Online management method and system for human resources

Abstract

The invention relates to the technical field of data mining, in particular to an online management method and system for human resources. The method comprises the following steps: collecting enterprise recruitment data, evaluating post fitness of interviews, making a post allocation strategy, and uploading the post allocation strategy to a human resource management platform to execute an allocation task; the employee work log is analyzed to identify a work behavior mode, and then an abnormal work behavior mode is found; correcting abnormal working behavior mode data by using an interviewer post allocation strategy, and constructing an abnormal behavior detection model; the employee work logs are analyzed through the detection model, the employees with abnormal behaviors are identified, and the post fitness of the employees is evaluated; and based on the post fitness data and the post fitness data, analyzing and formulating a post transfer strategy, and uploading the post transfer strategy to the human resource management platform to execute a post transfer task. The personnel loss rate and the recruitment cost can be reduced based on data mining.

Description

Technical Field

[0001] The present invention relates to the technical field of data mining, and in particular, to an online management method and system for human resources. Background Art

[0002] Human resource management (HRM), as a core component of enterprise management, traditionally relies on paper records and local computer systems to handle tasks such as employee recruitment, training, assessment, and salary management. With the development of information technology, enterprises have gradually realized that the traditional human resource management method is not only inefficient but also difficult to meet the complex human resource needs of modern enterprises in a globalized and rapidly changing environment. In the past few decades, with the rapid development of Internet technology, online management systems have gradually been introduced into the field of human resource management. Initially, online systems were limited to basic data storage and access functions. With the progress of technology, these systems have gradually evolved into integrated and comprehensive human resource management solutions. The emergence of these online systems has significantly improved the efficiency of human resource management, reduced human errors, and provided enterprises with more powerful data analysis and decision-making support capabilities. Traditional online management methods for human resources usually rely on manual processing and simple rules and do not have data analysis and prediction capabilities. Manual processing and simple rules can lead to imperfect recruitment processes and make it difficult to effectively screen and evaluate candidates. The recruitment efficiency is low, which may result in missing excellent talents and may also hire unsuitable employees, thus increasing the staff turnover rate and recruitment costs. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide an online management method and system for human resources to solve at least one of the above technical problems.

[0004] To achieve the above object, an online management method for human resources includes the following steps: Step S1: Obtain enterprise recruitment data, and evaluate the job fit degree of interview candidates based on the enterprise recruitment data to obtain job fit degree data of interview candidates; analyze the enterprise job assignment strategy based on the job fit degree data of interview candidates to obtain an interview candidate job assignment strategy, and upload it to the human resource management platform to execute the job assignment task; Step S2: Obtain the work logs of enterprise employees, and analyze the work behavior patterns of enterprise employees to obtain work behavior pattern data of employees; analyze abnormal work behavior patterns based on the work behavior pattern data of employees to obtain abnormal work behavior pattern data of employees; Step S3: Correct the abnormal work behavior patterns of interns for the data of abnormal work behavior patterns of employees according to the interviewer's position assignment strategy, so as to obtain the data of abnormal work behavior patterns; construct an employee abnormal work behavior detection model based on the data of abnormal work behavior patterns; Step S4: Detect abnormal work behaviors in the enterprise employee work logs through the employee abnormal work behavior detection model, so as to obtain the data of employees with abnormal work behaviors; evaluate the job fitness of employees with abnormal work behaviors for the data of the interviewer's job fit and the data of employees with abnormal work behaviors, so as to obtain the job fitness data of employees with abnormal work behaviors; Step S5: Analyze the job transfer strategy for employees with abnormal work behaviors based on the data of the interviewer's job fit and the job fitness data of employees with abnormal work behaviors, so as to obtain the job transfer strategy for employees with abnormal work behaviors, and upload it to the human resource management platform to execute the job allocation task.

[0005] By analyzing the data on the suitability of interviewees for positions, the present invention can more accurately match candidates to suitable positions. This helps improve the quality of recruitment, ensuring that the skills and experience of new employees match the job requirements. Data-driven assessment can automate the screening process, thereby reducing manual intervention, saving time, and improving recruitment efficiency. The job assignment strategy derived from data analysis can ensure that each position has the most suitable candidate, thus enhancing employee job satisfaction and efficiency. By analyzing employees' work logs, work behavior patterns of employees can be identified, including work habits, efficiency, and workload. This helps understand the actual performance and work methods of employees. Analysis of abnormal work behavior patterns can promptly identify employees who do not conform to normal behavior patterns, which helps take early measures to prevent potential problems or negative impacts from spreading. By correcting abnormal employee behavior patterns, those abnormal behavior patterns that may occur due to interns or probationary employees can be corrected and adjusted. This helps maintain the overall stability of the team and the working environment. The employee abnormal work behavior detection model constructed based on abnormal behavior data can provide a reliable tool for future abnormal behavior detection, enhancing the ability of early warning and intervention. Using the detection model to detect abnormal behavior in employees' work logs can promptly identify problems and reduce their negative impacts on work efficiency and team atmosphere. By evaluating the job fitness of employees with abnormal behavior, it can be understood which positions are more suitable for these employees, helping to formulate targeted improvement strategies. By analyzing and formulating job transfer strategies for employees with abnormal behavior, these employees can be transferred to more suitable positions, thereby enhancing their work performance and satisfaction. A reasonable job transfer strategy can reduce employee turnover caused by job inadaptability and lower the costs of recruiting and training new employees. Through systematic analysis of job transfer strategies, human resources can be better allocated, improving the overall operational efficiency of the enterprise and employee satisfaction. Overall, these steps, through data-driven analysis and model construction, help solve problems in traditional human resource management, such as the inefficiency of the recruitment process, employee-job mismatch, and the handling of abnormal behavior. Data analysis and model construction can not only improve recruitment efficiency and employee fitness but also enhance the enterprise's early warning and handling capabilities for abnormal behavior, thereby improving the overall effect of human resource management.

[0006] Optionally, step S1 is specifically as follows: Step S11: Obtain enterprise recruitment data, and extract enterprise job recruitment data and interview resume data from the enterprise recruitment data, so as to obtain enterprise job recruitment data and interviewee resume data; Step S12: Perform text conversion on the enterprise job recruitment data to obtain enterprise job recruitment text data, and extract recruitment requirement characteristics from the enterprise job recruitment text data to obtain enterprise job recruitment requirement data; Step S13: Extract the characteristics of the interviewer's work experience years from the interviewer's resume data to obtain the interviewer's work experience years data, and classify the interviewer's resume data according to the interviewer's work experience years data to obtain the interviewer's experience level classification resume data; Step S14: Evaluate the suitability of the interviewer for the position based on the enterprise's job recruitment requirement data and the interviewer's experience level classification resume data to obtain the interviewer's position suitability data; Step S15: Analyze the enterprise's job allocation strategy for the enterprise's job recruitment data according to the interviewer's position suitability data to obtain the interviewer's job allocation strategy, and upload it to the human resource management platform to execute the job allocation task.

[0007] By extracting the enterprise recruitment data and interview resume data, the present invention can accurately obtain the recruitment requirements and the detailed information of the interviewees, laying a foundation for subsequent analysis. Converting the recruitment data into text and extracting features helps to understand the specific recruitment requirements, thereby improving the pertinence and effectiveness of recruitment. By extracting the work experience years and performing level classification, the experience level of the candidates can be more accurately evaluated, providing a basis for job matching. Evaluating the position suitability helps to determine the matching degree between the interviewee and the position, thereby optimizing the recruitment decision and improving the recruitment success rate. Analyzing the enterprise's job allocation strategy based on the position suitability can improve the rationality and efficiency of job allocation, and ultimately achieve more accurate personnel arrangement.

[0008] Optionally, step S14 is specifically as follows: Step S141: Extract the skill requirement characteristics and the experience years requirement characteristics from the enterprise's job recruitment requirement data to obtain the job recruitment skill requirement data and the job recruitment experience years requirement data; Step S142: Screen the interviewer's experience level classification resume data according to the job recruitment experience years requirement data to obtain the interviewer's experience screening resume data; Step S143: Convert the interviewer's experience screening resume data into resume description text data, and extract the experience skill description characteristics from the interviewer's resume description text data to obtain the resume skill description data and the resume experience description data; Step S144: Calculate the similarity of the description skills between the job recruitment skill requirement data and the resume skill description data to obtain the description skill similarity data, and classify the interviewer's resume description text data according to the description skill similarity data to obtain the high similarity skill personnel data and the low similarity skill personnel data; Step S145: Calculate the descriptive experience similarity between the job recruitment experience years requirement data and the resume experience description data to obtain the descriptive experience similarity data, and classify the interviewer's resume description text data based on the descriptive experience similarity data to obtain the high-similarity experience personnel data and the low-similarity experience personnel data; Step S146: Perform an intersection operation on the personnel numbers of the high-similarity skill personnel data and the high-similarity experience personnel data to obtain the high-job-fit personnel data; perform an intersection operation on the personnel numbers of the low-similarity skill personnel data and the low-similarity experience personnel data to obtain the low-job-fit personnel data; Step S147: Merge the high-job-fit personnel data and the low-job-fit personnel data to obtain the interviewer's job-fit data.

