A multi-module collaborative management system and method fusing attendance and business approval

Abstract

The present application relates to the technical field of data processing, in particular to a multi-module collaborative management system and method fusing attendance and business approval. In the system: an event access module listens to and captures state data of different attendance terminals and business approval systems, standardizes the state data into attendance life cycle events and publishes the attendance life cycle events to an event distribution module; the event distribution module distributes the events to a conflict state buffer module in which a first cache set and a second cache set are located; the conflict state buffer module synchronizes a mapping relationship between a punch parameter change amount and an approval activity time period in real time; after determining a conflict trigger, a conflict resolution module adaptively decides to execute a fast and direct state ruling or quantitative confidence score calculation according to event characteristics and business urgency, and resolves the conflict of the event. The present application realizes deep decoupling of physical attendance and approval systems, greatly improving the real-time performance and accuracy of conflict self-healing determination.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a multi-module collaborative management system and method that integrates attendance and business approval. Background Technology

[0002] In large enterprises, modern manufacturing, and highly flexible retail sectors, attendance management systems are core tools for ensuring intelligent calculation of working hours, automatic anomaly detection, and payroll linkage. With the diversification of enterprise office models, the sources of attendance data have expanded from traditional single physical clock-in terminals to multi-terminal, multi-source environments, including GPS-based field location clock-in on mobile terminals, passive clock-in based on workplace Wi-Fi recognition, facial recognition clock-in based on smart cameras, and various business approval data synchronized by collaborative office systems.

[0003] This multi-source attendance environment presents serious challenges in data conflict and state consistency. Traditional attendance calculation systems often employ a timed batch processing model. These systems typically index and associate the raw attendance data collected that day with "attendance personnel and attendance date" late at night or at a preset settlement time, and then uniformly schedule batch calculations according to preset attendance rules. In this traditional request-response or timed batch processing model, the system cannot perceive real-time state changes of multi-source data.

[0004] When physical attendance data collection and business approval processes occur asynchronously, conflicts frequently arise. For example, an employee may be on a temporary business trip and appear "missed" on the attendance hardware terminal, while their business trip approval process is still in progress within collaborative office software (such as OA or workflow approval software) and has not yet been finalized. In the traditional timed calculation model, this period would be judged as abnormal absence until the final approval process is completed several days later, at which point historical corrections would be made through manual intervention or a global recalculation. This delayed processing mechanism compromises the real-time nature and accuracy of attendance data, as well as the stability of payroll-linked calculations.

[0005] Furthermore, in complex business scenarios (such as three-shift work in manufacturing, flexible scheduling in retail, and on-call and duty-based work in the healthcare industry), attendance rules not only vary greatly but are also highly complex. For example, complex constraints such as "at least one senior technician must be on duty on weekends" create a strong coupling between attendance calculations, production scheduling, and shift planning. Existing attendance systems struggle to handle such complex scheduling constraints and approval conflicts in real time, even with a high volume of multi-source data inflow. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a multi-module collaborative management system and method that integrates attendance and business approval.

[0007] A multi-module collaborative management system integrating attendance and business approval includes an event access module, an event distribution module, a conflict state buffer module, and a conflict resolution module;

[0008] The event access module is used to receive raw attendance and status change data from different physical attendance terminals and business approval systems in real time, and format them into standardized attendance lifecycle events.

[0009] The event distribution module is connected to the event access module and is based on the publish-subscribe pattern to realize the asynchronous transmission and routing distribution of the attendance lifecycle events in various modules of the system.

[0010] The conflict state buffer module is connected to the event distribution module and includes a first cache set and a second cache set; wherein the first cache set is used to dynamically accumulate and update the check-in parameter values ​​from the physical attendance terminal in real time; the second cache set is used to synchronously maintain the mapping relationship between the approval workflow status from the business approval system and the corresponding personnel and time activity identifiers.

[0011] The conflict resolution module is connected to the conflict state buffer module. It performs cross-matching queries on the first and second cache sets. If it is determined that an employee has both a physical attendance anomaly record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered. This logic utilizes a fast, direct adjudication based on the conflict decision rule matrix or performs state decision resolution after calculating a quantitative confidence score. Wherein:

[0012] The triggering condition for executing a fast and direct decision is: within the same assessment period, the attendance parameters of the first cache set and the approval status of the second cache set have a unique hard mapping relationship in the decision rule matrix without competition; or, the approval status is a non-final approval status, and the current calculation time is greater than or equal to the attendance settlement deadline by a preset grace period threshold.

