Tobacco enterprise human resource management auxiliary calibration method based on big data

Abstract

The invention relates to the field of human resource management, and discloses a tobacco enterprise human resource management auxiliary calibration method based on big data, and the method comprises the steps: obtaining multi-source heterogeneous data related to enterprise internal human resources, and constructing a structured human resource data model in combination with a data standardization processing mechanism and an abnormality elimination strategy; post portrait modeling is carried out on the structured human resource data model, a capability dimension nesting analysis method is introduced, key capability factors and weight distribution required by each post are extracted, and a post capability demand graph is constructed; based on the post capability demand map, fusing the staff portraits and the historical job data, and identifying the deviation between posts and the staff through a multi-dimensional feature matching algorithm to form a preliminary calibration suggestion set; and a dynamic service association analysis method is introduced, and key matching parameters in the preliminary calibration suggestion set are dynamically corrected in combination with latest service demand data and real-time task assignment information. The method has the advantage of improving the management efficiency.

Description

Technical Field

[0001] The present invention relates to the field of human resource management, and in particular to a tobacco enterprise human resource management auxiliary calibration method based on big data. Background Art

[0002] Tobacco companies have gradually introduced information management systems into their human resource management processes and are attempting to leverage big data for performance evaluation, job optimization, and recruitment demand forecasting. However, due to diverse data sources and inconsistent standards, particularly the mismatch between historical personnel data and current business needs, errors can occur in decision-making support. For example, the actual workload of certain positions is not accurately reflected in system records, resulting in staffing recommendations that are inconsistent with actual conditions. This data misalignment is difficult to effectively identify and calibrate using traditional statistical analysis, thus compromising the scientific nature and accuracy of human resource allocation. Therefore, it is necessary to design a big data-based human resource management calibration method for tobacco companies to improve management efficiency. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides a tobacco enterprise human resources management auxiliary calibration method based on big data, which has the advantage of improving management efficiency and solves the problems in the above-mentioned background technology.

[0004] To achieve the above-mentioned purpose of improving management efficiency, the present invention provides the following technical solution: a tobacco enterprise human resources management auxiliary calibration method based on big data, comprising the following steps:

[0005] Acquire multi-source heterogeneous data related to internal human resources of the enterprise, combine data standardization processing mechanism with anomaly elimination strategy, and build a structured human resources data model;

[0006] Conduct job profile modeling on the structured human resources data model and introduce the capability dimension nested analysis method to extract the key capability factors and weight distribution required for each position and construct a job capability requirement map;

[0007] Based on the job competency requirements map, employee profiles and historical employment data are integrated, and a multi-dimensional feature matching algorithm is used to identify the deviations between jobs and employees, forming a preliminary calibration recommendation set.

[0008] Introducing a dynamic business correlation analysis method that combines the latest business demand data with real-time task assignment information to dynamically correct key matching parameters in the initial calibration recommendation set and identify potential imbalances in job configurations;

[0009] Based on the potential imbalance points in job configuration, an intelligent recommendation mechanism is used to generate an actionable human resource configuration optimization plan and output an auxiliary calibration report to support organizational decision-making.

[0010] Preferably, the process of building a structured human resources data model is:

[0011] Collect multi-source heterogeneous data and establish a data source identification mechanism;

[0012] Use standardized data conversion modules to clean, normalize, and format raw data, and build unified field standards that meet analysis needs;

[0013] Combined with the industry knowledge base, the ontology mapping mechanism is used to identify and merge data fields with the same semantics but different representations;

[0014] Outliers and missing values ​​are processed by combining rule elimination with machine learning interpolation algorithms;

[0015] Finally, a structured human resources data model is generated to support job modeling and capability analysis.

[0016] Preferably, the process of performing job profile modeling on the structured human resources data model is as follows:

[0017] Based on the job description fields in the structured data model, natural language processing technology is combined to perform keyword extraction and semantic clustering to generate a set of job function labels;

[0018] Introducing employee employment history and job rotation records, and determining the typical employment path and growth cycle for each position through statistical analysis;

[0019] Integrate multiple dimensions to build a multi-level job portrait model.

