Artificial intelligence-based employee grading human resource modeling management system and device

Abstract

The application provides an employee hierarchical human resource modeling management system and device based on artificial intelligence, and relates to the field of human resource management.The employee hierarchical human resource modeling management system comprises a data acquisition module, the data acquisition module is used for acquiring employee information, the employee information comprises identity information, performance information, salary information, skill information and evaluation information; a knowledge processing module, the knowledge processing module constructs a human resource knowledge graph after processing the acquired data.The employee hierarchical human resource modeling management system based on artificial intelligence provided by the application constructs an intelligent grading module based on the human resource knowledge graph and a machine learning model, intelligently grades employees according to multiple index data of the employees, can determine key data in the multiple data of the employees, can more accurately perform rating, and thus the intelligent grading system can realize objective and efficient grading of employees.

Description

Technical Field

[0001] This invention relates to the field of human resource management, and more particularly to an employee grading human resource modeling management system and device based on artificial intelligence. Background Technology

[0002] The enterprise employee hierarchical management system uses employees' abilities, job value, performance, and professional qualities as the core basis to divide employees into different levels / tiers. It matches each level with a differentiated human resource management system of compensation and benefits, training and development, promotion channels, authority and responsibility, and assessment standards. This achieves matching people to positions, determining salaries based on abilities, tiered training, and precise incentives, making human resource allocation more in line with the company's development needs.

[0003] Traditional employee grading relies heavily on managers' experience and subjective evaluations, which leads to problems such as vague standards, biased results, and low efficiency. There is still room for improvement in how to intelligently provide employee grading information based on multi-dimensional data analysis.

[0004] Therefore, it is necessary to provide an AI-based employee tiered human resource modeling management system and device to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention provides an AI-based employee grading human resource modeling and management system, which solves the problem that there is still room for improvement in intelligently generating employee grading information after analyzing multi-dimensional employee data.

[0006] To solve the above-mentioned technical problems, the present invention provides an employee hierarchical human resource modeling and management system, comprising:

[0007] The data acquisition module is used to collect employee information, including identity information, performance information, salary information, skill information, and evaluation information.

[0008] The knowledge processing module processes the collected data to construct a human resources knowledge graph.

[0009] Among them, an intelligent hierarchical module is built based on human resource knowledge graphs and machine learning models;

[0010] The intelligent grading module includes a weighting module, a rating module, a training module, and a job matching module. The weighting module classifies the importance of multiple evaluation indicators for employees based on their job type. The rating module provides an employee rating based on the input employee information. The training module recommends training programs based on the weaknesses in the employee's multiple evaluation indicators. The job matching module provides suitable job selections based on the employee's information.

[0011] The present invention also provides an artificial intelligence-based employee tiered human resource modeling and management device, used in the aforementioned artificial intelligence-based employee tiered human resource modeling and management system, comprising:

[0012] Fingerprint unlocking devices;

[0013] A protective cover structure is slidably disposed on the fingerprint unlocking device and covers the fingerprint sensing area of ​​the fingerprint unlocking device. A cleaning tank is provided on the protective cover structure, and a water inlet pipe is connected to the bottom of the protective cover structure. The water inlet pipe is connected to the cleaning tank.

[0014] A push-pull device, used to drive the protective cover structure to move horizontally;

[0015] A water supply device, wherein the water supply device is installed on the fingerprint unlocking device;

[0016] When the push-pull device drives the protective cover structure to move a preset distance and separates it from the fingerprint sensing area, the water inlet pipe is connected to the water outlet of the water supply device.

[0017] Compared with related technologies, the employee hierarchical human resource modeling and management system based on artificial intelligence provided by this invention has the following beneficial effects:

[0018] This invention provides an AI-based employee grading human resource modeling and management system. It constructs an intelligent grading module based on a human resource knowledge graph and machine learning model. The system intelligently grades employees according to multiple performance indicators, identifying key data points for more accurate ratings. It also recommends training programs to address employee weaknesses, improving their skills, and suggests more suitable positions based on employee information. Therefore, this intelligent grading system enables objective and efficient employee grading. Attached Figure Description

[0019] Figure 1 A schematic diagram of the principle of the AI-based employee hierarchical human resource modeling and management system provided by the present invention;

[0020] Figure 2 A block diagram illustrating the composition of the intelligent hierarchical module provided by this invention;

[0021] Figure 3 A block diagram illustrating the composition of the knowledge processing module provided by this invention;

[0022] Figure 4 A flowchart outlining the steps involved in building the intelligent grading module;

[0023] Figure 5 A block diagram illustrating the composition of the key module provided by this invention;

[0024] Figure 6 A flowchart illustrating the steps of the key module operation method provided by the present invention;

[0025] Figure 7 A schematic diagram of the structure of the device for employee hierarchical human resource modeling and management based on artificial intelligence provided by the present invention;

[0026] Figure 8 A schematic diagram of the structure of the AI-based employee hierarchical human resource modeling and management device after removing part of the second cover, as provided in this invention.

[0027] Figure 9 A schematic diagram of the structure for removing the protective cover of the equipment used in the AI-based employee hierarchical human resource modeling and management system provided by the present invention;

[0028] Figure 10 A partial cross-sectional view of the AI-based employee hierarchical human resource modeling and management device provided by the present invention.