[0009] By extracting the skill requirements and years of experience requirements characteristics from the enterprise job recruitment demand data, the specific requirements of each position for skills and experience can be clearly understood. This helps to ensure the accuracy of recruitment requirements, thereby improving recruitment efficiency. With clear skill and years of experience requirements, enterprises can more targeted design recruitment criteria and screening conditions, making recruitment decisions more data-driven. Screening resumes according to the years of experience required for the position can effectively filter out candidates who do not meet the experience requirements, saving the time of recruiters. Through screening by years of experience, it can be ensured that the selected candidates have the required work experience, thus improving the matching degree between candidates and positions. Converting the resume description text into structured data and extracting skill and experience characteristics helps to standardize the information extraction process, making subsequent analysis more efficient and accurate. By extracting skill and experience descriptions, the actual capabilities and experiences of candidates can be more deeply understood, which helps to accurately evaluate their job matching degree. By calculating the skill similarity between the job recruitment skill requirement data and the resume skill description data, the matching degree between the candidate's skills and the job requirements can be evaluated. Classifying resumes according to skill similarity can give priority to candidates who highly match the job skill requirements, thereby improving the recruitment effect. By calculating the experience similarity between the job recruitment years of experience requirement data and the resume experience description data, the matching degree between the candidate's experience and the job requirements can be evaluated. Classifying resumes according to experience similarity can more effectively screen out candidates who meet the experience requirements, further improving the recruitment quality. By performing an intersection operation on highly similar skill personnel and highly similar experience personnel, candidates who highly meet the job requirements in both skills and experience can be identified, which helps to focus on evaluating candidates who best match the job requirements, thus improving the accuracy and efficiency of recruitment decisions. By merging the data of personnel with high job fit and low job fit, the job fit situation of all candidates can be comprehensively understood. The merged data can help recruiters quickly identify the best candidates and arrange more targeted interviews, and at the same time can also identify candidates who do not meet the requirements, thus improving recruitment efficiency.

[0010] Optionally, step S15 is specifically as follows: Step S151: Classify personnel according to the job fit data of the interview personnel, so as to obtain high job fit personnel data and low job fit personnel data; Step S152: Allocate the enterprise job recruitment data and the high job fit personnel data to the positions applied by the personnel, so as to obtain the allocated data of the high fit personnel for the positions applied, and extract the vacant positions from the enterprise job recruitment data according to the allocated data of the high fit personnel for the positions applied, so as to obtain the vacant position recruitment data; Step S153: Extract the descriptive experience features of the data of personnel with low job fit, so as to obtain the descriptive experience data of the interviewees, and perform job recruitment description matching on the descriptive experience data of the interviewees and the recruitment data of vacant positions, so as to obtain the job matching data of personnel with low fit; Step S154: Perform hierarchical sorting job assignment on the job matching data of personnel with low fit according to the resume data divided by the experience level of the interviewees, so as to obtain the job assignment data of the interviewees with low fit; Step S155: Integrate the job assignment strategies according to the job assignment data of the personnel with high fit and the job assignment data of the personnel with low fit, so as to obtain the job assignment strategy of the interviewees, and upload it to the human resource management platform to execute the job assignment task.

[0011] By classifying interviewees according to job fit, the recruitment team can more quickly identify which personnel are most suitable for the job requirements. This reduces the time wasted on unsuitable candidates. The recruitment team can focus more resources on candidates with high job fit, improving the recruitment quality. Clearly distinguishing the data of personnel with high fit and low fit can help recruiters conduct subsequent job assignment and matching work more targeted. Assigning jobs according to the data of personnel with high job fit can significantly improve the matching degree between the applicant and the job, thus increasing the recruitment success rate. By docking the personnel with high fit with the enterprise job recruitment data, the vacant positions can be effectively filled, reducing the vacancies in enterprise recruitment. High-matching job assignment helps to improve the job satisfaction and retention rate of employees because they are more likely to perform well in the jobs suitable for them. Even personnel with low job fit may have potential value for other jobs. Through the extraction of descriptive experience features, the jobs suitable for these personnel can be discovered. Through job recruitment description matching, the personnel with low fit can be more flexibly assigned to the jobs suitable for their experience, thus maximizing their work potential. By effectively matching the experience of personnel with low fit and the requirements of vacant positions, all recruitment resources can be better utilized. The classification and sorting of experience levels can help recruiters systematically assign jobs to personnel with low fit, making the matching more reasonable. By sorting personnel according to their experience levels, it can be ensured that the personnel with low fit are assigned to the jobs most suitable for their experience, improving the job suitability. This orderly assignment method can reduce the randomness in the recruitment process and ensure that each job gets the suitable personnel. By integrating the job assignment data of personnel with high fit and low fit, a more comprehensive and optimized recruitment strategy can be formed. The integration strategy helps to ensure that every step in the recruitment process is fully considered and optimized, thus enhancing the overall recruitment effect. Uploading the strategy to the human resource management platform can realize the systematic management and execution of the recruitment strategy, improving the management efficiency.

[0012] Optionally, step S2 is specifically as follows: Step S21: Obtain the work logs of enterprise employees, and extract the employee attendance logs and the project work logs of enterprise employees from the work logs of enterprise employees, so as to obtain the employee attendance logs and the project work logs of employees; Step S22: Extract the project volume characteristics and the project work time series characteristics from the project work logs of employees, so as to obtain the project volume data and the project work time series data; Step S23: Calculate the project completion time according to the project work time series data, so as to obtain the project completion time data, and integrate the employee project completion modes according to the project completion time data and the project volume data, so as to obtain the employee project completion mode data; Step S24: Extract the employee attendance missing characteristics from the employee attendance logs, so as to obtain the employee absenteeism data and the employee field work data; Step S25: Extract the project secondment log characteristics from the project work logs of employees, so as to obtain the project secondment data, and correct the work field trips of employees according to the project secondment data, so as to obtain the abnormal employee field work data; Step S26: Integrate the employee absenteeism data and the abnormal employee field work data to obtain the employee work time mode data; Step S27: Conduct descriptive statistical identification of the employee work behavior modes on the employee project completion mode data and the employee work time mode data, so as to obtain the employee work behavior mode data; Step S28: Analyze the abnormal work behavior modes according to the employee work behavior mode data, so as to obtain the abnormal employee work behavior mode data.

[0013] The present invention centrally collects and extracts the attendance and project work logs of employees, ensuring the integrity and consistency of data and providing a reliable data basis for subsequent analysis. It clearly records the attendance of employees, including information such as going to work, getting off work, and taking leave, helping the enterprise understand the attendance of employees. It clarifies the work logs of employees in each project, helping the enterprise monitor the progress of projects and the allocation of employees' working hours. By extracting the project volume characteristics, it is possible to analyze indicators such as the scale and workload of the project, providing a quantitative basis for project management. Extracting the project work time series characteristics helps to understand the progress and time distribution of the project, thereby effectively monitoring the execution of the project and discovering potential progress problems or delays. Calculating the actual completion time of the project and comparing it with the planned time helps to evaluate the accuracy of project progress and the work efficiency of employees. By integrating the project volume data and the completion time data, it is possible to identify the work patterns of employees in different projects, helping the enterprise optimize resource allocation and work processes and improve project management efficiency. Extracting the absenteeism characteristics of employees can identify the absenteeism frequency and reasons, providing decision-making support for management to improve the attendance policy or provide employee support. Obtaining the field work data of employees helps to manage the expatriate tasks and business trips, ensuring that the work of employees during field work is effectively recorded and managed. Recording the project expatriate data helps to track the work situation of employees at different project and customer sites, improving the transparency of expatriate management. By correcting the field work data of employees, it is possible to identify and correct abnormal field work situations, such as false reporting or unauthorized field work, ensuring the accuracy and integrity of the data. Integrating the absenteeism data and abnormal field work data of employees can comprehensively understand the work time patterns of employees and identify abnormal work time behaviors. By analyzing the work time patterns of employees, the enterprise can formulate more effective work arrangements and attendance policies, improving the overall work efficiency and employee satisfaction. Through descriptive statistics of the employee project completion mode data and work time mode data, the work habits and behavior patterns of employees can be systematically understood. This statistical identification can reveal the typical behavior characteristics of employees in the process of completing tasks, including work efficiency, time allocation, and task completion. This can help the enterprise identify the behavior patterns of highly efficient workers and provide data support for formulating more effective work processes and performance evaluation criteria. Conducting abnormal behavior analysis on the employee work behavior pattern data helps to identify those work behaviors that deviate from the normal pattern. The identification of this abnormal pattern can reveal potential problems, such as low work efficiency, procrastination, or potential employee problems. By timely discovering and analyzing these abnormal patterns, the enterprise can take targeted measures for intervention to improve the overall work performance and employee satisfaction, while reducing potential risks and losses.

[0014] Optionally, step S28 is specifically as follows: Step S281: Extract the employee project cycle characteristics and employee absenteeism time characteristics from the employee work behavior pattern data, so as to obtain the employee project cycle data and the employee absenteeism time data; Step S282: Perform project volume clustering on the employee project cycle data, so as to obtain the project volume clustering cycle data, and perform project abnormal cycle threshold statistics on the project volume clustering cycle data, so as to obtain the project abnormal cycle threshold; Step S283: Perform abnormal cycle classification calculation on the employee project cycle data according to the project abnormal cycle threshold, so as to obtain the employee abnormal project cycle data; Step S284: Perform absenteeism time threshold statistics on the employee absenteeism time data, so as to obtain the abnormal absenteeism time threshold, and perform abnormal absenteeism time classification calculation on the employee absenteeism time data according to the abnormal absenteeism time threshold, so as to obtain the employee abnormal absenteeism time data; Step S285: Construct an employee abnormal work mode detection model according to the employee abnormal project cycle data and the employee abnormal absenteeism time data; Step S286: Use the employee abnormal work mode detection model to obtain the abnormal employee work behavior pattern data from the employee work behavior pattern data.