[0020] Preferably, the process of extracting the key competency factors and weight distribution required for each position is as follows:

[0021] Standardize the competency labels in the job profile model, unify the naming of competency factors, and remove redundancies and duplications;

[0022] The analytic hierarchy process is used to classify various capability factors into knowledge-based, execution-based, and management-based dimensions, and an expert scoring mechanism is introduced to determine the initial weight coefficients;

[0023] Integrate historical performance evaluation and employee competency assessment results to conduct regression modeling on the performance contribution of different ability factors in actual work and optimize the weight distribution model;

[0024] The final output is a set of key capability factors, along with corresponding job matching weights.

[0025] Preferably, the process of constructing a job capability requirement map is as follows:

[0026] The extracted key competency factors of positions are classified according to position levels and functional categories, and a model of capability sharing and differentiation between positions is established;

[0027] Construct a dependency diagram between ability factors to identify the order and interaction intensity of different factors in the task completion process;

[0028] Based on the graph structure, set up a many-to-many mapping relationship between positions and ability factors;

[0029] The graph database is used to store and query graphs, forming a job competency requirement graph from a competency perspective.

[0030] Preferably, the process of identifying the deviations between positions and employees through a multi-dimensional feature matching algorithm to form a preliminary calibration suggestion set is as follows:

[0031] Build an employee portrait model covering multiple dimensions;

[0032] Quantify employee profiles and job competency requirements into multi-dimensional vectors, and use the Euclidean distance metric to assess the match between employees and jobs.

[0033] Conduct deviation analysis on job-employee combinations whose matching degree is lower than the set threshold, identify capability gaps and job redundancy points, and output a preliminary calibration recommendation set.

[0034] Preferably, the process of dynamically correcting the key matching parameters in the preliminary calibration suggestion set is:

[0035] Collect the latest strategic goals and task breakdown data of the company's current business segments, and extract new demands for human resources in key business scenarios;

[0036] Based on the data flow of the task assignment platform, track the real-time task execution status and completion indicators of the positions in actual operations;

[0037] Adopt a dynamic weight adjustment mechanism to update key parameters in job competency requirements based on task urgency and job activity;

[0038] For employee-position pairs in the marginal matching area recommended by the preliminary calibration, the deviation is recalculated and the matching value is updated.

[0039] Preferably, the process of identifying potential imbalance points in job configuration is:

[0040] Analyze the revised employee-job matching matrix to identify jobs where the matching values ​​for key competency dimensions are chronically low or fluctuate frequently;

[0041] Combining organizational structure change data with historical job adjustment records, identify job units that are frequently deployed but have not formed a stable capacity match;

[0042] Introducing a job importance scoring mechanism to score and rank the risks caused by capability gaps in key positions;

[0043] Mark all job sets with scores above a set threshold as potential imbalance points.

[0044] Preferably, the process of outputting auxiliary calibration reports to support organizational decision-making includes:

[0045] Organize identified job imbalances, capability gap types, and potential job mismatch risks into a structured report format and generate a risk heat map visualization interface;

[0046] Integrate the job transfer costs, training cycles, and business impact indexes involved in each job transfer plan to establish a multi-dimensional influencing factor analysis model;

[0047] Use intelligent recommendation algorithms to generate several alternative human resource allocation optimization plans, each with a key indicator comparison table, and output an auxiliary calibration report.

[0048] Compared with the existing technology, the present invention provides a tobacco enterprise human resources management auxiliary calibration method based on big data, which has the following beneficial effects:

[0049] 1. By integrating multi-source heterogeneous human resources data and combining standardized processing mechanisms with outlier elimination strategies, we can effectively improve the consistency, integrity, and analyzability of data, and build a high-quality structured human resources data foundation to support subsequent modeling and matching analysis.

[0050] 2. Generate job function labels through natural language processing and semantic clustering, extract key competency factors by combining job history data and performance, and determine competency weights through competency dimension nesting analysis and hierarchical analysis method to form a structured, multi-level job competency requirement map, providing a clear and quantifiable competency framework for job matching and competency assessment.

[0051] 3. By converting employee portraits and job maps into multi-dimensional vectors and using Euclidean distance for similarity calculation, it is possible to identify deviations such as capability gaps and job redundancies, output preliminary calibration suggestions, and enhance the adaptability and pertinence of human-job configuration.

[0052] 4. By dynamically analyzing real-time business task flows and organizational structure adjustments, we can update the importance of positions and the status of competency requirements, identify job units that are chronically inefficient, frequently transferred, and have mismatched competencies, and locate job imbalance risks. This will effectively improve the responsiveness of human resource allocation to business changes and its proactive adjustment capabilities.

