Employee departure management system

Abstract

The embodiment of the invention relates to the field of attendance management, and discloses an employee leaving management system. The system comprises a departure request receiving module used for determining employee identity information, departure time and departure cause if an employee departure request of an employee is received; the identity verification module is used for acquiring face image data of the employee and determining whether the identity information of the employee is authorized identity information; the storage module is used for generating an employee departure record according to the employee identity information, the departure time and the departure reason if the identity information is authorized, and storing the employee departure record to the database; and the regular transmission module is used for transmitting the staff leaving records to the management system every time a preset transmission time interval is reached. According to the scheme, the identity verification, record generation, storage and transmission processes of staff leaving are automated and integrated with a management system, so that the security, efficiency, compliance and transparency can be improved, human errors and management vulnerabilities are reduced, and the management decision quality is improved.

Description

Technical Field

[0001] This application relates to the technical field of attendance management, and in particular, to an employee off-duty management system. Background Art

[0002] In some specific places, the off-duty management of staff (such as management staff, etc.) is crucial. Especially in some special safety environments, it is required that all staff must be on duty to perform their duties to ensure safety and order. However, due to various reasons, such as work requirements, health reasons or emergencies, staff may need to be off duty temporarily. In order to ensure the continuity and safety of management, an efficient off-duty management system is needed to track and record the off-duty situation of employees.

[0003] Existing management systems usually rely on manual records and reports for the off-duty management of employees. When an employee needs to be off duty, they may need to fill out a paper form or an electronic form, recording their identity information, off-duty time and reasons for leaving. Then, these records are manually input into the management system by the management staff.

[0004] Existing management systems rely on manual records and reports, which leads to low management efficiency and inaccurate data due to human errors. And the lack of automation and real-time nature results in low management efficiency and easy errors. Summary of the Invention

[0005] An object of this application is to provide an employee off-duty management system, which is at least used to solve the problem that existing management systems rely on manual records and reports, resulting in low management efficiency and inaccurate data due to human errors. And the lack of automation and real-time nature leads to low management efficiency and easy errors.

[0006] To achieve the above object, some embodiments of this application provide the following aspects:

[0007] In a first aspect, some embodiments of this application further provide an employee off-duty management system, characterized in that the system includes:

[0008] An off-duty request receiving module, configured to, if receiving an employee off-duty request from an employee, determine the first employee identity information, off-duty time and reasons for leaving according to the employee off-duty request;

[0009] An identity verification module, configured to obtain the first employee facial image data and determine whether the first employee identity information is authorized identity information according to the first employee facial image data;

[0010] A storage module, configured to generate an employee departure record based on the first employee identity information, departure time, and reason for departure if the first employee identity information is authorized identity information, and store the employee departure record in a database;

[0011] A regular transmission module, configured to transmit the employee departure record to a management system every time a preset transmission time interval is reached.

[0012] Compared with the related art, in the solution provided by the embodiments of the present application, a departure request receiving module is configured to determine the first employee identity information, departure time, and reason for departure based on the employee departure request if an employee departure request is received; an identity verification module is configured to obtain the first employee facial image data and determine whether the first employee identity information is authorized identity information based on the first employee facial image data; a storage module is configured to generate an employee departure record based on the first employee identity information, departure time, and reason for departure if the first employee identity information is authorized identity information, and store the employee departure record in a database; a regular transmission module is configured to transmit the employee departure record to a management system every time a preset transmission time interval is reached. Through the above employee departure management system, the processes of identity verification, record generation, storage, and transmission of employee departure are automated and integrated with the management system, which can significantly improve security, efficiency, compliance, and transparency, reduce human errors and management loopholes, and improve the quality of management decisions. Description of the Drawings

[0013] One or more embodiments are exemplarily illustrated by the pictures in the corresponding drawings. These exemplary illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements, unless otherwise stated, and the drawings in the drawings do not constitute a proportional limitation.

[0014] Figure 1 It is an exemplary structural diagram of an employee departure management system provided according to some embodiments of the present application;

[0015] Figure 2 It is an exemplary structural diagram of an employee departure management system provided according to some embodiments of the present application;

[0016] Figure 3 It is an exemplary structural diagram of an employee departure management system provided according to some embodiments of the present application. Detailed Embodiments

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Apparently, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0018] First Embodiment

[0019] The first embodiment of this application relates to an employee off - duty management system. As Figure 1 shown, the system may include:

[0020] An off - duty request receiving module 101, configured to, if an off - duty request of an employee is received, determine first employee identity information, off - duty time, and off - duty reason according to the off - duty request of the employee;

[0021] An identity verification module 102, configured to obtain first employee facial image data and determine whether the first employee identity information is authorized identity information according to the first employee facial image data;

[0022] A storage module 103, configured to, if the first employee identity information is authorized identity information, generate an employee off - duty record according to the first employee identity information, off - duty time, and off - duty reason, and store the employee off - duty record in a database;

[0023] A regular transmission module 104, configured to transmit the employee off - duty record to a management system every time a preset transmission time interval is reached.

[0024] In this solution, the off - duty request of an employee may be a request actively made by the employee to the system, indicating a desire to leave the post or work location.

[0025] The first employee identity information may refer to the detailed identity data of an employee verified and confirmed by the system. This information generally includes: Employee number: The unique identifier of each employee. Name: The name of the employee. Position / post information: The work position or department where the employee is located. Employee photo: The identity can be verified through technologies such as facial recognition. Employment information: The employment time, contract information, etc. of the employee.

[0026] The off - duty time may be the specific time period when the employee makes an off - duty request, and may include: Off - duty start time: The specific time when the employee requests to be off - duty. Off - duty end time: The time when the employee plans to return to the post, or the duration of being off - duty.

[0027] The off - duty reason may be the specific reason provided by the employee when making an off - duty request. For example, the employee takes leave due to physical discomfort or leaves due to personal reasons.