[0015] By extracting the project cycle and absenteeism time characteristics of employees, the present invention can systematically organize the work data of employees. This makes the data more structured and facilitates subsequent analysis. Obtaining the project cycle data and absenteeism time data of employees can help managers better understand the work and absenteeism patterns of employees, thereby providing a basis for adjusting work arrangements and optimizing human resource allocation. Through project volume clustering, the project cycle data can be classified by volume to identify the cycle characteristics of different types of projects. This helps to understand the cycle rules of different project types. By statistically calculating the abnormal cycle threshold, those projects with abnormal cycles can be identified. This helps to discover potential problems in project management, such as schedule delays or improper resource allocation, and thus take corresponding corrective measures. Through classification calculation, the abnormal project cycles can be accurately identified. This helps enterprises to discover those projects with significant deviations in time or progress, providing data support for further investigation and improvement. The work performance of employees can be evaluated by combining the abnormal cycle data to identify employees who need additional support or training. Statistically calculating the absenteeism time threshold and classifying can help identify employees with abnormal absenteeism time. This is very important for understanding absenteeism patterns, managing absenteeism problems, and formulating relevant policies (such as sick leave or leave management). By analyzing the abnormal absenteeism data, employee groups that may need additional support or improved benefits can be identified, thereby improving overall employee satisfaction and productivity. The model can comprehensively consider the data in both aspects of the project cycle and absenteeism time to provide comprehensive detection of abnormal work patterns. This helps enterprises to more accurately identify potential problems in work behaviors. Through the abnormal work patterns detected by the model, enterprises can take early intervention measures to improve work processes, optimize resource allocation, enhance employee performance and work satisfaction. Through the abnormal work patterns detected by the model, enterprises can take early intervention measures to improve work processes, optimize resource allocation, enhance employee performance and work satisfaction. Provide data-driven decision support to help management formulate more accurate management strategies and improvement measures, and enhance the operational efficiency and employee performance of the organization.

[0016] Optionally, step S3 is specifically as follows: Step S31: Extract the assigned position number characteristics of the interviewer's position assignment strategy to obtain the assigned position number data; Step S32: Perform position number association on the assigned position number data and the abnormal employee work behavior pattern data to obtain the abnormal work behavior pattern data of intern employees; Step S33: Correct the abnormal work behavior pattern data of intern employees according to the abnormal work behavior pattern data of intern employees to obtain the abnormal work behavior pattern data; Step S34: Iteratively optimize the parameters of the employee abnormal work pattern detection model according to the abnormal work behavior pattern data to obtain the employee abnormal work behavior detection model.

[0017] By extracting the job number features, the present invention can accurately understand the requirements and demands of each position, so as to make reasonable job assignments. This helps to ensure that the skills and job responsibilities of employees match the job requirements, improving work efficiency and employee satisfaction. The extracted features can help integrate information from different data sources, establish a systematic database, facilitate subsequent data analysis and processing, provide basic data support for subsequent analysis of abnormal behavior patterns, and contribute to the construction of an accurate behavior pattern detection model. By associating the assigned job number data with the abnormal work behavior pattern data, abnormal behavior patterns related to specific positions can be identified. This helps to discover which positions are more prone to abnormal behavior and formulate targeted intervention measures. Associating the job number with the abnormal behavior pattern helps to improve the accuracy of data matching and provide a more accurate data basis for subsequent analysis. By identifying abnormal behavior patterns, measures can be taken in advance to prevent potential risks and enhance the scientificity and effectiveness of employee management. By correcting the abnormal work behavior pattern data of interns, the accuracy of the data can be improved, ensuring that the prediction and detection results of the model are more reliable. The corrected data helps to optimize the behavior pattern detection model, making it more adaptable to the actual work environment and employee behavior, thus enhancing the performance and effect of the model. Provide precise analysis of the behavior of interns, help the company better understand and manage the performance of interns during the internship period, and reduce misjudgment and missed judgment. By iteratively optimizing and adjusting the parameters of the abnormal work mode detection model, the detection accuracy and efficiency of the model can be significantly improved, so as to more effectively identify abnormal work behaviors. The optimized model can better adapt to the actual situation of the enterprise and the employee behavior pattern, enhancing the flexibility and applicability of the detection. More precise detection of abnormal behaviors can help the enterprise timely discover and respond to potential risks, improve the management level, and reduce the negative impacts brought by abnormal behaviors.

[0018] Optionally, step S4 is specifically as follows: Step S41: Detect abnormal work behaviors in the enterprise employee work logs through the employee abnormal work behavior detection model, so as to obtain abnormal work behavior employee data; Step S42: Extract the employee start time features from the abnormal work behavior employee data, so as to obtain abnormal employee start time data; Step S43: Perform clustering of employees with high job fit based on the interviewer job fit data and the interviewer job assignment strategy, so as to obtain high job fit abnormal employee data; Step S44: Conduct statistical analysis of the start time based on the abnormal employee start time data, so as to obtain high-amount start time abnormal employee data and low-amount start time abnormal employee data; Step S45: Perform an intersection operation on the employee numbers of the employees with abnormal high job fitness and the employees with abnormal low entry time to obtain the data of employees with abnormal low fitness; perform an intersection operation on the employee numbers of the employees with abnormal high job fitness and the employees with abnormal high entry time to obtain the data of employees with abnormal high fitness; Step S46: Perform an evaluation and combination of the job fitness of the employees with abnormal low fitness and the employees with abnormal high fitness to obtain the job fitness data of the employees with abnormal work behaviors.

[0019] Through detecting abnormal work behaviors, an enterprise can discover potential employee problems at an early stage, such as bad work habits, low productivity or other behaviors that do not meet the company's standards. Detecting based on actual data can provide objective records of abnormal behaviors and help management make data-driven decisions. Improvement measures and training plans can be formulated for the detected abnormal behaviors, thereby improving the overall performance of employees. By extracting the characteristics of the entry time, it can be identified whether the entry time has an impact on the performance of employees. For example, the adaptation period at the beginning of employment may affect work performance, providing basic data for further analyzing employee performance and helping to understand the relationship between the entry time and employees' abnormal behaviors. Through cluster analysis, employees with high job fitness but abnormal performance can be identified, which helps the enterprise evaluate the effectiveness of interview and job assignment strategies, and can improve the recruitment and job assignment strategies targeted to improve employees' job fitness and reduce the incidence of abnormal behaviors. Grouping abnormal employees according to the entry time can help understand the performance characteristics of employees in different entry time periods, provide more detailed management suggestions, identify trends or patterns between the entry time and abnormal behaviors, and help optimize the entry training and support strategies. Through the intersection operation, abnormal employees can be more accurately classified as having low fitness or high fitness, so as to conduct more targeted management, and it can be identified which employees, even with high job fitness, perform abnormally due to other reasons (such as entry time problems), helping to conduct targeted interventions. By combining and evaluating the data of employees with abnormal low fitness and high fitness, a more comprehensive analysis result of employees' job fitness can be obtained, helping the enterprise formulate more accurate management strategies, such as improving on-the-job training, adjusting job matching, and enhancing the overall fitness and work performance of employees.

[0020] Optionally, Step S5 is specifically as follows: Step S51: Adjust the job fitness of the job fitness data of the interviewees according to the job fitness data of the employees with abnormal work behaviors to obtain the actual job fitness data of the interviewees; Step S52: Extract the job fitness adjustment of the job assignment strategy of the interviewees according to the actual job fitness data of the interviewees to obtain the job assignment data of the employees to be transferred; Step S53: Extract the job assignment description from the job assignment data of the employee to be transferred, so as to obtain the job assignment description data to be allocated; Step S54: Perform job description matching based on the actual job fit data of the interviewees and the job assignment description data to be allocated, so as to obtain the job transfer strategy for employees with abnormal work behaviors, and upload it to the human resource management platform to execute the job allocation task.

[0021] By adjusting the job fit data of the interviewees, the present invention can more accurately evaluate their actual adaptability. This helps to ensure that the performance of employees in the new positions meets expectations, thereby improving work efficiency and satisfaction. Using the actual job fit data to adjust the job assignment strategy can optimize the personnel allocation and reduce the problem of improper job assignment. This can improve the job matching degree and reduce the risks and costs of job transfers. Extracting the job description of the job assignment data of the employee to be transferred helps to clarify the job requirements and responsibilities. This can provide accurate information for subsequent job matching and ensure that the job description is consistent with the actual needs. By matching the actual job fit data and the job assignment description data to be allocated, a more effective job transfer strategy can be formulated and uploaded to the human resource management platform to promote the automation and systematization of job allocation. This not only improves the allocation efficiency but also enhances the rationality and scientificity of personnel allocation.

[0022] Optionally, the present invention also provides an online management system for human resources, which is used to execute an online management method for human resources as described above. The online management system for human resources includes: An interviewee job assignment module, which is used to obtain enterprise recruitment data, evaluate the job fit of interviewees based on the enterprise recruitment data, so as to obtain the job fit data of interviewees; analyze the enterprise job assignment strategy according to the job fit data of interviewees, so as to obtain the job assignment strategy of interviewees, and upload it to the human resource management platform to execute the job assignment task; An abnormal work behavior pattern analysis module, which is used to obtain the work logs of enterprise employees, analyze the work behavior patterns of enterprise employees based on the work logs of enterprise employees, so as to obtain the work behavior pattern data of employees; analyze the abnormal work behavior patterns according to the work behavior pattern data of employees, so as to obtain the abnormal work behavior pattern data of employees; An abnormal work behavior pattern correction module, which is used to correct the abnormal work behavior pattern data of interns according to the job assignment strategy of interviewees to obtain the abnormal work behavior pattern data; construct an employee abnormal work behavior detection model based on the abnormal work behavior pattern data; The employee position fitness evaluation module is used to detect abnormal work behaviors in the work logs of enterprise employees through an employee abnormal work behavior detection model, so as to obtain data on employees with abnormal work behaviors; evaluate the position fitness of employees with abnormal work behaviors based on the position fitness data of interviewees and the data of employees with abnormal work behaviors, so as to obtain the position fitness data of employees with abnormal work behaviors. The employee job transfer strategy analysis module is used to analyze the job transfer strategies of employees with abnormal work behaviors based on the position fitness data of interviewees and the position fitness data of employees with abnormal work behaviors, so as to obtain the job transfer strategies of employees with abnormal work behaviors and upload them to the human resource management platform to execute the job allocation task.