[0013] The triggering conditions for performing quantitative confidence score calculation are: receiving two or more physical check-in events from different devices with contradictory representation states within the same assessment period; or, the approval status is a non-final approval status, and the current calculation time is less than the preset grace period threshold from the attendance settlement deadline; or, the overlap ratio between physical check-in duration and approval duration is within the critical fluctuation range of the judgment threshold.

[0014] Preferably, the standardized attendance lifecycle event includes a set of attributes to describe the characteristics of a specific event, and its data structure includes: a unique event ID, an employee user ID, a physical timestamp of the event occurrence, a data source type, an event behavior type, and load data for carrying state load.

[0015] Preferably, the load data is a structured key-value pair with nested fields; when the data source type is a physical attendance terminal, the load data includes the physical device location, hardware MAC address, client network IP, and liveness detection feature hash value; when the data source type is a business approval system, the load data includes the approval form ID, activity time range, approval activity type, and current workflow status.

[0016] Preferably, the conflict decision rule matrix specifically includes:

[0017] If the physical attendance status in the first cache set is "absent", and the approval status in the second cache set is "leave application or business trip approval in progress", when the current time is greater than or equal to the preset grace period threshold, the final resolution decision status is output as "pending and suspended"; when the time is less than the preset grace period threshold, the quantitative confidence score is calculated.

[0018] If the physical attendance status in the first cache set is absent, and the approval status in the second cache set is approved for leave or business trip, and the activity time period corresponding to the approval form covers the current attendance time period, the final resolution decision status is planned leave.

[0019] If the physical attendance status in the first cache set is late, and the approval status in the second cache set is card replacement or flexible appeal approval in progress, and the corresponding employee belongs to the preset flexible scheduling sequence, the final resolution decision status is pending suspension; if the approval status is approved and the card replacement time matches the scheduled shift, the final resolution decision status is normal attendance.

[0020] If the physical attendance status in the first cache set is Wi-Fi connected, and the approval status in the second cache set is "leave approved", and the proportion of the employee's actual on-duty time covering the leave time reaches a preset threshold, the final resolution decision status is output as "actual attendance to cover the leave status".

[0021] Preferably, when the quantitative confidence score calculation is triggered, the conflict resolution module calculates the conflict resolution confidence score of the candidate attendance state s at the conflict calculation time T by introducing a time decay function of the event occurrence and a data source confidence weight: ;

[0022] In the formula, C(s, T) is the conflict resolution confidence score, N is the number of events received during the assessment period related to the employee and the assessment period, W is the baseline trust weight coefficient of the data source that generated the event, and R(type, s) is the correlation function between the attendance event behavior type and the candidate state s. t is the preset time decay factor, and t is the physical timestamp of the actual occurrence of the event.

[0023] Preferably, the step of performing state decision resolution after calculating the quantitative confidence score specifically includes:

[0024] Select the state S1 with the highest confidence score and the state S2 with the second highest score, and compare the difference between the two with the preset decision safety threshold.

[0025] If the difference is greater than or equal to the preset decision safety threshold, then S1 will be output as the final attendance judgment decision state.

[0026] If the difference is less than the preset decision security threshold, the conflict resolution module marks the attendance status of that period as pending and suspends it, and triggers a reminder event which is routed to the approval system through the event distribution module for node approval reminder.

[0027] A multi-module collaborative management method integrating attendance and business approval includes the following steps:

[0028] The event access module listens for and captures status change data from various physical attendance terminals and business approval systems in real time, formats it into standardized attendance lifecycle events, and then publishes it to the event distribution module.

[0029] The event distribution module asynchronously distributes the received standardized attendance events to the conflict state buffer module; the first cache set updates the target parameter value of the corresponding employee on the assessment date according to the target parameter change carried in the physical attendance event; the second cache set synchronously stores and maintains the mapping relationship between the corresponding approval form status and the activity time period identifier according to the received approval workflow status event.

[0030] The conflict resolution module performs cross-matching queries on the first cache set and the second cache set. If it is determined that an employee has both an abnormal physical attendance record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered.

