Method, device, equipment, medium and program for recommending employment combination scheme

Abstract

The invention relates to the technical field of computers, in particular to a employment combination scheme recommendation method and device, equipment, a medium and a programmer.The method comprises the steps that at least one type of internal data and external data are obtained; generating an enterprise portrait vector representing enterprise attributes according to the internal data, and generating a market situation vector representing a labor market environment according to the external data; according to the enterprise portrait vector and the market situation vector, calculating a comprehensive benefit evaluation value of at least one target post in various different employment forms; and constructing a multi-target optimization model by taking the enterprise employment structure as an optimization object, inputting the comprehensive benefit evaluation values of different employment forms of each post into the multi-target optimization model, and outputting at least one recommended employment combination scheme by the multi-target optimization model. Therefore, the problems that in the related technology, human resource planning only depends on internal historical data, the model is single, the employment recommendation result is one-sided, and dynamic business requirements cannot be responded are solved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, device, medium and program for recommending a workforce combination scheme. Background Technology

[0002] With the rapid development of the digital economy and the diversification of corporate employment models, more and more companies are adopting hybrid employment structures that include full-time employees, part-time staff, outsourced services, gig workers, and expert consultants to improve organizational agility, control labor costs, and cope with business fluctuations. However, companies currently face core challenges in developing and optimizing such complex employment structures, including subjective decision-making, static solutions, and fragmented approaches.

[0003] In related technologies, existing mainstream human resource management systems or human resource information systems mainly focus on the transactional management of full-time employees, such as organizational structure maintenance, attendance, payroll, and recruitment process automation. However, their analytical models are usually based on simple extrapolation or budget allocation of internal historical data, which has obvious limitations: on the one hand, these models rarely incorporate refined cost and risk quantification of flexible employment forms (such as part-time, outsourcing, and gig work); on the other hand, they lack the ability to perceive the dynamics of the external labor market (such as changes in regional talent supply and demand, fluctuations in salary levels, and adjustments in policies and regulations) in real time, resulting in output planning solutions that are essentially static and lagging, making it difficult to support enterprises' forward-looking decisions in a rapidly changing environment. Summary of the Invention

[0004] This application provides a method, apparatus, equipment, medium, and program for recommending employment combination schemes, in order to solve the problems in related technologies where human resource planning relies solely on internal historical data and has a single model, resulting in one-sided employment recommendation results that cannot respond to dynamic business needs.

[0005] The first aspect of this application provides a method for recommending employment combination schemes, comprising the following steps: acquiring at least one type of internal data and external data, wherein the internal data comes from internal employment data sources of the enterprise, and the external data comes from market employment data sources; generating an enterprise profile vector representing enterprise attributes based on the internal data, and generating a market situation vector representing the labor market environment based on the external data; calculating a comprehensive benefit evaluation value for at least one target position under multiple different employment forms based on the enterprise profile vector and the market situation vector; constructing a multi-objective optimization model with the enterprise employment structure as the optimization object, wherein the optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility; inputting the comprehensive benefit evaluation value of different employment forms for each position into the multi-objective optimization model, and the multi-objective optimization model outputting at least one recommended employment combination scheme.

[0006] Optionally, the comprehensive benefit evaluation value includes a comprehensive cost value, a risk assessment value, and an agility value; wherein, calculating the comprehensive benefit evaluation value of the target position under multiple different employment forms includes: obtaining the explicit cost of the market situation vector and the operational efficiency characteristics of the enterprise profile vector, wherein the explicit cost includes salary or service quotation data; calculating the corresponding implicit cost based on the operational efficiency characteristics of the enterprise profile vector; and calculating the comprehensive cost value of the target position under multiple different employment forms based on the explicit cost and the implicit cost.

[0007] Optionally, the calculation of the comprehensive benefit evaluation value of the target position under multiple different employment forms also includes: obtaining the target risk factors of the enterprise profile vector and the market situation vector, wherein the target risk factors include turnover rate, project complexity and data security risk; calculating the corresponding risk weights based on the target risk factors of the enterprise profile vector and the market situation vector; and calculating the agility value of the target position under multiple different employment forms based on the target risk factors and the corresponding risk weights.

[0008] Optionally, the employment form includes at least one of full-time, part-time, outsourcing and gig work. The calculation of the comprehensive benefit evaluation value of the target position under multiple different employment forms also includes: obtaining the recruitment cycle and average arrival time in multiple different employment forms; and calculating the agility value of the target position under multiple different employment forms based on the recruitment cycle and the average arrival time.

[0009] Optionally, the processing method of the multi-objective optimization model is as follows: the employment form selection and number configuration of each position in the target position set are encoded as population individuals, and the population is iteratively evolved based on the multi-objective optimization algorithm; in each generation of the population, the individuals are selected based on the non-dominated ranking and crowding distance, and the individuals with the best trade-off among the optimization objectives are retained; a Pareto optimal solution set is generated based on the individuals with the best trade-off among the optimization objectives in each generation of the population, where each solution corresponds to a global employment structure scheme that achieves different trade-offs among multiple optimization objectives.

[0010] Optionally, it also includes: receiving user feedback on adjustments to the recommended workforce combination scheme, and / or collecting actual operational data after the workforce combination scheme is implemented; and optimizing the multi-objective optimization model based on the adjustment feedback and / or the actual operational data.