[0023] The online management system for human resources of the present invention can implement any one of the online management methods for human resources of the present invention, and is a medium for combining the operations and signal transmissions between various modules to complete the online management method for human resources. The internal modules of the system cooperate with each other, thereby reducing the personnel turnover rate and recruitment costs. Brief Description of the Drawings

[0024] By reading the detailed description of the non-restrictive embodiments with reference to the following drawings, other features, objectives, and advantages of the present invention will become more apparent: Figure 1 It is a schematic flow chart of the steps of the online management method for human resources of the present invention; Figure 2 It is a detailed schematic flow chart of step S1 in the present invention; Figure 3 It is a detailed schematic flow chart of step S14 in the present invention; The realization, functional characteristics, and advantages of the objectives of the present invention will be further described in conjunction with the embodiments with reference to the drawings. Detailed Embodiments

[0025] The technical method of the present invention will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0026] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0027] It should be understood that although terms such as "first" and "second" may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the associated listed items.

[0028] To achieve the above object, please refer to Figures 1 to 3 , the present invention provides an online management method for human resources, and the method includes the following steps: Step S1: Obtain enterprise recruitment data, and evaluate the job fit degree of the interviewed personnel based on the enterprise recruitment data, so as to obtain the job fit degree data of the interviewed personnel; analyze the enterprise job allocation strategy based on the job fit degree data of the interviewed personnel, so as to obtain the job allocation strategy of the interviewed personnel, and upload it to the human resources management platform to execute the job allocation task; In this embodiment, the resumes, interview feedback, and skills test results of the interviewed personnel are extracted from the enterprise recruitment system. These data include educational background, work experience, professional skills, interview scores, etc. A machine learning model (such as a decision tree or a random forest) is used to evaluate the job fit degree of the interviewed personnel. The model inputs include job requirements (such as required skills and experience) and the qualification data of the interviewed personnel. The model output is the fitness score of the interviewed personnel for each job. Based on the job fit degree score, the best job allocation strategy is determined through an optimization algorithm (such as linear programming) to ensure that each job is assigned to the most suitable person. The final job allocation strategy is uploaded to the human resources management platform through the API to ensure that the human resources system can execute the job allocation task according to these strategies.

[0029] Step S2: Obtain the work logs of enterprise employees, and analyze the work behavior patterns of enterprise employees based on the work logs of enterprise employees, so as to obtain the work behavior pattern data of employees; analyze the abnormal work behavior patterns based on the work behavior pattern data of employees, so as to obtain the abnormal work behavior pattern data of employees; In this embodiment, work log data of employees is extracted from the enterprise's employee management system, including task completion time, work activity records, attendance data, etc. Data mining techniques (such as clustering analysis) are applied to analyze the work behavior patterns of employees. By extracting behavior features (such as work time distribution, task type, etc.), common behavior patterns are identified. An anomaly detection model (such as an isolation forest-based model) is established to analyze the work behavior patterns of employees and identify anomaly behavior data that deviates from the normal pattern.

[0030] Step S3: According to the interview candidate position assignment strategy, correct the abnormal work behavior patterns of interns for the abnormal work behavior pattern data of employees, so as to obtain the abnormal work behavior pattern data; construct an employee abnormal work behavior detection model based on the abnormal work behavior pattern data; In this embodiment, the identified abnormal work behaviors are audited and corrected in combination with the employee's entry time in the position assignment strategy. For example, if an employee's behavior is abnormal and the employee's position does not match their skills, adjustment is required if the employee has just been assigned a position. The anomaly detection model is trained using work behavior data marked as normal or abnormal. Select a model (such as a support vector machine or a deep learning model) and use this data for training to improve the model's ability to identify abnormal behaviors.

[0031] Step S4: Detect abnormal work behaviors in the enterprise employee work log through the employee abnormal work behavior detection model, so as to obtain data on employees with abnormal work behaviors; evaluate the job fitness of employees with abnormal work behaviors for the interview candidate position fitness data and the data on employees with abnormal work behaviors, so as to obtain data on the job fitness of employees with abnormal work behaviors; In this embodiment, the constructed employee abnormal work behavior detection model is applied to analyze new work log data. The model will identify potential abnormal behaviors, such as frequent absences or irregular working hours. Combining the position fitness data of interview candidates and the detected abnormal behaviors, a weighted scoring model is used to evaluate the job fitness of each employee with abnormal behaviors. The evaluation criteria may include changes in work performance, the matching degree between position requirements and actual performance, etc.

[0032] Step S5: Analyze the job transfer strategy for employees with abnormal work behaviors based on the interview candidate position fitness data and the job fitness data of employees with abnormal work behaviors, so as to obtain the job transfer strategy for employees with abnormal work behaviors and upload it to the human resource management platform to execute the position allocation task.

[0033] In this embodiment, based on the evaluation results of job fitness, an optimization algorithm (such as a genetic algorithm) is used to analyze the job transfer strategy for employees with abnormal work behaviors. The algorithm takes into account the skills of employees, the types of abnormal behaviors, and job requirements to generate the optimal job transfer strategy. The final job transfer strategy is uploaded to the human resource management platform through the API to ensure that the system can execute the job adjustment tasks for employees according to these strategies.

[0034] Optionally, step S1 is specifically as follows: Step S11: Obtain enterprise recruitment data, and extract enterprise job recruitment data and interview resume data from the enterprise recruitment data, so as to obtain enterprise job recruitment data and interviewer resume data; In this embodiment, recruitment and resume data are extracted from enterprise recruitment platforms (such as LinkedIn, Indeed, etc.) and resume databases. Automated data scraping is performed using web crawler technology or API interfaces to ensure recruitment information including job titles, job descriptions, required skills, company names, recruitment dates, etc., as well as resume content such as the educational background, work experience, and skills of interviewees. Parse the job information fields to extract key information from the job description, such as job responsibilities, required skills, educational requirements, etc. Extract information such as work experience, educational background, and skills from the resume. Use natural language processing (NLP) techniques for information extraction, such as entity recognition technology to extract company names, job names, and work time periods.

[0035] Step S12: Perform text conversion on the enterprise job recruitment data to obtain enterprise job recruitment text data, and extract recruitment requirement features from the enterprise job recruitment text data to obtain enterprise job recruitment requirement data; In this embodiment, the extracted recruitment data is preprocessed, including removing noise characters, unifying formats, etc. Then the recruitment information is converted into a structured text data format, such as JSON or CSV. Use text analysis techniques, such as TF-IDF (Term Frequency-Inverse Document Frequency) or BERT (Bidirectional Encoder Representations from Transformers) models, to extract features from the recruitment text. The extracted features include core skills required for the job, required years of experience, educational requirements, etc. For example, from a recruitment description of "Requires more than 3 years of marketing experience", "marketing" is extracted as a skill requirement and "3 years" as the required years of experience.

[0036] Step S13: Extract the working experience years features of the interviewers from the interviewer resume data to obtain the interviewer working experience years data, and classify the interviewer resume data according to the interviewer working experience years data to obtain the interviewer experience level classification resume data; In this embodiment, the work experience part in the interview resume is analyzed. Using the time extraction tool in NLP technology, "5 years" is extracted from "worked in Company X for 5 years" as the work experience duration. According to the experience duration, the interviewees are divided into different levels. For example, interviewees with 0 - 2 years of experience are classified as junior, those with 3 - 5 years of experience are classified as intermediate, and those with more than 6 years of experience are classified as senior. If the resume mentions "has 6 years of market research experience", then this interviewee is marked as "senior".

[0037] Step S14: Evaluate the job - fitting degree of the interviewees based on the enterprise job recruitment requirement data and the resume data divided by the experience levels of the interviewees, so as to obtain the job - fitting degree data of the interviewees; In this embodiment, matching algorithms such as cosine similarity and semantic matching of the BERT model are used to calculate the matching degree between the interviewee's resume and the job requirements. The scoring system can divide the matching degree into high, medium, and low. For example, for a position with the job requirement of "being familiar with market analysis", if the interviewee's resume contains relevant market analysis experience, the matching degree will be higher.

[0038] Step S15: Analyze the enterprise job allocation strategy based on the job - fitting degree data of the interviewees for the enterprise job recruitment data, so as to obtain the job allocation strategy for the interviewees and upload it to the human resource management platform to execute the job allocation task.

[0039] In this embodiment, based on the job - fitting degree data, optimization algorithms (such as linear programming or genetic algorithms) are used to analyze the best job allocation strategy. For example, considering the different requirements of multiple positions and the matching degrees of the interviewees, an optimal job allocation plan is formulated. For example, if a certain position requires senior experience and there are multiple senior interviewees, then the interviewee who best meets the requirements of this position is selected for job allocation. The analyzed job allocation strategy is imported into the human resource management platform through an API interface or in a batch upload manner.

[0040] Optionally, step S14 is specifically as follows: Step S141: Extract the skill requirement characteristics and the experience - duration requirement characteristics according to the enterprise job recruitment requirement data, so as to obtain the job recruitment skill requirement data and the job recruitment experience - duration requirement data; In this embodiment, skills and experience - duration characteristics are extracted from the enterprise job recruitment requirement data. For example, if a position requires "proficient in Python programming" and "at least 5 years of work experience", then these two characteristics are extracted. Using natural language processing technology, skill keywords and experience requirements are identified from the job description and a structured requirement data table is formed.

[0041] Step S142: Screen the resume data of the classified resume data of the interview candidates according to the required years of recruitment experience for the position, so as to obtain the screened resume data of the interview candidates' experience; In this embodiment, the resume data of the interview candidates is screened according to the required years of experience. For example, if the position requires at least 5 years of experience, all resumes that meet this requirement are screened out. Use regular expressions or an algorithm for extracting years of experience to extract work experience data from the resume and compare its duration with the position requirements. It is also possible to perform a quick screening based on the classified resume data of the interview candidates with predefined experience levels to obtain interview candidates whose years of recruitment experience requirements for the position are similar.

[0042] Step S143: Convert the screened resume data of the interview candidates into resume description text data, and extract the feature of the experience and skill description from the resume description text data of the interview candidates, so as to obtain the resume skill description data and the resume experience description data; In this embodiment, the screened resumes are converted into text, and the skills and experience descriptions are extracted. For example, convert the work experience description in the resume into structured data, and identify the skills of "project management" and "3 years of management experience". Use a text analysis tool to perform word segmentation and word frequency analysis on the resume to extract the skills and experience features.