[0031] If the event has low ambiguity, low urgency, or originates from a single data source, the conflict resolution module directly adjudicates and outputs the result based on a preset conflict decision rule matrix. If the event data stream contains conflicts from multiple data sources, has high urgency, or strong ambiguity, the conflict resolution module performs state decision resolution after calculating a quantitative confidence score.

[0032] Compared with the prior art, the advantages of this invention are:

[0033] By using an event-driven architecture, the multi-source attendance physical terminals and the business approval system can achieve complete asynchronous, non-blocking, and loose coupling. When new attendance devices need to be connected or new business approval workflows need to be launched, only new event generators or subscribers need to be registered on the distribution bus. There is no need to reconstruct the existing core conflict resolution module, which improves the system's agility and elastic scalability.

[0034] The conflict resolution module adaptively adopts two different processing strategies based on the ambiguity of the scenario conflict and the urgency of the business: "fast direct state adjudication" and "quantitative confidence calculation". In the majority of conventional, non-competitive, low-ambiguity scenarios, it quickly executes direct hard rule matching adjudication, avoids complex calculations, reduces server load, and achieves efficient processing of concurrent attendance streams.

[0035] To address the competitive conflicts among multiple devices (such as the physical competition between Wi-Fi on-duty and field GPS), strong conflicts in settlement and clearing timeliness, or highly ambiguous critical states such as high network latency, this system uses quantitative confidence calculation and a time decay function and a trusted weight mechanism for multi-dimensional adjudication. This can effectively filter out "edge noise" caused by network congestion, device offline jitter, GPS signal offset, and asynchronous delays in workflow approval. This allows for the smooth and autonomous resolution of state conflicts and the completion of a closed loop for attendance and payroll linkage, eliminating disputes caused by misjudgments of state. Attached Figure Description

[0036] Figure 1 This is a system architecture diagram of a multi-module collaborative management system that integrates attendance and business approval proposed in this invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0038] refer to Figure 1 This solution is a multi-module collaborative management system integrating attendance and business approval, including an event access module, an event distribution module, a conflict state buffer module, and a conflict resolution module, wherein:

[0039] The event access module is used to receive raw attendance and status change data from different physical attendance terminals and business approval systems in real time, and format them into standardized attendance lifecycle events.

[0040] The event distribution module is connected to the event access module and is based on the publish-subscribe pattern to realize the asynchronous transmission and routing distribution of the attendance lifecycle events in various modules of the system.

[0041] As an edge access layer, the event access module receives raw attendance or status change data from various physical devices and collaborative software in real time through Restful API, Websocket, and MQTT protocol.

[0042] To ensure standardized transmission of event streams, the event access module captures this data and encapsulates it into a standardized attendance lifecycle event, which specifically includes the following attributes:

[0043] Unique Event ID: Used to globally and uniquely identify the event in a distributed network, such as "evt_309af8bc92d1".

[0044] Employee User ID: Used to associate with specific employees who clock in or approve, such as "usr_9011832".

[0045] Physical timestamp of the event: Used to record the actual millisecond time when the event occurred, for example, "1780312800000".

[0046] Data source type: Used to distinguish different attendance devices or collaborative software sources, such as Wi-Fi hotspots, smart camera facial recognition, GPS positioning, OA approval, physical fingerprint machines, etc.

[0047] Event behavior type: Used to define the specific action of the event, such as entering a Wi-Fi hotspot area, successful facial recognition, submission of an approval process, etc.

[0048] Load data: Used to carry personalized extended fields brought about by specific event sources. For example, when the data source type is a physical attendance terminal, it carries the physical device location, hardware MAC address, client network IP, and liveness detection feature hash value; when the data source type is a business approval system, it carries the approval form ID, activity time range, approval activity type, and current workflow status.

[0049] Because different types of physical terminals and systems differ in data acquisition accuracy, transmission latency, and anti-cheating measures, this embodiment pre-configures the attributes of each data source:

[0050] Camera-based facial recognition check-in: Its baseline trust weight coefficient is set to 0.95, with extremely low latency tolerance (less than or equal to 1 second), and it adopts binocular liveness detection and device binding verification as a mechanism to prevent cheating during data collection.

[0051] Physical wall-mounted fingerprint attendance machine: Its baseline trust weight coefficient is set to 0.90, with low latency tolerance (less than or equal to 5 seconds), and it uses a hardware security chip for encrypted fingerprint transmission to prevent cheating.