[0028] When an employee submits a request for leaving the post, the system records the key information in the request: The employee submits a request for leaving the post through the system (such as through a self-service system, APP, or human resource management system). The employee fills in the necessary information: Name / number (identification information), leaving time (start time and end time), reason for leaving the post (the employee selects or manually enters the reason). Then the system parses the employee's request for leaving the post, determines the identification information, leaving time, and reason for leaving the post filled in by the employee, and looks up the first employee identity information of the employee in the database according to the identification information filled in by the employee.

[0029] The first employee facial image data may refer to the facial image of the employee captured by an image acquisition device (such as a camera, facial recognition system). This image data is used to verify the identity of the employee. Specifically, the facial image data includes the facial features of the employee (such as the shape and relative position of the eyes, nose, and mouth) and other biometric features, which are unique and can be used to identify individuals.

[0030] The authorized identity information may be the employee identity information that has been authenticated, valid, recognized by the authorized system, and stored in the database. This information may include the basic information of the employee, such as: Employee number: The unique identifier of the employee. Name: The name of the employee. Position information: The position, department, etc. that the employee is responsible for. Facial feature data: The key biometric feature data used for facial recognition, such as facial templates, calibration data, etc.

[0031] When an employee submits a request for leaving the post or performs other operations that require identity verification, the system activates the camera device to collect facial images. The system can use a fixed or mobile camera (such as a computer camera, access control system camera, or dedicated facial recognition device) to obtain the real-time facial image data of the employee. The collected facial images may have noise or be blurred and need to be preprocessed, such as face alignment, facial feature point calibration, and image enhancement, etc., to improve the accuracy of comparison. Then the facial area is extracted from the complete image, and key facial feature points (such as eyes, nose, mouth, facial contour, etc.) are extracted. These feature points will be used for subsequent comparison and matching. The system compares the real-time collected facial image of the employee with the facial feature template of the employee stored in the first employee identity information in the database through a facial recognition algorithm (such as a deep learning algorithm, convolutional neural network, etc.). The facial feature data pre-entered at the time of the employee's entry is stored in the database. Then the system calculates the similarity between the collected facial image and the facial feature data of the employee in the database. Common facial recognition technologies, such as "deep neural network + feature vector comparison", will calculate the vector difference of the image features and generate a matching score. If the matching degree of the comparison result is higher than a preset threshold (such as 95% or 98%), it can be confirmed that the employee's identity is valid and is the authorized identity information.

[0032] An employee off - duty record can refer to record data containing an employee's off - duty request and related information.

[0033] A database can be a system used to store and manage employee off - duty records and other related data. It usually has efficient storage, retrieval, update, and deletion functions to ensure the integrity and security of off - duty data.

[0034] If the first employee's facial image data is consistent with the facial image data in the first employee's identity information, then the first employee's identity information is authorized identity information. An employee off - duty record is created based on the first employee's identity information, off - duty time, and off - duty reason. A connection is established with the database through an API, database connection pool, or ORM (Object - Relational Mapping) tool. And the generated employee off - duty record is inserted into an appropriate table in the database through SQL or the corresponding query language.

[0035] A preset transmission time interval can refer to a fixed time interval or period set by the system when transmitting employee off - duty records. For example, the system may automatically transmit employee off - duty records to the management system every hour, every day, or every week.

[0036] A management system can be a system used to manage and monitor the environment, monitor the behavior of the monitored, employee activities, and off - duty records.

[0037] A scheduled task or periodic task can be implemented to transmit employee off - duty records to the management system at the preset time interval. Specifically, the scheduled task function of the operating system can be used, or a timer can be implemented in the application. For example, use a Cron job (in a Linux system) or Windows Task Scheduler to periodically trigger the transmission operation to transmit employee off - duty records to the management system.

[0038] In the embodiment of the present application, the off-duty request receiving module is used to determine the first employee identity information, off-duty time, and off-duty reason according to the employee's off-duty request if the off-duty request of the employee is received; the identity verification module is used to obtain the first employee facial image data and determine whether the first employee identity information is authorized identity information according to the first employee facial image data; the storage module is used to generate an employee off-duty record according to the first employee identity information, off-duty time, and off-duty reason and store the employee off-duty record in the database if the first employee identity information is authorized identity information; the regular transmission module is used to transmit the employee off-duty record to the management system every time a preset transmission time interval is reached. Through the above employee off-duty management system, the processes of identity verification, record generation, storage, and transmission of employee off-duty are automated and integrated with the management system, which can significantly improve security, efficiency, compliance, and transparency, reduce human errors and management loopholes, and improve the quality of management decisions.

[0039] Based on the above technical solution, optionally, the system further includes an off-duty exception reminder module, and the off-duty exception reminder module is used for:

[0040] If the first employee identity information is unauthorized identity information, generate an off-duty exception reminder information according to the first employee facial image data and the first employee identity information, and transmit the off-duty exception reminder information to the management system.

[0041] In this solution, the unauthorized identity information may be that the employee's identity information has not been verified or authorized. In this context, the system detects through facial recognition that the employee's identity information does not belong to the list of authorized employees or valid identities in the database.

[0042] The off-duty exception reminder information may be a warning or reminder information generated by the system when an employee attempts to go off duty and their identity information is unauthorized, notifying relevant management personnel or the management system that the employee's off-duty behavior is abnormal and may involve risks such as identity fraud and unauthorized off-duty.

[0043] If the first employee facial image data does not match the facial image data in the first employee identity information, the first employee identity information is unauthorized identity information, and the system will trigger an abnormal situation: the comparison result indicates that the employee is not an authorized employee (such as facial data mismatch or matching an unregistered identity information). The system determines that the off-duty request is an abnormal request and generates an off-duty exception reminder information, including the following content: Employee identity information: For example, identification information such as employee name and employee number. Employee facial image data: The actually collected facial image, as evidence of the failure of employee identity verification, and then transmit the off-duty exception reminder information to the management system through a secure communication protocol (such as HTTPS).

[0044] In this solution, through automated facial recognition, anomaly alerts, and information traceability, the security, efficiency, and transparency of management can be significantly improved, while reducing the possibility of manual intervention and errors, achieving intelligent and standardized management of employees' leaving work.