[0043] Step S144: Calculate the similarity of the described skills between the required skill data for the position and the resume skill description data, so as to obtain the described skill similarity data, and classify the interview candidates' resume description text data according to the described skill similarity data, so as to obtain the high-similarity skill personnel data and the low-similarity skill personnel data; In this embodiment, the similarity between the required skill data for the position and the resume skill description data is calculated. Apply the vector space model or the cosine similarity algorithm to calculate the similarity between skills and classify the resumes according to the skill similarity. First, extract the keywords in the required skill data for the position and convert these keywords into vector representations of words; then, perform the same word vector conversion on the skill descriptions in the interview candidates' resumes. Next, use the cosine similarity algorithm to calculate the similarity score between the word vectors of the required skills for the position and the word vectors of the resume skills. For example, assume that the position requirements include "data analysis" and "Python programming", while the resume contains "data processing" and "Python development". The word vector similarity between "data analysis" and "data processing" and the word vector similarity between "Python programming" and "Python development" can be calculated. After integrating these similarity scores, the resumes are classified according to skills. Resumes with high scores will be classified as high-similarity skill personnel data, and those with low scores will be classified as low-similarity skill personnel data.

[0044] Step S145: Calculate the descriptive experience similarity between the job recruitment experience years requirement data and the resume experience description data to obtain the descriptive experience similarity data, and classify the resume description text data of the interviewees according to the descriptive experience similarity data to obtain the high-similarity experience personnel data and the low-similarity experience personnel data; In this embodiment, the similarity between the job recruitment experience years requirement data and the resume experience description data is calculated. Using a similarity algorithm, the experience years are compared, and the resumes are classified according to the experience similarity. Extract the specific years requirement in the job recruitment experience years requirement, such as "more than 5 years of project management experience", and convert it into an experience years range. Then, parse the skill experience description in the interviewee's resume, extract the years information and standardize it. Next, use a threshold matching algorithm to compare the job requirement years and the resume experience years. For example, if the job requirement is "more than 5 years of marketing experience" and the resume shows "4 to 6 years of marketing experience", this resume is classified as high-similarity experience personnel data using threshold matching. If the resume shows "2 years of marketing experience" or "4 to 6 years of Python programming experience", it is classified as low-similarity experience personnel data.

[0045] Step S146: Perform a personnel number intersection operation on the high-similarity skill personnel data and the high-similarity experience personnel data to obtain the high-job-fit personnel data; perform a personnel number intersection operation on the low-similarity skill personnel data and the low-similarity experience personnel data to obtain the low-job-fit personnel data; In this embodiment, an intersection operation is performed on the high-similarity skill personnel and the high-similarity experience personnel to obtain the high-job-fit personnel data; a similar operation is performed on the low-similarity personnel to obtain the low-job-fit personnel data. For example, screen the personnel who have both high skill and high experience similarity.

[0046] Step S147: Merge the high-job-fit personnel data and the low-job-fit personnel data to obtain the interviewee's job-fit data.

[0047] In this embodiment, the high-job-fit personnel data and the low-job-fit personnel data are merged to form the interviewee's job-fit data. After merging the data, the overall personnel adaptation situation can be analyzed and support can be provided for the recruitment decision.

[0048] Optionally, step S15 is specifically: Step S151: Classify the personnel according to the interviewee's job-fit data to obtain the high-job-fit personnel data and the low-job-fit personnel data; In this embodiment, the interviewees are divided into two categories: high job - fit and low job - fit based on the job - fit data. Machine learning classification algorithms, such as K - means clustering, can be used to divide the job - fit scores (e.g., comprehensive scores based on skills and experience) of each interviewee into two clusters. Suppose the job - fit score of interviewee A is 85 and that of interviewee B is 45, and the threshold is set at 60. Then, interviewee A is classified as high job - fit data, while interviewee B is classified as low job - fit data.

[0049] Step S152: Perform personnel job assignment for the enterprise job recruitment data and the data of high job - fit personnel, so as to obtain the job assignment data of high - fit personnel, and extract the vacant job data from the enterprise job recruitment data according to the job assignment data of high - fit personnel, so as to obtain the vacant job recruitment data; In this embodiment, job assignment is performed on high job - fit personnel to determine suitable jobs and extract vacant jobs therefrom. The data of high job - fit personnel is docked with the enterprise job recruitment data, and an optimization algorithm (such as linear programming) is used to assign jobs. Suppose job A requires 3 employees, and high - fit personnel X, Y, and Z meet this requirement, then they are assigned to job A. Then, the unfilled jobs are extracted from the data after the assignment of job A to form vacant job data, which is recorded as job A needing additional personnel.

[0050] Step S153: Extract the described experience features from the data of low job - fit personnel to obtain the described experience data of interviewees, and perform job recruitment description matching on the described experience data of interviewees and the vacant job recruitment data, so as to obtain the job matching data of low - fit personnel; In this embodiment, feature extraction is performed on the experience descriptions of low job - fit personnel and matched with the vacant job recruitment data. The work experience descriptions of low job - fit personnel can be extracted, and text analysis tools such as TF - IDF are used to extract keywords. Then, these keywords are matched with the requirement descriptions of vacant jobs, and cosine similarity calculation is applied. Suppose vacant job B requires "customer service experience", and the experience description of interviewee D contains "sales support experience", then matching is performed to generate the job matching data of low - fit personnel.

[0051] Step S154: Perform hierarchical sorting job assignment on the job matching data of low - fit personnel according to the resume data divided by the experience level of interviewees, so as to obtain the job assignment data of low - fit personnel; In this embodiment, the experience levels of personnel with low job fit are ranked to allocate the remaining positions. The sorting algorithm is applied to the experience data of personnel with low fit, such as sorting by years of experience from high to low. Suppose personnel E with low fit has 3 years of experience and personnel F has 1 year of experience, then personnel E will be preferentially allocated to the vacant position B. The sorting basis can be years of experience or the detail of relevant experience.

[0052] Step S155: Integrate the job allocation strategies for high-fit personnel and low-fit personnel based on their job application allocation data, so as to obtain the job allocation strategy for the interview personnel and upload it to the human resource management platform to execute the job allocation task.

[0053] In this embodiment, the job allocation data of high-fit and low-fit personnel are integrated to form the final job allocation strategy and uploaded to the human resource management platform. The allocation results of high-fit personnel are combined with those of low-fit personnel, and a data integration tool (such as an ETL tool) is used to generate a comprehensive allocation strategy. Suppose the high-fit personnel have been allocated to position A, and the low-fit personnel are to be allocated to the remaining positions B and C. After generating the complete allocation plan, it is uploaded to the HR system to execute the task.

[0054] Optionally, step S2 is specifically as follows: Step S21: Obtain the work logs of enterprise employees, and extract the employee attendance logs and the employee project work logs from the work logs of enterprise employees, so as to obtain the employee attendance logs and the employee project work logs; In this embodiment, the work logs of employees are extracted from the enterprise human resource management system using the API. These logs include the employees' clock-in times, project participation records, and task completion status. The original log data is cleaned and classified using a data processing tool (such as the pandas library in Python). For example, the work logs are divided into attendance logs and project work logs. The attendance logs are parsed to record the employees' arrival times, departure times, and leave situations; the project work logs are parsed to record the projects the employees participate in, task times, and progress. For example, the work logs of employee A extracted from the HR system include the clock-in time record. Through data processing, we divide these records into an attendance log (arrival time: 08:00, departure time: 17:00) and a project work log (project X, task Y, working hours: 3 hours).

[0055] Step S22: Extract the project volume characteristics and the project work time series characteristics from the employee project work logs, so as to obtain the project volume data and the project work time series data; In this embodiment, the total working hours, the number of tasks, and the completion status of each project are analyzed. For example, the working hours of each project, the complexity of tasks, etc. are counted. Time series data is extracted from the project work logs to record the change of the project progress over time, such as the change curve of the task completion time. For example, for project X, the extracted volume features include a total working time of 40 hours and 10 tasks. The working time series features show that the rate of task completion accelerates in the first two weeks and then slows down.

[0056] Step S23: Calculate the project completion time based on the project working time series data to obtain project completion time data, and integrate the employee project completion mode according to the project completion time data and the project volume data to obtain employee project completion mode data; In this embodiment, according to the project working time series data, the actual completion time of each project is calculated and compared with the planned completion time. The project completion time data is combined with the project volume data to analyze the employee's project completion mode, such as whether it is completed on time or overdue. For example, for project X, the actual completion time is 20 days and the planned completion time is 15 days. After integrating the data, it is analyzed that employee A's project completion mode is "overdue completion".

[0057] Step S24: Extract the employee attendance missing features from the employee attendance logs to obtain employee absenteeism data and employee field work data; In this embodiment, absenteeism situations are identified from the attendance logs, including not punching in, being late, leaving early, etc. The employee absenteeism records are marked as "full attendance", "absenteeism", or "field work". For example, employee B has 5 days of not punching in records in a certain month and is marked as "absenteeism"; employee A punches in at an abnormal location (neither the company location nor the project location) for 3 days in a certain month and is marked as "field work".

[0058] Step S25: Extract the project secondment log features from the employee project work logs to obtain project secondment data, and correct the employee field work data according to the project secondment data to obtain employee abnormal field work data; In this embodiment, the records of project secondment are extracted from the project work logs, such as the time and tasks when an employee is assigned to work at other locations. The project secondment data is corrected with the employee's field work data to detect abnormalities in the field work records. For example, employee C is seconded to work at project Y in another place for 10 days. Through correction, it is found that the actual field work record does not match the secondment task time and is recorded as "abnormal field work".