[0052] Automatic scanning and connection to Wi-Fi hotspots in the office area: The baseline trust weight coefficient is set to 0.70, the latency tolerance is medium (less than or equal to 30 seconds), and anti-cheating measures are taken by using a two-way handshake verification between SSID and client MAC.

[0053] Mobile APP field attendance location: Its baseline trust weight coefficient is set to 0.60, the latency tolerance is medium (less than or equal to 10 seconds), and anti-cheating mechanisms such as virtual location detection and multiple people from the same IP address attendance warning are adopted.

[0054] Collaborative office approval workflow change event: its baseline trust weight coefficient is set to 0.85. Since it is an asynchronous approval, its delay tolerance is high and not fixed. The anti-fraud mechanism is guaranteed by CA digital certificate and digital signature of workflow node.

[0055] After preprocessing the events (e.g., filtering out multiple attendance events with abnormal IP addresses), the multi-source event access gateway transmits the standardized events to the event distribution module. The distribution module uses a publish-subscribe-based routing mechanism to distribute the events to the conflict state buffer module according to the event's data source type and event behavior type.

[0056] The conflict state buffer module is connected to the event distribution module and includes a first cache set and a second cache set; wherein the first cache set is used to dynamically accumulate and update the check-in parameter values ​​from the physical attendance terminal in real time; and the second cache set is used to synchronously maintain the mapping relationship between the approval workflow status from the business approval system and the corresponding personnel and time activity identifiers.

[0057] To eliminate the I / O pressure on the underlying relational database caused by massive concurrent computation of events, the conflict state buffer module uses a high-performance in-memory database (such as Redis Cluster) to maintain hot data. Its core synchronization mechanism includes:

[0058] When an employee clocks in at a physical time clock or when an employee is detected connecting to a Wi-Fi hotspot, a Type I attendance event is generated. The event access module uses the event distribution module to send this event (carrying parameter changes, such as clock-in count +1, work hours change +8 hours, etc.) to the first cache set. The first cache set establishes a hash index association using the user ID as the key, and continuously updates the corresponding target parameter values ​​(such as the current cumulative clock-in count, the physical attendance status value for the day, etc.) in real time.

[0059] Meanwhile, when employees submit approval applications for leave, business trips, or card replacements, and when the approval process transitions in the OA system (generating a second type of approval event), the event access module uses the event distribution module to send the "Approval Form ID" and "Activity Identifier" (including leave start and end times, leave type, associated employee ID, etc.) to the second cache set. The second cache set stores the correspondence between the "Approval Form ID" and the "Activity Identifier" and maintains the current status of each approval form (e.g., under review, approved, rejected, etc.).

[0060] The conflict resolution module is connected to the conflict state buffer module and performs cross-matching queries on the first cache set and the second cache set. If it is determined that an employee has both an abnormal physical attendance record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered. The conflict resolution judgment includes a quick direct adjudication based on the conflict decision rule matrix or a state decision resolution after calculating a quantitative confidence score.

[0061] The conflict resolution module monitors the first cache set in real time. When it detects an anomaly in the physical attendance status (e.g., the first cache set shows no attendance records for the morning), it retrieves activity identifiers (i.e., approval form information in progress) related to the user ID from the second cache set that overlap in time period. By performing an intersection query between the "actual attendance parameter values" in the first cache set and the "correspondence between approval status and time period" in the second cache set, a conflict can be quickly determined.

[0062] After determining the existence of a conflict, the conflict resolution logic is triggered to resolve the state decision. Specifically, to balance computational accuracy and system resource consumption in a high-concurrency environment, the conflict resolution module, when resolving state conflicts, adaptively evaluates event attributes, urgency, and the state of competition among multiple data sources to determine whether to "perform a fast and direct state decision based on the configured rule matrix" or "perform quantitative calculations by calling a confidence formula."

[0063] When an event is low in ambiguity, low in urgency, or originates from a single data source, the conflict resolution module directly runs the decision matrix and directly determines the output result, avoiding the consumption of system computing power. Triggering conditions include:

[0064] Low ambiguity: Within the overlapping assessment timeframe, the attendance parameters (such as absence or lateness) of the first cache set and the approval status of the second cache set have a unique, non-competitive, and exclusive correspondence in the rule decision matrix. For example, if the attendance record is "absent" and the corresponding leave application is already in the "approved" final review status, there is no conflict or ambiguity, and the decision is directly "planned leave".