[0011] A second aspect of this application provides an apparatus for recommending employment combination schemes, comprising: an acquisition module for acquiring at least one type of internal data and external data, wherein the internal data comes from internal employment data sources of an enterprise, and the external data comes from market employment data sources; a generation module for generating an enterprise profile vector representing enterprise attributes based on the internal data, and generating a market situation vector representing the labor market environment based on the external data; a calculation module for calculating a comprehensive benefit evaluation value for at least one target position using multiple different employment forms based on the enterprise profile vector and the market situation vector; and a processing module for constructing a multi-objective optimization model with the enterprise employment structure as the optimization object, wherein the optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility, and inputting the comprehensive benefit evaluation value of different employment forms for each position into the multi-objective optimization model, and the multi-objective optimization model outputting at least one recommended employment combination scheme.

[0012] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to perform the employment combination scheme recommendation method as described in the above embodiments.

[0013] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to perform the employment combination scheme recommendation method as described in the above embodiments.

[0014] A fifth aspect of this application provides a computer program product, including a computer program or instructions, which, when executed, implement the employment combination scheme recommendation method as described in the above embodiments.

[0015] Therefore, this application has at least the following beneficial effects: This application embodiment can integrate internal enterprise employment data with external market employment data to generate an enterprise profile vector reflecting enterprise attributes and a market situation vector reflecting the labor market environment. Based on these two types of vectors, the comprehensive benefit evaluation value of the target position under various employment forms is calculated. Thus, a multi-objective optimization model is constructed with the goal of minimizing total cost, minimizing total risk, or maximizing total agility. This model recommends the optimal employment combination scheme for each position based on the comprehensive benefit evaluation value, ultimately achieving more scientific and reasonable enterprise employment structure decision support, enhancing the enterprise's competitiveness and adaptability in a complex and ever-changing market environment while reducing operating costs and risks.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart illustrating a method for recommending a workforce combination scheme according to an embodiment of this application; Figure 2 This is a flowchart illustrating a method for recommending employment combination schemes according to embodiments of this application. Figure 3 This is a block diagram of a labor combination scheme recommendation device provided according to an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0019] Currently, companies primarily rely on two fragmented systems when allocating human resources: traditional HR management software, whose core functions focus on information management, attendance and payroll, and recruitment process automation for full-time employees; and various independent gig / part-time platforms, outsourcing platforms, or expert consulting platforms, which only serve specific types of flexible employment scenarios. Neither of these systems operates independently and provides a comprehensive planning and optimization solution that can bridge and coordinate multiple employment models.

[0020] Some advanced HR software on the market offers human resource planning or human resource cost analysis modules, forming existing solutions most similar to this invention. However, these modules typically rely solely on historical full-time employee data within the company for simple trend forecasting or budgeting, exhibiting fundamental limitations: the analysis models are primarily geared towards full-time employees, rarely incorporating quantitative models for refined cost and risk assessment of flexible employment forms such as part-time, outsourcing, and gig work; they lack the ability to perceive and input real-time, dynamic external labor market data (such as regional salary levels, talent supply and demand, and policies and regulations); and they cannot automatically generate structured, optimized mixed employment combination recommendations based on multi-source data and complex models, requiring significant manual intervention and subjective judgment from managers in the decision-making process.

[0021] Because existing planning tools rely solely on internal historical data and employ a single model (centered on full-time employees), they suffer from a dual deficiency: they cannot perceive dynamic changes in the external market environment, nor can they accurately model and compare diverse flexible employment models. Consequently, the planning results they produce are inevitably static and one-sided, failing to respond to rapidly changing business needs and unable to achieve optimal balance and control over overall human resource costs, risks, and agility at a global level.

[0022] This application constructs a comprehensive data perception system that can simultaneously integrate the internal state of an enterprise with the external market environment: it establishes a unified, computable model for accurately assessing and comparing the costs, risks, and agility benefits of using various employment forms for different positions. Based on the above data and model, it automatically seeks optimization through intelligent algorithms and outputs a data-driven, dynamically adjustable, globally optimal employment combination recommendation scheme, thereby elevating human resource allocation from experience-based decision-making to data-intelligent decision-making.

[0023] The following description, with reference to the accompanying drawings, outlines a method, apparatus, device, medium, and procedure for recommending labor combination schemes according to embodiments of this application.

[0024] Specifically, Figure 1 This is a flowchart illustrating a method for recommending a workforce combination scheme provided in an embodiment of this application.

[0025] like Figure 1 As shown, the recommended method for this employment combination scheme includes the following steps: In step S101, at least one type of internal data and external data are acquired, wherein the internal data comes from the enterprise's internal employment data source, and the external data comes from the market employment data source.

[0026] It is understood that the embodiments of this application can deeply explore the relationship between enterprise employment patterns and market trends by integrating internal enterprise employment data and market employment data, providing enterprises with accurate market positioning and employment decision support, enhancing the enterprise's competitiveness in the industry, effectively responding to market changes, optimizing human resource allocation and cost control, thereby promoting sustainable development while maintaining the enterprise's flexibility and agility.

[0027] It should be noted that internal data comes from the company's internal employment data sources, such as data from the company's own operations and management systems. This data is used to characterize the company's internal features, current status, and historical behavior, and mainly includes: (1) Basic enterprise information and organizational data: industry, size (number of employees, revenue), development stage (startup, growth, maturity), organizational structure, department setup, and job system. (2) Historical and current employment data: current and historical employment forms for each position (full-time, part-time, outsourcing, etc.), employee compensation and benefits structure, historical turnover rate, recruitment cycle data, and details of human resource costs (salary, benefits, recruitment fees, management fees, etc.). (3) Business operation and efficiency data: seasonal fluctuations in business (such as e-commerce promotional seasons), project complexity, data security level requirements, maturity of the enterprise's outsourcing management, and internal process efficiency.