[0059] Step S26: Integrate the employee absenteeism data and the employee abnormal field work data to obtain employee working time mode data; In this embodiment, the employee absenteeism data and abnormal field work data are integrated to analyze the actual working time pattern of employees. The absenteeism data, including the absenteeism date and period, is extracted from the employee attendance log. The abnormal field work data, such as the field work time, location, and project type, is extracted from the project assignment log. The missing values in the absenteeism data are filled, and the reasons such as holidays and sick leave are marked. The time of the abnormal field work data is standardized to ensure alignment with the normal working hours. The time series analysis technique (such as the time window sliding method) is used to integrate the data. The working time pattern of each employee is calculated, the daily working time distribution map of the employees is drawn, and the total working hours of each employee are calculated.

[0060] Step S27: Conduct descriptive statistical identification of the employee work behavior pattern on the employee project completion pattern data and the employee working time pattern data, so as to obtain the employee work behavior pattern data; In this embodiment, the time, type, and volume data of the projects completed by employees are extracted from the employee project completion pattern data. The employee working time pattern data is summarized, including the working hours, frequencies, and absenteeism patterns. The basic statistical indicators of employee work behavior are calculated, such as the mean, standard deviation, maximum, and minimum of the working hours. The periodicity of project completion and the peak working periods are analyzed. The clustering analysis technique (such as K-means clustering) is used to group the employee work behavior patterns, and the common work patterns (such as efficient, stable, and fluctuating types) are identified. A behavior pattern report is generated and visualized as charts (such as bar charts and pie charts).

[0061] Step S28: Conduct analysis of abnormal work behavior patterns based on the employee work behavior pattern data, so as to obtain the abnormal employee work behavior pattern data.

[0062] In this embodiment, various indicators (such as working hours and project completion time) are extracted from the employee work behavior pattern data. The data is standardized to ensure data consistency. Abnormal detection algorithms, such as Isolation Forest or Z-score method, are applied to identify work behaviors that deviate from the normal pattern. A threshold (such as Z-score greater than 3) is set to mark abnormal behaviors. A report on abnormal work behaviors is generated, including the list of abnormal employees and detailed descriptions of their abnormal behaviors (such as frequent absenteeism or abnormal working hours). The distribution of abnormal patterns is displayed through data visualization (such as heat maps and scatter plots).

[0063] Optionally, step S28 is specifically: Step S281: Extract the employee project cycle characteristics and the employee absenteeism time characteristics from the employee work behavior pattern data, so as to obtain the employee project cycle data and the employee absenteeism time data; In this embodiment, the working cycle data of each employee in a specific project is extracted from the employee's work log system, such as the project start date, end date, working hours, etc. In addition, the absenteeism data of the employees is extracted from the attendance system, including the absenteeism date, absenteeism duration, etc. Next, feature extraction is performed on this data, such as calculating the total number of working days of each project, the average daily working hours, the total absenteeism duration of the employees and its distribution in a specific project. Finally, the extracted data will construct an employee project cycle data table and an employee absenteeism time data table to prepare for subsequent analysis.

[0064] Step S282: Perform project volume clustering on the employee project cycle data to obtain project volume clustering cycle data, and perform statistics on the project abnormal cycle threshold for the project volume clustering cycle data to obtain the project abnormal cycle threshold; In this embodiment, a clustering algorithm, such as K-means or DBSCAN, is applied to the extracted employee project cycle data for project volume classification. The volume clustering can be based on indicators such as the total working hours and the number of working days of the project. After clustering, statistical analysis is performed on the cycle data of each cluster, and the working cycle distribution of each cluster, including the mean, standard deviation, etc., is calculated to determine the threshold of the project abnormal cycle. For example, if the project cycle of a certain cluster deviates significantly from its mean, the corresponding cycle will be marked as an abnormal cycle.

[0065] Step S283: Perform abnormal cycle classification calculation on the employee project cycle data according to the project abnormal cycle threshold to obtain employee abnormal project cycle data; In this embodiment, the obtained project abnormal cycle threshold is used to classify the employee project cycle data. For example, if the project cycle of an employee is more than twice the threshold, this cycle will be marked as abnormal. When applying this classification method, it is necessary to compare the project data of all employees to generate marked data of abnormal project cycles. Finally, these marked data are summarized to form an employee abnormal project cycle data set.

[0066] Step S284: Perform statistics on the absenteeism time threshold for the employee absenteeism time data to obtain the abnormal absenteeism time threshold, and perform abnormal absenteeism time classification calculation on the employee absenteeism time data according to the abnormal absenteeism time threshold to obtain employee abnormal absenteeism time data; In this embodiment, threshold statistics are performed on the employee absenteeism time data, the absenteeism duration distribution of each employee is calculated, and the abnormal threshold of the absenteeism time is calculated according to the distribution. For example, the Z-score method is used to standardize the absenteeism duration of each employee, and then a threshold is set, and the absenteeism records exceeding this threshold will be marked as abnormal. This classification method is applied to the absenteeism data of all employees to generate employee abnormal absenteeism time data.

[0067] Step S285: Construct an employee abnormal work pattern detection model based on the employee abnormal project cycle data and the employee abnormal absenteeism time data; In this embodiment, an employee abnormal work pattern detection model is constructed based on the obtained abnormal project cycle data and abnormal absenteeism time data. A machine learning model such as a decision tree or a random forest can be used. The abnormal cycle data and absenteeism data of employees are input as features, and the model is trained through historical data (employee work behavior pattern data) to identify the abnormal work patterns of employees. After the model is trained, it is used to predict the abnormal patterns in new data to generate the detection results of the employee abnormal work pattern.

[0068] Step S286: Use the employee abnormal work pattern detection model to process the employee work behavior pattern data to obtain the abnormal employee work behavior pattern data.

[0069] In this embodiment, the constructed abnormal work pattern detection model is applied to predict the new employee work behavior pattern data. The model will analyze the work data of each employee and output the abnormal detection results, such as whether the employee has an abnormal work pattern and the specific category of the abnormality. Finally, an employee work behavior pattern data set containing the abnormal detection results is generated to provide a basis for further decision-making support.

[0070] Optionally, step S3 is specifically as follows: Step S31: Extract the assigned position number features from the interviewer position assignment strategy to obtain the assigned position number data; In this embodiment, using the feature engineering technology in machine learning, the number features of each position included in the interviewer position assignment strategy are extracted. Assuming that the decision tree algorithm is used, the features of each position number (such as position requirements, skill matching degree, interview score, arrival time, etc.) are extracted and quantified into numerical features to generate a set of position number data. Taking the position number A123 as an example, the features include "skill matching degree 0.8", "experience years 5 years", etc. These feature data can be stored in the database for subsequent processing.

[0071] Step S32: Perform position number association on the assigned position number data and the abnormal employee work behavior pattern data to obtain the abnormal work behavior pattern data of the intern employees; In this embodiment, the obtained job number data is associated with the work behavior pattern data of actual employees (for example, the abnormal behavior records collected by the monitoring system). There is an abnormal behavior pattern data table, which includes "employee ID", "abnormal behavior type", and "behavior time". Through a data matching algorithm (such as a joint probability model or association rule learning), each initially assigned job number is matched with the abnormal behavior pattern of the employee. For example, if an employee with job number A123 shows the behavior of "frequent lateness", this behavior pattern will be associated with the abnormal pattern of the employee with job number A123. The generated result is that the association degree between job number A123 and the "frequent lateness" behavior is 0.7. Then, the employees with high association degree with the abnormal work behavior pattern and being initially assigned job numbers are marked as interns, and the abnormal work behavior pattern data of interns is obtained.

[0072] Step S33: Correct the abnormal work behavior pattern data of employees according to the abnormal work behavior pattern data of interns, so as to obtain the abnormal work behavior pattern data; In this embodiment, statistical analysis and machine learning algorithms are used to analyze the abnormal work behavior pattern data of interns to correct the abnormal behavior pattern. Specifically, the behavior pattern of interns is compared with the behavior pattern of normal employees to identify behaviors that do not conform to the normal pattern. For example, if it is found that interns frequently leave their posts during working hours while regular employees do not have this behavior, this abnormal pattern can be marked as "to be corrected". Apply a correction algorithm (such as an adaptive filter or a model correction algorithm) to adjust the detection threshold and behavior characteristics of the abnormal pattern to generate accurate abnormal work behavior pattern data. For example, adjust the leaving-post frequency threshold from 10% to 15%.

[0073] Step S34: Iteratively optimize and tune the parameters of the employee abnormal work mode detection model according to the abnormal work behavior pattern data, so as to obtain the employee abnormal work behavior detection model.

[0074] In this embodiment, the generated abnormal work behavior pattern data is used to optimize the employee abnormal work mode detection model. First, the abnormal pattern data is used as training data, and an optimization algorithm (such as grid search or random search) is used to tune the model parameters. For example, assuming that a random forest model is used, by adjusting parameters such as the number and depth of trees, multiple experiments are carried out to find the best parameter combination. Assume that after iterative optimization, the detection accuracy of the model is improved from 85% to 90%. The finally obtained optimized model has higher prediction accuracy and better generalization ability.

[0075] Optionally, step S4 is specifically: Step S41: Detect abnormal work behaviors in the enterprise employee work logs through the employee abnormal work behavior detection model, so as to obtain data of employees with abnormal work behaviors; In this embodiment, the abnormal work behavior detection model uses machine learning algorithms (such as random forest or neural network) to analyze the employee work logs (including login time, operation frequency, etc.) to identify employees with abnormal behaviors. For example, the model finds that the login frequency of some employees is abnormally high or low and marks them as "abnormal work behaviors".

[0076] Step S42: Extract the characteristics of the employee's start time from the data of employees with abnormal work behaviors, so as to obtain the data of the abnormal employees' start time; In this embodiment, the extracted abnormal employee data will include the start time of the employees. Feature engineering techniques are used to extract the start time (such as the number of days since employment), and these data can be processed through the pandas library in Python to calculate the data of the abnormal employees' start time, such as extracting employees whose start time is within the past 6 months.