[0065] Low urgency: When a physical attendance anomaly in the first cache set collides with an approval order in transit in the second cache set, and the current calculation time is greater than or equal to the preset grace period threshold (set to 3 days in this embodiment), the judgment process has ample time to converge naturally without initiating quantitative calculation, and is directly decided as "pending suspension".

[0066] Single data source: During a specific assessment period, only physical attendance events generated by a single high-trust data source (such as facial liveness attendance with a baseline trust weight coefficient greater than 0.95) are received, and no approval workflows of any collaborative office systems are in circulation or overlap during this period. The decision is made directly based on the high-trust data source event.

[0067] In a preferred embodiment, the conflict decision rule is defined by the following logic:

[0068] Rule 1: When the status of the first cache set is absent and the status of the second cache set is leave or business trip approval in progress, if the current calculation time is greater than or equal to 3 days away from the end date of the current attendance settlement, the conflict resolution decision status is determined to be "pending suspension".

[0069] Rule 2: When the first cache set status is absent and the second cache set status is leave or business trip approval in progress, if the current calculation time is less than 3 days away from the current attendance settlement deadline, the abnormal degradation strategy will be triggered, and the conflict resolution strategy will be adjusted to quantitative confidence score calculation.

[0070] Rule 3: When the first cache set status is absent, and the second cache set status is approved for leave or business trip, and the start and end times of the approval form cover the current attendance period, the conflict resolution decision status is directly resolved and determined as "planned leave".

[0071] Rule 4: When the first cache set status is late, and the second cache set status is card replacement or flexible appeal approval, and the employee belongs to the flexible scheduling sequence, the conflict resolution decision status is determined as "pending suspension".

[0072] Rule 5: When the first cache set status is late, and the second cache set status is card replacement or flexible appeal has been approved, and the card replacement time matches the scheduled shift, the conflict resolution decision status is determined as "normal attendance".

[0073] Rule 6: When the first cache set is in Wi-Fi connection state, the second cache set is in leave approved state, and the physical on-duty time covers the leave time by more than or equal to 50%, the conflict coverage logic is triggered, and the conflict resolution decision state is determined as "actual attendance".

[0074] When multiple data source conflicts, high urgency, or strong ambiguity occur in the event data stream, and a direct determination cannot be made using the pre-configured rule matrix, a quantitative confidence score calculation is initiated. The triggering conditions include:

[0075] Multiple data source conflicts: Two or more attendance events from different physical terminals with contradictory statuses are captured within the same assessment period. For example, at 09:10, a network connection event from WiFi (indicating employee attendance) and an abnormal attendance event from GPS outside the factory area (indicating employee field attendance) are received simultaneously. Due to the conflict, confidence calculations must be performed to determine the specific conflict resolution.

[0076] High urgency: When the first cache set shows a physical attendance anomaly, and the second cache set contains a corresponding approval form in progress, but the current calculation time is less than the preset grace period threshold (set to 3 days in this embodiment) from the current attendance settlement deadline. Since settlement is imminent, the suspended state must converge. At this time, the direct suspension decision is no longer executed. Instead, the confidence score of the current approval form in progress at each approval node is calculated and quantitatively compared with the absentee confidence score.

[0077] Strong ambiguity: The event distribution module detects severely out-of-order events caused by network congestion or device offline retransmission, meaning the captured event's physical timestamp lags significantly behind the current system time, and the lag exceeds the preset maximum latency tolerance threshold. In this case, quantitative calculation must be initiated to accurately quantify the residual weight of the out-of-order event at the current assessment time using a time decay function.

[0078] When it is determined that quantitative calculation is required, the conflict resolution module extracts all events associated with the employee during the assessment period on that day and performs conflict resolution confidence score calculation:

[0079]

[0080] In the formula, C(s, T) is the conflict resolution confidence score of candidate attendance state s at conflict calculation time T, N is the number of events received during the assessment period related to the employee and the assessment period, W is the baseline trust weight coefficient of the data source that generated the event, and R(type, s) is the support or rejection strength of the attendance event behavior type for candidate state s (the value ranges from -1 to 1, with positive values ​​representing support and negative values ​​representing rejection). t is the preset time decay factor, and t is the physical timestamp of the actual occurrence of the event.