[0053] 5. Generate multiple sets of optional optimization solutions through an intelligent recommendation mechanism and present them in structured reports and visualizations, providing organizations with intuitive, quantitative, and executable human resource calibration suggestions, thereby enhancing the scientific nature and effectiveness of management decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 Schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] Example 1: Please refer to Figure 1 As shown, the tobacco enterprise human resource management auxiliary calibration method based on big data according to an embodiment of the present invention includes the following steps:

[0057] S1: Obtain multi-source heterogeneous data related to internal human resources of the enterprise, combine data standardization processing mechanism with anomaly elimination strategy, and build a structured human resources data model.

[0058] The process of constructing a structured human resources data model in S1 is as follows:

[0059] Collect multi-source heterogeneous data including employee performance records, job descriptions, training feedback information, and department workload statistics, and establish a data source identification mechanism to track the data traceability path;

[0060] A standardized data conversion module is used to clean, normalize, and format raw data, establishing a unified field standard that meets analytical requirements. Field mapping strategies are implemented on collected raw data, mapping each system field to a unified standard field for human resources data. During the cleaning process, duplicate records, illegal characters, and logically conflicting data are removed, and field units are uniformly converted to ensure horizontal comparability. Regular expressions and natural language processing techniques are applied to uniformly categorize text fields into terminology. Field default values ​​and default logical fill rules are set, and key indicators are output in a standardized format. Ultimately, a unified field template is formed, providing a standardized data interface for model building and analysis algorithm input.

[0061] In conjunction with the industry knowledge base, the ontology mapping mechanism is used to identify and merge data fields with the same semantics but different representations. An enterprise human resources domain ontology library containing core terms such as position, competency, performance, and training is constructed, and terminology is normalized based on industry standards. Terms in structured fields and free text fields are matched and converted to vector representations, and synonymous fields are identified through semantic similarity calculations. The ontology relationship map is used to determine the hierarchical relationship between terms and to assist in field merging and classification. For ambiguous data items, a manual review and expert verification process is set up to ensure the accuracy and business rationality of field merging. After the semantic merging is completed, the unified field standard is updated to achieve dual guarantees of structural consistency and semantic clarity.

[0062] Outliers and missing values ​​are processed using a combination of rule-based elimination and machine learning interpolation algorithms to ensure the integrity and consistency of the data model. Empirical rules and statistical thresholds are used to directly eliminate obviously erroneous data. For missing numeric fields, rule-based methods such as mean filling and median interpolation are used to handle low-impact variables to improve processing efficiency. For highly sensitive or key variables, the most reasonable missing values ​​are predicted. Confidence assessments are performed on interpolation results, and if they are below the set threshold, they are marked as requiring manual confirmation. After outlier processing is completed, processing logs and marked fields are generated.

[0063] Finally, a structured human resources data model is generated to support job modeling and capability analysis.

[0064] S2: Model job profiles for the structured human resource data model, and introduce a nested analysis method of capability dimensions to extract the key capability factors and weight distribution required for each position, and construct a job capability requirement map.

[0065] The process of performing job profile modeling on the structured human resource data model in S2 is as follows:

[0066] Based on the job description field in the structured data model, natural language processing technology is used to extract keywords and perform semantic clustering to generate a set of job function labels. The standardized "job description" field text data is extracted, and Chinese word segmentation and stop word filtering are performed on it to remove invalid words and redundant semantics. The TF-IDF algorithm or TextRank method is used to extract keywords, retaining high-weight words that can represent the core functions of the position. The keywords are semantically encoded using word vector technology, and semantically similar job keywords are merged and clustered using K-means or hierarchical clustering algorithms. The clustering results are manually verified and labeled, and a unified set of job function labels is output. Each position can be associated with multiple function labels, and finally a job label matrix is ​​formed.

