A human resource intelligent analysis system based on big data

Abstract

The present application relates to the technical field of human resources, in particular to a human resource intelligent analysis system based on big data, which comprises a behavior rhythm capturing module, a post load identification module, a task connection analysis module, a post transfer tracking module and a behavior rhythm adaptation evaluation module.In the present application, the time corresponding relationship between behavior and task is constructed by extracting the behavior frequency of employees in the differentiated period and combining the start and end time of the task, the continuous bearing relationship between the repeated task type and the post number along the time axis is identified, the post replacement characteristics and the frequently occurring boundary position in the task handover process are located, the continuous replacement path of the task between the differentiated posts is tracked, and the adaptation level of the post behavior rhythm is measured in combination with the time overlap of the post responsibility period and the task behavior section, which helps to strengthen the bearing identification ability of the post in the task dynamic distribution process and improve the timing coordination and path integrity of the task coordination connection between multiple posts.

Description

Technical Field

[0001] This invention relates to the field of human resources technology, and in particular to a human resources intelligent analysis system based on big data. Background Technology

[0002] The field of human resource technology involves core aspects of an organization's recruitment, allocation, training, assessment, incentives, and management of personnel. It typically utilizes methods such as job analysis, performance management, human resource inventory, and talent assessment to achieve dynamic control and optimized allocation of personnel structure and capabilities. Process-based management methods include personnel needs identification, information collection, evaluation standard formulation, evaluation result application, and feedback correction. With the deepening development of information management, this field is gradually adopting data-driven management methods to improve the scientific nature and operability of personnel decisions. Traditional human resource intelligent analysis systems target specific matters such as personnel competency assessment, recruitment effectiveness judgment, and performance development trend identification. They obtain basic information by setting evaluation dimensions based on job competency models, combining manual entry and organization of historical performance data with methods such as scoring tables, structured questionnaires, and evaluation scoring. Statistical analysis methods, such as frequency analysis and regression analysis, are then used to summarize and compare the data, ultimately assisting in personnel selection and performance evaluation management activities.

[0003] Existing technologies establish structured evaluation models based on job evaluation dimensions, and form personnel capability assessment results through static indicator collection and manual scoring. This approach fails to consider the dynamic connection relationship of employee task behavior on the time axis, resulting in the inability to capture the consistency and transmission path of job behavior rhythm in high-frequency task scheduling scenarios such as changes in task distribution, job critical transitions, and frequent alternation of responsibility cycles. This limits the ability to judge the task collaboration and connection mode between jobs, and consequently, problems such as ambiguous job load division and decreased task division efficiency are prone to occur in task configuration, affecting the execution efficiency and smooth connection of multi-job collaborative tasks. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a human resources intelligent analysis system based on big data. The technical solution is as follows:

[0005] On the one hand, a big data-based intelligent human resources analysis system is provided, which includes:

[0006] The behavior rhythm capture module obtains daily employee behavior records in different time periods, extracts hourly behavior activities, the corresponding task start and end time positions, filters overlapping time periods of tasks, and obtains the corresponding segments of behavior tasks.

[0007] The job load identification module extracts repetitive task types, corresponding task and job number based on the segment corresponding to the behavioral task, tracks the time distribution of job task content, compares the task execution between jobs, and obtains functional load job groups.

[0008] Based on the functional load job group, the task connection analysis module extracts the execution order of upstream and downstream tasks, finds the task connection position, identifies the continuous task relationship between differentiated jobs, and obtains the task boundary link.

[0009] Based on the task boundary link, the job transfer tracking module extracts job flow records during the boundary time period, analyzes job information that is uninterrupted and whose responsibilities have not changed, promotes the task relationship between jobs, and obtains job connection path groups.

[0010] The behavioral rhythm adaptation assessment module extracts the employee's task start time and job responsibility cycle based on the job connection path group, cross-compares the time periods, analyzes the time overlap between the task behavior interval and the responsibility cycle, and obtains the job behavior rhythm adaptation index.

[0011] As a further aspect of the present invention, the behavioral task corresponding segment includes the frequency of behavior occurrence, the start and end positions of task execution, and the continuous interval of behavior; the functional load job group includes repetitive task types, the correspondence between tasks and jobs, and the job task undertaking sequence; the task boundary link includes the acceptance time point, the frequency of job task alternation, and the position of continuous task connection; the job connection path group includes job flow records, task uninterrupted identifiers, and job task order; and the job behavior rhythm adaptation index includes the high-frequency behavior interval of tasks, the job responsibility bearing cycle, and the number of times the time intersection position is repeated.

[0012] As a further aspect of the present invention, the differentiated positions refer to positions that are distinguished by differences in task type, time distribution, and responsibility characteristics during task execution.

[0013] The "uninterrupted and unchanged responsibilities" job information refers to job records where, during task handover and job rotation, personnel and time change, but the job responsibilities remain consistent and the task is not interrupted.

[0014] As a further aspect of the present invention, the job responsibility cycle refers to the time period during which the job is responsible, and the start and end time range of the tasks undertaken by the job.

[0015] The overlapping time intervals between the task behavior time intervals and the job responsibility cycle reflect the overlap of tasks and responsibilities in time.

[0016] As a further aspect of the present invention, the behavior rhythm capture module includes:

[0017] The time-segment behavior collection submodule acquires daily employee behavior records in different time periods, divides time segments by hour, extracts the time of occurrence of behavior within a time segment, locates the position of behavior within a time segment, and obtains the time sequence of behavior activities.