[0029] Figure 11 This is a schematic diagram illustrating the working principle of the protective cover structure provided by the present invention, wherein... Figure 11 (a) is a schematic diagram showing the state after the protective cover structure is separated from the fingerprint sensing area. Figure 11 (b) is a schematic diagram showing the state of the cleaning solution entering the cleaning tank;

[0030] Figure 12 This is a schematic diagram illustrating the working principle of the push-pull device provided by the present invention, wherein... Figure 12 (a) is a schematic diagram showing the state in which the push-pull device drives the protective cover structure to move and separates it from the fingerprint sensing area. Figure 12 (b) is a schematic diagram of the push-pull device driving the second cover to separate from the first cover, causing the cleaning tank to leak out.

[0031] Numbering on the map:

[0032] 1. Fingerprint unlocking device; 101. Fingerprint sensing area; 102. Water-absorbing block;

[0033] 2. Stop;

[0034] 3. Protective cover structure; 31. Driven cover; 32. Active cover;

[0035] 311. First cover; 312. Connector; 313. Sliding rod; 303. Water inlet pipe;

[0036] 321. Second cover; 322. Moving plate; 301. Cleaning tank; 302. Positioning block;

[0037] 4. Push-pull device; 41. Drive device; 42. Drive shaft; 43. Rotating plate; 44. Protruding shaft; 421. Rotating component;

[0038] 5. Water supply device; 51. Cylinder; 52. Piston block; 53. Positioning pipe. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0040] This invention provides an employee grading human resource modeling management system and device based on artificial intelligence.

[0041] Please refer to the following: Figure 1 and Figure 2 In one embodiment of the present invention, the AI-based employee tiered human resource modeling and management system includes:

[0042] The data acquisition module is used to collect employee information, including identity information, performance information, salary information, skill information, and evaluation information.

[0043] The knowledge processing module processes the collected data to construct a human resources knowledge graph.

[0044] Among them, an intelligent hierarchical module is built based on human resource knowledge graphs and machine learning models;

[0045] The intelligent grading module includes a weighting module, a rating module, a training module, and a job matching module. The weighting module classifies the importance of multiple evaluation indicators for employees based on their job type. The rating module provides an employee rating based on the input employee information. The training module recommends training programs based on the weaknesses in the employee's multiple evaluation indicators. The job matching module provides suitable job selections based on the employee's information.

[0046] Based on the construction of a human resources knowledge graph and machine learning model, an intelligent grading module is built. This module intelligently grades employees according to multiple indicator data and can identify key data points for more accurate rating. It can also recommend corresponding training programs based on the weaknesses in employee evaluation indicators to improve employee capabilities. Furthermore, it can suggest more suitable positions for employees based on their information. Thus, this intelligent grading system can achieve objective and efficient employee grading.

[0047] The employee information includes: identity information such as education, age, and position; performance information such as KPI completion rate, project results, and attendance data; salary information such as current salary data and expected salary data; skills information such as skills certificates, training results, job competency assessment results, and professional ability examination data; and evaluation information such as feedback data from colleagues and superiors and subordinates, and work behavior data.

[0048] This intelligent rating module can identify core talent, high-potential talent, and talent needing improvement for enterprises by rating employees multiple times. If the core data improves in each rating, the employee is identified as high-potential talent. If the core data is consistently at a high level with minimal fluctuations in each assessment, the employee is identified as core talent. If the assessment data has obvious shortcomings or significant fluctuations, the employee is identified as talent needing improvement.

[0049] Once the management system is built, it will be presented as an app. When using it, users can input employee indicator information, and the intelligent rating module will provide rating results.

[0050] Among the various data points for employees, key data points are determined as follows: for entry-level positions, the focus is on execution ability and performance (total percentage ≥ 70%); for middle-level positions, the percentage of management and coordination ability is increased; and for senior-level positions, the focus is on strategy, decision-making ability, and potential. For example, when evaluating technical employees, the graded indicators are "professional ability (40%), project performance (30%), learning potential (20%), and professional qualities (10%)".

[0051] Please see Figure 3 In this embodiment, the knowledge processing module includes a knowledge extraction module, a knowledge fusion module, a knowledge processing module, and a graph construction module. The knowledge extraction module is used to transform the collected employee data into quantifiable and computable structured data. The knowledge fusion module enables the structured data to form a unified human resources knowledge standard. The knowledge processing module is used to transform the data processed by the knowledge fusion module into standardized, computable hierarchical rules and quantitative indicators, and form a quantitative hierarchical indicator library. The graph construction module visualizes and structures the knowledge processed by the knowledge processing module to form a human resources knowledge graph.

[0052] The knowledge extraction module incorporates data collection and data cleaning to extract structured data from employees' unstructured data.

[0053] Includes: Text data: feedback from superiors and subordinates, promotion defense records, employee work summaries, interview evaluations, etc. Core information is extracted through NLP entity recognition, relation extraction, and attribute extraction (such as extracting from comments: strong communication skills, low project implementation efficiency, leading the team to complete 2 core projects, etc.).

[0054] For audio and video data: defense audio and offline assessment videos are first converted to text using ASR, and then the above text extraction is performed;

[0055] Semi-structured data: Employee skill certificates, training records, etc., are used to extract key attributes (such as certificate name, level, assessment results, etc.) through rules and deep learning.

[0056] Because human resources data comes from diverse sources, has different formats, and inconsistent definitions (e.g., different departments have different definitions of "excellent performance"; different standards for "competency rating"), knowledge integration is the solution to this problem of data silos and inconsistent definitions.