[0081] In a preferred embodiment, the correlation function R(type, s) is assigned scores in real time using a two-dimensional discrete mapping matrix. A specific assignment example is shown in the table below:

[0082] Entering the Wi-Fi hotspot area 0.8 −0.5 0.2 −0.9 Leave the Wi-Fi hotspot area −0.2 0 0.5 0.1 Facial recognition successful (at work location) 1 −1.0 −0.8 −1.0 GPS check-in outside the factory area (field work) 0.4 −0.3 0.1 −0.6 Approval workflow submission (leave / business trip) −0.4 0.7 −0.2 0.5 Approval process approved (card replacement appeal) 0.9 −0.5 −0.7 −0.9

[0083] After calculating the confidence score of each candidate state s, the conflict resolution module executes the following decision logic:

[0084] Find the state S1 with the highest confidence score and the state S2 with the second highest confidence score.

[0085] If the difference between S1 and S2 is greater than the set decision safety threshold (set to 0.35 in this embodiment), the resolution decision state is directly output as S1.

[0086] If the difference between S1 and S2 is less than the set decision security threshold, it indicates that the current data conflict still has a high degree of ambiguity. At this time, the conflict resolution module sets the decision status to "pending suspension", does not count it as an anomaly, and uses the event distribution module to send an asynchronous reminder notification, which is routed to the approval software by the bus for reminder.

[0087] Based on the aforementioned multi-module collaborative management system integrating attendance and business approval, this solution provides a multi-module collaborative management method integrating attendance and business approval, comprising the following steps:

[0088] The event access module listens for and captures status change data from various physical attendance terminals and business approval systems in real time, formats it into standardized attendance lifecycle events, and then publishes it to the event distribution module.

[0089] The event distribution module asynchronously distributes the received standardized attendance events to the conflict state buffer module; the first cache set updates the target parameter value of the corresponding employee on the assessment date according to the target parameter change carried in the physical attendance event; the second cache set synchronously stores and maintains the mapping relationship between the corresponding approval form status and the activity time period identifier according to the received approval workflow status event.

[0090] The conflict resolution module performs cross-matching queries on the first cache set and the second cache set. If it is determined that an employee has both an abnormal physical attendance record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered.

[0091] If the event has low ambiguity, low urgency, or originates from a single data source, the conflict resolution module directly adjudicates and outputs the result based on a preset conflict decision rule matrix. If the event data stream contains conflicts from multiple data sources, has high urgency, or strong ambiguity, the conflict resolution module performs state decision resolution after calculating a quantitative confidence score.

[0092] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0093] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A multi-module collaborative management system integrating attendance and business approval, characterized in that, It includes an event access module, an event distribution module, a conflict state buffer module, and a conflict resolution module; The event access module is used to receive raw attendance and status change data from different physical attendance terminals and business approval systems in real time, and format them into standardized attendance lifecycle events. The event distribution module is connected to the event access module and is based on the publish-subscribe pattern to realize the asynchronous transmission and routing distribution of the attendance lifecycle events in various modules of the system. The conflict state buffer module is connected to the event distribution module and includes a first cache set and a second cache set; wherein the first cache set is used to dynamically accumulate and update the check-in parameter values ​​from the physical attendance terminal in real time; The second cache set is used to synchronously maintain the mapping relationship between the approval workflow status from the business approval system and the corresponding personnel and time activity identifiers. The conflict resolution module is connected to the conflict state buffer module. It performs cross-matching queries on the first and second cache sets. If it is determined that an employee has both a physical attendance anomaly record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered. This logic utilizes a fast, direct adjudication based on the conflict decision rule matrix or performs state decision resolution after calculating a quantitative confidence score. Wherein: The triggering condition for executing a fast and direct decision is: within the same assessment period, the attendance parameters of the first cache set and the approval status of the second cache set have a unique hard mapping relationship in the decision rule matrix without competition; or, the approval status is a non-final approval status, and the current calculation time is greater than or equal to the attendance settlement deadline by a preset grace period threshold. The triggering conditions for performing quantitative confidence score calculation are: receiving two or more physical check-in events from different devices with contradictory representation states within the same assessment period; or, the approval status is a non-final approval status, and the current calculation time is less than the preset grace period threshold from the attendance settlement deadline; or, the overlap ratio between physical check-in duration and approval duration is within the critical fluctuation range of the judgment threshold.

