Enterprise system integration method based on micro-service architecture

Abstract

The invention relates to the technical field of enterprise system integration, and discloses a micro-service architecture-based enterprise system integration method, which comprises the following steps of: executing an overtime office degree evaluation strategy, calculating terminal indexes by combining click times and an office interface, and identifying two types of spatio-temporal behaviors of employees for calibrating weights; performing weighted average on the terminal indexes to obtain an overtime office degree; executing a behavior rationality evaluation strategy, comparing the position trajectory with the reference path to evaluate the trajectory deviation degree, executing a consumption authenticity evaluation strategy, evaluating the authenticity of the consumption amount according to the per capita meal fee, and analyzing a card swiping interval to identify a card swiping behavior; executing a pseudo overtime behavior scoring strategy, identifying a pseudo overtime behavior according to the position track of leaving the restaurant, identifying a secondary overtime behavior, and calculating an overtime office degree for calculating a behavior score; and executing a task checking overtime identification strategy, and identifying a pseudo overtime behavior according to the task demand score and the behavior score, thereby improving the fairness and transparency of enterprise welfare issuing.

Description

Technical Field

[0001] This invention relates to the field of enterprise system integration technology, specifically to an enterprise system integration method based on a microservice architecture. Background Technology

[0002] To ensure employees' basic living needs during overtime work, companies generally provide benefits such as meals at the company cafeteria and transportation allowances. These measures not only provide necessary energy replenishment during overtime but also alleviate employees' stress. By providing comprehensive overtime support services, companies can also improve employee work experience and enhance the overall humanistic care of the company.

[0003] Existing systems for overtime pay and meal expense management still have significant shortcomings. First, most companies rely solely on clock-in records or self-reported overtime hours to determine overtime behavior, making it difficult to accurately distinguish between genuine and perfunctory overtime, and increasing the likelihood of fraudulent overtime claims. Second, existing systems cannot comprehensively track employee work status during overtime, such as whether employees are truly engaged in work, remain at their workstations, and are working according to the actual task progress, making it difficult for managers to assess the effectiveness of overtime. Third, regarding meal expense payments, existing systems cannot verify the authenticity of spending; employees may use proxy swiping or have restaurant staff assist with card swiping to claim subsidies, or even swipe very small amounts (such as 0.1 yuan) to be recorded, increasing resource waste and management complexity. Finally, existing methods lack comprehensive evaluation and rationality analysis of employee behavior, failing to consider multiple dimensions of information such as work behavior, overtime hours, and meal expenses, resulting in a lack of scientific basis for overtime management and making it difficult to implement fair performance appraisals.

[0004] This solution proposes an enterprise system integration method based on a microservice architecture. Summary of the Invention

[0005] This invention provides an enterprise system integration method based on a microservice architecture, which helps to solve the problems mentioned in the background art.

[0006] This invention provides the following technical solution: an enterprise system integration method based on a microservice architecture, comprising:

[0007] The testing period is set after the company's closing time, and testing takes place during that period.

[0008] For any given employee, locate the employee's position.

[0009] Record the number of keystrokes and mouse clicks on employees' office computers and identify the office interface;

[0010] Implement an overtime work evaluation strategy, combine click count and office interface to calculate terminal indicators, identify two types of employee spatiotemporal behaviors for weighting, and calculate the overtime work score by weighted average of terminal indicators.

[0011] Set an office activity threshold;

[0012] Obtain employee attendance records for the testing period, filter out employees who did not attend during the testing period, and classify employees whose overtime work level is lower than the work level threshold as a category of employees.

[0013] For one type of employee, real-time location detection generates location trajectories.

[0014] The implementation of a behavior rationality assessment strategy involves comparing the location trajectory with a reference path to assess the degree of trajectory deviation, and calculating the speed of adjacent locations based on the location trajectory to assess the degree of speed deviation.

[0015] The reasonableness of the behavior is calculated based on the degree of deviation from the trajectory and the degree of deviation from the speed. If the reasonableness is less than the set reasonableness threshold, the employees of Category I are marked as Category II employees.

[0016] When Category 2 employees dine at the restaurant, record the amount spent.

[0017] Implement a consumption authenticity assessment strategy, assess the authenticity of consumption amount based on the average meal cost per person, and combine location matching verification to analyze card swipe intervals to identify proxy swipe behavior;

[0018] Based on the strategy for assessing the authenticity of consumption, employees in category two were reclassified as employees in category three.

[0019] Implement a pseudo-overtime behavior scoring strategy, identify pseudo-overtime behavior based on the location trajectory of leaving the restaurant, identify secondary overtime behavior, and recalculate the overtime work degree for the purpose of calculating the behavior score;

[0020] The system implements a task verification and overtime identification strategy, obtains pending tasks and project durations to calculate task requirement scores, and identifies pseudo-overtime behavior based on task requirement scores and behavior scores.