[0057] This includes entity integration: unifying the naming of core entities such as employees, positions, capabilities, and performance; for example, merging "R&D position" and "Technical Development position" into "Technical R&D position";

[0058] Attribute integration: unify the evaluation criteria for the same attribute (e.g., unify the "performance A, B, C" and "ability excellent, good, average" ratings of different departments into a standardized enterprise-level rating system);

[0059] Knowledge Source Integration: Integrate multi-source data from HR systems, performance systems, training systems, and office collaboration systems to link the full-dimensional knowledge of the same employee, such as linking an employee's training results with performance outcomes and competency evaluations;

[0060] The knowledge processing module transforms "fuzzy talent evaluation knowledge" into standardized, computable grading rules and quantitative indicators, providing a clear training basis for AI grading models;

[0061] This includes knowledge reasoning: uncovering hidden knowledge connections through logical reasoning and ontology reasoning (e.g., reasoning that employees with strong project implementation capabilities are likely to have high execution and problem-solving abilities; reasoning that employees with consistently excellent performance and minimal disruption are core talents of the company).

[0062] Knowledge quantification: transforming qualitative knowledge into quantitative scores, such as converting excellent, good, and average communication skills into "90-100 points, 70-89 points, and 60-69 points", and converting team leadership in completing projects into specific performance bonuses based on project difficulty and results;

[0063] Knowledge screening: Combining association rule mining and factor analysis, we screen out the core knowledge that has the greatest impact on employee classification from a massive amount of basic knowledge; for example, we remove irrelevant attributes such as employee height and retain core attributes such as professional ability, project performance, and learning potential.

[0064] The knowledge graph construction module uses entities (employees, positions, abilities, performance, training, etc.) as nodes and relationships as edges (such as the abilities possessed by employees, the positions supported by those abilities, and the relationship between performance and those abilities), to build a visualized knowledge network for enterprise human resources. It supports hierarchical indicator modeling, hierarchical algorithm training, talent matching, and the application of hierarchical results.

[0065] Please see Figure 4 In this embodiment, the construction of the intelligent grading module specifically includes the following steps:

[0066] S01. Acquire employee data and perform preprocessing such as labeling, noise reduction, and dimensionality reduction on the data. The sources of employee data include a quantitative indicator library after knowledge processing, and extended feature data extracted from the human resources knowledge graph through inference engine or association query.

[0067] S02. Divide the preprocessed data feature set into a training set, a validation set, and a test set, which are used for model training, parameter tuning, and performance verification, respectively.

[0068] S03. Input the training set into the initial model to complete the basic training;

[0069] S04. Find the optimal combination of hyperparameters for the model through tuning;

[0070] S05. Apply the optimal combination of hyperparameters to the model, and retrain it using the full dataset of the training set and validation set to obtain the final model with the best performance.

[0071] S06. Input the test set into the final model for technical verification.

[0072] S01. Extract core features from the quantitative indicator library after knowledge processing, and label the features according to the indicator weights. This includes hard features: education level, job level, attendance compliance rate, etc., which are quantified into numerical / level codes; performance features: project result scores, KPI completion rate, annual and quarterly performance levels, which are quantified into 0-100 points; and ability features: professional ability and soft skills ability in various dimensions are quantified.

[0073] Extended features are extracted from the human resources knowledge graph through reasoning engines and related queries, and quantified into numerical values. These include: job matching features: the degree of matching between the employee's competency graph and the level standard of the job (0-100%); talent value features: the degree of fit between the employee's core competencies and the core job requirements of the enterprise (0-100%); and competency association features: the achievement rate of related competencies of the employee's high-scoring competencies (such as the average achievement rate of related competencies such as project implementation competency of 80 points and execution and planning competency).

[0074] For feature annotation, the basic features and graph extended features are integrated into a total feature set. A feature data table is constructed using Pandas, with rows representing employee samples, columns representing feature dimensions, and values ​​representing quantified feature values.

[0075] The data is labeled, denoised, and dimensionality reduced, including:

[0076] Missing value handling:

[0077] Numerical features: filled with mean, median, and mode;

[0078] Business-related characteristics: If the job matching degree is missing as shown in the knowledge graph, use knowledge graph reasoning to complete it or mark it as "0" (representing no match); Outlier handling: Use box plot method or 3σ principle to identify outliers (e.g., an employee's performance score is 100 points, and all others are below 60 points), and combine with business judgment: if it is real data, keep it; if it is erroneous data, correct or delete it.

[0079] Standardize or normalize the feature scale to eliminate dimensional differences;

[0080] Feature dimensionality reduction: When there are too many feature dimensions (e.g., more than 50 dimensions), use PCA (principal component analysis) or LDA (linear discriminant analysis) to reduce dimensionality, retain principal components with a cumulative variance contribution rate ≥95%, reduce model computation, and avoid overfitting;

[0081] Feature selection: Combining business logic and statistical methods to remove irrelevant features improves model accuracy.

[0082] Statistical methods: Calculate the correlation coefficient between features and hierarchical labels (supervised), perform analysis of variance (ANOVA), and remove features with a correlation coefficient < 0.2;

[0083] Business approach: Combine enterprise classification rules to eliminate features that are irrelevant to classification (such as employee height and place of origin).

[0084] Output: Clean feature set after preprocessing;

[0085] In S02, the ratio of the training set, validation set, and test set for supervised classification models is 7:2:1 or 8:1:1. Supervised classification models provide enterprises with clear historical classification labels, and then use AI to replicate and optimize the classification logic.

