Intelligent salary domain management method and system, electronic equipment and storage medium

Abstract

The invention provides an intelligent salary domain management method and system, electronic equipment and a storage medium, and relates to the technical field of human resource information management, and the method comprises the steps: obtaining a structured salary group rule corresponding to an employee salary grouping request; according to the structured salary group rule, generating a plurality of candidate salary group division schemes, and screening out a target salary group division scheme with the lowest conflict rate through an SAT solver; generating a salary group snapshot and binding an employee affiliation relationship; performing risk scoring on the high-risk operation, and controlling the operation authority based on a scoring result; writing an operation event of the high-risk operation into a Hash chain type auditing structure, and periodically generating a zk-SNARK proof; and analyzing salary-related indexes in real time based on a streaming computing engine. In this way, intelligent division of salary groups, intelligent screening of salary group division schemes based on the conflict rate, dynamic risk control for high-risk operation, non-tampering operation auditing and real-time analysis of salary related indexes are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of human resource information management, and in particular to an intelligent salary domain management method and system, an electronic device and a storage medium. BACKGROUND

[0002] With the acceleration of enterprise digital transformation, the salary management system as the core module in the HRIS (Human Resource Information System) has been widely used in various large enterprises and multinational organizations. The existing salary management system generally adopts a management mode based on "wage group + static permission role", that is, the HR (Human Resources) manually configures the wage group according to the attributes of the employees such as department, post and job level, and assigns the corresponding operation permission and data access range for each wage group.

[0003] However, with the increasing complexity of enterprise organizational structure, frequent personnel flow and increasing compliance supervision requirements, the existing salary management system gradually exposes a series of problems. First, since the permission configuration depends on manual grouping, the HR needs to manually maintain a large number of rules, which is not only inefficient, but also after the adjustment of the organizational structure, it is easy to cause permission conflicts or incomplete coverage, resulting in data leakage or limited operation. Secondly, the existing system lacks a dynamic control mechanism for high-risk operations, and cannot perform real-time risk assessment and permission adjustment according to the context of operation behavior (such as time, place, operation content, user identity, etc.), thereby there is a potential security risk. In addition, the existing system relies on a relational database to record operation logs, but such logs are easy to be tampered with by system administrators or DBAs (Database Administrators), so that the authenticity of the operation records cannot be guaranteed, affecting the effectiveness of enterprise audit and compliance review.

[0004] In terms of exception handling, once the existing system encounters an exception such as the employee not matching the wage group, the operation and maintenance personnel usually need to be involved in troubleshooting, which affects the execution efficiency of the entire salary calculation process. At the same time, due to the lag of salary data analysis and display, the HR and finance departments are difficult to real-time grasp the fairness of the salary structure, the budget use, the salary adjustment trend and other key indicators, which restricts the ability of the enterprise to make data-driven decisions. Therefore, the existing salary management system has been difficult to meet the efficient, compliant and intelligent operation needs of modern enterprises for salary management in terms of flexibility, security, compliance and intelligence. SUMMARY

[0005] The present application aims to provide an intelligent salary domain management method and system, an electronic device and a storage medium to at least solve one of the above problems existing in the existing salary management system.

[0006] In a first aspect, the present application provides a smart salary domain management method, comprising: When receiving a request for grouping employee salaries, obtaining a structured salary group rule corresponding to the request for grouping employee salaries; According to the structured salary group rule, a plurality of candidate salary group division schemes are generated, and a SAT solver is used to screen a target salary group division scheme with the lowest conflict rate; wherein different candidate salary group division schemes correspond to different combinations of salary group division conditions; According to the target salary group division scheme, a salary group snapshot is generated and the employee affiliation is bound; When a high-risk operation related to the target salary group division scheme is monitored, the high-risk operation is scored, and the operation permission is controlled based on the scoring result; Write the operation event of the high-risk operation into a hash chain audit structure, and periodically generate a zk-SNARK proof; Real-time analysis of salary-related indicators based on a streaming computing engine, including one or more of salary fairness, budget utilization rate, and performance-related salary.

[0007] In an optional implementation, obtaining a structured salary group rule corresponding to the request for grouping employee salaries comprises: By combining a large language model with a preset prompt word template, the grouping statement in the request for grouping employee salaries is parsed into a rule array in JSON format; The rule array is verified for legality, and the rule array that passes the verification is determined as the structured salary group rule; wherein the legality verification includes one or more of field white list comparison, field value verification, and structure integrity verification.

[0008] In an optional implementation, the SAT solver is used to screen the target salary group division scheme with the lowest conflict rate, comprising: The affiliation of employees and salary groups in each candidate salary group division scheme is mapped to a Boolean expression; Using a SAT solver to detect conflicts of unique affiliation constraints for the Boolean expression corresponding to each candidate salary group division scheme, to obtain a feasible allocation scheme and a conflict rate; The feasible allocation scheme with the lowest conflict rate is determined as the target salary group division scheme.