[0018] The task time matching submodule extracts the task start time and end time based on the behavior activity time sequence, the corresponding behavior time point and the task time segment time position, determines whether the time point is within the task, and obtains a task behavior correspondence table according to the behavior and task correspondence.

[0019] The behavior task interval extraction submodule calls the task behavior correspondence table, extracts the part of continuous behavior time period that overlaps with the task time, locates the start and end time of the behavior and the task boundary position, and delineates the segment edge according to the time sequence to obtain the segment corresponding to the behavior task.

[0020] As a further aspect of the present invention, the job load identification module includes:

[0021] The task type extraction submodule extracts task name data within a time period based on the segment corresponding to the behavior task, sequentially combines and locates the task name with the time point, associates the repeated task name with the task number, arranges the task type in the order of the task number, and obtains a duplicate task type reference table.

[0022] The job task tracking submodule calls the repetitive task type lookup table, extracts the job number associated with each task number, collects the task content corresponding to the job in the time period along the time axis, and continuously extracts the positional relationship between time and task number to obtain the job task time sequence table.

[0023] The task content comparison submodule extracts the task numbers associated with differentiated positions within a continuous time range based on the job task time series table, compares the consecutive repetition order of the task numbers, and extracts the job numbers that undertake the same type of task in the time period to obtain the functional load job group.

[0024] As a further aspect of the present invention, the task connection analysis module includes:

[0025] The task time extraction submodule extracts the start and end times of each task from the job task schedule based on the functional load job group, arranges the task time periods in chronological order, determines the time connection interval between consecutive tasks, and obtains the task start and end time sequence.

[0026] Based on the task start and end time sequence, the task connection and positioning submodule filters task combinations with adjacent time positions according to the positional relationship between tasks with different end and start times. It compares the cases where there is no time overlap and no interval between tasks, and obtains a task time connection comparison table according to the task time boundary matching order.

[0027] The task node evaluation submodule extracts the corresponding job number in the task pair based on the task time connection lookup table, identifies the position of continuous association of task pairs under the same job number, filters out the repeated connection relationship in the continuous time period, extracts the corresponding task number and time segment, and obtains the task boundary link.

[0028] As a further aspect of the present invention, the job transfer tracking module includes:

[0029] The job flow extraction submodule collects job flow records within the boundary time period based on the task boundary link, extracts the job number and task start and end time corresponding to each flow record, and outputs the job number sequence corresponding to the succession of consecutive tasks according to the time order to obtain the job time connection sequence.

[0030] The responsibility continuity judgment submodule calls the job time connection sequence, analyzes the job description field in the time period before and after the task, compares the job task description text, removes job numbers whose responsibility content has changed, and retains the set of job numbers with consistent responsibility description fields to obtain the job number group with consistent responsibilities.

[0031] The task succession path extraction submodule extracts the number pairs of positions that appear in adjacent positions on the time axis based on the job number group with consistent responsibilities, analyzes the task succession relationship, forms a linear sequence of task connection paths according to the number pairs, and outputs the correspondence between positions and tasks in the path to obtain the job connection path group.

[0032] As a further aspect of the present invention, the process of collecting job flow information within the boundary time period in the job flow extraction submodule is as follows: the boundary time range is divided into continuous segments according to a preset time interval length, the associated job number and task start and end time are extracted in each time segment, and the job number sequence corresponding to the task is arranged in chronological order.

[0033] The process of comparing the job task description text in the continuous responsibility judgment submodule is as follows: the description content of each job number within the time range before and after the task is decomposed into responsibility fields, the job responsibility fields of adjacent time periods are compared for text consistency, job numbers with obvious changes in description are filtered out according to the character overlap ratio, and job numbers with consistent description content are retained.

[0034] The process of extracting job number pairs with adjacent time positions in the task succession path extraction submodule is as follows: among the job numbers with consistent job descriptions, the numbers are arranged in chronological order, consecutive number pairs are extracted, and the corresponding task information for each number is combined to list the correspondence between job numbers and task numbers in chronological order, forming a complete task succession path sequence.

[0035] As a further aspect of the present invention, the behavior rhythm adaptation assessment module includes:

[0036] The time information extraction submodule extracts the start time of job tasks and the corresponding responsibility time period based on the job connection path group. It then categorizes the time information of differentiated jobs into data sequences with self-numbering, and filters out time intervals where there is an overlap between the task time sequence and the responsibility time period to obtain the job task time segment.

[0037] The behavior overlap recognition submodule calls the job task time segment, extracts the time position of behavior data from the task execution time, compares the time of each behavior with the job responsibility time segment, determines whether it appears repeatedly in the same time interval, and imports it into the result table according to the job number to obtain the behavior interval overlap sequence group;

[0038] The rhythm adaptation assessment submodule extracts the number of behaviors for each position at the time intersection position based on the overlapping sequence group of the behavior intervals. Combining the time segment order corresponding to the job responsibilities, it compares the occurrence time of the behaviors with the number of behaviors and classifies them into the task carrying relationship table for each position to obtain the job behavior rhythm adaptation index.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In this invention, by extracting the frequency of employee behaviors within differentiated time periods and combining them with the start and end times of tasks to construct a temporal correspondence between behaviors and tasks, the continuous undertaking relationship between repetitive task types and job numbers is identified along the time axis. The job replacement characteristics and frequent boundary positions in the task handover process are located, and the continuous succession path of tasks between differentiated jobs is tracked. By combining the time overlap between job responsibility cycles and task behavior segments, the adaptability level of job behavior rhythm is measured. This helps to strengthen the job's ability to bear and identify tasks in the dynamic distribution process and improve the temporal coordination and path integrity of task collaboration between multiple jobs. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a system flowchart of the present invention;