[0086] In S03, a tree model is preferred, with random forest being the best option. XGBoost or LightGBM can also be used.

[0087] In S04, the training set features and labels are input into the classification model, and training is performed according to the preset hyperparameters. The model parameters are optimized through forward propagation and backpropagation. During the training process, the training set loss value / accuracy is monitored to ensure that the loss value continues to decrease and the accuracy gradually increases without training interruption. When the preset number of iterations is reached, or the loss value converges, or the training set accuracy reaches the preset threshold (such as 80%), the initial training model is output.

[0088] In S05, the optimal combination of hyperparameters for the model is found through tuning, allowing the model to achieve its best performance on the validation set and balancing overfitting (accurate on the training set, poor on the test set) and underfitting.

[0089] Optimization methods include grid search: enumerating preset hyperparameter combinations, iterating and verifying them, which has the best effect but is time-consuming and is suitable for small feature sets / small sample sizes;

[0090] Random search: Random sampling and verification within the hyperparameter range is highly efficient and suitable for large feature sets / large sample sizes.

[0091] Among them, overfitting can be addressed by: increasing the regularization term, decreasing the tree depth, reducing the number of iterations, and increasing the number of training samples.

[0092] Underfitting: Increase tree depth, number of iterations, enrich feature set, and reduce regularization strength.

[0093] When the validation set metric reaches its optimal value and no longer increases, record the optimal combination of hyperparameters at this point.

[0094] The optimal hyperparameters are applied to the model, and the model is retrained using the full dataset of the training and validation sets to obtain the final model with the best performance, thus maximizing the value of the data.

[0095] Merge the training and validation sets to form a full training set; assign the optimal hyperparameters to the model, perform final training using the full training set, and train until the loss value converges; monitor the full training set metrics during training to ensure that the metrics are basically consistent with the validation set metrics during tuning (deviation < 5%), and output the final model.

[0096] Please see Figure 1 As a preferred embodiment, the AI-based employee rating human resource modeling management system also includes an annotation module, which is used to provide the basis information for each employee's rating.

[0097] By annotating the rating criteria information and sending it to the relevant employees, the rating information becomes more credible.

[0098] After the intelligent grading module outputs the grading results, managers can make adjustments based on the actual situation of the employees.

[0099] Please see Figure 5 As a preferred embodiment, the AI-based employee hierarchical human resource modeling and management system further includes a key module, which includes a password module, a fingerprint module, and a verification module. The password module is used to set a symbolic key, the fingerprint module is used to input a fingerprint key, and the verification module is used to input a verification code.

[0100] Because the system stores a large amount of employee data, including basic data, skills data, and business capability data, it is necessary to strengthen the system's security protection to reduce the likelihood of other enterprises or individuals stealing the company's employee data or employees tampering with the data.

[0101] This invention requires multiple verification methods, including setting a key, fingerprint, and verification code, before login to the management system. This multi-layered verification approach further enhances the protection of the system.

[0102] Please see Figure 6 The operation method of the key module specifically includes the following steps:

[0103] S11. Set the symbolic key and collect the fingerprint key;

[0104] S12. When unlocking, enter a symbolic key and enter a fingerprint key n times within a preset time; wherein, the n fingerprint keys are the same fingerprint or different fingerprints;

[0105] S13. When both the symbolic key and the fingerprint key are correct, input the data of a multiplied by n into the verification module as the verification code.

[0106] When logging into the system, you enter your account and symbol key. If the symbol key is correct, you enter your fingerprint key. When entering the fingerprint key, you need to sense your fingerprint n times within a preset time (e.g., if n is 2, you need to enter your fingerprint twice, with an interval of three seconds between adjacent fingerprint inputs. The two fingerprint inputs can be the same finger used for two consecutive inputs, or two different fingers used for one input each). If both inputs are correct, the system will pop up a verification code input window. Enter the preset number 'a' multiplied by 'n' to log in (e.g., if 'a' is 7, then the verification code is 2 multiplied by 7, which is 14). Enter '14' in the verification code to complete the login.

[0107] When the verification code pops up, reserve 3 to 10 character positions. During verification, it is not necessary to fill all the character positions. For example, in the above example, only two positions need to be filled. It is preferable to set six character positions.

[0108] The symbol key consists of one or more of the following: numbers, letters, punctuation marks, and English symbols.

[0109] Please see Figures 7 to 9 The present invention also provides an artificial intelligence-based employee hierarchical human resource modeling and management device for the aforementioned artificial intelligence-based employee hierarchical human resource modeling and management system, comprising: a fingerprint unlocking device 1;

[0110] A protective cover structure 3 is slidably disposed on the fingerprint unlocking device 1 and covers the fingerprint sensing area 101 of the fingerprint unlocking device 1. A cleaning groove 301 is provided on the protective cover structure 3, and a water inlet pipe 303 is connected to the bottom of the protective cover structure 3. The water inlet pipe 303 is connected to the cleaning groove 301.

[0111] Push-pull device 4, which is used to drive the protective cover structure 3 to move horizontally;

[0112] Water supply device 5, which is installed on the fingerprint unlocking device 1;

[0113] When the push-pull device 4 drives the protective cover structure 3 to move a preset distance and separate from the fingerprint sensing area 101, the water inlet pipe 303 is connected to the water outlet of the water supply device 5.