[0009] In an optional implementation, the high-risk operation is scored, and the operation permission is controlled based on the scoring result, comprising: According to the operation attribute data corresponding to the high-risk operation, the scoring result corresponding to the high-risk operation is calculated using a preset risk scoring formula; wherein the operation attribute data includes one or more of operation time, data sensitivity, batch operation scale, and operator level. When the score result is greater than a preset score threshold, an authorization verification process is triggered, and a temporary token is issued after authorization.

[0010] In an optional embodiment, the operation events of high-risk operations are written into a hash chain audit structure, and a zk-SNARK proof is periodically generated, including: After the operation events of each high-risk operation are encrypted and summarized, they are stored in a block structure in chronological order; For every preset number of operation events of high-risk operations, a zk-SNARK proof is generated using SnarkJS.

[0011] In an optional embodiment, the above method further includes: Real-time monitoring of changes in employee attribute data; When changes in the attribute data of the target employee are detected, the target employee's salary group matching is performed according to the changed attribute data of the target employee.

[0012] In an optional embodiment, when changes in the attribute data of the target employee are detected, the target employee's salary group matching is performed according to the changed attribute data of the target employee, and the above method further includes: Abnormal monitoring of the salary group matching result; When an abnormal event is detected, a repair suggestion for the abnormal event is generated.

[0013] In a second aspect, the present application provides an intelligent salary domain management system, comprising: A rule acquisition module for acquiring structured salary group rules corresponding to an employee salary grouping request when the employee salary grouping request is received; A scheme generation module for generating a plurality of candidate salary group division schemes according to the structured salary group rules, and screening out a target salary group division scheme with the lowest conflict rate through a SAT solver; wherein different candidate salary group division schemes correspond to different combinations of salary group division conditions; A snapshot generation and attribution binding module for generating a salary group snapshot and binding employee attribution according to the target salary group division scheme; An operation risk control module for performing risk scoring on high-risk operations when high-risk operations related to the target salary group division scheme are detected, and controlling operation permissions based on the score result; An operation audit module for writing operation events of high-risk operations into a hash chain audit structure, and periodically generating a zk-SNARK proof; A real-time analysis display module is configured to analyze, in real time based on the stream computing engine, a compensation-related indicator, the compensation-related indicator including one or more of compensation fairness, budget usage, and performance-compensation correlation.

[0014] In a third aspect, the present application provides an electronic device, comprising a memory and a processor, the memory storing a computer program capable of running on the processor, and the processor executes the computer program to implement the method of any one of the preceding embodiments.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium, the computer-readable storage medium storing a computer program, and the computer program is run by the processor to execute the method of any one of the preceding embodiments.

[0016] The intelligent compensation domain management method, system, electronic device and storage medium provided by the present application can, when receiving a request for grouping employee compensation, obtain a structured wage group rule corresponding to the request for grouping employee compensation; generate a plurality of candidate wage group division schemes according to the structured wage group rule, and filter out a target wage group division scheme with the lowest conflict rate through a SAT solver; wherein different candidate wage group division schemes correspond to different combinations of wage group division conditions; generate a wage group snapshot and bind the employee affiliation according to the target wage group division scheme; when a high-risk operation related to the target wage group division scheme is monitored, the high-risk operation is scored, and the operation permission is controlled based on the scoring result; write the operation event of the high-risk operation into a hash chain audit structure, and periodically generate a zk-SNARK proof; analyze, in real time based on a stream computing engine, a compensation-related indicator, the compensation-related indicator including one or more of compensation fairness, budget usage, and performance-compensation correlation. In this way, the intelligent division of the wage group, the intelligent filtering of the wage group division scheme based on the conflict rate, the dynamic risk control for the high-risk operation, the tamper-proof operation audit, and the real-time analysis of the compensation-related indicators are realized, thereby meeting the efficient, compliant and intelligent operation requirements of the enterprise for compensation management to a certain extent. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0018] Figure 1 A flowchart of an intelligent compensation domain management method provided by an embodiment of the present application; Figure 2Another flowchart of an intelligent salary domain management method provided by an embodiment of the present application is shown in the figure. Figure 3 A structural diagram of an intelligent salary domain management system provided by an embodiment of the present application is shown in the figure. Figure 4 A structural diagram of an electronic device provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0019] The technical solutions of the present application will be described clearly and completely in combination with embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0020] The current salary management system based on "wage group + static permission role" has the following defects: 1. Low efficiency and easy conflict in manual grouping: HR needs to manually configure rules (department, post, job level, etc.), and the cost of manual maintenance is very high after the organizational structure changes; 2. Lack of dynamic permission control: high-risk operations (such as exporting salary data at midnight or batch salary adjustment, etc.) cannot be dynamically controlled and authorized according to the context; 3. Audit logs can be tampered with, and compliance is poor: traditional database logs can be modified or deleted by DBA, and there is a lack of verifiable operation chain; 4. Abnormal processing depends on manual operation: in the case where the employee is not matched to the wage group, the operation and maintenance personnel need to manually troubleshoot, affecting the entire salary calculation process; 5. Data insight lags: HR and financial personnel have difficulty in real-time viewing of key indicators such as salary fairness and budget usage.