[0077] Step S43: Cluster employees with high job fit according to the job fit data of the interviewees and the job assignment strategy for the interviewees, so as to obtain data of abnormal employees with high job fit; In this embodiment, the situation of employees with abnormal work behaviors and the job fit score (such as the matching degree between job requirements and employee skills) in the job assignment strategy for the interviewees are used to divide the employees into multiple categories through the K-means clustering algorithm. Employees with high job fit will be clustered together to generate a dataset of abnormal employees with high job fit. For example, through cluster analysis, those employees with high fit but abnormal performance are identified.

[0078] Step S44: Conduct statistical analysis of the start time based on the data of the abnormal employees' start time, so as to obtain data of abnormal employees with high start time and data of abnormal employees with low start time; In this embodiment, statistical analysis is performed on the start time data (for example, calculating the standard deviation and mean), and the employees are divided into two categories: abnormal employees with high start time and abnormal employees with low start time. For example, abnormal employees with high start time may refer to employees whose start time exceeds the normal range of the company.

[0079] Step S45: Perform an intersection operation on the employee numbers of the data of abnormal employees with high job fit and the data of abnormal employees with low start time to obtain data of abnormal employees with low fitness; perform an intersection operation on the employee numbers of the data of abnormal employees with high job fit and the data of abnormal employees with high start time to obtain data of abnormal employees with high fitness; In this embodiment, the intersection operation is performed on the data of employees with abnormal high job fit and the data of employees with abnormal low entry time to find employees with abnormal low fitness. Similarly, the intersection of the data of employees with abnormal high job fit and the data of employees with abnormal high entry time is calculated to find employees with abnormal high fitness. For example, the set operation in Python is used to find the intersection of the two data sets.

[0080] Step S46: The job fitness of employees with abnormal low fitness and employees with abnormal high fitness is evaluated and merged to obtain the job fitness data of employees with abnormal work behaviors.

[0081] In this embodiment, the obtained data of employees with abnormal low fitness and abnormal high fitness are evaluated and merged for job fitness to generate the final job fitness data of employees with abnormal work behaviors. The data can be merged using the weighted average method. For example, the job fit score and the degree of abnormal behavior are comprehensively calculated to evaluate the overall job fitness of employees.

[0082] Optionally, step S5 is specifically as follows: Step S51: The job fit of the job fit data of the interview candidates is adjusted according to the job fit data of employees with abnormal work behaviors to obtain the actual job fit data of the interview candidates. In this embodiment, by analyzing the abnormal work behavior data of employees and combining with the job fitness model, the job fit data of the interview candidates is adjusted. Machine learning algorithms (such as random forest or support vector machine) can be used to identify and adjust the fitness of the interview candidates in different positions. For example, the system compares the abnormal work behaviors (such as high turnover rate or low performance) with the job fitness data and adjusts the job fit score of the interview candidates to more accurately reflect their actual job adaptability. The abnormal work behavior data and the job fitness data are input into the machine learning model. After training, the model can adjust the job fitness data of the interview candidates. For example, if a certain interview candidate has performed poorly in the previous position (showing abnormal work behaviors), the model will reduce the fitness score of this person in the new position to obtain the actual job fit data.

[0083] Step S52: The job fit of the job assignment strategy of the interview candidates is adjusted to extract the job assignment of the employees to be transferred according to the actual job fit data of the interview candidates, so as to obtain the job assignment data of the employees to be transferred. In this embodiment, based on the actual job fit data of the interviewed personnel, the job assignment strategy is adjusted. Optimization algorithms (such as genetic algorithms or linear programming) can be used to improve the job assignment strategy. For example, the actual job fit data of the interviewed personnel is input into the optimization algorithm, and the model generates an optimal job assignment plan according to the data. The algorithm will generate job assignment data for the employees to be transferred. For example, the system will assign employees with high skill matching degrees to positions with high demands, and adjust employees with low matching degrees to positions suitable for their skills.

[0084] Step S53: Extract the job assignment descriptions for the employees to be transferred to obtain the job assignment description data to be allocated; In this embodiment, for the job assignment data of the employees to be transferred, the corresponding job descriptions are extracted. This can be achieved through natural language processing (NLP) technology. For example, using text classification and information extraction algorithms, relevant job responsibilities and requirements are automatically extracted from the job description database to form the detailed description data of the job to be allocated. Through the NLP algorithm, specific job descriptions are extracted from the job assignment data of the employees to be transferred. For example, the system automatically extracts descriptions from the job library, such as "3 years of project management experience required" and "familiar with data analysis tools" and other information to generate the detailed description data of the job to be allocated. This step ensures that the job descriptions of the employees to be transferred match their skills and fitness.

[0085] Step S54: Perform job description matching based on the actual job fit data of the interviewed personnel and the job assignment description data to be allocated, so as to obtain the job transfer strategy for employees with abnormal work behaviors and upload it to the human resource management platform to execute the job transfer task.

[0086] In this embodiment, the actual job fit data of the interviewed personnel and the job assignment description data to be allocated are used for job description matching. Matching algorithms (such as cosine similarity or deep learning models) can be used to compare the job fit degree of the interviewed personnel with the job description. The finally generated job transfer strategy for employees with abnormal work behaviors will be uploaded to the human resource management platform, and the job transfer task will be automatically executed through the system to ensure the effective implementation and recording of the job transfer strategy. Specifically, the actual job fit data of the interviewed personnel is compared with the description data of the job to be allocated using a job matching algorithm (such as cosine similarity or deep learning model), and the most suitable job transfer strategy is generated through the algorithm. For example, positions with a high matching degree between the skills of the interviewed personnel and the job requirements will be recommended. After the job transfer strategy is generated, the system automatically uploads the strategy to the human resource management platform and initiates the corresponding job transfer task.

[0087] Optionally, the present invention further provides an online management system for human resources, which is used to execute an online management method for human resources as described above. The online management system for human resources includes: An interviewer position assignment module, which is used to obtain enterprise recruitment data, evaluate the suitability of interviewers for positions based on the enterprise recruitment data, so as to obtain interviewer position suitability data; analyze the enterprise position assignment strategy based on the interviewer position suitability data, so as to obtain the interviewer position assignment strategy, and upload it to the human resources management platform to execute the position assignment task; An abnormal work behavior pattern analysis module, which is used to obtain the work logs of enterprise employees and analyze the work behavior patterns of enterprise employees based on the work logs of enterprise employees, so as to obtain employee work behavior pattern data; analyze the abnormal work behavior patterns based on the employee work behavior pattern data, so as to obtain abnormal employee work behavior pattern data; An abnormal work behavior pattern correction module, which is used to correct the abnormal work behavior patterns of interns based on the interviewer position assignment strategy for the abnormal employee work behavior pattern data, so as to obtain abnormal work behavior pattern data; construct an employee abnormal work behavior detection model based on the abnormal work behavior pattern data; An employee position fitness evaluation module, which is used to detect abnormal work behaviors in the work logs of enterprise employees through the employee abnormal work behavior detection model, so as to obtain abnormal work behavior employee data; evaluate the position fitness of abnormal work behavior employees based on the interviewer position suitability data and the abnormal work behavior employee data, so as to obtain abnormal work behavior employee position fitness data; An employee position transfer strategy analysis module, which is used to analyze the employee position transfer strategy for abnormal work behaviors based on the interviewer position suitability data and the abnormal work behavior employee position fitness data, so as to obtain the employee position transfer strategy for abnormal work behaviors, and upload it to the human resources management platform to execute the position transfer task.

[0088] The online management system for human resources of the present invention can implement any online management method for human resources of the present invention, and is used as a medium for coordinating the operations and signal transmissions between various modules to complete the online management method for human resources. The internal modules of the system cooperate with each other, thereby reducing the personnel turnover rate and recruitment costs.

[0089] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be included in the present invention.

[0090] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An online management method for human resources, characterized in that: The following steps are involved: Step S1: Obtain enterprise recruitment data, and conduct job fit assessment on interviewees based on the enterprise recruitment data, thereby obtaining job fit data on interviewees; Analyze the company's job allocation strategy based on the job fit data of the interviewees, thereby obtaining the job allocation strategy for the interviewees and uploading it to the human resources management platform to execute the job allocation task; Step S2: obtaining the work logs of the enterprise employees, and performing work behavior pattern analysis on the work logs of the enterprise employees, thereby obtaining employee work behavior pattern data; performing abnormal work behavior pattern analysis based on the employee work behavior pattern data, thereby obtaining abnormal employee work behavior pattern data; Step S3: Correct the abnormal work behavior pattern of the intern employees according to the job allocation strategy of the interviewers, thereby obtaining abnormal work behavior pattern data; and construct an employee abnormal work behavior detection model based on the abnormal work behavior pattern data; Step S4: Detect abnormal work behaviors of enterprise employees’ work logs through the employee abnormal work behavior detection model, so as to obtain abnormal work behavior employee data; evaluate the job adaptability of employees with abnormal work behaviors on the job fit data of interviewees and the data of employees with abnormal work behaviors, so as to obtain the job adaptability data of employees with abnormal work behaviors; Step S5: Analyze the job transfer strategy for employees with abnormal work behavior based on the job fit data of the interviewees and the job adaptability data of employees with abnormal work behavior, so as to obtain the job transfer strategy for employees with abnormal work behavior, and upload it to the human resources management platform to execute the job transfer task.

2. The online management method for human resources according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: Obtain enterprise recruitment data, and extract enterprise job recruitment data and interview resume data from the enterprise recruitment data, thereby obtaining enterprise job recruitment data and interviewer resume data; Step S12: performing text conversion on the enterprise job recruitment data to obtain enterprise job recruitment text data, and extracting recruitment demand features from the enterprise job recruitment text data to obtain enterprise job recruitment demand data; Step S13: extracting the interviewer's work experience years feature from the interviewer's resume data, thereby obtaining the interviewer's work experience years data, and classifying the interviewer's resume data according to the interviewer's work experience years data, thereby obtaining the interviewer's experience level classification resume data; Step S14: evaluating the job suitability of interviewees according to the enterprise job recruitment demand data and the resume data of interviewees’ experience levels, thereby obtaining job suitability data of interviewees; Step S15: Analyze the enterprise job allocation strategy of the enterprise job recruitment data according to the job fit data of the interviewees, so as to obtain the interviewee job allocation strategy, and upload it to the human resources management platform to execute the job allocation task.