[0043] Figure 2 This is a system block diagram of the present invention;

[0044] Figure 3 This is a flowchart of the behavior rhythm capture module in this invention;

[0045] Figure 4 This is a flowchart of the job load identification module in this invention;

[0046] Figure 5 This is a flowchart of the task connection analysis module in this invention;

[0047] Figure 6 This is a flowchart of the job transfer tracking module in this invention;

[0048] Figure 7 This is a flowchart of the behavior rhythm adaptation evaluation module in this invention. Detailed Implementation

[0049] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0050] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0051] This invention provides a big data-based intelligent human resource analysis system, such as... Figures 1-2 The diagram shown illustrates a big data-based intelligent human resources analysis system, which includes:

[0052] The behavior rhythm capture module acquires daily employee behavior records in different time periods, extracts the behavior activities within each hour, compares the time periods before and after the task is triggered item by item, the frequency of the corresponding behavior and the start and end time positions of the task execution, finds the time range that overlaps with the task in the time period when the behavior occurs continuously, and separates the time range positions to obtain the segment corresponding to the behavior task.

[0053] The job load identification module extracts the recurring task types within the corresponding segments of behavioral tasks, identifies each task and associated job, tracks the task content undertaken by the job along the time sequence, identifies whether the task content is executed by the same job in different time periods, compares the continuity of the task content according to the job order, and obtains the functional load job group.

[0054] The task connection analysis module is based on the functional workload job group. It extracts the start and end time sequence of upstream and downstream tasks from the job task schedule, locates the connection time point between tasks, compares the alternation phenomenon in the job task arrangement, identifies the frequency between the connection time points of differentiated job tasks, locates the task positions where continuous connection occurs, and obtains the task boundary link.

[0055] The job transfer tracking module is based on the task boundary link, extracts job flow records within the boundary time period, identifies job numbers whose tasks have not been interrupted and whose responsibilities have not changed, advances the task sequence corresponding to the job number according to the time sequence, tracks the continuous task succession between adjacent jobs, and obtains job connection path groups.

[0056] The behavioral rhythm adaptation assessment module is based on the job connection path group. It extracts the start time of employee tasks and the cycle of job responsibilities, cross-compares time information segments, locates the degree of overlap between the high-frequency behavior interval of tasks and the job's time, and judges the behavioral tendency based on the number of repetitions at the time intersection to obtain the job behavioral rhythm adaptation index.

[0057] The behavioral task corresponding segment includes the frequency of behavior occurrence, the start and end positions of task execution, and the continuous interval of behavior. The functional load job group includes repetitive task types, the correspondence between tasks and jobs, and the job task undertaking sequence. The task boundary link includes the acceptance time point, the frequency of job task alternation, and the position of continuous task connection. The job connection path group includes job flow records, task uninterrupted indicators, and job task sequence. The job behavior rhythm adaptation index includes the high-frequency behavior interval of tasks, the job responsibility carrying cycle, and the number of times the time intersection position is repeated.

[0058] Specifically, such as Figure 2 , Figure 3 As shown, the behavior rhythm capture module includes:

[0059] The time-segment behavior collection submodule acquires daily employee behavior records in different time periods, divides time segments by hour, extracts the time of occurrence of behavior within a time segment, locates the position of behavior within a time segment, and obtains the time sequence of behavior activities.

[0060] First, logs are extracted from the personnel behavior monitoring platform. These logs contain employee identifiers, event triggers, action codes, and their corresponding locations and instructions. Each record is arranged chronologically according to the actions performed. For example, actions performed by employee A within a certain time period, such as "entering a page," "clicking a button," and "submitting a form," are listed sequentially. Each action is accompanied by a sequence number and the corresponding job identifier. Then, the complete records are divided into hourly groups, and all actions belonging to that hourly segment are extracted to form hourly behavior fragment sets. Next, for each extracted action within an hourly segment, the corresponding action time is extracted. By comparing the action execution identifier with the task start marker in the job task list, the corresponding position of the actual action sequence is extracted and matched with the job activity sequence in the task assignment record, thus identifying the action within the hourly timeframe. The actions are selected from the beginning, middle, or end of a time segment and paired with their position values ​​(e.g., the action is ranked 3rd or 5th in a certain hour segment) and their type of action. Then, the position sequences corresponding to all actions in that hour segment are sequentially arranged to form the employee's action activity sequence record for that time segment. This process is repeated for all hour segments of the day. By traversing the action activity position sequences of all time periods, the distribution of actions in different time periods can be obtained according to the job position or employee identifier. For example, if employee B performed "data entry, data verification, interface submission" and "log export, task identification, module jump" in two consecutive hours, the action sequence shows that the first time period focused more on data operations and the second time period focused more on interface operations. This indicates that the task behavior cycle changed between the two hour segments, ultimately yielding the action activity time sequence.

[0061] The task time matching submodule extracts the task start time and end time based on the behavior activity time sequence, the corresponding behavior time point and the task time segment time position, determines whether the time point is within the task, and obtains a task behavior correspondence table according to the behavior and task correspondence.