[0021] Therefore, a new type of intelligent, automated, secure and compliant salary management solution is needed. Based on this, the embodiments of the present application provide an intelligent salary domain management method, system, electronic device and storage medium, which is oriented to multi-level organizations, uses natural language configuration, dynamic permission evaluation, tamper-proof audit chain and automatic remediation mechanism, and can improve the flexibility, security, compliance and intelligence of salary management, and is suitable for salary management scenarios in large organizations with multiple laws, multiple organizations and multiple roles.

[0022] The embodiments of the present application aim to achieve: Provide an intelligent salary domain division mechanism for multiple organizations and multiple roles, and support natural language generation of wage group division scheme; Realize unique matching judgment and conflict detection of employee belonging to a wage group; Introduce a dynamic permission control mechanism to automatically adjust the authorization mode according to the risk; Provide a salary chain audit structure to realize the non-tamperability and verifiability of operation records; Build a remediation function to realize automatic diagnosis and repair of salary group exceptions; Provide real-time data analysis capabilities to assist HR and financial personnel in fairness and budget management.

[0023] For the convenience of understanding the present embodiment, first, a kind of intelligent salary domain management method disclosed in the present embodiment is introduced in detail.

[0024] The present embodiment provides an intelligent salary domain management method, which can be executed by an electronic device with data processing capability. Referring to Figure 1 The flowchart of the intelligent salary domain management method is shown in the figure, which mainly includes the following steps S110 to S160: Step S110, when receiving the employee salary group request, obtain the structured salary group rule corresponding to the employee salary group request.

[0025] Administrators and other users can input oral group statement (i.e. oral group command) expressed in natural language through user interface to trigger employee salary group request. For example, the user inputs the group statement "please group the outsourcing employees in Guangzhou and Shenzhen into one salary group, and the rest according to job level" through the management background. The present embodiment can automatically parse the oral group statement in the employee salary group request into structured conditions by calling language parsing service to obtain structured salary group rule. This allows managers to express salary group requirements in natural language.

[0026] The language parsing service can use LLM (Large Language Model) to parse the group statement, and the LLM can be a general AI (Artificial Intelligence) or a self-built model. For example, OpenAI. Optionally, the LLM can be, but not limited to, a bean bag model, GPT-4, Claude or LLaMA2 local service. LLM can output a unified JSON structure, and the content can be group field and value.

[0027] In specific implementation, the front end can use forms or input boxes to collect oral text (i.e. group statement), and the back end can provide interface POST / salary-group / parse-instruction to send natural language input (i.e. group statement) into LLM, and LLM converts natural language input into structured fields.

[0028] In some possible embodiments, a large language model + prompt word template can be used to limit the output format to JSON; to further improve the usability of the structured salary group rule, a field constraint dictionary can also be added, such as allowing only preset fields for "city". Based on this, the above step S110 can include: parsing the grouping statement in the employee salary grouping request into a rule array in JSON format by using a large language model combined with a preset prompt word template; performing legality verification on the rule array, and determining the rule array that passes the verification as a structured salary group rule; wherein the legality verification includes one or more of field white list comparison, field value checking, and structure integrity checking.

[0029] The above field white list comparison refers to comparing the fields (such as "city" and "post") appearing in the rule array output by the model with the system field white list to prevent spelling errors or illegal fields. The field value checking refers to checking whether the field values (such as "Guangzhou" and "sales department") exist in the system data dictionary. For structure integrity checking, a preset JSONSchema template can be used for structure integrity checking. When the rule array fails the verification, a prompt can be returned to the user to modify or re-enter the grouping statement.

[0030] Step S120: generating a plurality of candidate salary group division schemes according to the structured salary group rule, and screening out a target salary group division scheme with the lowest conflict rate by a SAT solver.

[0031] Different candidate salary group division schemes correspond to different combinations of salary group division conditions.

[0032] According to different combinations of salary group division conditions, a plurality of different candidate salary group division schemes can be generated. A SAT (Boolean Satisfiability Problem) solver can be used to detect conflicts for each candidate salary group division scheme, and then screen out a target salary group division scheme with the lowest conflict rate.

[0033] In some possible embodiments, the step S120 of screening out a target salary group division scheme with the lowest conflict rate by a SAT solver can include: mapping the affiliation relationship between employees and salary groups in each candidate salary group division scheme into a Boolean expression; using a SAT solver to detect conflicts of the unique affiliation constraint corresponding to each candidate salary group division scheme, to obtain a feasible allocation scheme and a conflict rate; and determining the feasible allocation scheme with the lowest conflict rate as the target salary group division scheme.

[0034] Map the salary group condition and all employee attribution into a Boolean expression, ensuring that each employee is uniquely attributed to a certain salary group; use the SAT solver to filter the solution with the lowest conflict rate. Specifically, the SAT solver can use CP-SAT Solver (Google OR-Tools) or MiniSAT, where CP-SAT Solver is a constraint programming solver in Google OR-Tools that combines the advantages of constraint programming and Boolean satisfiability techniques, capable of handling complex combinatorial optimization problems; MiniSAT is a lightweight but efficient open-source SAT solver. Each employee constructs a Boolean variable x ij , meaning "employee i belongs to the salary group j "; add the constraint "one person can only belong to one group": , output the legal solution (i.e., the feasible allocation scheme with conflicting employees removed) and the conflict rate.