[0045] Based on the above technical solution, optionally, the system further includes a return-to-work management module, and the return-to-work management module is used for:

[0046] If a return-to-work request from an employee is received, determine the second employee identity information and the return-to-work time according to the return-to-work request from the employee;

[0047] Obtain the second employee's facial image data, and determine whether the second employee identity information is authorized identity information according to the second employee's facial image data;

[0048] If the second employee identity information is authorized identity information, generate an employee return-to-work record according to the second employee identity information and the return-to-work time, and store the employee return-to-work record in the database;

[0049] Transmit the employee return-to-work record to the management system every time a preset transmission time interval is reached.

[0050] In this solution, a return-to-work request from an employee can be an application submitted by the employee to return to the post after leaving work. Usually, this request includes the employee's identity information, the return time, and possibly the reason or explanation for returning to work.

[0051] The second employee identity information can be the unique identity identification information of the employee (such as employee number, name, position, etc.). The role of the identity information is to confirm whether the identity of the employee is authorized identity, ensuring that the employee making the return-to-work request is a legitimate employee and that their identity has not been tampered with.

[0052] The return-to-work time can be the time when the employee returns to the workplace after leaving work.

[0053] The second employee's facial image data can be the facial photo or image data of the employee collected through facial recognition technology. These images are used to verify whether the employee is the authorized person corresponding to their identity information, preventing impersonation.

[0054] The employee return-to-work record is generated and stored by the system, recording the relevant information of the employee from leaving work to returning to work. This record includes the employee's identity information, return-to-work time, etc.

[0055] The system receives the return-to-work request submitted by the employee. This request usually includes the employee's identity information, return-to-work time, and other possible details. Then, facial image data of the employee is obtained through cameras, monitoring devices, etc. This is used to verify the employee's identity to ensure that the employee's identity information matches the facial data. Then, the obtained facial image data is compared with the authorized identity information in the employee database. If the verification passes, the identity is confirmed as legal and authorized. If the employee identity verification is successful, the system generates a return-to-work record based on the employee's identity information and return-to-work time. The generated return-to-work record will be stored in the system's database to ensure the integrity and traceability of the information. Then, when the preset transmission time interval is reached, the system will automatically transmit the return-to-work record to the management system for further data analysis, report generation, and compliance checks.

[0056] In this solution, through facial recognition technology, automatically generating return-to-work records, and regularly transmitting data, it is ensured that the employee's return-to-work behavior complies with the regulations and safety requirements, improves management efficiency, reduces human intervention, and at the same time ensures the transparency and traceability of the data.

[0057] Based on the above technical solution, optionally, the system further includes a return-to-work exception reminder module, and the return-to-work exception reminder module is used for:

[0058] If the second employee identity information is unauthorized identity information, a return-to-work exception reminder message is generated according to the second employee's facial image data and the second employee's identity information, and the return-to-work exception reminder message is transmitted to the management system.

[0059] In this solution, the return-to-work exception reminder message can refer to the warning or reminder message generated when the system identifies certain abnormal situations during the employee's return to work. These abnormal situations can include: Unauthorized employee identity: That is, the second employee's identity information does not match the identity information registered in the system, or their facial image fails the authentication. Mismatch between facial image and identity information: That is, after comparison through facial recognition technology, the facial image does not match the employee information stored in the system.

[0060] If the facial image data of the second employee does not match the facial image data in the second employee's identity information, the second employee's identity information is unauthorized identity information, and the system will trigger an abnormal situation: The comparison result indicates that the employee is not an authorized employee (such as the facial data does not match, or an unregistered identity information is matched). The system determines that the return-to-work request is an abnormal request and generates a return-to-work exception reminder message, which includes the following content: Employee identity information: For example, identification information such as employee name, employee number, etc. Employee facial image data: The actually collected facial image, as evidence of the failure of the employee identity verification, and then the return-to-work exception reminder message is sent to the management system through a secure communication protocol (such as HTTPS).

[0061] In this solution, the benefits of generating and transmitting reminder messages for abnormal return to work are to improve the safety of management in specific places, enhance management efficiency, strengthen risk early warning capabilities, ensure employee compliance, reduce human errors and omissions, and thus promote a more efficient and safe management system.

[0062] Second Embodiment

[0063] The second embodiment of this application relates to a system for managing employees' leaving the post. The second implementation is an improvement based on the first embodiment. The second implementation is substantially the same as the first implementation, and the main differences are as follows:

[0064] The system further includes a leaving-post prediction module 105, and the leaving-post prediction module 105 is used for:

[0065] Obtain the historical leaving-post times and historical violation times of each employee during a preset collection period, and determine the penalty factor corresponding to the historical violation times according to the historical violation times and the preset penalty factor setting rules;

[0066] Calculate the predicted leaving-post times of each employee during a preset prediction period according to the historical leaving-post times, the preset collection period, the penalty factor, the preset prediction period, and the preset leaving-post prediction formula;

[0067] If there are predicted leaving-post times greater than a preset leaving-post times threshold, use the employees corresponding to the predicted leaving-post times greater than the preset leaving-post times threshold as target employees, obtain the third employee identity information of the target employees, generate a leaving-post risk alarm according to the third employee identity information and the predicted leaving-post times, and transmit the leaving-post risk alarm to the management system.

[0068] In this embodiment, the preset collection period can be a time interval defined by the system. During this time period, the system will record and analyze the historical leaving-post data of employees. This period may be a fixed time period in units of days, weeks, months, etc., used to count the leaving-post behaviors of employees. For example, the collection period may be "the past month", or "the past week", etc.

[0069] The historical leaving-post times can refer to the actual leaving-post times of employees during the specified collection period. These leaving-post behaviors can be legitimate (such as sick leave, annual leave, etc.), or cases of leaving the post without reason or not following the prescribed procedures.

[0070] The historical violation times refer to the number of times employees leave the post in violation of regulations during the collection period. For example, violations such as not notifying in advance or leaving the post without approval.