[0114] The device provided by this invention is used in the fingerprint module of the key module, that is, for collecting, storing and unlocking the user's fingerprint.

[0115] The terminal device that stores and executes the management system is connected to the fingerprint unlocking device 1 via USB; the user's finger contacts the fingerprint sensing area 101 of the fingerprint unlocking device 1 to collect the user's fingerprint and the fingerprint used for subsequent verification.

[0116] In this embodiment, by setting a protective cover structure 3 on the fingerprint unlocking device 1, it can further play a protective role, preventing unauthorized personnel from moving the user's finger to the fingerprint sensing area 101 to unlock the device when the user is not mentally awake, and using the protective cover structure 3 to cover the fingerprint sensing area 101 to prevent dust and other foreign objects from falling into the fingerprint sensing area 101, thereby causing the fingerprint sensing area 101 to become insensitive, and at the same time extending the service life of the fingerprint unlocking device 1.

[0117] Furthermore, by setting a cleaning groove 301 on the protective cover structure 3, when the fingerprint unlocking device 1 is used, the push-pull device 4 pulls the protective cover structure 3 to move a preset distance and separate it from the fingerprint sensing area 101, and the water inlet pipe 303 connected to the cleaning groove 301 is connected to the water outlet of the water supply device 5.

[0118] When a user's fingers are covered with dust, food residue, or oil, the water supply device 5 injects a small amount of cleaning solution into the cleaning tank 301. The user can then wet their fingers and wipe them with a tissue, or wet a tissue before wiping their fingers to clean them. If the fingerprint sensing area 101 is also covered with foreign objects, it can be wiped in the same way, making the fingerprint unlocking device 1 more convenient and efficient to use.

[0119] After cleaning, it is preferable to use paper towels or similar materials to wipe away the water stains in the cleaning tank 301.

[0120] To open the protective cover structure 3, a separate symbolic password can be set. A password input area can be set on the login page of the management system or on the fingerprint unlocking device 1. When the password is entered, the protective cover structure 3 will open; or a symbolic login key set in the password module can be used.

[0121] As an optional embodiment, the protective cover structure 3 can be a rectangular box with an opening at the bottom.

[0122] Please see Figure 7 , Figure 8 and Figure 10 As another optional embodiment, the protective cover structure 3 includes a driven cover 31 and an active cover 32. The driven cover 31 includes a first cover body 311 and a connector 312. The cleaning groove 301 is formed on the first cover body 311. The active cover 32 includes a second cover body 321. The first cover body 311 is detachably connected to the second cover body 321 through the connector 312. One end of the second cover body 321 covers the cleaning groove 301. The output end of the push-pull device 4 is connected to the second cover body 321.

[0123] By using the passive cover 31 and the active cover 32, where the second cover 321 covers the cleaning tank 301, the cleaning tank 301 can be protected, preventing dust and other foreign objects from falling into the cleaning tank 301 and contaminating the cleaning tank 301, as well as from falling onto the fingerprint sensing area 101 through the cleaning tank 301 and the water inlet pipe 303 in sequence.

[0124] In one embodiment, the connector 312 comprises multiple magnetic blocks and multiple adsorption blocks. The multiple magnetic blocks are embedded in the first cover 311 and located on both sides of the cleaning tank 301. The adsorption blocks are embedded in the second cover 321 and are positioned corresponding to the magnetic blocks. The adsorption blocks are metal blocks that can be attracted to the magnetic blocks. Positioning blocks 302 are installed on both sides of the first cover 311. A stop block 2 is installed on the fingerprint unlocking device 1, and the stop block 2 is spaced at a preset distance from the positioning block 302.

[0125] When the push-pull device 4 pulls the second cover 321 to move, the second cover 321 drives the first cover 311 to move along with it through the action of the magnetic block and the adsorption block. When the first cover 311 separates from the fingerprint sensing area 101, the positioning block 302 abuts against the stop block 2. When the push-pull device 4 drives the second cover 321 to move away from the fingerprint sensing area 101 again, the first cover 311 cannot move because the stop block 2 limits the positioning block 302. At this time, the second cover 321 separates from the first cover 311, and the cleaning tank 301 leaks out. Figure 12 (a) and Figure 12 (b)

[0126] Among them, such as Figure 7 A stop block 2 is provided on the left side of the first cover 311 to position the first cover 311 so that the subsequent second cover 321 can overlap with the first cover 311 again, and the adsorption block and the magnetic block are aligned and adsorbed.

[0127] In another embodiment, the connector 312 includes multiple sets of elastic structures, each including a spring and a trapezoidal abutment plate. The trapezoidal abutment plate is mounted on the bottom of the second cover 321 by the spring. The trapezoidal block is pushed against the first cover 311 by the elastic member in a compressed state. Through friction, the first cover 311 can move together with the second cover 321 without being restricted.

[0128] Subsequently, the first cover 311 presses against the inclined surface of the trapezoidal abutment plate, causing it to compress the spring and abut against the first cover 311 again.

[0129] Among them, such as Figure 8 Both ends of the first cover 311 are equipped with sliding rods 313, and both ends of the second cover 321 are correspondingly sleeved on the sliding rods 313, so that the first cover 311 and the second cover 321 form a sliding connection.