[0067] Introducing employee tenure history and job rotation records, through statistical analysis, determine the typical tenure path and growth cycle of each position; using employee historical personnel file data, construct an employee job migration trajectory table, recording the time, duration and type of job change; statistically analyzing the frequency of migration between positions, constructing a job migration network diagram, identifying high-frequency transfer paths and typical career development sequences; calculating the average tenure cycle, median tenure duration and average growth cycle from initial position to next position for each position, and assessing the degree of experience accumulation required for position maturity; marking job levels or channels for job paths with obvious hierarchical relationships, and outputting a job development pathway model; and finally, embedding the tenure path and growth cycle as portrait parameters into the job portrait model;

[0068] By integrating multiple dimensions such as job functions, tenure cycles, and performance, a multi-level job profile model is constructed to reflect the core role and capability requirements of the position in the organization. A three-layer structure of the job profile is established: foundation layer, structure layer, and capability layer. Job labels are mapped to corresponding capability items in the capability library, establishing a many-to-many mapping relationship between positions and capabilities, and quantifying the degree to which positions require capability dimensions. Combined with historical performance data, the typical performance curve and performance fluctuation range of the position are analyzed to measure the sensitivity of the position to performance levels. The position influence indicator is introduced as an assessment dimension for the strength of the role of the position in the organization. Ultimately, a job profile model is formed that can be used for comparison, recommendation, and modeling.

[0069] Visualize the portrait model to generate a set of job capability descriptions with clear structure and strong interpretability.

[0070] The process of extracting the key competency factors and weight distribution required for each position in S2 is as follows:

[0071] Standardize the competency labels in the job profile model, unify the naming of competency factors, and remove redundancies and duplicates; extract all competency labels from the constructed job profile model, including skill descriptors, performance drivers, key behavioral performance, etc. mapped from job function labels; establish a unified competency factor standard vocabulary, and semantically normalize competency labels based on the general competency model, for example, unifying "communication expression," "communication ability," and "language expression" into "communication ability"; use natural language processing technology to perform semantic aggregation of synonyms, detect and eliminate synonymous labels and redundant items; manage competency labels in layers, distinguishing between first-level competency factors and second-level sub-items; and output a standardized set of competency factors.

[0072] The analytic hierarchy process (AHP) was used to classify various competency factors into knowledge-based, execution-based, and management-based dimensions, and an expert scoring mechanism was introduced to determine the initial weight coefficients. Competency factors were divided into three categories based on the competency dimensions: knowledge-based, execution-based, and management-based. A hierarchical analysis structure model was constructed, with job objectives at the top level, competency dimensions at the middle level, and competency factors at the bottom level. A scoring expert panel consisting of HR experts, team leaders, and high-performing employees was assembled to compare and score the importance of each pair of competency factors in achieving job objectives on a scale of 1 to 9. A judgment matrix was constructed and consistency tested to ensure the logical rationality of the weighted scoring. The eigenvector method was used to calculate the initial weights of each competency factor in the structure, outputting a position-competency initial weight matrix.

[0073] By integrating historical performance evaluations with employee competency assessment results, regression modeling is performed on the performance contributions of different competency factors in actual work, optimizing the weight distribution model. A data set of employee competency factors and performance results is constructed, including employee competency scores and corresponding performance levels in multiple positions. Feature selection and normalization preprocessing are performed on the data to control for interfering variables such as position type and years of experience. A multivariate linear regression method is used to construct a model for the impact of competency factors on performance, outputting the regression coefficient or importance score of each competency factor. The competency impact coefficients obtained in the model are weighted averaged or Bayesian fused with the initial weights of the hierarchical analysis method to correct for deviations caused by subjective judgments of experts. A data-validated and optimized position competency factor weight distribution model is output to improve the objectivity and practicality of weight configuration.

[0074] The final output is a set of key competency factors covering dimensions such as knowledge, skills, attitudes, and experience, along with corresponding job matching weights. Standardized competency factors are divided into dimensions to construct a four-dimensional framework model of "knowledge (K)-skills (S)-attitudes (A)-experience (E)". Each type of position outputs its corresponding set of KSAE key factors, with numerical matching weights assigned to each factor. A job competency profile template is output, including the position name, a list of competency factors, and a radar chart or table display of weight distribution. Competency weights can be used in multiple application scenarios, including employee competency assessment comparison, job matching scoring, and training path optimization. Ultimately, a job key competency structure library is formed.