[0071] The preset penalty factor setting rule can be a penalty factor set according to the number of violations committed by an employee within a specific time period. Specifically, the more violations an employee has, the higher the penalty factor corresponding to the employee. This rule is used to quantify the penalty degree that an employee needs to bear due to violations and serves as an important parameter for predicting off-the-job behavior and risk assessment. For example, if an employee has two historical violation counts, the system may set the penalty factor to 2.

[0072] The penalty factor can be a numerical value assigned based on an employee's historical violation behavior. This value represents the degree of "penalty" that an employee needs to bear due to violations. The higher the penalty factor, the more serious the violation degree of the employee and the higher the future off-the-job risk.

[0073] The preset prediction period can refer to the time period during which the system predicts an employee's future off-the-job behavior. The length of this time period is set in advance, usually in weeks or months, with the aim of evaluating the off-the-job possibility of the employee during this period.

[0074] The predicted off-the-job count refers to the number of times an employee is predicted to be off the job within the preset prediction period based on information such as historical off-the-job data, violation records, and penalty factors. This is a numerical prediction made by the system based on historical behavior and rules. For example, the predicted off-the-job count of an employee in the next week is 5 times.

[0075] The preset off-the-job count threshold can be an upper limit set by the system for the maximum number of off-the-job times allowed within a certain time period. When an employee's predicted off-the-job count exceeds this threshold, the system will mark the employee as a high-risk target and prompt the management to pay attention.

[0076] Target employees can refer to those employees whose predicted off-the-job count exceeds the preset off-the-job count threshold. Such employees are considered to have a relatively high off-the-job risk and need special attention, and further monitoring or intervention measures may be taken.

[0077] The third employee identity information can refer to the identity information of those employees who are determined to be target employees. This identity information is used to generate off-the-job risk alerts and is transmitted to the management system together with other data of the employee (such as the predicted off-the-job count) for further processing.

[0078] The off-the-job risk alert can be a warning message generated based on the off-the-job prediction result of an employee. When an employee's predicted off-the-job count exceeds the preset threshold, the system will generate this alert to remind the management that the off-the-job behavior of this employee poses a risk. The alert can include the employee's identity information, the predicted off-the-job count, and other relevant data.

[0079] Count the number of times an employee leaves the post within a preset collection period through the employee's historical data records (such as attendance systems, leave request records, etc.). The collection period can be a fixed cycle such as a month or a quarter. By monitoring the employee's violation records (for example, leaving the post without reason, being late, absent from work, etc.), count the number of violations of the employee within the same preset collection period. According to the preset rules, link the number of violations with the corresponding penalty factors. For example: 1 violation → penalty factor = 1. Then substitute the historical number of times of leaving the post, the preset collection period, the penalty factor, and the preset prediction period into the preset leaving-post prediction formula to calculate the predicted number of times of leaving the post for each employee within the preset prediction period. For employees whose predicted number of times of leaving the post exceeds the threshold, mark them as target employees. Then obtain the personal information of the employee (such as name, position, employee number, etc.) through the employee identity information system. Generate a leaving-post risk alert based on the third employee identity information and the predicted number of times of leaving the post, and transmit the leaving-post risk alert to the management system through wireless communication technology.

[0080] In this embodiment, by predicting, evaluating, and monitoring the leaving-post behavior of employees, potential risks can be identified more effectively, measures can be taken in advance to ensure the smooth progress of the work process, while improving the efficiency of resource allocation and the transparency of employee management, and enhancing overall security. This not only helps to reduce the complexity of management, but also improves the overall operation efficiency and emergency handling ability of the organization.

[0081] On the basis of the above technical solution, optionally, the system further includes a risk management module, and the risk management module is used for:

[0082] Determine the historical violation behaviors corresponding to the historical number of violations of the target employee, and determine the historical reasons for leaving the post corresponding to the historical number of times the target employee leaves the post;

[0083] Input the historical number of violations, historical violation behaviors, historical number of times of leaving the post, and historical reasons for leaving the post of the target employee into a preset risk management model to obtain a risk management plan;

[0084] Transmit the third employee identity information, the predicted number of times of leaving the post, and the risk management plan to the management system.

[0085] In this solution, historical violation behaviors can be the behaviors of employees violating the company's or organization's rules and regulations within a past period of time. These behaviors can include coming to work late and leaving early, leaving the post without reason, not following safety regulations at work, and violating discipline or behavior norms.

[0086] The historical reasons for leaving the post refer to the specific reasons corresponding to the leaving-post behaviors of employees within a past period of time. For example, an employee leaves the post due to health problems, an employee leaves the post due to urgent matters, an employee takes leave due to family problems, and an employee leaves the post due to other personal matters.

[0087] The preset risk management model can be a mathematical or algorithmic model designed based on historical data and business rules, used to analyze the historical behaviors of employees (such as leaving the post, violating regulations, etc.), predict the future behavior risks of employees, and provide decision-making support for management.

[0088] The risk management plan can be the result calculated based on the preset risk management model, used to evaluate the risks of target employees and provide countermeasures. It can include the following: Risk level: Based on the historical data of employees and the model output, evaluate their future risks of leaving the post, which may be different risk levels such as low, medium, and high. Countermeasures: For example, if an employee is judged to be at high risk, strategies such as stricter monitoring, arranging substitute personnel, and strengthening employee communication may need to be taken. Improvement measures: If an employee has a high risk due to frequent leaving the post or violating regulations, improvement measures such as training, disciplinary action, and job adjustment may be proposed.

[0089] The historical leaving post records and reasons for leaving the post of employees can be obtained from the employee management system, and then the historical violation behaviors of each violation of employees can be obtained from the employee management system. Data such as the historical number of violations, historical violation behaviors, historical number of times of leaving the post, and historical reasons for leaving the post of the target employee are input into the preset risk management model. Based on the input data and model prediction, a risk management plan is generated. The plan may include risk level assessment and specific countermeasures (such as whether additional monitoring is required, whether the post needs to be adjusted, whether communication needs to be strengthened, etc.). Then, the identity information of the target employee (the third employee identity information), the predicted number of times of leaving the post, and the generated risk management plan are transmitted to the management system through wireless communication technology.