[0062] First, when extracting the task start and end times, it's necessary to access each employee's action time list and job task allocation list. The action time list records the action number, employee ID, and trigger time for each action. The job task allocation list lists the start and end time segments for tasks in different positions, with each task number corresponding to a job number. First, the start and end times corresponding to each task number must be extracted to form a task time segment. Then, the time points corresponding to all of the employee's actions are retrieved from the action time sequence. Each action time point is compared to see if it falls within any task time segment. The judgment operation uses the action time point as a benchmark, comparing it sequentially with the start and end values ​​of each task time segment. If the time point is earlier than the task start time or later than the task end time, the action is excluded; if the time point is between the start and end times, the action is retained. The behaviors are marked as actions within the task segment. After performing the above filtering process on all behaviors, all behavior events falling within the task time period can be retained. Then, based on the time point of each behavior event, it is cross-matched with its employee number and job number to establish the correspondence between behavior and task segment, forming a one-to-one correspondence set between tasks and behaviors. For example, if employee Z's start time for task T1 is the 20th time unit and the end time is the 40th time unit, and the time points of his behavior B1 are the 25th unit, B2 is the 38th unit, and B3 is the 43rd unit, then it is determined that B1 and B2 are within the task segment, and B3 is outside the segment. The behavior information of B1 and B2 is retained and matched with task T1, and the correspondence between their behavior number and task number is registered in the reference set. In this way, the matching relationship between employee behavior and task in terms of time can be filtered one by one, and finally a task behavior correspondence reference table is obtained.

[0063] The behavior task interval extraction submodule calls the task behavior correspondence table, extracts the part of continuous behavior time period that overlaps with the task time, locates the start and end time of behavior and the task boundary position, and delineates the segment edge according to the time sequence to obtain the segment corresponding to the behavior task.

[0064] First, obtain the list of behavior time points and task time period boundaries for each task. Record the task number, behavior number, and time point position in the table. Group all corresponding behavior time points according to the task number, arranging them chronologically. Gradually find the interval between adjacent behavior events. If the time interval between two consecutive behavior events does not exceed the set behavior interval judgment threshold, then the two time points are determined to belong to a continuous behavior segment. Proceed forward in this manner until an interruption occurs, then start a new behavior set. Finally, divide the sequence into multiple continuous behavior segments. Then, extract the start and end points of each behavior segment to form a behavior time period. Next, determine the positional relationship between each behavior time period and its corresponding task time period. Compare the order between the start point of the behavior segment and the start boundary of the task time period. If the start point of the behavior segment is earlier than the task start boundary, correct it to the task start point; otherwise, retain the behavior. Similarly, the relationship between the starting point and the ending point of the behavior and the task ending boundary is compared. If the ending point of the behavior is later than the ending point of the task, it is corrected to the ending point of the task; otherwise, the ending point of the behavior is retained. This yields a time segment of behavior that overlaps within the task interval. For example, if the time range of task T1 is from the 50th to the 80th time unit, and the corresponding behavior action of this task occurs consecutively in the 48th, 52nd, 56th, 59th, and 63rd time units, then a time segment of behavior can be identified as the 52nd to 63rd time unit. The starting point of the 52nd time unit is within the task boundary and does not require correction. The ending point is similarly determined. This segment can be used as the overlapping interval of the behavior and task of task T1. Then, the time segments of behavior corresponding to each task are classified and stored according to the task number, and the start and end boundaries of each segment are arranged in chronological order. Through this process, the mapping relationship between the start and end times of the behavior and the task boundary can be obtained, and finally, the segment corresponding to the behavior task is obtained.

[0065] Specifically, such as Figure 2 , Figure 4 As shown, the job load identification module includes:

[0066] The task type extraction submodule extracts task name data within a time period based on the corresponding segment of the behavior task, combines and locates the task name and time point in sequence, associates the repeated task name with the task number, arranges the task type in the order of task number, and obtains a duplicate task type reference table.

[0067] First, extract the time segment content associated with each task and load the corresponding list of task names within that segment. Task names are typically recorded as strings in the behavior-task mapping table. During execution, extract the associated task names from each segment one by one and combine them with the start and end times corresponding to that segment to form task location pairs. For example, if task name A occurs once between time periods T1 and T3, then that action is combined into a data item consisting of task name A and time points T1 and T3. Simultaneously, record the task number corresponding to that task and sort the task number and the combined task name in a list format. For instance, if the behavior-task mapping table records task name A appearing in numbers T001 and T004, corresponding to time segments 1 and 4 respectively, then A will be assigned to the corresponding number... The task names are listed as two sub-items and added to the task name and number mapping set. Then, the task name duplication identification operation is performed. The combined task name and task number sets are compared. Entries with the same name but different task numbers are retained and included in the task duplication set. Then, they are arranged in ascending order of task number to ensure that the duplication tasks are in the same order in time. For example, if task A appears repeatedly in numbers T001, T004, and T008, then its arrangement order is T001→T004→T008. At the same time, the distribution pattern of tasks can be judged based on the time sequence of task numbers. Then, the sorted name and number contents are combined and stored to ensure that the correspondence between task names can be tracked and mapped. Finally, a duplication task type lookup table is obtained.

[0068] The job task tracking submodule calls the repetitive task type lookup table, extracts the job number associated with each task number, collects the task content corresponding to the job in the time period along the time axis, and continuously extracts the positional relationship between time and task number to obtain the job task time sequence table.