[0035] Step S130, according to the target salary group division scheme, generate salary group snapshot and bind employee attribution.

[0036] Select the optimal solution to generate the salary group snapshot snapshot_id (snapshot ID), write it into the salary_group and emp_group_map table (i.e., the salary group and employee group mapping table), and complete the employee attribution binding.

[0037] Step S140, when monitoring high-risk operations related to the target salary group division scheme, risk score the high-risk operations, and control the operation authority based on the scoring results.

[0038] High-risk behavior (such as exporting high-level salary at midnight) can trigger the risk control engine to calculate the risk score, and if the score exceeds the preset score threshold, secondary verification or approval authorization is required.

[0039] In some possible embodiments, the step S140 can include: calculating a score result corresponding to the high-risk operation according to operation attribute data corresponding to the high-risk operation by using a preset risk scoring formula; wherein the operation attribute data includes one or more of operation time, data sensitivity, batch operation scale, and operator level; triggering an authorization verification process when the score result is greater than a preset score threshold; and issuing a temporary token after authorization passes; wherein the authorization verification process includes a secondary verification process or an approval process. The temporary token is only valid within a preset time period, and such a ZSP (Zero-Staying Permission) can realize zero-staying permission control and ensure security. Zero-staying permission control aims to minimize the permissions possessed by an application or a user to reduce potential security risks. The preset time period can be set according to actual needs, which is not limited here, for example, the preset time period is 30 minutes.

[0040] In a possible implementation manner, the risk scoring formula can be as follows: ; wherein, represents an operation time factor, such as a higher weight at night; represents data sensitivity, such as a data sensitivity of 1.0 for the salary of a high-ranking official; represents a batch operation scale, such as high when >100 people; represents a user historical trust score, and the initial value can be determined based on the operator level; represents a weight, which supports dynamic configuration, i =1, 2, 3, 4, .

[0041] When the score result is greater than a preset score threshold, an approval module can be called, or two-factor authentication can be required; after the two-factor authentication passes, the user can be allowed to perform the high-risk operation, or a temporary token can be issued to allow the user to perform the high-risk operation within a preset time period; when the two-factor authentication fails, the user is prohibited from performing the high-risk operation; after the approval process passes, a temporary token can be issued to allow the user to perform the high-risk operation within a preset time period; when the approval process fails, the user is prohibited from performing the high-risk operation. When the score result is less than or equal to the preset score threshold, the user is allowed to perform the high-risk operation. The score threshold can be set according to actual needs, which is not limited here. R R When the score result is greater than a preset score threshold, an approval module can be called, or two-factor authentication can be required; after the two-factor authentication passes, the user can be allowed to perform the high-risk operation, or a temporary token can be issued to allow the user to perform the high-risk operation within a preset time period; when the two-factor authentication fails, the user is prohibited from performing the high-risk operation; after the approval process passes, a temporary token can be issued to allow the user to perform the high-risk operation within a preset time period; when the approval process fails, the user is prohibited from performing the high-risk operation. When the score result is less than or equal to the preset score threshold, the user is allowed to perform the high-risk operation. The score threshold can be set according to actual needs, which is not limited here.

[0042] In step S150, the operation event of the high-risk operation is written into a hash chain audit structure, and a zk-SNARK proof is periodically generated.

[0043] ​The high-risk operation event can be encapsulated as an unforgeable data block, written into an event chain through a hash chain structure, and periodically generate a zk-SNARK (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge) proof for audit verification.

[0044] In some possible embodiments, the above step S150 can include: after encrypting the operation event of each high-risk operation, storing the operation event in a block structure in chronological order; and for every preset number of operation events of high-risk operations, generating a zk-SNARK proof using SnarkJS. The preset number can be set according to actual needs, which is not limited herein, for example, the preset number is 1000. SnarkJS is a JavaScript library for generating and verifying zero-knowledge proofs.

[0045] In a specific implementation, after encrypting the operation event of each operation (for example, SHA-256 or Keccak256), the operation event is chained in chronological order to form a block Block(prev_hash, merkle_root, payload); a zk-SNARK zero-knowledge proof is generated every 1000 times, for an auditor to verify that the data has not been modified. The prev_hash refers to the hash value of the previous block; the merkle_root is the root hash of a Merkle tree constructed based on the hash values of all events in the block; and the payload is the payload of the block, including the actual event information.

[0046] In step S160, a streaming computing engine is used to analyze, in real time, a salary-related indicator, which includes one or more of salary fairness, budget utilization rate, and performance-salary correlation.

[0047] The Kafka stream processing + OLAP (Online Analytical Processing) engine can be used to construct real-time salary fairness, budget utilization rate, and performance-salary correlation indicators, and the indicators can be visualized and displayed.