[0028] External data refers to labor market and macroeconomic environment data from outside the enterprise, used to understand the external environment and constraints of the enterprise. It mainly includes: (1) Labor market supply and demand and cost data: salary levels of positions in the same region and industry (such as quantiles and medians), resume density on talent platforms (such as No.1 Zhipin and Kuaijiejian), number of active job seekers, hourly wages of flexible employment (part-time and casual work), project quotations, and service rates. (2) Talent mobility and efficiency data: average recruitment cycle, arrival time, order response time for casual / part-time workers, average delivery quality score, industry average turnover rate, and talent mobility trends for each position. (3) Policy, regulations and compliance data: national and local labor laws and regulations, social security policies, tax and compliance requirements related to flexible employment. (4) Macroeconomic and industry insight data: industry white papers, salary survey reports, regional economic development data, and competitive situation information.

[0029] In step S102, an enterprise profile vector representing enterprise attributes is generated based on internal data, and a market situation vector representing the labor market environment is generated based on external data.

[0030] It is understood that the embodiments of this application can construct enterprise profile vectors through internal data to accurately depict the attribute characteristics of enterprises in terms of scale, industry, development stage and employment preferences. At the same time, a market situation vector is generated based on external data to dynamically reflect key environmental factors such as the supply and demand relationship of the labor market, salary levels, compliance requirements and the availability of flexible employment. The combination of the two provides a high-dimensional, quantitative and context-aware data foundation for subsequent employment form assessment and structure optimization, which significantly improves the scientificity and adaptability of human resource decision-making.

[0031] It should be noted that enterprise attributes refer to a set of structured information that can objectively describe and quantify an enterprise's inherent characteristics, state, capabilities, and constraints. It is a collection of static and dynamic features extracted from internal enterprise data to characterize the enterprise's identity and state. These mainly include the following dimensions: Basic attributes: industry classification, establishment time, registered capital, financing stage, company size (number of employees / revenue). Organizational and human resource attributes: organizational structure complexity, job type distribution, historical employment preferences, human resource cost structure. Operational attributes: business volatility (seasonal / cyclical), revenue growth rate, profit margin, efficiency of key business processes. Risk and compliance attributes: core position dependence, historical turnover rate, data security level, maturity of internal control. Strategic attributes: development stage (startup / growth / mature / transformation), strategic goals (expansion / contraction / transformation), intensity of technology investment.

[0032] The enterprise profile vector is a high-dimensional, structured numerical vector formed by feature engineering and numerical processing of the above-mentioned multi-dimensional enterprise attributes. It is a quantitative expression of enterprise attributes in a computer-processable form.

[0033] The labor market environment refers to the external talent supply and demand ecosystem and related macroeconomic conditions of enterprises. It is a system of external constraints and opportunities that influences enterprises' employment decisions. It mainly includes the following elements: Supply and demand relationship: talent supply, demand, supply-demand ratio, and competition intensity for specific positions. Cost level: median and percentile market salaries for various positions, flexible employment quotes, and welfare level trends. Talent mobility: recruitment cycle, onboarding time, turnover rate trends, and cross-industry mobility index. Policies and regulations: labor laws, social security policies, individual income tax regulations, and flexible employment compliance requirements. Regional characteristics: salary differences between different cities / regions, talent pool characteristics, and cost of living index. Technology / skills trends: demand for emerging skills, traditional skills attrition rate, and skills training costs.

[0034] The market situation vector is a dynamic numerical vector formed by real-time or periodic collection and structured processing of the labor market environment. It is a quantitative snapshot of the external environment in a computer.

[0035] In step S103, the comprehensive benefit evaluation value of at least one target position under multiple different employment forms is calculated based on the enterprise profile vector and the market situation vector.

[0036] The comprehensive benefit evaluation value includes the comprehensive cost value, risk assessment value, and agility value; the employment form includes at least one of full-time, part-time, outsourcing, and gig work.

[0037] It is understood that, based on enterprise profile vectors and market situation vectors, the system can quantitatively evaluate the comprehensive benefits of at least one target position under various employment forms such as full-time, part-time, outsourcing, and gig work. This evaluation value integrates multi-dimensional factors such as cost, risk, and agility, and fully considers the interaction between the enterprise's own characteristics and the external market environment. This provides a comparable, interpretable, and data-driven decision-making basis for different employment strategies, significantly improving the accuracy and rationality of job-level employment selection.

[0038] Specifically, for each job Ji in the target job set, the built-in model is invoked to calculate the comprehensive cost value (Cost), risk value (Risk), and agility value (Agility) for different forms such as full-time (F), part-time (P), outsourcing (C), gig work (G), and consulting (S), forming the benefit-risk matrix Ri for that job.

[0039] In this embodiment of the application, the comprehensive benefit evaluation value of at least one target position under multiple different employment forms is calculated based on the enterprise profile vector and the market situation vector. This includes: obtaining the explicit costs of the market situation vector and the operational efficiency characteristics of the enterprise profile vector, wherein the explicit costs include salary or service quotation data; calculating the corresponding implicit costs based on the operational efficiency characteristics of the enterprise profile vector; and calculating the comprehensive cost value of the target position under multiple different employment forms based on the explicit and implicit costs.

[0040] It is understood that the embodiments of this application can quantify the implicit costs under different employment forms by obtaining explicit cost data such as salary or service quotes from the market situation vector and combining them with the operational efficiency characteristics reflected by the enterprise profile vector. The system can comprehensively calculate the comprehensive cost value of the target position under various employment forms by integrating explicit and implicit costs, thereby more comprehensively assessing the actual economic burden of each employment plan, effectively avoiding decision-making bias caused by relying solely on surface prices, and improving the refinement and scientific level of enterprise employment cost management.