[0069] First, extract the sorted task numbers and then query the job number associated with each task number. Job numbers typically appear in the task record table as a unique combination of characters. During extraction, iterate through the task numbers, retrieving the job number information for each task and adding it to the job-task mapping table. Simultaneously, search the task execution database for the tasks undertaken by that job number at different times. The fields for retrieving task content are usually task descriptions or task codes. Compare these with the original task numbers to confirm whether the content belongs to the same task. Then, add the confirmed content to the task sequence of the current job number, forming a task content set based on job number. For example, job number P001 in task sequence T0... Task 01 is responsible for inspection tasks, and the task time period is interval 1. The recorded content is P001-T001-inspection-interval 1. At the same time, extract the next task number T002. When its associated job number is still P001, repeat the above process, record the task content and time of T002 and attach it to the task set of P001. After ensuring that the content is continuous and unbroken, fill it into the time axis data frame in the order of number. Continue to perform the task number extraction operation to form a corresponding structure between time and task number. When a job number no longer appears in the subsequent task numbers, stop the content filling operation for that job. This extraction method ensures that each job is fully covered in the entire task number chain, and finally obtains the job task time sequence table.

[0070] The task content comparison submodule extracts the task numbers associated with differentiated positions within a continuous time range based on the job task time sequence table, compares the consecutive repetition state order of the task numbers, and extracts the job numbers that undertake the same type of tasks in the time period to obtain the functional load job group.

[0071] First, locate the task time period corresponding to each job number and extract the task number sequence within each time period. Based on this, mark the arrangement of task numbers for each job. By identifying whether the adjacent order of task numbers is continuous and whether they are repeated, determine whether the task is in a stable and continuous state. At the same time, extract the task number sequence of other jobs within the same time interval and compare whether there are duplicate or parallel task number sequences between different jobs. For example, if job P003 performs tasks T008, T009, and T010 in interval A, which is the same as job P007 performing tasks T008, T009, and T010 in the same interval, then it is determined that the two have the same type of task in that time period. Further filter out job combinations with high task number repetition rates, perform correlation statistics on job numbers in the combination, and exclude job combinations that have different task content labels but the same substantive operation. For example, if P004 has different task labels but the same operation description field, and the task content is the same as P003, then the operation tasks are considered equivalent and can be classified into the same functional group. After comparing multiple groups of jobs, the job number groups with similar task continuity characteristics are classified to obtain functional load job groups.

[0072] Specifically, such as Figure 2 , Figure 5 As shown, the task connection analysis module includes:

[0073] The task time extraction submodule is based on the functional load job group. It extracts the start time and end time of each task from the job task schedule, arranges the task time periods in chronological order, and determines the time connection interval between consecutive tasks to obtain the task start and end time sequence.

[0074] First, when extracting the start and end times of each task from the job task schedule, we first extract the corresponding task identifier and associated time point for each job number. After separating the start and end times of each task into time pairs, we arrange the task time pairs in chronological order. Through this order, we establish the timeline for each job. In this process, we determine whether there is an interval between adjacent task time pairs. By comparing the time difference between the end time of the previous task and the start time of the next task, we distinguish whether they belong to a continuous connection state. For example, in job number P002, the end time of task T015 is 12:00, and the start time of task T016 is 12:01. At this time, the time interval is less than the set tolerance, so the two tasks are determined to be continuous. If the start time of task T016 is 12:10, it is determined to be an interval state. All continuous connection and interval states are marked with their positions. Then, by aggregating all time series information under the job number, we identify the connection characteristics of different jobs on the timeline, and form a one-to-one correspondence between repeated connection phenomena and interval phenomena to obtain the task start and end time series.

[0075] The task connection and positioning submodule is based on the task start and end time sequence. According to the positional relationship between tasks with different end and start times, it filters task combinations with adjacent time positions, compares the cases where there is no time overlap and no interval between tasks, and obtains the task time connection comparison table according to the task time boundary matching order.

[0076] First, the start and end times of each task are extracted and used as boundary points on the timeline. All tasks are then numbered according to their chronological order. By judging the time boundary relationship between each pair of tasks, the time difference between the start time of the subsequent task and the end time of the preceding task is extracted as the basis for judgment. Task combinations whose start time immediately follows the end time of the preceding task are selected. Then, the time boundaries of the two tasks in the selected task pairs are reconfirmed to ensure that there is no overlap in time, while eliminating combinations with obvious intervals. For example, if the end time of task T005 is 13:00, the two tasks are only considered connected if the start time of the subsequent task T006 is also 13:00. If the start time of T006 is earlier than 13:00 or later than 13:02, it is excluded. This method completes the screening of tasks with temporal sequential relationships. After completing the comparison of all task pairs, the tasks are numbered according to their boundary positions on the timeline, and their correspondences are mapped to complete the connection identification of each task pair on the timeline, resulting in a task time connection lookup table.

[0077] The task node evaluation submodule extracts the corresponding job number in the task pair based on the task time connection lookup table, identifies the position of continuous association of task pairs under the same job number, filters out the repeated connection relationship in the continuous time period, extracts the corresponding task number and time segment, and obtains the task boundary link.

[0078] First, the job number associated with each record in the task pair is extracted, and the job number is regarded as the unique identifier of the task execution responsibility. By comparing the position of the same job number in different task pairs, the continuity and repetition of the timeline are sorted out. Then, the same job number is repeatedly assigned to tasks within continuous time areas. For example, job P003 appears continuously in tasks T007 and T008, and then appears again in tasks T010 and T011. It can be determined that it maintains the task connection attribute in different time periods. For each such repeated association point, the job number and task number are cross-matched, and the tasks are divided according to the start and end positions of the time segment. The time periods after each division are sorted, and the numbers of repeated task connections within these time periods are extracted. The connection status between adjacent tasks is judged. When the task number is continuously assigned by the same job in different time periods and its time boundary is continuous, it is identified as a task boundary point. Finally, all the identified task numbers and time segments are organized into a set to obtain the task boundary link.