[0048] In a specific implementation, the salary group event stream is written to Kafka, Apache Flink SQL is used to calculate the indicators in real time, and the data is written to an OLAP engine (for example, Pinot or ClickHouse). A front-end dashboard can display charts, such as salary difference, Gini coefficient, and budget occupancy rate. The Gini coefficient, also known as the Gini coefficient or Gini coefficient, has a value between 0 and 1, where 0 represents absolute fairness and 1 represents absolute unfairness.

[0049] The intelligent salary domain management method provided by the embodiment of the present application realizes intelligent division of a wage group, intelligent screening of a wage group division scheme based on a conflict rate, dynamic risk control for high-risk operations, unalterable operation auditing, and real-time analysis of salary-related indicators, thereby meeting the efficient, compliant, and intelligent operation requirements of enterprises on salary management to a certain extent.

[0050] Further, when the post, position, city, etc. of the employee changes, the re-matching logic can be automatically triggered. Based on this, the above method further includes: monitoring the changes of the attribute data of the employee in real time; when it is monitored that the attribute data of the target employee changes, performing the wage group matching of the target employee according to the changed attribute data of the target employee. The attribute data of the employee can include post, position, city, etc. If the target employee fails to match (i.e., matches 0 wage groups) or matches multiple wage groups, it can be marked as an exception and written into an exception table.

[0051] Further, the above method further includes: performing exception monitoring on the wage group matching result; when an exception event is monitored, generating a repair suggestion for the exception event.

[0052] The exception event can be sent to a remediation agent module to analyze the exception reason and generate a repair suggestion (such as a field supplement or re-grouping), and an administrator can repair it with one key. The exception event can be monitored through Kafka, a diagnostic model such as ChatGPT can be called to analyze the exception type (such as missing fields or contradictory conditions), and a repair suggestion of SQL statement and / or configuration adjustment (i.e., rule adjustment) can be automatically generated; the embodiment also provides a one-key execution interface, and the administrator can click to execute or perform approval to complete the repair.

[0053] For ease of understanding, the embodiment of the present application also provides a specific implementation scheme of the above intelligent salary domain management method. Referring to Figure 2 Another flowchart of an intelligent salary domain management method is shown in FIG. 2, which includes the following steps S201 to S210: Step S201, receiving a natural language input.

[0054] An example input can be "Please divide the outsourcing employees in Guangzhou and Shenzhen into one wage group, and the rest into groups according to the position".

[0055] Step S202, parsing the natural language input into a structured rule.

[0056] A preset prompt word template can be used to issue a parsing request to a natural language understanding model. An example of the prompt word template is as follows: "Please parse the following grouping instruction into a standard JSON format rule array, each rule contains group_name and several field matching conditions. For example, city, post, employment form, etc. The output format is as follows: [ {"group_name": "xxx", "city": ["xxx"], "employment type": ["xxx"]}, ... ] Grouping requirement: {User input text} Here is an example of the structured rules (JSON) returned by the model: [ { "group_name": "South China Outsourcing Group", "City": ["Guangzhou", "Shenzhen"], Employment type: ["Outsourcing"] }, { "group_name": "Other job level groups", Rule: Automatically group by job level } ] After the model output is returned, the following processing flow will be executed: 1. Field dictionary validation: Compare fields appearing in the JSON (such as "city" and "job title") with the system's field whitelist to prevent spelling errors or illegal fields; 2. Field value validation: Verify whether field values ​​(such as "Guangzhou" or "Sales Department") exist in the system's data dictionary; 3. Structure Validation: Perform structural integrity verification using a preset JSON Schema template; 4. Result Display: Rules with correct structure (i.e., rules that have passed verification) can be directly displayed as a group preview; if there are any anomalies (i.e., verification failed), the administrator can be prompted to correct or re-enter the results.

[0057] Step S203: Generate multiple candidate salary group division schemes.

[0058] Multiple candidate salary group partitioning schemes can be generated based on structured rules, each representing a different combination of salary group partitioning conditions.

[0059] By constructing a list of solutions through rule combinations, it can support: multi-dimensional condition combinations (such as city × job level), special combination methods (cross grouping, condition coverage, etc.), and each solution includes a mapping relationship table.

[0060] Step S204: Detect conflicts in candidate wage group partitioning schemes using the SAT solver.

[0061] A Boolean constraint condition that "an employee can only belong to one salary group" can be constructed, and the SAT solver is used to detect whether there is a conflict and calculate the conflict rate.

[0062] The algorithm variables are defined as follows: the employee set is ; the salary group set is ; the proposition variable is , indicating that the employee e i belongs to the salary group g j .

[0063] The constraint formula is as follows: 1. Unique belonging constraint (each employee can only belong to one group): ; 2. If the employee e i does not meet the condition of g j , then x ij =0.

[0064] The SAT CNF (Conjunctive Normal Form) conversion idea is as follows: Convert the above unique belonging constraint to CNF: for each pair of x ij , x ik , add a mutual exclusion term: ; All x ij constitute a 0-1 Boolean variable input SAT solver (such as MiniSAT, CP-SAT Solver).

