score: the feedback emotion score reflects the employees' subjective satisfaction with the service; the psychological counseling appointment emotion score of the traditional human resources intelligent terminal is 0.55, while the new method reaches 0.85, an increase of 54.5%; this shows that the new method can not only improve service efficiency, but also significantly improve employee satisfaction.

[0161] Regarding the average score: Among all service types, the average score of the new method is higher than that of the traditional human resources intelligent terminal; for example, the score of the psychological counseling appointment service increased from 3.8 to 4.7, and the score of the welfare application service increased from 4.0 to 4.6; this shows that the new method can provide higher quality services.

[0162] To sum up, traditional human resources intelligent terminals only rely on static data and cannot analyze employee status in real time. The present invention realizes real-time prediction through a dynamic rule engine combined with time series data, expanding the management capability boundary of human resources intelligent terminals. The service push in traditional human resources intelligent terminals is a fixed template and lacks pertinence. The new method generates personalized content based on employee portraits, which significantly improves service accuracy and employee experience. Traditional human resources intelligent terminals cannot optimize rules based on feedback data, while the new method dynamically adjusts the rule engine parameters through a reinforcement learning mechanism, so that service accuracy and employee satisfaction continue to improve.

[0163] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An employee service management method for a human resources intelligent terminal, characterized in that: include, Collect employee static data and employee dynamic data through human resources intelligent terminals and generate employee data sets; Generate employee profiles containing employee characteristic information based on employee data sets; Based on the employee profile, the dynamic rule engine matches the rule candidate set that the employee currently corresponds to, and predicts the employee's service needs based on the matched rule candidate set; Based on the predicted service needs, the service linkage mechanism is triggered to call corresponding resources to generate personalized service content, and the personalized service content is sent to the employee's human resources smart terminal; Collect employees’ feedback on the personalized services they push, analyze the feedback results, and optimize the matching logic of the dynamic rule engine; Update employee portraits and service rules based on the optimized dynamic rule engine.

2. The employee service management method for a human resources intelligent terminal according to claim 1, characterized in that: The specific steps of collecting employee static data and employee dynamic data through the human resources intelligent terminal are as follows: Deploy human resources intelligent terminals as data collection carriers; Collect employee static data in real time through human resources intelligent terminals and clean the employee static data; employee static data includes: personal information and job information; The cleaned employee static data is stored in a relational database to form an initial employee data set; Use human resources intelligent terminals to collect employee dynamic data in real time; employee dynamic data includes clock-in records, heart rate, and workstation usage status; The employee dynamic data is transmitted to the backend server through the MQTT protocol and stored in the time series database to form a time series dynamic data stream; Utilizing the employee initial data set and time series dynamic data stream, the employee static data and employee dynamic data are associated based on the employee number field, and the integrity and timeliness of the data are verified through timestamp matching to generate an employee data set.

3. The employee service management method for a human resources intelligent terminal according to claim 2, characterized in that: The specific steps of generating an employee portrait containing employee characteristic information based on the employee data set are as follows: Extract employee static data associated with the employee number field from the employee dataset to generate basic information features of the employee; Analyze employee dynamic data associated with the employee number field from the employee data set and calculate employee behavior pattern characteristics; Statistical indicators are extracted from employees' historical records, and employee static data and employee dynamic data associated with the employee number field in the employee dataset are combined to generate employee portraits containing employee characteristic information using machine learning models.

4. The employee service management method for a human resources intelligent terminal according to claim 3, characterized in that: Based on the employee portrait, the dynamic rule engine is used to match the rule candidate set corresponding to the employee, and the employee's service needs are predicted according to the matched rule candidate set. The specific steps are as follows: Standardize and embed the data in the employee portrait into time series to generate the feature matrix X; The feature matrix X is input into the generator of the adversarial network for dynamic rule engine matching to generate the rule candidate set corresponding to the employee. The expression is: Among them, R'(t) represents the rule candidate set dynamically generated at time point t, T represents the length of the time window, is the integral symbol, D(G(X,W g )) is the output of the discriminator model, D represents the discriminator function whose input is the generation rule, G(X,W g ) is the generator function, W g represents the generator weight parameter, λ is the regularization coefficient, n is the total number of rules in the historical rule set, i is the index of the rule, P i (R) is the frequency of occurrence of the ith rule R in the historical records, and dt represents a small increment in the time dimension; The generated rule candidate set R'(t) is screened and optimized through reinforcement learning, and decisions are generated to predict service demand; According to the decision of predicting service demand, select the rule that maximizes the total benefit from the candidate rule set R'(t); According to the definition of the largest rule, the corresponding service demand type is extracted as the employee's service demand.