[0075] The process of constructing the job capability requirement map in S2 is as follows:

[0076] The extracted key competency factors for positions are classified by position level and functional category, and a model of capability sharing and differentiation between positions is established. The extracted standardized key competency factor set is preliminarily grouped according to the hierarchical structure of the position. Positions are classified by function, such as technical, operational, marketing, and human resources, and the business unit to which the position belongs is marked. The overlap of competency factors for positions with the same function or the same level is counted to construct a competency sharing matrix. Differential competency factors are identified to form a comparison table of position competency differences. A three-dimensional classification table of position level × functional category × competency factor is output to lay the foundation for the node attribute structure of the composition.

[0077] Construct a dependency graph between capability factors to identify the order and interaction strength of different factors in the task completion process; extract the task completion chain from data such as job descriptions, process descriptions, and performance indicator systems; construct directed dependency relationships between capability factors based on the order in which capabilities are called in the task process; introduce work behavior event logs to mine the frequency and combination patterns of capability use and quantify the interaction strength; construct a directed graph of capability factors, where nodes are capability factors, edges represent dependency relationships, and edge weights represent interaction strength; the capability dependency graph serves as the basic network input for capability structure modeling;

[0078] Based on the graph structure, a many-to-many mapping relationship between positions and capability factors is established to support the reuse and optimization of different capability factors across multiple positions. With positions and capability factors as nodes, a bidirectional mapping edge between position and capability is established to indicate that a position depends on a certain capability. The mapping relationship supports one capability factor serving multiple positions, and the edge weight can be adjusted based on the position weight. Attributes such as capability importance, frequency, hierarchy, and capability maturity are annotated in the graph structure. A bidirectional graph structure of position and capability factor is output to support horizontal capability alignment and job commonality mining.

[0079] The graph database is used to store and query graphs, support subsequent job matching and dynamic adjustment analysis, and form a job capability requirement graph from a capability perspective.

[0080] The technical solution of this embodiment is as follows: through natural language processing technology, keywords are extracted and semantic clustered from the job responsibility field to generate job function labels; combining employee job track and performance data, a multi-level job portrait model is constructed, and a capability dimension nested analysis method is introduced to standardize and classify the capability labels in the portrait, and the hierarchical analysis method and performance regression modeling are used to extract key capability factors and their weight distribution; finally, a many-to-many mapping relationship is established between positions and capability factors, and a job capability requirement map with correlation strength and reuse structure is constructed, which is stored in a graph database to support efficient query and dynamic maintenance. Accurate modeling of job capability portraits and weighted extraction of key capability factors are achieved, and the correspondence between job responsibilities and actual capability requirements is improved. The constructed job capability map has high interpretability and structuredness, which helps to realize intelligent decision-making in core businesses such as person-job matching, capability assessment, and talent inventory, and significantly enhances the scientificity, agility, and visualization of the organization's human resource allocation.

[0081] Example 2: Figure 1 As shown, a tobacco enterprise human resource management auxiliary calibration method based on big data also includes the following steps:

[0082] S3: Based on the job competency requirement map, integrating employee portraits and historical employment data, a multi-dimensional feature matching algorithm is used to identify the deviations between jobs and employees to form a preliminary calibration recommendation set.

[0083] The process of identifying the deviation between positions and employees by using a multi-dimensional feature matching algorithm in S3 to form a preliminary calibration suggestion set is as follows:

[0084] Build an employee profile model that covers multiple dimensions, including educational background, years of work experience, performance level, and competency assessment results. Preprocess the raw data, including missing value filling, outlier removal, and format standardization, to ensure data consistency. Convert various types of information into structured feature variables according to unified employee profile field specifications. For example, convert educational background into hierarchical values ​​(junior college = 1, bachelor's degree = 2, master's degree = 3, etc.) and competency assessment into a Likert scale. Output a standardized profile vector for each employee to form a structured employee profile model.

[0085] The employee profile and job competency requirement map are represented by multi-dimensional vectors, and the Euclidean distance index is used to evaluate the matching degree between employees and jobs. The formula is:

[0086]

[0087] Where n is the total number of dimensions of the capability factor, e i is the employee's score on the i-th ability factor, p i is the demand value of the position at the i-th capability factor, w iis the weight coefficient of the i-th capability factor;

[0088] Deviation analysis is performed on job-employee combinations with a matching degree below the set threshold to identify capability gaps and job redundancy points, and a preliminary calibration recommendation set is output; a minimum allowable threshold for job matching is set to screen the matching results of all employee-job combinations; deviation analysis is performed on combinations with insufficient matching: job capability factors are compared with employee capability vectors to locate low matching or missing capabilities; parts of employee capabilities that are not required for a job or exceed the requirements are discovered for job migration recommendations; a deviation analysis report is output, which includes the employee number and current job; a list of insufficient capability factors and recommended improvement paths; excess capability factors and possible matching jobs; and a preliminary calibration recommendation set is generated.