[0090] The training process of the preset risk management model includes:

[0091] Data such as the number of historical violations, historical violation behaviors, the number of historical departures from work, and the reasons for historical departures from work need to be collected and organized into a dataset with a large number of samples for training and validating a risk management model. Specifically, all the number of historical violations, historical violation behaviors, the number of historical departures from work, and the reasons for historical departures from work can be read from the database, and each record corresponds to the data within a time period or a cycle. Each record will correspond to a historical risk management plan, which may be formulated manually or generated by a machine, but has been put into practice, and the risk management plan is used as the label for each record. Organize the original data into a format suitable for model input. For example, convert each record into a feature vector, including features such as the number of violations and the number of departures from work, and attach the corresponding label (historical risk management plan). Feature columns: the number of historical violations, historical violation behaviors, the number of historical departures from work, reasons for historical departures from work, etc. Label column: historical risk management plan (usually a classification label, such as low risk, medium risk, high risk, or management measures, etc.). Then handle missing data, and missing values can be filled with the mean, median, most common value, or by interpolation. For non-numeric types (such as violation behaviors, reasons for leaving work, etc.), encoding is required. Common methods are: Use one-hot encoding to convert categorical data into binary representation (for example: being late, leaving work without permission, etc. as different columns). Use label encoding to convert categorical data into integer values (for example: different violation behaviors are marked as 0, 1, 2, etc.). Standardize or normalize numerical data (such as the number of violations, the number of departures from work) so that they can be compared on the same scale. Select features that are helpful for prediction (such as the number of historical violations, the number of historical departures from work, the type of violation behavior, etc.), and remove irrelevant features. Technologies that can be used are: Correlation analysis: such as Pearson correlation coefficient, chi-square test, etc., to find features that are closely related to the label. Recursive Feature Elimination (RFE): Remove unimportant features and retain the most important features to improve model performance. Then select a suitable machine learning algorithm according to the task requirements. If there is a linear relationship between the features and the target, logistic regression can be selected. If there are complex non-linear relationships in the data, decision trees or random forests can be selected. If the dataset is very large and high-dimensional data needs to be processed, Support Vector Machine (SVM) can be selected. If the data volume is very large and there are complex feature interactions, neural networks can be used. If the task requires multiple model integrations to improve prediction accuracy, Gradient Boosting Machine (GBM) can be considered. Then divide the dataset into a training set and a test set (a common ratio is 80% for the training set and 20% for the test set) for model training and evaluation. Input the training data into the selected machine learning algorithm for training. The model will learn the relationship between the features in the data and the risk management plan label. Adjust the hyperparameters of the model (such as the depth of the decision tree, the kernel function of SVM, etc.) to improve the prediction ability of the model. Use the test set to evaluate the trained model.Judge whether the model has sufficient predictive ability according to evaluation metrics (such as accuracy, recall rate, F1 score). Cross-validation techniques can be used to divide the data into multiple subsets and repeat training and validation to obtain a more robust model evaluation. Then, use grid search or random search methods to adjust the hyperparameters of the model to find the best parameter combination. Combine the prediction results of multiple models (such as through ensemble methods like random forest, AdaBoost, etc.) to improve the prediction accuracy. After the model is trained and optimized, deploy it to actual applications. Integrate the model with the management system through API interfaces and other means to achieve real-time prediction and risk warning.

[0092] In this solution, by analyzing historical violation behaviors and reasons for leaving the post, employees who may pose risks can be identified more accurately, avoiding biases in human judgment, and corresponding management measures can be taken in a timely manner.

[0093] On the basis of the above technical solution, optionally, the system further includes a continuous monitoring module, and the continuous monitoring module is used for:

[0094] If there is no predicted number of times of leaving the post greater than the preset threshold of the number of times of leaving the post, after each preset monitoring time interval, re-obtain the historical number of times of leaving the post and the historical number of violations of each employee during the preset collection period, and update the penalty factor corresponding to the historical number of violations according to the historical number of violations and the preset penalty factor setting rules;

[0095] Update the predicted number of times of leaving the post of each employee during the preset prediction period according to the historical number of times of leaving the post, the preset collection period, the penalty factor, the preset prediction period, and the preset formula for predicting the number of times of leaving the post;

[0096] If there is a predicted number of times of leaving the post greater than the preset threshold of the number of times of leaving the post, take the employees corresponding to the predicted number of times of leaving the post greater than the preset threshold of the number of times of leaving the post as target employees, obtain the third employee identity information of the target employees, generate a leaving-the-post risk alarm according to the third employee identity information and the predicted number of times of leaving the post, and transmit the leaving-the-post risk alarm to the management system.

[0097] In this solution, the preset monitoring time interval can refer to that after a fixed time period, the system will re-collect and update the data of the historical number of times of leaving the post and the historical number of violations of employees.

[0098] Whenever the monitoring time interval is reached, the system will update the penalty factor corresponding to the number of violations according to the latest historical number of violations and the preset penalty factor setting rules. Then, based on the updated historical number of off - post times, penalty factor, preset collection period, and off - post prediction formula, recalculate the number of off - post times of each employee within the preset prediction period. If the predicted number of off - post times exceeds the preset off - post times threshold, the system will automatically identify these employees as target employees, obtain their third - party identity information, generate an off - post risk alarm based on the predicted number of off - post times and identity information, and finally transmit the alarm information to the management system. Then repeat the above steps when the next monitoring time interval is reached.

[0099] In this solution, by regularly monitoring the off - post and violation data of employees, potential risk employees can be identified in advance before the number of employees' off - post times reaches the preset threshold, so as to take necessary preventive or management measures to reduce the occurrence of emergencies.

[0100] Based on the above - mentioned technical solution, optionally, the preset off - post prediction formula is:

[0101]

[0102] Among them, F Leave is the predicted number of off - post times; N Leave is the historical number of off - post times; T period is the preset collection period; H penalty is the penalty factor; T prediction period is the preset prediction period.