5. The employee service management method for a human resources intelligent terminal according to claim 4, characterized in that: The service linkage mechanism is triggered based on the predicted service demand to call the corresponding resources to generate personalized service content, and the personalized service content is sent to the employee's human resources smart terminal. The specific steps are as follows: Based on the predicted service requirements, the corresponding resources in the resource mapping table are matched and verified to see if they pass. Specific verification indicators include the running status of the resources, external API availability, and permission configuration. After the verification is passed, the resources that have passed the verification are marked as available resources; Based on the matched available resources, combined with the employee static data and employee dynamic data in the employee data set, resource call request parameters are generated, and resource calls are initiated with the adapted interface protocol according to the resource call request parameters; Call the interface of the matching resource, parse the response data returned by the resource to extract the service data, and organize the service data into structured data; Generate personalized service content based on the structured data returned by the resources and the type of service requirements; Choose a push method based on employees’ notification channel preferences; Through the push interface, personalized service content is sent to the employee's human resources smart terminal according to the selected push method.

6. The employee service management method for a human resources intelligent terminal according to claim 5, characterized in that: The specific steps of collecting employees' feedback data on the pushed personalized services, analyzing the feedback results, and optimizing the matching logic of the dynamic rule engine are as follows: The feedback data in the personalized service content is collected and pushed in real time through the interface, and the employee number, service ID and timestamp are attached. After being transmitted to the backend server, it is stored in the document database; Clean the feedback data stored in the document database, verify the data's association integrity, and generate a standardized structured feedback data set; Aggregate and analyze the structured feedback dataset by service type, time, and employee characteristics, and calculate the average rating, reception rate, and sentiment score of different service types; Identify the performance differences of rule matching through the average ratings, reception rates, and sentiment scores of different service types, and then evaluate the fit between push content and employee needs; Optimize the matching logic of the dynamic rule engine based on the degree of fit.

7. The employee service management method for a human resources intelligent terminal according to claim 6, characterized in that: According to the optimized dynamic rule engine, the employee portrait and service rules are updated. The specific steps are as follows: According to the optimized dynamic rule engine, the rules are associated with employee portraits through employee numbers to form a corresponding relationship between rules and portraits; Based on the correspondence between rules and portraits, combined with dynamic employee data collected in real time, the behavioral regularity characteristics in employee portraits are updated; Based on the updated employee portrait, extract the static feature fields involved in rule matching, compare them one by one with the static features in the employee portrait, and improve the employee portrait; Adjust rule weights based on improved employee profiles, generate new rules based on service feedback data, and update service rules; Use the updated employee portraits and service rules to verify the matching effect of the rule output and check the compatibility of the portraits and rules.

8. An employee service management system for a human resources intelligent terminal, based on the employee service management method for a human resources intelligent terminal according to any one of claims 1 to 7, characterized in that: Including, data set generation module, portrait generation module, demand prediction module, service push module, feedback collection module and rule update module; The data set generation module is used to collect employee static data and employee dynamic data through the human resources intelligent terminal and generate an employee data set; The portrait generation module is used to generate an employee portrait containing employee characteristic information based on the employee data set; The demand prediction module is used to match the rule candidate set currently corresponding to the employee through the dynamic rule engine based on the employee portrait, and predict the employee's service demand according to the matched rule candidate set; The service push module is used to generate personalized service content and send the personalized service content to the employee's human resources smart terminal; The feedback collection module is used to collect employees' feedback data on the pushed personalized services, analyze the feedback results, and optimize the matching logic of the dynamic rule engine; The rule updating module is used to update employee portraits and service rules according to the optimized dynamic rule engine.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the employee service management method for a human resources intelligent terminal according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the employee service management method for a human resources intelligent terminal according to any one of claims 1 to 7 are implemented.

Images