[0130] Please see Figure 9 A water-absorbing block 102 is provided on the top of the fingerprint unlocking device 1 and between the fingerprint sensing area 101 and the water supply device 5. When the protective cover structure 3 moves to the driving position to cover and protect the fingerprint sensing area 101, the bottom end of the water inlet pipe 303 can interact with the water-absorbing block 102 to absorb any water droplets that may be attached to the bottom end of the water inlet pipe 303, thus preventing them from dripping onto the fingerprint sensing area 101.

[0131] The absorbent block 102 is made of sponge or cloth, etc.

[0132] Please refer to it again. Figure 10In this embodiment, the push-pull device 4 includes a drive device 41, a drive shaft 42, a rotating plate 43, and a convex shaft 44. The drive device 41 is installed on the fingerprint unlocking device 1 and is used to drive the drive shaft 42 to rotate. One end of the rotating plate 43 is installed on the drive shaft 42, and the convex shaft 44 is installed on the other end of the rotating plate 43.

[0133] The second cover 321 also includes two driving plates 322, which are installed at intervals at the bottom of the second cover 321, and the top end of the convex shaft 44 extends between the two driving plates 322.

[0134] Please refer to the following: Figure 12 (a) and Figure 12 In (b), when the push-pull device 4 is working, the drive device 41 drives the drive shaft 42 to rotate counterclockwise, the drive shaft 42 drives the convex shaft 44 to rotate eccentrically, the convex shaft 44 moves between the two drive plates 322, and pushes the drive plates 322 to move the second cover 321 away from the fingerprint sensing area 101. When the rotating plate 43 rotates 90 degrees, the second cover 321 drives the first cover 311 to separate from the fingerprint sensing area 101. When the rotating plate 43 rotates 90 degrees again, the second cover 321 separates from the first cover 311, and the cleaning groove 301 leaks out. Subsequently, it rotates 180 degrees again, pushing the second cover 321 to move to the original position. At the same time, the second cover 321 drives the first cover 311 to the original position, realizing the movement of the drive protective cover structure 3.

[0135] Of course, in other instances, the push-pull device 4 can also be an electric push cylinder, which is mounted on the fingerprint unlocking device 1 via a mounting plate, and the output end of the electric push cylinder is connected to the second cover 321.

[0136] Please see Figure 10 In this example, the water supply device 5 includes a cylinder 51, a piston block 52, and a positioning tube 53. The cylinder 51 is installed at one end of the fingerprint unlocking device 1. The top end of the drive shaft 42 passes through the cylinder 51 and is connected to the rotating plate 43 through a rotating member 421. The drive shaft 42 can only be driven to rotate in one direction through the rotating member 421. The piston block 52 is threaded to the drive shaft 42 and is located inside the cylinder 51. The top end of the positioning tube 53 passes through and is fixed to the top of the cylinder 51. The bottom end of the positioning tube 53 passes through the piston block 52 and leaves a gap between it and the bottom of the cylinder 51.

[0137] In this embodiment, the surface of the drive shaft 42 is provided with threads, and the piston block 52 is threadedly connected to the drive shaft 42.

[0138] In the initial state, such as Figure 10 A gap is left between the piston block 52 and the surface of the cleaning fluid. When the protective cover structure 3 moves to the point where the positioning block 302 on the first cover 311 abuts against the stop block 2, that is, when the first cover 311 separates from the fingerprint sensing area 101, the drive shaft 42 rotates, causing the piston block 52 to move down, and its bottom just contacts the surface of the cleaning fluid. The water inlet pipe 303 is aligned and connected with the top of the positioning pipe 53. Figure 11 (a) and Figure 11 (b)

[0139] If cleaning fluid is needed later, the drive device 41 continues to rotate 90 degrees counterclockwise to separate the second cover 321 from the first cover 311. At the same time, the piston block 52 continues to descend, allowing the cleaning fluid to enter the cleaning tank 301 through the positioning pipe 53 and the water inlet pipe 303.

[0140] After use, the drive unit 41 continues to drive the drive shaft 42 to rotate counterclockwise, so that the protective cover structure 3 moves to its original position and covers the fingerprint sensing area 101. Then, the drive unit 41 rotates 90 degrees clockwise, so that the piston block 52 moves up and leaves a gap with the liquid surface again.

[0141] Furthermore, when the drive shaft 42 rotates clockwise, it will not cause the rotating plate 43 to rotate, so the protective cover structure 3 will not move at this time.

[0142] Thus, the push-pull device 4 can be used to drive the protective cover structure 3 to move in one state, and can be used to drive the piston block 52 to move in another state, so as to realize the function of supplying water to the cleaning tank 301;

[0143] Furthermore, during the state switching process, the first cover 311 and the second cover 321 are separated, allowing the cleaning tank 301 to leak out.

[0144] The water inlet pipe 303 has a flange at its bottom end, and a sealing ring is bonded to the bottom of the flange. When the water inlet pipe 303 is connected to the positioning pipe 53, the sealing ring is tightly attached to the positioning pipe 53, thereby ensuring the sealing of the connection.

[0145] In one embodiment, the rotating component 421 is a one-way bearing, with the inner ring of the one-way bearing mounted on the top of the drive shaft 42 and the rotating plate 43 mounted on the outer ring of the one-way bearing.

[0146] In another embodiment, the rotating component 421 includes an annular sleeve, a ratchet, and a driving block. The ratchet is elastically connected to the inner wall of the annular sleeve and can only rotate in one direction. The driving block is mounted on the drive shaft 42. The annular sleeve is fitted on the drive shaft 42 and rotatably connected to the cylinder 51. The driving block is located inside the annular sleeve. The rotating plate 43 is fitted on and fixed to the annular sleeve.