[0041] Specifically, the overall cost value (Cost) is calculated as the sum of explicit and implicit costs.

[0042] Cost(Ji,F) = [Market benchmark salary × (1 + corporate welfare coefficient)] + per capita recruitment and management amortization cost + fixed equipment / site cost for the position; Cost(Ji,C) = outsourcing quote × (1 + management fee coefficient) + conversion and collaboration cost.

[0043] Among them, the market benchmark salary / quote is derived from the market situation vector M (including salary percentiles and outsourcing service quotes in the same region and industry); the company-specific coefficient is derived from the company profile vector E. For example, the "welfare coefficient" is related to company size and profitability; the "management fee coefficient" is related to the company's outsourcing management maturity; the implicit cost model is predicted by training a regression model using historical data. For example, the "recruitment cost" can be estimated based on the company's historical "average cost per job invitation" and "average number of interview rounds" for similar positions on the Yihaozhipin platform.

[0044] In this embodiment of the application, the comprehensive benefit evaluation value of at least one target position under multiple different employment forms is calculated based on the enterprise profile vector and the market situation vector. The method further includes: obtaining target risk factors from the enterprise profile vector and the market situation vector, wherein the target risk factors include turnover rate, project complexity, and data security risk; calculating the corresponding risk weights based on the target risk factors from the enterprise profile vector and the market situation vector; and calculating the agility value of the target position under multiple different employment forms based on the target risk factors and the corresponding risk weights.

[0045] It is understood that the embodiments of this application can obtain target risk factors in the enterprise profile vector and market situation vector, including turnover rate, project complexity and data security risk, and dynamically calculate the corresponding risk weights based on these factors. The system can further combine the risk factors and their weights to quantitatively evaluate the agility value of the target position under various employment forms. Thus, based on the consideration of enterprise characteristics and market environment, it can comprehensively reflect the overall performance of different employment methods in terms of response speed, adaptability and execution stability, and improve the applicability and reliability of employment strategies in complex business scenarios.

[0046] Specifically, Risk Value Calculation: Risk Value is a quantitative assessment of management complexity, reliance on core capabilities, and compliance and delivery risks.

[0047] Calculation formula: Risk(Ji,F) = Turnover risk weight × Turnover rate prediction + Core technology dependence risk weight; Risk(Ji,G) = Delivery quality fluctuation risk weight × Project complexity coefficient + Data security risk weight × Data sensitivity coefficient.

[0048] Risk factor identification and weighting are obtained through expert scoring (AHP) or by training a classification model (such as a decision tree) on historical employment issues. Specific risk levels are derived from enterprise profile vectors and market situation vectors. For example, "turnover rate prediction" can be based on industry trends (M) and historical turnover data (E); "data sensitivity" is determined by the nature of the job (Ji).

[0049] In this embodiment of the application, the comprehensive benefit evaluation value of at least one target position under multiple different employment forms is calculated based on the enterprise profile vector and the market situation vector. It also includes: obtaining the recruitment cycle and average arrival time in multiple different employment forms; and calculating the agility value of the target position under multiple different employment forms based on the recruitment cycle and average arrival time.

[0050] It is understood that by obtaining the recruitment cycle and average arrival time corresponding to various employment forms, and combining the enterprise profile vector and market situation vector, the system can quantify and calculate the agility value of the target position under various employment forms, truly reflecting the response efficiency from the generation of employment demand to the actual arrival of personnel. Thus, in the comprehensive benefit evaluation, it can effectively reflect the differences in the capabilities of different employment forms in terms of rapid deployment, flexible adjustment and business continuity support, and significantly enhance the agility of human resource allocation and the accuracy of decision-making of enterprises in dynamic business environments.

[0051] Specifically, Agility score calculation: Agility score measures the speed and flexibility of an employment model in responding to business fluctuations.

[0052] Calculation formula: Agility(Ji,F) = - Recruitment cycle (days) (negative correlation, the longer the cycle, the worse the agility); Agility(Ji,G) = +1 / Average arrival time (days) + Scalability coefficient.

[0053] The core data comes directly from historical data on the Zhipin / Kuaijie referral platforms. For example, the historical "average recruitment cycle" and "average order response time for gig workers" for each job type. The scalability coefficient is set by rules based on the contract flexibility of the employment form (e.g., gig work > part-time > full-time).

[0054] In step S104, a multi-objective optimization model is constructed with the enterprise's employment structure as the optimization object. The optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility. The comprehensive benefit evaluation value of different employment forms for each position is input into the multi-objective optimization model, and the multi-objective optimization model outputs at least one recommended employment combination scheme.

[0055] It is understood that the embodiments of this application construct a multi-objective optimization model with the enterprise's employment structure as the optimization object. It comprehensively considers at least one of the optimization objectives, namely minimizing total cost, minimizing total risk, and maximizing total agility. The comprehensive benefit evaluation value of each position under different employment forms is used as input. The model can automatically generate one or more recommended employment combination schemes that take into account economy, robustness and responsiveness, effectively realize the coordinated allocation of global resources, and improve the systematicness, scientificity and feasibility of enterprise employment decisions while meeting business needs.