[0103] In this solution, if the preset collection period is the working days of the past week, then T period = 5. T prediction period If it is the working days of the next week, then T prediction period = 5.

[0104] Third Embodiment

[0105] The third embodiment of this application relates to an employee off - post management system. The third implementation is an improvement based on the first embodiment. The third implementation is substantially the same as the first implementation. The main differences are as follows:

[0106] The system further includes an off - post analysis module 106, and the off - post analysis module 106 is used for:

[0107] Obtain the off - post frequency, off - post behavior pattern, and key position data of employees, and determine the off - post type and specific entity of the off - post reason;

[0108] Input the reason for leaving the post, the frequency of leaving the post, the leaving-behavior pattern, the key data of the post, the type of leaving the post, and the specific entity into a preset leaving-post compliance analysis model to obtain a compliance analysis report on the reason for leaving the post, and transmit the compliance analysis report to the management system.

[0109] In this embodiment, the frequency of leaving the post can be the number or frequency of times an employee leaves the post within a period of time, usually the frequency of leaving the post counted by day, week, or month.

[0110] The leaving-behavior pattern can be the pattern or rule of an employee leaving the post. For example, whether the employee frequently leaves the post at a specific time or after a specific event, and whether there is a seasonal or cyclical trend.

[0111] The key data of the post can refer to the core data related to the work content of the employee's post, such as the importance of the work content, the scope of responsibilities, the working hours, the workload, etc. These data determine the impact degree of the employee leaving the post.

[0112] The type of leaving the post can be determined by classifying the text of the reason for leaving the post to identify the specific reason for the employee to leave the post. It can include sick leave: the employee leaves the post due to health problems and usually requires a medical certificate or relevant health documents. Personal leave: the employee leaves the post due to personal affairs (such as family problems, weddings, funerals, etc.) and usually needs to explain the reason. Emergency situation: such as a sudden illness or accident of a family member, and the employee needs to leave the work post immediately. Other types: such as annual leave, compensatory leave, training, etc., may also be part of leaving the post.

[0113] The specific entity can refer to the key information extracted from the text of the reason for leaving the post for further analysis and evaluation of the compliance of leaving the post. It can include the leaving duration: referring to the length of time an employee leaves the post, usually involving the start and end times of a certain period. For example, an employee taking sick leave may only need one day or several days. The leaving time: referring to the specific date and time when the employee actually leaves the post, which helps to track whether the leaving occurs during a critical period or when it has a greater impact on the work. The urgency of the reason: identifying the urgency of the reason for leaving the post. For example, sick leave may require a medical certificate, and personal leave may also require additional verification if it is due to a family emergency. The type of disease: if an employee leaves the post due to illness, the system can obtain the type of disease (such as cold, fracture, long-term illness, etc.) through text analysis to provide a basis for further management. Sudden events at home: for example, a sudden illness or other emergency of a family member at home, and the employee may need to leave the post immediately. The system needs to extract this information and evaluate the compliance according to its urgency.

[0114] The preset off - the - job compliance analysis model can be a trained and validated model used to analyze and evaluate whether an employee's off - the - job behavior complies with company regulations or legal requirements. It may use machine learning algorithms (such as classification models, regression models, etc.) to analyze the employee's off - the - job data, thereby evaluating whether the off - the - job behavior meets the compliance standards.

[0115] The compliance analysis report can be obtained by analyzing the reasons for an employee's off - the - job situation and evaluating whether the off - the - job reason complies with the regulations and compliance requirements of the organization and management system.

[0116] The off - duty records can be obtained from the employee's attendance system or monitoring system. The records include the off - duty time, off - duty duration, reasons for leaving, etc. By counting the number of times an employee leaves the post within a certain period (such as a week, a month, etc.), the off - duty frequency can be obtained. A high off - duty frequency may indicate potential risk signals. By analyzing the time points when the off - duty occurs, it is identified whether there is a pattern in the off - duty, for example, whether employees often leave the post during certain time periods (such as in the morning or at noon on weekdays). Evaluate the pattern of off - duty duration. Frequent and long - term off - duty may indicate non - compliant behavior. If the reasons for leaving are unclear or lack sufficient rationality, it may form an irregular behavior pattern. Analyze the combination of the reasons for leaving and the time period to determine if there is a specific pattern (for example, frequent off - duty on a specific date of each month). Then collect the employee's job information, including job responsibilities, working - time requirements, the importance of the job in the project, etc. By analyzing the influence of the job on the organization's operation, evaluate the impact of the off - duty on aspects such as project progress and work efficiency. Off - duty in key positions (such as safety positions, management positions, etc.) may have a greater impact on the overall business. Then classify the text descriptions of the reasons for leaving to determine the types of off - duty. Using natural language processing techniques (such as text - classification algorithms), automatically classify the reasons for leaving into different types, such as sick leave, personal leave, emergency situations, etc. The reasons for leaving can be identified through classification annotation tools or machine - learning models, such as using a pre - trained classification model to classify the text of the reasons for leaving. Using the named - entity recognition (NER) model in natural language processing technology, extract key information (specific entities) from the reasons for leaving, such as off - duty duration, off - duty time, urgency of the reason, type of disease, etc. Extract the key entities in the reasons for leaving and label them as specific entities. For example, extract "Employee A took sick leave due to a cold, left the post for 3 days, and the off - duty time was from February 5th to February 7th, 2025". In this text, "cold", "3 days", and "from February 5th to February 7th, 2025" are specific entities. Input the collected data such as reasons for leaving, off - duty frequency, off - duty behavior patterns, job - criticality data, off - duty types, specific entities, etc. into a preset off - duty compliance - analysis model. The model analyzes based on these input data to evaluate whether the reasons for leaving comply with the company's policies and whether there are potential risks. The model may include a rule engine, decision tree, machine - learning model, etc. The analysis model generates a compliance - analysis report based on the input data. The report contains the compliance evaluation of the employee's off - duty. Then transmit the generated compliance - analysis report through a secure channel (such as an API interface) to the management system for relevant personnel to perform subsequent processing, such as whether warning, punishment, or improvement measures are needed.