[0079] Specifically, such as Figure 2 , Figure 6 As shown, the job transfer tracking module includes:

[0080] The job flow extraction submodule is based on the task boundary link. It collects job flow records within the boundary time period, extracts the job number and task start and end time corresponding to each flow record, and outputs the job number sequence corresponding to the succession of consecutive tasks according to the time order to obtain the job time connection sequence.

[0081] First, the identified task boundary time range is used as input. The corresponding time period's flow data is retrieved from the job history records. All personnel movement details within that time period are extracted. Each record must clearly indicate the personnel number, original job number, target job number, and the start and end time range of the corresponding task. In practice, if an employee moves from P02 to P03 between the end of T005 and the beginning of T006, and the task number changes, this can be considered valid flow data. Then, the flow records between adjacent tasks are compared chronologically. Multiple consecutive job flow records are arranged sequentially. During this process, it is necessary to identify whether there is a continuous personnel succession relationship between jobs. For example, if P03 takes over the task of the personnel leaving P02, a continuous transfer can be formed in the sequence. Each transfer relationship is then sorted according to the task time succession order, and the corresponding job number at the continuous task switching node is extracted. This further forms a job transition path sequence within the time period. This path fully reflects the relative position and succession order between jobs in continuous tasks, ultimately yielding a job time connection sequence.

[0082] The responsibility continuity judgment submodule calls the job time connection sequence, analyzes the job description fields in the time period before and after the task, compares the job task description text, removes job numbers whose responsibility content has changed, and retains the set of job numbers with consistent responsibility description fields, thus obtaining the job number group with consistent responsibilities.

[0083] First, the job description field is extracted by using each job number and its task start and end time as input conditions. The description field is then split into comparable datasets by phrases, key verbs, and scope of responsibilities. In practice, for example, if job A's job description field includes "review contracts and archive documents" during tasks T101 to T103, and still contains "review contracts and archive documents" during tasks T104 to T106, this can be considered a case of consistent job description fields. Subsequently, for each job number, the job description field is compared one by one with the job description fields of the same job number in the preceding and following time periods, focusing on identifying changes in verb or noun combinations in the job description field. For example, if job B's job description field is "calculate wages and organize invoices" during T201 to T203, but "calculate wages and approve invoices" during T204 to T205, this is considered a change in the job description field. Then, job numbers with consistent job description fields are extracted, and job numbers with changed job description fields are removed, finally resulting in a group of job numbers with consistent responsibilities.

[0084] The task succession path extraction submodule extracts the number pairs of positions that appear in adjacent positions on the timeline based on the job number group with consistent responsibilities, analyzes the task succession relationship, forms a linear sequence of task connection paths according to the number pairs, and outputs the correspondence between positions and tasks in the path to obtain the job connection path group.

[0085] First, arrange each job number sequentially along the timeline to form a continuous sorting result. Then, extract adjacent job number pairs sequentially and match them one by one with the task time segments associated with each pair. Within each pair, extract the start and end times of the tasks undertaken by job numbers A and B. Determine if the start time of task B is immediately following the end time of task A. If the times are adjacent and there are no gaps, the pair is considered to have a continuous succession relationship. For example, job numbers 030 and 031 are sequentially arranged between tasks T120 to T123 and tasks T124 to T126. Furthermore, the job descriptions remain consistent and can be used as successor pairs for subsequent processing. Then, according to the task number order, all number pairs with successor relationships are connected to form a linear chain of task paths, such as the path "030-031-032", which means that the task completed by position 030 is continued by 031 and then completed by 032. Further, the task numbers of each position number contained in each successor path are traced back to extract the corresponding tasks in the path, forming a one-to-one correspondence between positions and tasks. All such linear paths are output one by one, finally obtaining the position connection path group.

[0086] Specifically, such as Figure 2 , Figure 7 As shown, the behavioral rhythm adaptation assessment module includes:

[0087] The time information extraction submodule extracts the start time of job tasks and the corresponding responsibility time period based on the job connection path group. It categorizes the time information of differentiated jobs into data sequences with self-numbering, and filters out time intervals where there is an overlap between the task time sequence and the responsibility time period to obtain the job task time segment.

[0088] First, the start time of each task is retrieved according to its corresponding task number in the path, constructing a set of time period information for each job task. Then, by comparing the task time period and responsibility time period associated with each number, the task time period and responsibility time period are extracted from the two time sources. The job number is used as a sequence identifier, and its start and end positions for the task time period and responsibility time period are stored sequentially in the data structure for subsequent matching steps. Next, the execution time ranges of the task time period and responsibility time period are compared one by one, unit by unit, to determine if there is any overlap between the two time periods. For the combined part, the judgment method is to compare whether the task start time is earlier than the duty end time and whether the task end time is later than the duty start time. If they are satisfied, it means that the two time segments overlap. Further, the actual start and end times of the two overlapping intervals are extracted and recorded as the overlapping time period of the job task. For example, the task segment of job number 042 is from 12:00 to 14:00 and the duty segment is from 13:00 to 15:00. Then the overlapping segment is from 13:00 to 14:00. Then, in order of job number, all overlapping time periods are output in the path order to obtain the job task time segment.

[0089] The behavior overlap recognition submodule calls the job task time segment, extracts the time position of behavior data from the task execution time, compares the time of each behavior with the job responsibility time segment, determines whether it appears repeatedly in the same time interval, and imports it into the result table according to the job number to obtain the behavior interval overlap sequence group.