[0056] Specifically, it receives the matrix {Ri} of all positions, the enterprise profile vector E, the market situation vector M, and the user optimization objective O; it describes the company's overall employment structure planning as a multi-objective constrained optimization problem, uses an algorithm to solve it, and outputs a structured solution that clearly lists the suggested employment form, number of employees, expected cost savings, risk changes, and other quantitative results for each position.

[0057] 1. Definition of the optimization problem Decision variables: For each job position Ji, select an employment type T (F, P, C, G, S) and determine the number of employees Xi under that type. Employment types include: full-time (F), part-time (P), outsourcing (C), gig work (G), and consulting (S).

[0058] Objective function (multi-objective): Minimize total cost: MinΣ [Cost(Ji, T)] [Xi];Minimize total risk: Min Σ [Risk(Ji,T)] [Xi]; Maximize total agility: Max Σ [Agility(Ji, T)] Xi].

[0059] Where Cost(Ji, T) is the unit comprehensive cost of job Ji using employment form T, Xi is the number of people employed in job Ji under employment form T, Risk(Ji, T) is the unit risk score of job Ji using employment form T, and Agility(Ji, T) is the unit agility score of job Ji using employment form T.

[0060] Constraints: Enterprise hard constraints (from vector E): total budget limit, total number of full-time employees (limited by staffing or workstation), core positions must be full-time (T == F); business logic constraints: company departments dynamically set constraints according to actual management needs, such as: the ratio of full-time employees to part-time employees should be between 50% and 70%; market supply constraints (from vector M): the total demand for gig jobs in a specific region cannot exceed 30% of the active gig job supply on the platform in that region.

[0061] 2. Algorithm and Solution A multi-objective optimization algorithm, such as the improved NSGA-II (a non-dominated ranking genetic algorithm with an elitist strategy), is employed to encode the employment structure selection schemes for all positions across the company into a single "chromosome." An initial set of schemes (population) is randomly generated, and the algorithm iterates through a biological evolutionary process simulating selection, crossover, and mutation. Each generation of the population is evaluated using an objective function and constraints. The core of NSGA-II is to simultaneously select superior individuals based on the non-dominated ranking (Pareto rank) and crowding distance of the schemes, ensuring that the solution set possesses both convergence (closeness to the optimal frontier) and diversity (broad coverage of various trade-off schemes). Upon termination, the algorithm outputs a Pareto-optimal solution set, not a single solution. Each solution represents a global employment structure scheme that achieves different trade-offs among cost, risk, and agility, thereby finding a series of "optimal compromise schemes" that cannot be improved under existing constraints. These are then provided to decision-makers for final selection based on their preferences, avoiding the subjectivity of manually pre-setting weights and achieving objective global optimization.

[0062] In this embodiment, the processing method of the multi-objective optimization model is as follows: the employment form selection and number configuration of each position in the target position set are encoded as population individuals, and the population is iteratively evolved based on the multi-objective optimization algorithm; in each generation of the population, the individuals are selected based on the non-dominated ranking and crowding distance, and the individuals with the best trade-off among the optimization objectives are retained; a Pareto optimal solution set is generated based on the individuals with the best trade-off among the optimization objectives in each generation of the population, where each solution corresponds to a global employment structure scheme that achieves different trade-offs among multiple optimization objectives.

[0063] It is understood that the embodiments of this application can encode the employment form selection and number configuration of each position in the target job set into population individuals, and use a multi-objective optimization algorithm for iterative evolution. In each generation, individuals with good trade-offs among multiple optimization objectives are selected based on non-dominated sorting and crowding distance, and finally Pareto optimal solution set is generated. Each solution corresponds to a global employment structure scheme, thereby effectively supporting enterprises to make diversified and refined strategy choices between cost, risk and agility, and significantly improving the globality, robustness and practical deployability of employment structure optimization.

[0064] In this embodiment of the application, it further includes: receiving user feedback on adjustments to the recommended workforce combination scheme, and / or collecting actual operational data after the workforce combination scheme is implemented; and optimizing the multi-objective optimization model based on the adjustment feedback and / or the actual operational data.

[0065] It is understood that the embodiments of this application can receive user feedback on adjustments to the recommended employment combination scheme and collect actual operational data after the scheme is implemented. This information can be used to continuously optimize the multi-objective optimization model. The system can continuously calibrate the relationship between the comprehensive benefit evaluation value and the optimization objective, improve the model's adaptability to real business scenarios, and make the subsequently generated employment combination scheme more in line with the actual needs and operational effects of enterprises. This forms a closed-loop learning mechanism and enhances the intelligence, accuracy, and long-term availability of the decision support system.

[0066] According to the employment combination scheme recommendation method proposed in the embodiments of this application, by integrating internal enterprise employment data and external market employment data, it is possible to generate an enterprise profile vector reflecting enterprise attributes and a market situation vector reflecting the labor market environment. Based on these two types of vectors, the comprehensive benefit evaluation value of the target position under various employment forms is calculated. Thus, a multi-objective optimization model is constructed with the goal of minimizing total cost, minimizing total risk, or maximizing total agility. This model recommends the optimal employment combination scheme for each position based on the comprehensive benefit evaluation value, ultimately achieving more scientific and reasonable enterprise employment structure decision support, enhancing the enterprise's competitiveness and adaptability in a complex and ever-changing market environment while reducing operating costs and risks.

[0067] The following will combine Figure 2 The recommended method for the employment combination scheme in this application is described in detail below: Step 1: Multi-source data acquisition and input Data source and input method: (1) Internal enterprise data: Source data: Automatically obtain organizational structure, employee roster, historical salary and benefits, recruitment records, and business operation data (such as order volume and project information) through API interfaces with the enterprise's existing human resource management system (HRM), enterprise resource planning (ERP) and other internal systems. Input method: Automatic timed synchronization or triggered extraction by the system background.