[0065] The output is: a feasible allocation scheme (each employee is uniquely matched to a group) and the conflict rate; where the conflict rate = the number of conflicting employees / the total number of employees.

[0066] Step S205, generate a salary group snapshot and write it to the database.

[0067] The feasible allocation scheme with the lowest conflict rate is taken as the target salary group division scheme, and its salary group rules are archived in the form of a snapshot, and the employees are bound to unique ownership.

[0068] The data structure design is introduced as follows.

[0069] 1. An example of the snapshot structure is as follows: CREATE TABLE salary_group_snapshot ( snapshot_id VARCHAR PRIMARY KEY, group_rules JSONB, created_at TIMESTAMP ); 2. Employee affiliation binding example as follows: CREATE TABLE emp_group_map ( emp_id BIGINT, group_id BIGINT, snapshot_id VARCHAR, UNIQUE(emp_id) ); 3. Process description: a. For the SAT output of the employee salary group affiliation result, employees assigned as x ij =1 (i.e. only matching successful employees), write to emp_group_map (employee group mapping).

[0070] b. If the employee affiliation fails (multiple group matches / zero group matches, i.e. the employee matches 0 or multiple salary groups), write to the exception table group_conflict_log (group conflict log).

[0071] Step S206, when the employee attributes change, automatically re-match the salary group.

[0072] When the employee's position / seniority / city changes, the re-matching logic is automatically triggered. If the matching fails or multiple groups are matched, it is marked as an exception.

[0073] Specifically, real-time monitoring of data updates can be performed using Kafka or Webhook; change events trigger the matching engine to perform logical matching; if the matching result is multiple salary groups or zero salary groups, write to the exception table group_match_error.

[0074] Step S207, after matching fails, trigger exception repair.

[0075] When an exception such as an employee unable to match a salary group or a missing field occurs, a repair suggestion is automatically generated.

[0076] The processing flow can be: 1. Monitor Topic group_match_error in Kafka; 2. Call the diagnostic model (ChatGPT function-calling) to analyze the exception type; 3. Output repair suggestions: such as generating recommended SQL for missing fields, or generating recommended modified rules for rule conflicts; 4. Provide an "Immediate Repair" button that is pressed, and the background executes the script.

[0077] The example output is as follows: { "Exception Type": "Field Missing", "Suggested Repair": "Set employee city to 'Shanghai'", "SQL": "UPDATE employee SET city='Shanghai' WHERE emp_id=10088;" } After the administrator clicks "One-Click Repair", the recommended SQL can be automatically executed and the matching can be re-triggered.

[0078] Step S208, when performing high-risk operations related to the salary group division scheme, trigger the risk control engine.

[0079] When performing sensitive operations, you can dynamically decide whether to allow the operation, whether to require secondary verification or approval, based on the risk level. The risk level can be a score result calculated based on a risk scoring formula. Before the user performs a high-risk operation, you can call the risk calculation module (based on time, data sensitivity, operator level, etc.) to calculate the score result.

[0080] The strategy control logic can be: if the score result R is less than or equal to 0.5, then directly allow the high-risk operation. If the score result R value meets 0.5 R ≤0.8, then 2FA (Two-Factor Authentication) is required, and after authentication, the high-risk operation can be allowed, otherwise it is prohibited. If the score result R value is greater than 0.8, the supervisor's approval is required, and after approval, a temporary token with a validity period of 30 minutes can be issued, otherwise the high-risk operation is prohibited.

[0081] Control operation permissions based on score results, no need for long-term role authorization, consistent with the principle of least privilege.

[0082] Step S209, write to the salary chain audit structure.

[0083] The salary chain audit structure, referred to as the audit chain, includes event evidence and zk-SNARK proof, which is used to record all key salary operation behaviors and prevent tampering, supporting audit verification. The audit chain structure and zk-SNARK proof are introduced below.

[0084] An example of the record structure of each event is as follows: { "event_id": "ev-1011", "emp_id": 10088, "action": "EXPORT", "target": "high-level salary", "operator": "Zhang San", "timestamp": "2025-07-05T01:25:00", "payload_hash": "keccak256(...)" } An example of the event block structure (i.e., the structure of the block) is as follows: { "block_id": 413, "prev_hash": "fd93a...", "merkle_root": "ce80b...", "events": [...], "signature": "platform private key signature" } Generation of zk-SNARK proof: every 1000 events form a batch, and SnarkJS is used to generate zero-knowledge proof.

[0085] The audit party can confirm the integrity of the data by verifying the zk-SNARK proof and the block hash.

[0086] Step S210, real-time index calculation and display.

[0087] Key indicators such as internal salary fairness, budget usage, and performance-related salary can be calculated in real time. The following indicators are aggregated in real time through Kafka→Flink SQL→Apache Pinot: Gini coefficient of fairness; Department budget usage rate = actual expenditure / department budget upper limit; Performance-related salary, which can use Pearson coefficient wherein, X , Y respectively represent performance, salary, σ X , σ Y respectively represent X , Y the standard deviation of.