[0090] First, obtain the start and end times of the task segments in order of job number. Simultaneously, extract the content of the behavioral event corresponding to each task and the time position when the behavior occurred. Then, extract the time point or time period for each behavior record. For example, behavior number B103 occurs from 09:32 to 09:46. Next, retrieve its corresponding job number P08 from the data, with a corresponding duty time period of 09:30 to 09:50. Perform an overlap judgment between the behavior's time range and the job's duty time interval. The judgment logic is to compare the behavior's start and end times sequentially with the start and end times of the job's duty time interval, determining whether the behavior's start time is no earlier than the duty start time and the behavior's end time is no later than the duty end time. If satisfied, mark the behavior as occurring within the duty scope and record the behavior number and job number. The correspondence between the numbers is determined, and the result is written into the result table with P08 as the row label. This judgment operation is repeated in all job numbers. For cases where the behavior time interval spans multiple job responsibility time intervals, it is necessary to compare with multiple job time intervals and output multiple job numbers. For example, if the time interval of behavior B203 is from 10:05 to 10:25, and it appears in the responsibility time of both job P09 and P10, then the result must include the corresponding entries for both P09 and P10, with behavior number B203 and job numbers P09 and P10. At the same time, the time interval corresponding to each job in the result table must be clearly stated, which is used to determine the density of behavior and job matching later. By adding all job numbers associated with each behavior to the job behavior comparison sequence in sequence, a behavior interval overlap sequence group is obtained.

[0091] The rhythm adaptation assessment submodule extracts the number of behaviors for each position at the time intersection based on the overlapping sequence of behavior intervals. Combined with the time interval sequence corresponding to the job responsibilities, the module compares the occurrence time of the behaviors with the number of behaviors and classifies them into the task carrying relationship table for each position to obtain the job behavior rhythm adaptation index.

[0092] First, extract the behavioral event data for each job position by job number and determine the specific time intervals in which these behaviors occur. For example, the behavioral events corresponding to job position P07 include B101, B104, and B109, where B101 occurs from 08:45 to 08:52, B104 occurs from 08:57 to 09:03, and B109 occurs from 09:10 to 09:19. Next, according to the time period of 08:40 to 09:20 in the job description of job position P07, compare the time of each of the three behavioral events with the job time interval to determine whether each behavior falls entirely within the job time. If so, the behavior is considered a valid occurrence within the job's job time interval. Then, map the three behavior times to three positions within the job time interval, for example, 08:40 to 08:55, 08:55 to 09:05, and 09:05 to 09:20, corresponding to B101, B104, and B109, respectively, and denoted as job position P07. 7. For each behavior record in these three position segments, write the P07 number and the occurrence markers of the three behavior segments into the task carrying relationship table. Continue processing the next job number P08, and so on to complete the comparison of behaviors and responsibility segments for all jobs. If a behavior spans multiple job responsibility segments, for example, behavior B204 occurs from 09:30 to 09:45 and appears in responsibility segments P09 and P10, record it as a behavior occurring once in both P09 and P10. Then, compare its order in the time segment to determine the position of the behavior in the responsibility segment, whether it is in the beginning, middle or end, and classify it as appearing in the beginning, middle or end of the rhythm. Record these position markers and job numbers into the task carrying relationship table. Then, compare the number and distribution of behaviors for each job number in the table, and output the distribution characteristics of the number of behaviors in each job to form a behavior rhythm matching record for each job in each responsibility time segment. Finally, write the relationship between the job number and the number of behaviors in each segment into the indicator table to obtain the job behavior rhythm fit index.

[0093] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A big data-based human resource intelligent analysis system, characterized in that, The system includes: The behavior rhythm capture module obtains daily employee behavior records in different time periods, extracts hourly behavior activities, the corresponding task start and end time positions, filters overlapping time periods of tasks, and obtains the corresponding segments of behavior tasks. The job load identification module extracts repetitive task types, corresponding task and job number based on the segment corresponding to the behavioral task, tracks the time distribution of job task content, compares the task execution between jobs, and obtains functional load job groups. Based on the functional load job group, the task connection analysis module extracts the execution order of upstream and downstream tasks, finds the task connection position, identifies the continuous task relationship between differentiated jobs, and obtains the task boundary link. Based on the task boundary link, the job transfer tracking module extracts job flow records during the boundary time period, analyzes job information that is uninterrupted and whose responsibilities have not changed, promotes the task relationship between jobs, and obtains job connection path groups. The behavioral rhythm adaptation assessment module extracts the employee's task start time and job responsibility cycle based on the job connection path group, cross-compares the time periods, analyzes the time overlap between the task behavior interval and the responsibility cycle, and obtains the job behavior rhythm adaptation index. The differentiated positions refer to those that are distinguished by differences in task type, time distribution, and responsibilities during task execution. The "uninterrupted and unchanged responsibilities" job information refers to job records where, during task handover and job rotation, personnel and time change, but the job responsibilities remain consistent and the task is not interrupted. The corresponding segment of the behavioral task includes the frequency of behavior occurrence, the start and end positions of task execution, and the continuous interval of behavior. The functional load job group includes repetitive task types, the correspondence between tasks and jobs, and the job task undertaking sequence. The task boundary link includes the acceptance time point, the frequency of job task alternation, and the position of continuous task connection. The job connection path group includes job flow records, task uninterrupted indicators, and job task sequence. The job behavior rhythm adaptation index includes the high-frequency behavior interval of tasks, the job responsibility carrying cycle, and the number of times the time intersection position is repeated.