3. The online management method for human resources according to claim 2, characterized in that: Step S14 is specifically as follows: Step S141: extracting skill requirement features and experience years requirement features according to the enterprise job recruitment demand data, thereby obtaining job recruitment skill requirement data and job recruitment experience years requirement data; Step S142: screening the resume data of the interviewer experience level classification according to the job recruitment experience years requirement data, thereby obtaining the interviewer experience screening resume data; Step S143: converting the interviewer experience screening resume data into resume description text to obtain the interviewer resume description text data, and extracting experience skill description features from the interviewer resume description text data to obtain resume skill description data and resume experience description data; Step S144: Calculate the skill description similarity of the job recruitment skill requirement data and the resume skill description data to obtain the skill description similarity data, and classify the interviewer resume description text data into skilled personnel according to the skill description similarity data to obtain skilled personnel data with high similarity and skilled personnel data with low similarity; Step S145: Calculate the experience similarity of the job recruitment experience years and the resume experience description data, thereby obtaining the experience similarity data, and classify the interviewer resume description text data into experienced personnel according to the experience similarity data, thereby obtaining high-similarity experienced personnel data and low-similarity experienced personnel data; Step S146: Performing personnel number intersection operation on high-similarity skill personnel data and high-similarity experience personnel data, thereby obtaining high-position fit personnel data; performing personnel number intersection operation on low-similarity skill personnel data and low-similarity experience personnel data, thereby obtaining low-position fit personnel data; Step S147: Merge the data of personnel with high job fit and the data of personnel with low job fit, so as to obtain the job fit data of the interviewee.

4. The online management method for human resources according to claim 2, characterized in that: Step S15 is specifically as follows: Step S151: classifying the interviewees according to their job fit data, thereby obtaining data of interviewees with high job fit and data of interviewees with low job fit; Step S152: performing personnel job assignment for the enterprise job recruitment data and the high job fit personnel data, thereby obtaining high fit personnel job assignment data, and extracting vacant positions from the enterprise job recruitment data based on the high fit personnel job assignment data, thereby obtaining vacant position recruitment data; Step S153: extracting description experience features from the data of personnel with low job fit, thereby obtaining description experience data of interview personnel, and matching the description experience data of interview personnel and vacant job recruitment data with job recruitment descriptions, thereby obtaining job matching data for personnel with low job fit; Step S154: sorting the resume data of the low-fit candidates according to their experience levels and assigning them positions, thereby obtaining the position assignment data of the low-fit candidates; Step S155: Integrate the job allocation strategy based on the job allocation data of high-fit candidates and the job allocation data of low-fit candidates, so as to obtain the job allocation strategy for the interviewers, and upload it to the human resources management platform to execute the job allocation task.

5. The online management method for human resources according to claim 1, characterized in that: Step S2 is specifically as follows: Step S21: obtaining the work log of the enterprise employees, and extracting the employee attendance log and the employee project work log from the enterprise employee work log, thereby obtaining the employee attendance log and the employee project work log; Step S22: extracting project volume features and project work time series features from employee project work logs, thereby obtaining project volume data and project work time series data; Step S23: Calculate the project completion time according to the project work time series data, thereby obtaining the project completion time data, and integrate the employee project completion mode according to the project completion time data and the project volume data, thereby obtaining the employee project completion mode data; Step S24: extracting employee attendance missing features according to the employee attendance log, thereby obtaining employee absence data and employee field data; Step S25: extracting project dispatch log features according to the employee project work log, thereby obtaining project dispatch data, and performing work dispatch correction on the employee fieldwork data according to the project dispatch data, thereby obtaining abnormal fieldwork data of the employee; Step S26: Integrate the employee work time pattern of the employee absence data and the employee abnormal field data, thereby obtaining the employee work time pattern data; Step S27: performing descriptive statistical identification of employee work behavior patterns on the employee project completion pattern data and the employee working time pattern data, thereby obtaining employee work behavior pattern data; Step S28: performing abnormal work behavior pattern analysis based on the employee work behavior pattern data, thereby obtaining abnormal employee work behavior pattern data.

6. The online management method for human resources according to claim 5, characterized in that: Step S28 is specifically as follows: Step S281: extracting employee project cycle features and employee absence time features from employee work behavior pattern data, thereby obtaining employee project cycle data and employee absence time data; Step S282: performing project volume clustering on the employee project cycle data to obtain project volume clustering cycle data, and performing project abnormal cycle threshold statistics on the project volume clustering cycle data to obtain project abnormal cycle thresholds; Step S283: performing abnormal cycle classification calculation on employee project cycle data according to the project abnormal cycle threshold, thereby obtaining employee abnormal project cycle data; Step S284: performing absence time threshold statistics on the employee absence time data to obtain abnormal absence time thresholds, and performing abnormal absence time classification calculations on the employee absence time data according to the abnormal absence time thresholds to obtain abnormal absence time data of the employees; Step S285: constructing an employee abnormal work mode detection model based on the employee abnormal project cycle data and the employee abnormal absence time data; Step S286: Detect abnormal employee work behavior pattern data according to the employee abnormal work pattern detection model.

7. The online management method for human resources according to claim 1, characterized in that: Step S3 is specifically as follows: Step S31: extracting the assigned position number feature of the interviewer position allocation strategy, thereby obtaining the assigned position number data; Step S32: Associating the assigned job number data and the abnormal employee work behavior pattern data with the job number, thereby obtaining the abnormal work behavior pattern data of the intern employee; Step S33: Correcting the abnormal work behavior pattern of the intern according to the abnormal work behavior pattern data of the intern, thereby obtaining the abnormal work behavior pattern data; Step S34: Iteratively optimize and adjust the parameters of the employee abnormal work mode detection model according to the abnormal work behavior pattern data, so as to obtain the employee abnormal work behavior detection model.

8. The online management method for human resources according to claim 1, characterized in that: Step S4 is specifically as follows: Step S41: Detect abnormal work behavior of enterprise employees in their work logs through an employee abnormal work behavior detection model, thereby obtaining data on employees with abnormal work behavior; Step S42: extracting employee entry time features from employee data with abnormal work behaviors, thereby obtaining abnormal employee entry time data; Step S43: clustering employees with high job fit according to the interviewer job fit data and the interviewer job allocation strategy, thereby obtaining data of abnormal employees with high job fit; Step S44: Performing onboarding time statistical analysis based on the onboarding time data of abnormal employees, thereby obtaining data of employees with high-salary abnormal onboarding time and data of employees with low-salary abnormal onboarding time; Step S45: performing an employee number intersection operation on the abnormal employee data with high job fit and the abnormal employee data with low salary entry time, thereby obtaining the abnormal employee data with low fitness; performing an employee number intersection operation on the abnormal employee data with high job fit and the abnormal employee data with high salary entry time, thereby obtaining the abnormal employee data with high fitness; Step S46: merging the employee job fitness evaluation data of the low fitness abnormal employees and the high fitness abnormal employees to obtain the job fitness data of the employees with abnormal work behaviors.

9. The online management method for human resources according to claim 1, characterized in that: Step S5 is specifically as follows: Step S51: adjusting the job suitability data of the interviewee according to the job suitability data of the employee with abnormal work behavior, so as to obtain the actual job suitability data of the interviewee; Step S52: performing job fit adjustment on the job allocation strategy of the interviewee according to the actual job fit data of the interviewee, thereby extracting the job allocation data of the employee to be transferred; Step S53: extracting the assigned job description from the job assignment data of the employee to be transferred, thereby obtaining the description data of the job to be assigned; Step S54: Match the job description according to the actual job fit data of the interviewee and the job description data to be assigned, so as to obtain the job transfer strategy for employees with abnormal work behavior, and upload it to the human resources management platform to execute the job transfer task.

10. An online management system for human resources, characterized in that: For executing the online management method for human resources as claimed in claim 1, the online management system for human resources comprises: The interviewer job allocation module is used to obtain enterprise recruitment data, and conduct interviewer job fit assessment based on the enterprise recruitment data, so as to obtain interviewer job fit data; conduct enterprise job allocation strategy analysis based on interviewer job fit data, so as to obtain interviewer job allocation strategy, and upload it to the human resources management platform to execute job allocation tasks; The abnormal work behavior pattern analysis module is used to obtain the work logs of enterprise employees and perform employee work behavior pattern analysis on the enterprise employee work logs, thereby obtaining employee work behavior pattern data; perform abnormal work behavior pattern analysis based on the employee work behavior pattern data, thereby obtaining abnormal employee work behavior pattern data; The abnormal work behavior pattern correction module is used to correct the abnormal work behavior pattern of interns based on the job allocation strategy of the interviewers, thereby obtaining abnormal work behavior pattern data; and construct an employee abnormal work behavior detection model based on the abnormal work behavior pattern data; The employee job adaptability assessment module is used to detect abnormal work behaviors in the work logs of enterprise employees through the employee abnormal work behavior detection model, so as to obtain the data of employees with abnormal work behaviors; the job adaptability assessment of employees with abnormal work behaviors is performed on the job fit data of interviewees and the data of employees with abnormal work behaviors, so as to obtain the job adaptability data of employees with abnormal work behaviors; The employee transfer strategy analysis module is used to analyze the transfer strategies of employees with abnormal work behaviors based on the job fit data of interviewees and the job adaptability data of employees with abnormal work behaviors, so as to obtain the transfer strategies of employees with abnormal work behaviors and upload them to the human resources management platform to execute job transfer tasks.

Images