[0117] The training process of the preset off - duty compliance - analysis model includes:

[0118] Collect relevant data from sources such as the management system, employee records, leave requests, and off - duty logs. Specifically include: Reason for leaving: The specific reason for an employee to leave (such as sick leave, personal leave, emergency, etc.). Leave frequency: The number of times an employee leaves during a specific period. Leave behavior pattern: The leave pattern of an employee, such as the time period of frequent leave, the time interval of leave, etc. Key data of the position: The urgency or criticality of an employee's position. Leave type: The type of reason for leaving (such as sick leave, personal leave, etc.), extracted through a text classification model. Specific entity: Key information extracted from the reason for leaving, such as leave duration, leave time, urgency, disease type, family emergency, etc. Then clean and format the collected data to ensure data consistency and accuracy. Process the text data (such as the reason for leaving) to extract meaningful features. Each record corresponds to a compliance label for the reason for leaving, and the label can be "compliant" or "non - compliant". These labels are usually manually marked by experts, managers, or historical experience, or can be automatically generated according to actual business rules. Then divide the dataset into a training set, a validation set, and a test set, usually in the ratio of 70% (training set), 15% (validation set), and 15% (test set). Select a suitable machine learning algorithm according to the characteristics of the data (such as a combination of classification tasks and regression tasks). Common choices include: Classification model: If the goal is to predict whether the reason for leaving is compliant (e.g., compliant / non - compliant), a classification algorithm can be selected, such as decision tree, random forest, support vector machine (SVM), logistic regression, neural network, etc. Regression model: If it is necessary to predict continuous variables of leaving (such as leave duration, frequency, etc.), a regression algorithm can be used. Use the data in the training set to train the model through the selected algorithm. During the training process, the model will learn based on input features such as the reason for leaving, leave frequency, leave behavior pattern, etc., and the corresponding labels (compliant / non - compliant). During the training process, the model will continuously adjust parameters to minimize the error between the predicted value and the actual label. Use the validation set data to evaluate the performance of the model and check indicators such as accuracy, recall rate, F1 - score, etc. Adjust the hyperparameters of the model (such as learning rate, tree depth, regularization parameter, etc.) according to the evaluation results to optimize the performance of the model. If the model is overfitting or underfitting, methods such as cross - validation, feature selection, and dimensionality reduction can be used for further optimization. Then use the test set data to conduct a final test on the trained model and evaluate the performance of the model on unseen data. According to the test results, confirm the generalization ability of the model and ensure its reliability in practical applications.

[0119] In this embodiment, automated off - work compliance analysis helps managers process employees' off - work records more efficiently. The system can quickly identify potential problems, reduce the time for manual review, and improve management efficiency. By analyzing off - work frequency, behavior patterns, and job - criticality data, abnormal off - work behaviors can be detected. For example, frequent off - work or off - work at critical moments may be signs of shirking responsibilities. Timely detection of these behaviors helps take appropriate preventive or intervention measures.

[0120] Based on the above - mentioned technical solution, optionally, the off - work analysis module is further configured to:

[0121] Determine the off - work type of the off - work reason according to the NLP model;

[0122] Determine the specific entity of the off - work reason according to the named - entity recognition technology.

[0123] In this solution, the NLP (Natural Language Processing) model is a type of machine - learning model used to process and understand human language. In this scenario, the NLP model is typically used to analyze and understand employees' off - work reasons, extract meaningful information from them, and perform classification and inference. Common NLP models include the bag - of - words model and TF - IDF model based on traditional methods, as well as modern deep - learning - based models such as BERT and GPT. The goal of the NLP model in this solution is to automatically determine the off - work type (such as "sick leave", "personal leave", or "emergency situation", etc.) based on the given off - work reason text (such as "taking leave due to illness" or "off - work due to family emergency").

[0124] Named - entity recognition is a task in natural language processing aimed at identifying entities with specific meanings in text, such as person names, locations, organizations, dates, times, etc. In the scenario of off - work reasons, the named - entity recognition technology can extract specific key information from off - work reasons, such as the time of off - work, the duration of off - work, the type of disease, the type of emergency event, etc.

[0125] The reasons for leaving work can be cleaned and preprocessed, including operations such as removing punctuation marks, stop words, and converting to lowercase, so that the model can better understand the text. Then, the NLP model extracts features from the input text of the reasons for leaving work, such as word frequency, context information, etc. If a deep learning model, such as BERT, is used, the model can understand the semantics of the text through its pre-trained context understanding ability. Based on the features of the reasons for leaving work, the model classifies them. For example, the model may classify the text "taking leave due to illness" as "sick leave" and "having an urgent matter at home" as "personal leave", etc. During training, a dataset with labels is used, and the labels are the known types of leaving work. Through continuous training, the NLP model learns how to predict the type of leaving work based on the content of the reasons for leaving work. Then, the text of the reasons for leaving work is preprocessed, such as removing noise and standardizing the format. The processed text is input into the trained NER model. The NER model is usually trained to identify various entity categories, such as time (duration of leaving work, date), location (e.g., "office"), person (e.g., "family member"), event type (e.g., "illness" or "family emergency"), etc. The NER model identifies named entities in the text through context information. For example, in the text "taking leave due to illness for three days", the NER model may identify "illness" as the entity type and "taking leave for three days" as the duration of leaving work. Similar to the classification model, the NER model needs to be trained with labeled data. In the training data, the entities in the reasons for leaving work are manually labeled (e.g., marking the leaving work time, duration, disease type, etc.). Through training, the NER model can automatically identify these entities in new reasons for leaving work.

[0126] In this solution, by combining these two technologies, the system can accurately analyze the reasons for employees to leave work, not only knowing the type of leaving work, but also being able to extract relevant specific information to help the management make better decisions.