[0089] S4: Introduce a dynamic business correlation analysis method, combine the latest business demand data with real-time task assignment information, dynamically correct the key matching parameters in the preliminary calibration recommendation set, and identify potential imbalance points in job configuration.

[0090] The process of dynamically correcting the key matching parameters in the preliminary calibration suggestion set in S4 is as follows:

[0091] Collect the latest strategic goals and task breakdown data of the company's current business segments, and extract new demands for human resources in key business scenarios;

[0092] Based on the data flow of the task assignment platform, track the real-time task execution status and completion indicators of the positions in actual operations;

[0093] Adopt a dynamic weight adjustment mechanism to update key parameters in job competency requirements based on task urgency and job activity;

[0094] For employee-position pairs in the marginal matching area recommended by the preliminary calibration, the deviation is recalculated and the matching value is updated.

[0095] The process of identifying potential imbalance points in job configuration in S4 is as follows:

[0096] Analyze the revised employee-job matching matrix to identify jobs where the matching values ​​for key competency dimensions are chronically low or fluctuate frequently;

[0097] Combine organizational structure change data with historical job adjustment records to identify positions that are frequently deployed but lack stable competency alignment. Obtain organizational change data and job adjustment records. Calculate a deployment frequency index for each position, including metrics like the number of employee changes per unit time and the number of revisions to job responsibilities or competency requirements. Compare deployment frequency trends with competency alignment. If the alignment remains low or shows no significant improvement after frequent adjustments, mark the position as a failed competency alignment position.

[0098] Introducing a job importance scoring mechanism to score and rank the risks associated with capability gaps in key positions; constructing a job importance assessment model that considers factors such as: the degree to which the position is a critical path in organizational processes; the impact of the position on output quality; and the scarcity and difficulty of replacing personnel in the position; establishing a capability gap risk scoring model: Position risk score = Position importance weight × Capability Gap Index; ranking the job risk scores to identify positions where capability gaps significantly impact organizational operations;

[0099] Mark all job sets with scores above a set threshold as potential imbalance points.

[0100] S5: Based on the potential imbalance points in job configuration, an intelligent recommendation mechanism is used to generate an actionable human resource configuration optimization plan and output an auxiliary calibration report to support organizational decision-making.

[0101] The process of outputting the auxiliary calibration report to support organizational decision-making in S5 includes:

[0102] Organize identified job imbalances, capability gap types, and potential job mismatch risks into a structured report format and generate a risk heat map visualization interface;

[0103] By integrating the job transfer costs, training cycles, and business impact indexes involved in each job transfer plan, a multi-dimensional impact factor analysis model is established. The formula is:

[0104] Job deployment impact index = w1*C 调岗成本 +w2*T 培训周期 +w3*R 业务影响指数

[0105] Where w1, w2, and w3 are the weight coefficients of job transfer cost, training cycle, and business impact index, respectively;

[0106] Use intelligent recommendation algorithms to generate several alternative human resource allocation optimization plans, each with a key indicator comparison table, and output an auxiliary calibration report.

[0107] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0108] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A tobacco enterprise human resources management auxiliary calibration method based on big data, characterized in that: The following steps are involved: Acquire multi-source heterogeneous data related to internal human resources of the enterprise, combine data standardization processing mechanism with anomaly elimination strategy, and build a structured human resources data model; Conduct job profile modeling on the structured human resources data model and introduce the capability dimension nested analysis method to extract the key capability factors and weight distribution required for each position and construct a job capability requirement map; Based on the job competency requirements map, employee profiles and historical employment data are integrated, and a multi-dimensional feature matching algorithm is used to identify the deviations between jobs and employees, forming a preliminary calibration recommendation set. Introducing a dynamic business correlation analysis method that combines the latest business demand data with real-time task assignment information to dynamically correct key matching parameters in the initial calibration recommendation set and identify potential imbalances in job configurations; Based on the potential imbalance points in job configuration, an intelligent recommendation mechanism is used to generate an actionable human resource configuration optimization plan and output an auxiliary calibration report to support organizational decision-making.

2. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 1, characterized in that: The process of building a structured human resources data model is as follows: Collect multi-source heterogeneous data and establish a data source identification mechanism; Use standardized data conversion modules to clean, normalize, and format raw data, and build unified field standards that meet analysis needs; Combined with the industry knowledge base, the ontology mapping mechanism is used to identify and merge data fields with the same semantics but different representations; Outliers and missing values ​​are processed by combining rule elimination with machine learning interpolation algorithms; Finally, a structured human resources data model is generated to support job modeling and capability analysis.

3. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 2, characterized in that: The process of job profile modeling for structured human resources data model is as follows: Based on the job description fields in the structured data model, natural language processing technology is combined to perform keyword extraction and semantic clustering to generate a set of job function labels; Introducing employee employment history and job rotation records, and determining the typical employment path and growth cycle for each position through statistical analysis; Integrate multiple dimensions to build a multi-level job portrait model.

4. The method for assisting calibration of tobacco enterprise human resource management based on big data according to claim 3, characterized in that: The process of extracting the key competency factors and weight distribution required for each position is as follows: Standardize the competency labels in the job profile model, unify the naming of competency factors, and remove redundancies and duplications; The analytic hierarchy process is used to classify various capability factors into knowledge-based, execution-based, and management-based dimensions, and an expert scoring mechanism is introduced to determine the initial weight coefficients; Integrate historical performance evaluation and employee competency assessment results to conduct regression modeling on the performance contribution of different ability factors in actual work and optimize the weight distribution model; The final output is a set of key capability factors, along with corresponding job matching weights.

5. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 4 is characterized in that: The process of building a job capability requirement map is as follows: The extracted key competency factors of positions are classified according to position levels and functional categories, and a model of capability sharing and differentiation between positions is established; Construct a dependency diagram between ability factors to identify the order and interaction intensity of different factors in the task completion process; Based on the graph structure, set up a many-to-many mapping relationship between positions and ability factors; The graph database is used to store and query graphs, forming a job competency requirement graph from a competency perspective.

6. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 5, characterized in that: The process of identifying the deviations between positions and employees through a multi-dimensional feature matching algorithm to form a preliminary calibration suggestion set is as follows: Build an employee portrait model covering multiple dimensions; Quantify employee profiles and job competency requirements into multi-dimensional vectors, and use the Euclidean distance metric to assess the match between employees and jobs. Conduct deviation analysis on job-employee combinations whose matching degree is lower than the set threshold, identify capability gaps and job redundancy points, and output a preliminary calibration recommendation set.

7. The method for assisting calibration of tobacco enterprise human resource management based on big data according to claim 6, characterized in that: The process of dynamically correcting the key matching parameters in the initial calibration suggestion set is: Collect the latest strategic goals and task breakdown data of the company's current business segments, and extract new demands for human resources in key business scenarios; Based on the data flow of the task assignment platform, track the real-time task execution status and completion indicators of the positions in actual operations; Adopt a dynamic weight adjustment mechanism to update key parameters in job competency requirements based on task urgency and job activity; For employee-position pairs in the marginal matching area recommended by the preliminary calibration, the deviation is recalculated and the matching value is updated.

8. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 7 is characterized in that: The process of identifying potential imbalances in job configuration is as follows: Analyze the revised employee-job matching matrix to identify jobs where the matching values ​​for key competency dimensions are chronically low or fluctuate frequently; Combining organizational structure change data with historical job adjustment records, identify job units that are frequently deployed but have not formed a stable capacity match; Introducing a job importance scoring mechanism to score and rank the risks caused by capability gaps in key positions; Mark all job sets with scores above a set threshold as potential imbalance points.

9. The tobacco enterprise human resources management auxiliary calibration method based on big data according to claim 8, characterized in that: The process of outputting auxiliary calibration reports to support organizational decision-making includes: Organize identified job imbalances, capability gap types, and potential job mismatch risks into a structured report format and generate a risk heat map visualization interface; Integrate the job transfer costs, training cycles, and business impact indexes involved in each job transfer plan to establish a multi-dimensional influencing factor analysis model; Use intelligent recommendation algorithms to generate several alternative human resource allocation optimization plans, each with a key indicator comparison table, and output an auxiliary calibration report.

Images