2. The multi-module collaborative management system integrating attendance and business approval as described in claim 1, characterized in that, The standardized attendance lifecycle event includes a set of attributes to describe the characteristics of a specific event. Its data structure includes: a unique event ID, an employee user ID, a physical timestamp of the event occurrence, a data source type, an event behavior type, and load data for carrying state load.

3. The multi-module collaborative management system integrating attendance and business approval as described in claim 2, characterized in that, The load data is a structured key-value pair with nested fields; when the data source type is a time and attendance physical terminal, the load data contains nested physical device location, hardware MAC address, client network IP, and liveness detection feature hash value. When the data source type is a business approval system, the load data includes the approval form ID, activity time range, approval activity type, and current workflow status.

4. The multi-module collaborative management system integrating attendance and business approval as described in claim 1, characterized in that, The conflict decision rule matrix specifically includes: If the physical attendance status in the first cache set is "absent", and the approval status in the second cache set is "leave application or business trip approval in progress", when the current time is greater than or equal to the preset grace period threshold, the final resolution decision status is output as "pending and suspended"; when the time is less than the preset grace period threshold, the quantitative confidence score is calculated. If the physical attendance status in the first cache set is absent, and the approval status in the second cache set is approved for leave or business trip, and the activity time period corresponding to the approval form covers the current attendance time period, the final resolution decision status is planned leave. If the physical attendance status in the first cache set is late, and the approval status in the second cache set is card replacement or flexible appeal approval in progress, and the corresponding employee belongs to the preset flexible scheduling sequence, the final resolution decision status is pending suspension; if the approval status is approved and the card replacement time matches the scheduled shift, the final resolution decision status is normal attendance. If the physical attendance status in the first cache set is Wi-Fi connected, and the approval status in the second cache set is "leave approved", and the proportion of the employee's actual on-duty time covering the leave time reaches a preset threshold, the final resolution decision status is output as "actual attendance to cover the leave status".

5. A multi-module collaborative management system integrating attendance and business approval as described in claim 1, characterized in that, When the quantitative confidence score calculation is triggered, the conflict resolution module calculates the conflict resolution confidence score of the candidate attendance state s at the conflict calculation time T by introducing the time decay function of the event occurrence and the confidence weight of the data source: ； In the formula, C(s, T) is the conflict resolution confidence score, N is the number of events received during the assessment period related to the employee and the assessment period, W is the baseline trust weight coefficient of the data source that generated the event, and R(type, s) is the correlation function between the attendance event behavior type and the candidate state s. t is the preset time decay factor, and t is the physical timestamp of the actual occurrence of the event.

6. The multi-module collaborative management system integrating attendance and business approval according to claim 1, characterized in that, The process of resolving state decisions after calculating the quantitative confidence score specifically includes: Select the state S1 with the highest confidence score and the state S2 with the second highest score, and compare the difference between the two with the preset decision safety threshold. If the difference is greater than or equal to the preset decision safety threshold, then S1 will be output as the final attendance judgment decision state. If the difference is less than the preset decision security threshold, the conflict resolution module marks the attendance status of that period as pending and suspends it, and triggers a reminder event which is routed to the approval system through the event distribution module for node approval reminder.

7. A multi-module collaborative management method integrating attendance and business approval, based on the multi-module collaborative management system integrating attendance and business approval described in claims 1-6, characterized in that, Includes the following steps: The event access module listens for and captures status change data from various physical attendance terminals and business approval systems in real time, formats it into standardized attendance lifecycle events, and then publishes it to the event distribution module. The event distribution module asynchronously distributes the received standardized attendance events to the conflict state buffer module; the first buffer set updates the target parameter value of the corresponding employee on the assessment date according to the target parameter change carried in the physical attendance event. The second cache set synchronously stores and maintains the mapping relationship between the corresponding approval form status and the activity time period identifier based on the received approval workflow status events; The conflict resolution module performs cross-matching queries on the first cache set and the second cache set. If it is determined that an employee has both an abnormal physical attendance record and a business approval form in progress during a specific assessment period, the conflict resolution judgment logic is triggered. If the event is of low ambiguity, low urgency, or originates from a single data source, the conflict resolution module will directly adjudicate and output the result based on a preset conflict decision rule matrix. If multiple data source conflicts, high urgency, or strong ambiguity occur in the event data stream, the conflict resolution module performs state decision resolution after calculating the quantitative confidence score.

Images