[0088] It can be displayed by a front-end dashboard, supporting multi-dimensional filtering (departments, job levels and time periods, etc.).

[0089] The embodiment of the application realizes the intelligent division of the salary group, conflict detection, dynamic risk control, non-tamperable operation audit, automatic exception repair and real-time index analysis and other capabilities.

[0090] Corresponding to the above-mentioned intelligent salary domain management method, the embodiment of the application also provides an intelligent salary domain management system. Referring to Figure 3 The structure diagram of an intelligent salary domain management system is shown, the system comprises: The rule acquisition module 301 is used to acquire the structured salary group rule corresponding to the employee salary group division request when receiving the employee salary group division request; The scheme generation module 302 is used to generate a plurality of candidate salary group division schemes according to the structured salary group rule, and filter out the target salary group division scheme with the lowest conflict rate through the SAT solver; wherein different candidate salary group division schemes correspond to different salary group division condition combinations; The snapshot generation and attribution binding module 303 is used to generate the salary group snapshot and bind the employee attribution according to the target salary group division scheme; The operation risk control module 304 is used to perform risk scoring on the high-risk operation when monitoring the high-risk operation related to the target salary group division scheme, and control the operation permission based on the scoring result; The operation audit module 305 is used to write the operation event of the high-risk operation into the hash chain audit structure, and periodically generate the zk-SNARK proof; The real-time analysis and display module 306 is used to analyze the salary-related indicators in real time based on the stream computing engine, and the salary-related indicators include one or more of the salary fairness, the budget usage rate and the performance-salary correlation.

[0091] The intelligent salary domain management system provided by the embodiment of the application realizes the intelligent division of the salary group, the intelligent screening of the salary group division scheme based on the conflict rate, the dynamic risk control for the high-risk operation, the non-tamperable operation audit and the real-time analysis of the salary-related indicators, thereby meeting the efficient, compliant and intelligent operation requirements of the enterprise on the salary management to a certain extent.

[0092] Further, the rule obtaining module 301 is specifically configured to: parse the grouping statement in the employee salary grouping request into a rule array in JSON format by combining a large language model with a preset prompt word template; perform legality verification on the rule array, and determine the rule array that passes the verification as a structured wage group rule; wherein the legality verification includes one or more of field white list comparison, field value verification, and structure integrity verification.

[0093] Further, the scheme generation module 302 is specifically configured to: map the attribution relationship between employees and wage groups in each candidate wage group division scheme to a Boolean expression; use a SAT solver to perform conflict detection on the Boolean expression corresponding to each candidate wage group division scheme under the constraint of unique attribution, to obtain a feasible allocation scheme and a conflict rate; and determine the feasible allocation scheme with the lowest conflict rate as the target wage group division scheme.

[0094] Further, the operation risk control module 304 is specifically configured to: calculate a score result corresponding to a high-risk operation according to operation attribute data corresponding to the high-risk operation using a preset risk scoring formula; wherein the operation attribute data includes one or more of operation time, data sensitivity, batch operation scale, and operator level; and when the score result is greater than a preset score threshold, trigger an authorization verification process, and issue a temporary token after authorization.

[0095] Further, the operation audit module 305 is specifically configured to: chain store operation events of each high-risk operation in a block structure in chronological order after encrypting the operation events; and for every preset number of operation events of high-risk operations, generate a zk-SNARK proof using SnarkJS.

[0096] Further, the system further comprises: A change monitoring and matching module is configured to monitor changes in employee attribute data in real time, and perform wage group matching for a target employee according to the changed attribute data of the target employee when a change in the attribute data of the target employee is monitored.

[0097] Further, the system further comprises: An abnormality monitoring and repairing module is configured to monitor the wage group matching result for abnormalities, and generate a repair suggestion for an abnormal event when the abnormal event is monitored.

[0098] The intelligent salary domain management system provided in the embodiment has the same implementation principle and technical effects as the foregoing intelligent salary domain management method embodiment. For brevity, reference can be made to the corresponding content in the foregoing intelligent salary domain management method embodiment for parts not mentioned in the intelligent salary domain management system embodiment.

[0099] AsFigure 4 As shown, the electronic device 400 provided by the embodiment of the present application comprises a processor 401, a memory 402 and a bus, the memory 402 stores a computer program which can run on the processor 401, the processor 401 and the memory 402 communicate through the bus when the electronic device 400 runs, and the processor 401 executes the computer program to realize the smart salary domain management method described above.

[0100] Specifically, the memory 402 and the processor 401 can be general memory and processor, which are not specifically limited here.

[0101] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is run by a processor to execute the smart salary domain management method in the foregoing method embodiment. The computer readable storage medium comprises a U disk, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk and various storage program codes.

[0102] The term "and / or" in the present document is only used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. In addition, the term "at least one" in the present document means any one of the plurality or any combination of at least two of the plurality, for example, at least one of A, B and C includes any one or more elements selected from the set consisting of A, B and C.