[0147] Please see Figure 9The top of the cylinder 51 is provided with a liquid replenishment port, and a rubber stopper is provided on the liquid replenishment port. An annular sleeve is connected to the rubber stopper. By pulling the annular sleeve, the rubber stopper is separated from the cylinder 51, and cleaning fluid can be replenished into the cylinder 51.

[0148] The working principle of the AI-based employee tiered human resource modeling and management device provided by this invention is as follows:

[0149] The terminal device that stores and executes the management system is connected to the fingerprint unlocking device 1 via USB; the user's finger contacts the fingerprint sensing area 101 of the fingerprint unlocking device 1 to collect the user's fingerprint and the fingerprint used for subsequent verification.

[0150] In this embodiment, by setting a protective cover structure 3 on the fingerprint unlocking device 1, it can further play a protective role, preventing unauthorized personnel from moving the user's finger to the fingerprint sensing area 101 to unlock the device when the user is not mentally awake, and using the protective cover structure 3 to cover the fingerprint sensing area 101 to prevent dust and other foreign objects from falling into the fingerprint sensing area 101, thereby causing the fingerprint sensing area 101 to become insensitive, and at the same time extending the service life of the fingerprint unlocking device 1.

[0151] Furthermore, by setting a cleaning groove 301 on the protective cover structure 3, when the fingerprint unlocking device 1 is used, the push-pull device 4 pulls the protective cover structure 3 to move a preset distance and separate it from the fingerprint sensing area 101, and the water inlet pipe 303 connected to the cleaning groove 301 is connected to the water outlet of the water supply device 5.

[0152] When the user's fingers are covered with dust, food residue, or oil, the water supply device 5 will input a small amount of cleaning solution into the cleaning tank 301. The user can then wet their fingers and wipe them with a tissue, or wet the tissue before wiping their fingers to clean them. If the fingerprint sensing area 101 is also covered with foreign objects, it can be wiped in the same way.

[0153] For details, please refer to the following: Figure 12 (a) and Figure 12 In (b), when the push-pull device 4 is working, the drive device 41 drives the drive shaft 42 to rotate counterclockwise, the drive shaft 42 drives the convex shaft 44 to rotate eccentrically, the convex shaft 44 moves between the two drive plates 322, and pushes the drive plates 322 to move the second cover 321 away from the fingerprint sensing area 101. When the rotating plate 43 rotates 90 degrees, the second cover 321 drives the first cover 311 to separate from the fingerprint sensing area 101. When the rotating plate 43 rotates 90 degrees again, the second cover 321 separates from the first cover 311, and the cleaning groove 301 leaks out. Subsequently, it rotates 180 degrees again, pushing the second cover 321 to move to the original position. At the same time, the second cover 321 drives the first cover 311 to the original position, realizing the movement of the drive protective cover structure 3.

[0154] The push-pull device 4 pulls the second cover 321 to move. The second cover 321 drives the first cover 311 to move along with it through the action of the magnetic block and the adsorption block. When the first cover 311 separates from the fingerprint sensing area 101, the positioning block 302 abuts against the stop block 2. When the push-pull device 4 drives the second cover 321 to move away from the fingerprint sensing area 101 again, the first cover 311 cannot move because the stop block 2 limits the positioning block 302. At this time, the second cover 321 separates from the first cover 311, and the cleaning tank 301 leaks out.

[0155] The specific working principle of water supply device 5 is as follows: In the initial state, such as... Figure 10 A gap is left between the piston block 52 and the surface of the cleaning fluid. When the protective cover structure 3 moves to the point where the positioning block 302 on the first cover 311 abuts against the stop block 2, that is, when the first cover 311 separates from the fingerprint sensing area 101, the drive shaft 42 rotates, causing the piston block 52 to move down, and its bottom just contacts the surface of the cleaning fluid. The water inlet pipe 303 is aligned and connected with the top of the positioning pipe 53. Figure 11 (a) and Figure 11 (b)

[0156] If cleaning fluid is needed later, the drive device 41 continues to rotate 90 degrees counterclockwise to separate the second cover 321 from the first cover 311. At the same time, the piston block 52 continues to descend, allowing the cleaning fluid to enter the cleaning tank 301 through the positioning pipe 53 and the water inlet pipe 303.

[0157] After use, the drive unit 41 continues to drive the drive shaft 42 to rotate counterclockwise, so that the protective cover structure 3 moves to its original position and covers the fingerprint sensing area 101. Then, the drive unit 41 rotates 90 degrees clockwise, so that the piston block 52 moves up and leaves a gap with the liquid surface again.

[0158] Furthermore, when the drive shaft 42 rotates clockwise, it will not cause the rotating plate 43 to rotate, so the protective cover structure 3 will not move at this time.

[0159] Thus, the push-pull device 4 can be used to drive the protective cover structure 3 to move in one state, and can be used to drive the piston block 52 to move in another state, so as to realize the function of supplying water to the cleaning tank 301;

[0160] Furthermore, during the state switching process, the first cover 311 and the second cover 321 are separated, allowing the cleaning tank 301 to leak out.