[0068] (2) External market data: Source data: Job salary and resume supply and demand data from recruitment platforms (such as Yihaozhipin) through authorized partners or public channels; gig worker quotes and response time data from flexible employment platforms (such as Kuaijiejian); labor law and policy texts from government websites; and industry reports from third-party salary research institutions. Input method: Data is retrieved periodically through configured API interfaces, or structured collection of publicly available information is performed using web crawlers.

[0069] (3) User goals and constraints: Source data: set by enterprise decision-makers through the system front-end interface. Input content: includes core optimization goals (such as "cost minimization priority" or "agility maximization priority") and specific constraints (such as annual total budget limit, number of full-time employees, and rules that core positions must be full-time).

[0070] Step 2: Feature Extraction and Vectorization Construction Data preprocessing: Cleaning (duplicate removal, error correction, and filling in missing values) and standardization (unifying format, units, and coding) of the raw data input in step one.

[0071] Constructing the Enterprise Profile Vector (E): Processing Module: Enterprise Profile Analysis Module. Processing Logic: Extracting and calculating feature indicators that characterize the enterprise's intrinsic attributes from cleaned internal data. For example, quantifying "development stage" as values ​​such as "startup stage = 0.2, growth stage = 0.8, maturity stage = 1.0"; quantifying "business volatility" as the coefficient of variation of historical business data. After normalization, all features are sequentially combined into a multi-dimensional enterprise profile vector E.

[0072] Constructing a Market Situation Vector (M): Processing Module: Market Situation Awareness Module. Processing Logic: Extracting and calculating characteristic indicators reflecting the external labor market environment from cleaned external data. For example, calculating the "median salary" and "talent supply-demand ratio" (job postings / resume submissions) for specific regions and positions; converting policy and regulatory texts into "compliance risk scores" using natural language processing technology. All features are also combined into a multi-dimensional market situation vector M.

[0073] Step 3: Quantitative Assessment of the Benefits and Risks of Employment Models (1) Input reception: The benefit and risk assessment module receives vectors E and M from step two, as well as a list of target positions to be analyzed.

[0074] (2) Quantitative Calculation: For each position in the list, the module calls the built-in quantitative model and, in conjunction with the current enterprise characteristics (E) and market environment (M), calculates the three key indicator values ​​for each employment form adopted by the position: Total Cost: It not only calculates explicit salaries / offers, but also estimates implicit costs such as recruitment, management, and training through models.

[0075] Risk Assessment Value (Risk): This value comprehensively assesses the management risks, turnover risks, delivery risks, compliance risks, etc. that may arise under this employment arrangement, and provides a comprehensive risk score.

[0076] Agility score: This score assesses the employment type's ability to respond quickly, scale, and adjust.

[0077] Create an evaluation matrix: Generate a "benefit-risk matrix" for each position, clearly listing the cost, risk, and agility values ​​corresponding to various employment options for that position, providing basic data for overall optimization.

[0078] It should be noted that the benefit and risk assessment module involves a hybrid expert model architecture to efficiently achieve the fusion analysis of internal and external multi-source data and the accurate quantitative assessment of multi-dimensional employment benefits, thereby providing high-quality input for subsequent multi-objective optimization. Specifically: The structure of the hybrid expert model is as follows: (1) Multimodal input layer: receives and integrates the structured enterprise feature vector (E) from the enterprise profile analysis module, the encoded external environment feature vector (M) from the market situation awareness module, and the semantic representation of the user-defined optimization objective (O).

[0079] (2) The expert network consists of multiple independent feedforward neural network sub-models, each expert being trained to focus on a specific dimension or scenario for evaluating the effectiveness of employment. For example: cost-benefit experts are good at integrating market salary data and corporate welfare policies to accurately calculate explicit and implicit costs; compliance and risk experts focus on analyzing policy texts and job characteristics to assess the compliance and operational risks of different employment forms; agile configuration experts are good at combining market response speed and corporate business fluctuation needs to assess the flexibility and responsiveness of employment forms.

[0080] (3) Gated Network: As a dynamic routing controller. It calculates and outputs a sparse weight vector in real time based on the specific context of the current input (i.e., the specific enterprise profile E, market situation M, and user goal O). This weight vector is used to intelligently select and activate the top-k most relevant expert networks in the current scenario, rather than using all experts.

[0081] (4) Output layer: Aggregate the weighted output of a small number of activated experts, and finally map to generate a comprehensive benefit evaluation value (i.e. cost value, risk value, and agility value) for different employment forms for the target position, forming the benefit-risk matrix required for subsequent optimization steps.

[0082] Training and operation of hybrid expert models: Joint Training and Multi-Objective Learning: Supervised learning is employed, using historical decision data and results for training. The total loss function includes not only the task loss (such as mean squared error) that measures prediction accuracy but also an auxiliary load balancing loss. This design ensures that the gating network can evenly distribute different types of evaluation tasks among the experts, preventing individual experts from being overloaded. This allows all experts to be fully trained in their areas of expertise, collectively supporting the evaluation of multi-dimensional objectives.

[0083] Sparse Inference and Adaptive Generation: During the solution generation (inference) phase, the model dynamically activates the most relevant minority of experts for sparse computation based on the input context. For example, when the system is an assessment solution for a company in a heavily regulated industry, the gating network significantly increases the weight of compliance and risk experts; while when the core objective is rapid business expansion, agile configuration experts are prioritized for activation. This approach significantly improves computational efficiency while ensuring accurate and customized assessments.