[0127] It is worth mentioning that each module involved in this embodiment is a logical module. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, to highlight the innovative part of this application, units that are not closely related to solving the technical problems proposed in this application are not introduced in this embodiment, but this does not mean that there are no other units in this embodiment.

[0128] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the devices, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that, in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as the combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0129] The scope of the present application is defined by the appended claims rather than the above description. Therefore, all changes that fall within the meaning and scope of the equivalent elements of the claims are intended to be included in the present application. Any reference numerals in the claims should not be construed as limiting the claims involved. In addition, it is obvious that the term "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. The multiple elements or devices stated in the apparatus claims may also be implemented by one element or device through software or hardware. The terms "first", "second", etc. are only used for descriptive distinction and do not represent any specific order, nor can they be construed as indicating or implying relative importance.

[0130] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily mention changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims, and the above embodiments should be regarded as exemplary and non-limiting.

Claims

1. An employee off-duty management system, characterized in that, The system includes: A leaving-duty request receiving module, configured to, if receiving an employee's leaving-duty request, determine first employee identity information, leaving-duty time, and reasons for leaving duty according to the employee's leaving-duty request; An identity verification module, configured to obtain first employee facial image data and determine whether the first employee identity information is authorized identity information according to the first employee facial image data; A storage module, configured to, if the first employee identity information is authorized identity information, generate an employee leaving-duty record according to the first employee identity information, leaving-duty time, and reasons for leaving duty, and store the employee leaving-duty record in a database; A regular transmission module, configured to transmit the employee leaving-duty record to a management system every time a preset transmission time interval is reached.

2. The employee off-duty management system according to claim 1, wherein The system further includes a leaving-duty anomaly reminder module, and the leaving-duty anomaly reminder module is configured to: If the first employee identity information is unauthorized identity information, generate a leaving-duty anomaly reminder message according to the first employee facial image data and the first employee identity information, and transmit the leaving-duty anomaly reminder message to the management system.

3. The employee off-duty management system according to claim 1, characterized in that The system further includes a returning-to-duty management module, and the returning-to-duty management module is configured to: If receiving an employee's returning-to-duty request, determine second employee identity information and returning-to-duty time according to the employee's returning-to-duty request; Obtain second employee facial image data and determine whether the second employee identity information is authorized identity information according to the second employee facial image data; If the second employee identity information is authorized identity information, generate an employee returning-to-duty record according to the second employee identity information and the returning-to-duty time, and store the employee returning-to-duty record in the database; Transmit the employee returning-to-duty record to the management system every time a preset transmission time interval is reached.

4. The employee off - duty management system according to claim 1, characterized in that, The system further includes a leaving-duty prediction module, and the leaving-duty prediction module is configured to: Obtain the historical leaving-duty times and historical violation times of each employee during a preset collection period, and determine the penalty factor corresponding to the historical violation times according to the historical violation times and a preset penalty factor setting rule; Calculate the predicted leaving-duty times of each employee within a preset prediction period according to the historical leaving-duty times, the preset collection period, the penalty factor, the preset prediction period, and a preset leaving-duty prediction formula; If there are predicted leaving-duty times greater than a preset leaving-duty times threshold, use the employees corresponding to the predicted leaving-duty times greater than the preset leaving-duty times threshold as target employees, obtain third employee identity information of the target employees, generate a leaving-duty risk alarm according to the third employee identity information and the predicted leaving-duty times, and transmit the leaving-duty risk alarm to the management system.

5. The employee off - duty management system according to claim 4, wherein, The system further includes a risk management module, and the risk management module is configured to: Determine the historical violation behaviors corresponding to the historical violation times of the target employees, and determine the historical reasons for leaving duty corresponding to the historical leaving-duty times of the target employees; Input the historical violation times, historical violation behaviors, historical leaving-duty times, and historical reasons for leaving duty of the target employees into a preset risk management model to obtain a risk management plan; Transmit the third employee identity information, the predicted leaving-duty times, and the risk management plan to the management system.

6. The employee off - duty management system according to claim 4, characterized in that, The system further includes a continuous monitoring module, and the continuous monitoring module is used for: If there is no predicted number of departures greater than the preset threshold of the number of departures, after each preset monitoring time interval, re-obtain the historical number of departures and historical number of violations of each employee during the preset collection period, and update the penalty factor corresponding to the historical number of violations according to the historical number of violations and the preset penalty factor setting rule; Update the predicted number of departures of each employee within the preset prediction period according to the historical number of departures, the preset collection period, the penalty factor, the preset prediction period, and the preset departure prediction formula; If there is a predicted number of departures greater than the preset threshold of the number of departures, use the employees corresponding to the predicted number of departures greater than the preset threshold of the number of departures as target employees, obtain the third employee identity information of the target employees, generate a departure risk alert according to the third employee identity information and the predicted number of departures, and transmit the departure risk alert to the management system.

7. The employee off-duty management system according to claim 4, wherein The preset departure prediction formula is: Among them, F Leave is the predicted number of departures; N Leave is the historical number of departures; T period is the preset acquisition period; H penalty is the penalty factor; T predictionperiod is the preset prediction period.

8. The employee off-duty management system according to claim 1, characterized in that The system further includes a departure analysis module, and the departure analysis module is used for: Obtain the departure frequency, departure behavior pattern, and key position data of the employee, and determine the departure type and specific entity of the departure reason; Input the departure reason, departure frequency, departure behavior pattern, key position data, departure type, and specific entity into the preset departure compliance analysis model to obtain a compliance analysis report on the departure reason, and transmit the compliance analysis report to the management system.

9. The employee off - post management system according to claim 8, wherein, The departure analysis module is further used for: Determine the departure type of the departure reason according to the NLP model; Determine the specific entity of the departure reason according to the named entity recognition technology.

10. The employee off-duty management system according to claim 3, wherein, The system further includes a return-to-work anomaly reminder module, and the return-to-work anomaly reminder module is used for: If the second employee identity information is unauthorized identity information, generate a return-to-work anomaly reminder message according to the second employee facial image data and the second employee identity information, and transmit the return-to-work anomaly reminder message to the management system.

Images