[0161] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. An employee hierarchical human resource modeling and management system based on artificial intelligence, characterized in that, include: The data acquisition module is used to collect employee information, including identity information, performance information, salary information, skill information, and evaluation information. The knowledge processing module processes the collected data to construct a human resources knowledge graph. Among them, an intelligent hierarchical module is built based on human resource knowledge graphs and machine learning models; The intelligent grading module includes a weighting module, a rating module, a training module, and a job matching module. The weighting module classifies the importance of multiple evaluation indicators for employees based on their job type. The rating module provides an employee rating based on the input employee information. The training module recommends training programs based on the weaknesses of employees among multiple evaluation indicators, and the job matching module provides suitable job selections based on the employee's information.

2. The employee hierarchical human resource modeling and management system based on artificial intelligence according to claim 1, characterized in that, The knowledge processing module includes a knowledge extraction module, a knowledge fusion module, a knowledge processing module, and a graph construction module. The knowledge extraction module is used to transform the collected employee data into quantifiable and computable structured data. The knowledge fusion module enables structured data to form a unified human resources knowledge framework; The knowledge processing module is used to transform the data processed by the knowledge fusion module into standardized, computable hierarchical rules and quantitative indicators, and form a quantitative hierarchical indicator library; the graph construction module visualizes and structures the knowledge processed by the knowledge processing module to form a human resources knowledge graph.

3. The employee hierarchical human resource modeling and management system based on artificial intelligence according to claim 1, characterized in that, The construction of the intelligent hierarchical module specifically includes the following steps: S01. Acquire employee data and perform preprocessing such as labeling, noise reduction, and dimensionality reduction on the data. The sources of employee data include a quantitative indicator library after knowledge processing, and extended feature data extracted from the human resources knowledge graph through inference engine or association query. S02. Divide the preprocessed data feature set into a training set, a validation set, and a test set, which are used for model training, parameter tuning, and performance verification, respectively. S03. Input the training set into the initial model to complete the basic training; S04. Find the optimal combination of hyperparameters for the model through tuning; S05. Apply the optimal combination of hyperparameters to the model, and retrain it using the full dataset of the training set and validation set to obtain the final model with the best performance. S06. Input the test set into the final model for technical verification.

4. The employee hierarchical human resource modeling and management system based on artificial intelligence according to claim 1, characterized in that, It also includes an annotation module, which provides information on the basis for each employee's rating.

5. The employee hierarchical human resource modeling and management system based on artificial intelligence according to claim 1, characterized in that, It also includes a key module, which comprises a cryptographic module, a fingerprint module, and a verification module. The cryptographic module is used to set a symbolic key, the fingerprint module is used to input a fingerprint key, and the verification module is used to input a verification code.

6. The employee hierarchical human resource modeling and management system based on artificial intelligence according to claim 5, characterized in that, The operation method of the key module specifically includes the following steps: S11. Set the symbolic key and collect the fingerprint key; S12. When unlocking, enter a symbolic key and enter a fingerprint key n times within a preset time; wherein, the n fingerprint keys are the same fingerprint or different fingerprints; S13. When both the symbolic key and the fingerprint key are correct, input the data of a multiplied by n into the verification module as the verification code.

7. A device for employee hierarchical human resource modeling and management based on artificial intelligence, characterized in that, The AI-based employee tiered human resource modeling and management system as described in any one of claims 1-6 includes: Fingerprint unlocking devices; A protective cover structure is slidably disposed on the fingerprint unlocking device and covers the fingerprint sensing area of ​​the fingerprint unlocking device. A cleaning tank is provided on the protective cover structure, and a water inlet pipe is connected to the bottom of the protective cover structure. The water inlet pipe is connected to the cleaning tank. A push-pull device, used to drive the protective cover structure to move horizontally; A water supply device, wherein the water supply device is installed on the fingerprint unlocking device; When the push-pull device drives the protective cover structure to move a preset distance and separates it from the fingerprint sensing area, the water inlet pipe is connected to the water outlet of the water supply device.

8. The device for employee hierarchical human resource modeling and management based on artificial intelligence according to claim 7, characterized in that, The protective cover structure includes a driven cover and an active cover. The driven cover includes a first cover body and a connector. The cleaning groove is formed on the first cover body. The active cover includes a second cover body. The first cover body is detachably connected to the second cover body through the connector. One end of the second cover body covers the cleaning groove. The output end of the push-pull device is connected to the second cover body.

9. The device for employee hierarchical human resource modeling and management based on artificial intelligence according to claim 8, characterized in that, The push-pull device includes a drive device, a drive shaft, a rotating plate, and a convex shaft. The drive device is installed on the fingerprint unlocking device and is used to drive the drive shaft to rotate. One end of the rotating plate is installed on the drive shaft, and the convex shaft is installed on the other end of the rotating plate. The second cover also includes two drive plates, which are installed at intervals at the bottom of the second cover, and the top end of the convex shaft extends between the two drive plates.

10. The device for employee hierarchical human resource modeling and management based on artificial intelligence according to claim 9, characterized in that, The water supply device includes a cylinder, a piston block, and a positioning tube. The cylinder is installed at one end of the fingerprint unlocking device. The top end of the drive shaft passes through the cylinder and is connected to the rotating plate through a rotating component. The drive shaft can only be rotated in one direction through the rotating component. The piston block is threaded to the drive shaft and is located inside the cylinder. The top end of the positioning tube passes through and is fixed to the top of the cylinder. The bottom end of the positioning tube passes through the piston block and leaves a gap between it and the bottom of the cylinder.

Images