[0103] In all the examples shown and described herein, any specific value should be interpreted as merely exemplary and not as a limitation, therefore, other examples of the exemplary embodiments can have different values.

[0104] The flow diagrams and the block diagrams in the drawings are presented to illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks can sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flow diagrams, and combinations thereof, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

[0105] In several embodiments provided in the present application, it should be understood that the disclosed system and method can be implemented in other manners. The embodiments described above are merely schematic, and the division of the modules is merely logical function division, and there can be other division manners in actual implementation. For example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, and can be in electrical, mechanical or other forms.

[0106] The modules illustrated as separate components may or can not be physically separate, and the components illustrated as modules may or can not be physical modules, i.e., may be located in one place, or may be distributed on a plurality of network modules. Some or all of the modules can be selected according to actual needs to achieve the purposes of the embodiments.

[0107] In addition, the functional modules in each of the embodiments of the present application can be integrated into a processing module, or each module can exist physically, or two or more modules can be integrated into one module.

[0108] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for intelligent payroll domain management, characterized in that, The method comprises the following steps: When receiving an employee salary grouping request, obtain the structured salary group rule corresponding to the employee salary grouping request; According to the structured salary group rule, generate a plurality of candidate salary group division schemes, and screen out a target salary group division scheme with the lowest conflict rate through a SAT solver; wherein different candidate salary group division schemes correspond to different combinations of salary group division conditions; According to the target salary group division scheme, generate a salary group snapshot and bind the employee attribution relationship; When a high-risk operation related to the target salary group division scheme is monitored, risk score the high-risk operation, and control the operation authority based on the scoring result; Write the operation event of the high-risk operation into a hash chain audit structure, and periodically generate a zk-SNARK proof; Based on the stream computing engine, analyze the salary-related indicators in real time, including one or more of salary fairness, budget utilization rate, and performance-related salary.

2. The method of claim 1, wherein, The method comprises the following steps: By combining a large language model with a preset prompt word template, parse the grouping statement in the employee salary grouping request into a rule array in JSON format; Perform legality verification on the rule array, and determine the rule array that passes the verification as the structured salary group rule; wherein the legality verification includes one or more of field white list comparison, field value verification, and structure integrity verification.

3. The method of claim 1, wherein, The method comprises the following steps: Map the attribution relationship between employees and salary groups in each candidate salary group division scheme to a Boolean expression; Use a SAT solver to detect conflicts in the Boolean expression corresponding to each candidate salary group division scheme with unique attribution constraints to obtain a feasible allocation scheme and a conflict rate; Determine the feasible allocation scheme with the lowest conflict rate as the target salary group division scheme.

4. The method of claim 1, wherein, The method comprises the following steps: According to the operation attribute data corresponding to the high-risk operation, calculate the scoring result corresponding to the high-risk operation using a preset risk scoring formula; wherein the operation attribute data includes one or more of operation time, data sensitivity, batch operation scale, and operator level; When the scoring result is greater than a preset scoring threshold, trigger an authorization verification process, and issue a temporary token after authorization; wherein the authorization verification process includes a secondary verification process or an approval process.

5. The method of claim 1, wherein, The method comprises the following steps: After encrypting the operation event of each high-risk operation, store it in a block structure in chronological order; For every preset number of operation events of the high-risk operation, use SnarkJS to generate a zk-SNARK proof.

6. The method of claim 1, wherein, The method further comprises the following steps: Monitor the changes in employee attribute data in real time; When it is monitored that the attribute data of the target employee changes, the salary group matching of the target employee is performed according to the changed attribute data of the target employee.

7. The method of claim 6, wherein, When it is monitored that the attribute data of the target employee changes, the salary group matching of the target employee is performed according to the changed attribute data of the target employee, and the method further comprises: Abnormality monitoring is performed on the salary group matching result. When an abnormal event is monitored, a repair suggestion for the abnormal event is generated.

8. An intelligent compensation domain management system, characterized by, Comprise: A rule acquisition module is configured to acquire a structured salary group rule corresponding to an employee salary grouping request when the employee salary grouping request is received; A scheme generation module is configured to generate a plurality of candidate salary group division schemes according to the structured salary group rule, and to filter out a target salary group division scheme with the lowest conflict rate through a SAT solver; wherein different candidate salary group division schemes correspond to different salary group division condition combinations; A snapshot generation and attribution binding module is configured to generate a salary group snapshot and bind an employee attribution relationship according to the target salary group division scheme; An operation risk control module is configured to perform risk scoring on a high-risk operation related to the target salary group division scheme when the high-risk operation is monitored, and to control operation authority based on the scoring result; An operation audit module is configured to write operation events of the high-risk operation into a hash chain audit structure, and to periodically generate a zk-SNARK proof; A real-time analysis display module is configured to perform real-time analysis on salary-related indicators based on a stream computing engine, the salary-related indicators including one or more of salary fairness, budget usage rate, and performance-salary correlation.

9. An electronic device comprising a memory, a processor, the memory having stored therein a computer program executable on the processor, characterized in that, The processor executes the computer program to implement the method of any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-7.

Images