[0084] Step 4: Multi-objective optimization and solution generation (1) Defining the optimization problem: The intelligent recommendation engine module formalizes the enterprise employment structure planning into a multi-objective optimization problem.

[0085] Decision variables: Select an employment type for each position and determine the number of employees under that type.

[0086] The three main objectives are: to minimize total cost, to minimize total risk, and to maximize total agility (based on the matrix data calculation in step three).

[0087] Constraints: Integrate user-defined hard constraints (such as total budget, full-time staffing) and business logic constraints (such as core positions must be full-time).

[0088] (2) Intelligent optimization solution: Core algorithm: An improved multi-objective optimization algorithm (such as NSGA-II) is used to solve the problem.

[0089] Solution process: The algorithm treats countless possible combinations of labor as a "solution space" and efficiently searches the space by simulating the evolutionary process of "selection, crossover, and mutation". Its core mechanism is to simultaneously evaluate the performance of each solution on the three objectives and prioritize retaining those "excellent" solutions (i.e., non-dominated solutions) that cannot be completely surpassed by other solutions on the three objectives.

[0090] (3) Generating a set of recommended solutions: After the algorithm terminates, the output is not a single solution, but a Pareto optimal solution set. This set contains multiple solutions, each of which represents a different, non-improvable optimal trade-off between "cost, risk, and agility".

[0091] Step 5: Solution Delivery, Feedback, and System Iteration (1) Visualized delivery of the solution: The solution set generated in step four is presented to users in an intuitive format, such as charts and reports. Users can compare solutions with different trade-offs and fine-tune the solutions using interactive tools (for example, manually changing a position from outsourcing to full-time).

[0092] (2) Feedback data collection: Proactive feedback: Record users' adjustments to the plan and their final choices. Passive feedback: After the plan is implemented, the system continuously collects operational results data such as actual labor costs, personnel stability, and business response efficiency from the enterprise's internal systems.

[0093] (3) Closed-loop learning and optimization: Processing logic: The "user-adjusted plan" and "actual operational results" are used as new training data and fed back to the system's data pool and model layer. System iteration: Through these feedback data, the quantitative evaluation model (such as cost model and risk model) in step three is retrained or fine-tuned periodically to make its predictions increasingly closer to reality, thereby realizing the system's self-learning and continuous optimization.

[0094] Specific example: Background: An e-commerce company in a period of rapid growth (Series B financing) plans to expand into the East China market.

[0095] (1) Input: Internal data: The company is in the "growth stage" with seasonal business peaks (such as Double Eleven), and plans to add a "Marketing Promotion in East China" team. External data: The median monthly salary for full-time digital marketing specialists in East China is 15K, and the hourly wage for part-time specialists is 120 RMB. The quoted price for related outsourcing projects is 50,000-100,000 RMB per project. User goal: To quickly build a team and control initial fixed costs.

[0096] (2) Processing: The system generates a corporate profile vector E (labels: growth-oriented, e-commerce, strong seasonality). It also generates a market situation vector M (labels: East China, sufficient marketing talent, high cost).

[0097] For the "East China Market Promotion" position, the benefit-risk assessment module calculated that: full-time positions have high fixed costs but strong control; outsourced project-based positions have clear costs but slow response times; part-time positions offer greater flexibility and lower immediate costs but are more complex to manage. The intelligent recommendation engine performs optimization calculations based on vectors E, M, and target O.

[0098] (3) Output Plan: It is recommended to form a hybrid team consisting of "1 full-time promotion manager (core planning and control) + N flexible part-time promotion specialists (execution and field promotion)". Expected results: Compared with hiring full-time employees, the initial labor cost will be reduced by about 35%, the team size can be quickly adjusted according to monthly promotional activities, and the core capabilities will be preserved.

[0099] Next, the recommended apparatus for the labor combination scheme proposed according to the embodiments of this application is described with reference to the accompanying drawings.

[0100] Figure 3 This is a block diagram of a labor combination scheme recommendation device according to an embodiment of this application.

[0101] like Figure 3As shown, the recommended labor combination scheme device 10 includes: an acquisition module 100, a generation module 200, a calculation module 300, and a processing module 400.

[0102] The acquisition module 100 is used to acquire at least one type of internal data and external data, wherein the internal data comes from the enterprise's internal employment data source and the external data comes from the market employment data source; the generation module 200 is used to generate an enterprise profile vector representing the enterprise's attributes based on the internal data and a market situation vector representing the labor market environment based on the external data; the calculation module 300 is used to calculate the comprehensive benefit evaluation value of at least one target position under multiple different employment forms based on the enterprise profile vector and the market situation vector; the processing module 400 is used to construct a multi-objective optimization model with the enterprise's employment structure as the optimization object, wherein the optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility, and the comprehensive benefit evaluation value of each position under different employment forms is input into the multi-objective optimization model, and the multi-objective optimization model outputs at least one recommended employment combination scheme.

[0103] According to the employment combination scheme recommendation device proposed in the embodiments of this application, by integrating internal enterprise employment data and external market employment data, it can generate an enterprise profile vector reflecting enterprise attributes and a market situation vector reflecting the labor market environment. Based on these two types of vectors, it calculates the comprehensive benefit evaluation value of the target position under various employment forms, thereby constructing a multi-objective optimization model with the goal of minimizing total cost, minimizing total risk, or maximizing total agility. This model recommends the optimal employment combination scheme for each position based on the comprehensive benefit evaluation value, ultimately achieving more scientific and reasonable enterprise employment structure decision support, enhancing the enterprise's competitiveness and adaptability in a complex and ever-changing market environment while reducing operating costs and risks.