2. The big data based human resource intelligent analysis system according to claim 1, wherein, The job responsibility cycle refers to the time period during which the job is responsible, and the start and end time range of the tasks undertaken by the job. The overlapping time intervals between the task behavior time intervals and the job responsibility cycle reflect the overlap of tasks and responsibilities in time. 3.The big data based human resource intelligent analysis system according to claim 1, wherein, The behavior rhythm capture module includes: The time-segment behavior collection submodule acquires daily employee behavior records in different time periods, divides time segments by hour, extracts the time of occurrence of behavior within a time segment, locates the position of behavior within a time segment, and obtains the time sequence of behavior activities. The task time matching submodule extracts the task start time and end time based on the behavior activity time sequence, the corresponding behavior time point and the task time segment time position, determines whether the time point is within the task, and obtains a task behavior correspondence table according to the behavior and task correspondence. The behavior task interval extraction submodule calls the task behavior correspondence table, extracts the part of continuous behavior time period that overlaps with the task time, locates the start and end time of the behavior and the task boundary position, and delineates the segment edge according to the time sequence to obtain the segment corresponding to the behavior task.

4. The big data based human resource intelligent analysis system according to claim 1, wherein, The job load identification module includes: The task type extraction submodule extracts task name data within a time period based on the segment corresponding to the behavior task, sequentially combines and locates the task name with the time point, associates the repeated task name with the task number, arranges the task type in the order of the task number, and obtains a duplicate task type reference table. The job task tracking submodule calls the repetitive task type lookup table, extracts the job number associated with each task number, collects the task content corresponding to the job in the time period along the time axis, and continuously extracts the positional relationship between time and task number to obtain the job task time sequence table. The task content comparison submodule extracts the task numbers associated with differentiated positions within a continuous time range based on the job task time series table, compares the consecutive repetition order of the task numbers, and extracts the job numbers that undertake the same type of task in the time period to obtain the functional load job group. 5.The big data based human resource intelligent analysis system according to claim 1, wherein, The task connection analysis module includes: The task time extraction submodule extracts the start and end times of each task from the job task schedule based on the functional load job group, arranges the task time periods in chronological order, determines the time connection interval between consecutive tasks, and obtains the task start and end time sequence. Based on the task start and end time sequence, the task connection and positioning submodule filters task combinations with adjacent time positions according to the positional relationship between tasks with different end and start times. It compares the cases where there is no time overlap and no interval between tasks, and obtains a task time connection comparison table according to the task time boundary matching order. The task node evaluation submodule extracts the corresponding job number in the task pair based on the task time connection lookup table, identifies the position of continuous association of task pairs under the same job number, filters out the repeated connection relationship in the continuous time period, extracts the corresponding task number and time segment, and obtains the task boundary link.

6. The big data based human resource intelligent analysis system according to claim 1, wherein, The job transfer tracking module includes: The job flow extraction submodule collects job flow records within the boundary time period based on the task boundary link, extracts the job number and task start and end time corresponding to each flow record, and outputs the job number sequence corresponding to the succession of consecutive tasks according to the time order to obtain the job time connection sequence. The responsibility continuity judgment submodule calls the job time connection sequence, analyzes the job description field in the time period before and after the task, compares the job task description text, removes job numbers whose responsibility content has changed, and retains the set of job numbers with consistent responsibility description fields to obtain the job number group with consistent responsibilities. The task succession path extraction submodule extracts the number pairs of positions that appear in adjacent positions on the time axis based on the job number group with consistent responsibilities, analyzes the task succession relationship, forms a linear sequence of task connection paths according to the number pairs, and outputs the correspondence between positions and tasks in the path to obtain the job connection path group. 7.The big data based human resource intelligent analysis system according to claim 6, wherein, The process of collecting job flow information within the boundary time period in the job flow extraction submodule is as follows: the boundary time range is divided into continuous segments according to the preset time interval length, the associated job number and task start and end time are extracted in each time segment, and the job number sequence corresponding to the task is arranged in chronological order. The process of comparing the job task description text in the continuous responsibility judgment submodule is as follows: the description content of each job number within the time range before and after the task is decomposed into responsibility fields, the job responsibility fields of adjacent time periods are compared for text consistency, job numbers with obvious changes in description are filtered out according to the character overlap ratio, and job numbers with consistent description content are retained. The process of extracting job number pairs with adjacent time positions in the task succession path extraction submodule is as follows: among the job numbers with consistent job descriptions, the numbers are arranged in chronological order, consecutive number pairs are extracted, and the corresponding task information for each number is combined to list the correspondence between job numbers and task numbers in chronological order, forming a complete task succession path sequence.

8. The big data based human resource intelligent analysis system according to claim 1, wherein, The behavior rhythm adaptation assessment module includes: The time information extraction submodule extracts the start time of job tasks and the corresponding responsibility time period based on the job connection path group. It then categorizes the time information of differentiated jobs into data sequences with self-numbering, and filters out time intervals where there is an overlap between the task time sequence and the responsibility time period to obtain the job task time segment. The behavior overlap recognition submodule calls the job task time segment, extracts the time position of behavior data from the task execution time, compares the time of each behavior with the job responsibility time segment, determines whether it appears repeatedly in the same time interval, and imports it into the result table according to the job number to obtain the behavior interval overlap sequence group; The rhythm adaptation assessment submodule extracts the number of behaviors for each position at the time intersection position based on the overlapping sequence group of the behavior intervals. Combining the time segment order corresponding to the job responsibilities, it compares the occurrence time of the behaviors with the number of behaviors and classifies them into the task carrying relationship table for each position to obtain the job behavior rhythm adaptation index.

Images