[0104] Figure 4 is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 401, the processor 402, and the computer program stored on the memory 401 and capable of running on the processor 402.

[0105] When the processor 402 executes the program, it implements the labor combination scheme recommendation method provided in the above embodiments.

[0106] Furthermore, electronic devices also include: Communication interface 403 is used for communication between memory 401 and processor 402.

[0107] The memory 401 is used to store computer programs that can run on the processor 402.

[0108] Memory 401 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0109] If the memory 401, processor 402, and communication interface 403 are implemented independently, then the communication interface 403, memory 401, and processor 402 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0110] Optionally, in a specific implementation, if the memory 401, processor 402, and communication interface 403 are integrated on a single chip, then the memory 401, processor 402, and communication interface 403 can communicate with each other through an internal interface.

[0111] Processor 402 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0112] This application also provides a computer-readable storage medium storing a computer program or instructions thereon, which, when executed by a processor, implements the above-mentioned recommended method for combining labor resources.

[0113] This application also provides a computer program product, including a computer program or instructions, which, when executed, implement the above-mentioned method for recommending employment combination schemes.

[0114] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0115] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0116] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0117] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0118] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

Claims

1. A method for recommending employment combination schemes, characterized in that, Includes the following steps: Acquire at least one type of internal data and external data, wherein the internal data comes from the enterprise's internal employment data sources, and the external data comes from market employment data sources; Generate an enterprise profile vector representing enterprise attributes based on the internal data, and generate a market situation vector representing the labor market environment based on the external data. Calculate the comprehensive benefit evaluation value of at least one target position under multiple different employment forms based on the enterprise profile vector and the market situation vector; A multi-objective optimization model is constructed with the enterprise's employment structure as the optimization object. The optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility. The comprehensive benefit evaluation value of different employment forms for each position is input into the multi-objective optimization model, and the multi-objective optimization model outputs at least one recommended employment combination scheme.

2. The method for recommending employment combination schemes according to claim 1, characterized in that, The comprehensive benefit evaluation value includes a comprehensive cost value, a risk assessment value, and an agility value; wherein, based on the enterprise profile vector and the market situation vector, the comprehensive benefit evaluation value for at least one target position under multiple different employment forms is calculated, including: Obtain the explicit costs of market situation vectors and the operational efficiency characteristics of enterprise profile vectors, where explicit costs include salary or service quotation data; The corresponding implicit costs are calculated based on the operational efficiency characteristics of the enterprise profile vector. Based on the explicit costs and the implicit costs, calculate the comprehensive cost value of the target position under various employment forms.

3. The method for recommending employment combination schemes according to claim 2, characterized in that, Based on the enterprise profile vector and the market situation vector, calculate the comprehensive benefit evaluation value of at least one target position under multiple different employment forms, and also include: Obtain the target risk factors of the enterprise profile vector and the market situation vector, wherein the target risk factors include turnover rate, project complexity and data security risk; Calculate the corresponding risk weights based on the target risk factors of the enterprise profile vector and the market situation vector; Calculate the agility value of the target position under various employment forms based on the target risk factors and corresponding risk weights.

4. The method for recommending employment combination schemes according to claim 3, characterized in that, The employment forms include at least one of full-time, part-time, outsourcing, and gig work. Based on the enterprise profile vector and the market situation vector, a comprehensive benefit evaluation value is calculated for at least one target position under multiple different employment forms. This also includes: Obtain recruitment cycles and average onboarding times for various employment models; Calculate the agility value of the target position under various employment forms based on the recruitment cycle and the average arrival time.

5. The method for recommending employment combination schemes according to claim 1, characterized in that, The processing method for the multi-objective optimization model: The employment form selection and number of employees for each position in the target job set are encoded as population individuals, and the population is iteratively evolved based on a multi-objective optimization algorithm; In each generation of the population, individuals are selected based on their non-dominated ranking and crowding distance, retaining the individuals that balance the optimization objectives. A Pareto optimal solution set is generated based on the optimal individual in each generation of the population, where each solution corresponds to a global employment structure scheme that achieves different trade-offs among multiple optimization objectives.

6. The method for recommending employment combination schemes according to claim 1, characterized in that, Also includes: Receive user feedback on adjustments to the recommended workforce combination plan, and / or collect actual operational data after the workforce combination plan is implemented; The multi-objective optimization model is optimized based on the aforementioned adjustment feedback and / or actual operational data.

7. A device for recommending labor combination schemes, characterized in that, include: The acquisition module is used to acquire at least one type of internal data and external data, wherein the internal data comes from the enterprise's internal employment data source, and the external data comes from the market employment data source; The generation module is used to generate an enterprise profile vector representing enterprise attributes based on the internal data, and to generate a market situation vector representing the labor market environment based on the external data. The calculation module is used to calculate the comprehensive benefit evaluation value of at least one target position under multiple different employment forms based on the enterprise profile vector and the market situation vector; The processing module is used to construct a multi-objective optimization model with the enterprise's employment structure as the optimization object. The optimization objective is at least one of minimizing total cost, minimizing total risk, and maximizing total agility. The multi-objective optimization model is input with the comprehensive benefit evaluation value of different employment forms for each position, and the multi-objective optimization model outputs at least one recommended employment combination scheme.

8. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the employment combination scheme recommendation method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they are used to implement the employment combination scheme recommendation method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement the employment combination scheme recommendation method as described in any one of claims 1-6.

Images