[0021] Optionally, the overtime work evaluation strategy combines click counts and office interface data to calculate terminal metrics, identifies two types of employee spatiotemporal behaviors for weighting, and calculates the overtime work level by weighted averaging of terminal metrics, including:

[0022] Set up whitelists and blacklists for office-related websites;

[0023] If the office interface is on the whitelist, then set the interface relevance. If the office interface is on the blacklist, then If the office interface is not on the whitelist or blacklist, then ;

[0024] Obtain the relevance of all interfaces viewed by employees during the testing period, and calculate the mean, denoted as . ;

[0025] Get the number of keyboard and mouse clicks Set a maximum number of clicks. Map click counts to , , , This is the sensitivity coefficient;

[0026] The average number of interface relevance, keyboard clicks, and mouse clicks is recorded as the terminal metric. ;

[0027] Calculate collaborative terminal indicators This indicates an additional terminal metric added when the office interface is on the whitelist and the number of clicks increases. For the synergistic gain coefficient;

[0028] Correcting terminal indicators ;

[0029] The two types of spatiotemporal behaviors are being in the current employee position and leaving the current employee position, respectively.

[0030] Record the duration of each employee's time at and away from their workstation, and then divide each instance by the monitoring period to obtain a weight. Calculate employees' overtime work hours , These are basic terminal indicators.

[0031] Optionally, the reasonableness assessment strategy for the execution behavior compares the location trajectory with a reference path to assess the degree of trajectory deviation, and calculates the speed of adjacent positions based on the location trajectory to assess the degree of speed deviation, including:

[0032] A sampling interval is set, and the location points of employees are collected at each sampling interval. The reference path is obtained by fitting the location trajectories of multiple employees from the same office to the restaurant.

[0033] When the location trajectory does not pass through the designated restaurant area, set ;

[0034] When the location trajectory passes through the designated restaurant area, the degree of trajectory deviation is calculated, specifically as follows:

[0035] Get the length of the location trajectory of a type of employee Get the length of the reference path Calculate the distance deviation term , To prevent constants with a denominator of 0;

[0036] Calculate the position trajectory and the reference path at two position points at the same sampling time, calculate the interval between the two position points, calculate the standard deviation of the interval, and normalize the standard deviation to obtain the shape deviation term. ;

[0037] Assess the degree of trajectory deviation , As weight;

[0038] Obtain the displacement of adjacent points, and calculate the velocity of adjacent points by removing bits and using the sampling interval;

[0039] Obtain all speeds along the reference path and calculate the average speed. and variance ;

[0040] The mean velocity value at the position of the track is obtained. Consistent computation speed ;

[0041] Set a speed limit, determine if there is a return trajectory in the location trajectory after leaving the company, and if there is a return trajectory, count the number of speeds exceeding the speed limit in the return trajectory;

[0042] The proportion of burst speed is obtained by statistically analyzing and dividing the number by the total number of location points. ;

[0043] Assess the degree of speed deviation , is the scale factor.

[0044] Optionally, the calculation of the behavior's reasonableness based on the degree of trajectory deviation and the degree of speed deviation includes:

[0045] Calculate the reasonableness of behavior ,in, These are the weighting coefficients. The interaction coefficient represents the degree of behavioral rationality, indicating that the greater the deviation in trajectory and speed, the higher the probability of employees faking overtime.

[0046] Set half-death point Mapping the rationality of behavior to of Specifically:

[0047] , Used to control the steepness of the curve, when hour, The probability of employees faking overtime is low. hour, The probability of employees faking overtime is high.

[0048] Optionally, the implementation of the consumption authenticity assessment strategy, which assesses the authenticity of the consumption amount based on the average meal cost per person and combines location matching verification to analyze the card swiping interval to identify proxy swiping behavior, includes:

[0049] Obtain the restaurant's daily spending records, calculate the average, and record it as the average meal cost per person. ;

[0050] Obtain the consumption amount of Category II employees Set a benchmark consumption amount ;

[0051] like Then set the amount anomaly level. ;

[0052] like , ,otherwise ;

[0053] Get the prices of all dishes in the restaurant Calculate the price matching anomaly. ;

[0054] Get the employee to whom the current consumption record belongs, and get the location of the employee;

[0055] If the location is outside the restaurant area, retrieve the consumption records before this consumption record, calculate the time interval between adjacent consumption records, and set an interval threshold;

[0056] If the time interval is less than the interval threshold;

[0057] If the transaction is deemed to be a proxy transaction, then multiple subsequent transaction records will be obtained, and the time interval between adjacent transaction records will be calculated.

[0058] If the time interval between consecutive adjacent consumption records is less than the interval threshold, multiple consumption records after this consumption record will be identified as proxy spending.

[0059] Employees whose multiple spending records are linked to the proxy spending activity will not be eligible for overtime meal allowances.

[0060] Optionally, the implementation of the consumption authenticity assessment strategy, which assesses the authenticity of the consumption amount based on the average meal cost per person and combines it with location matching verification, further includes:

[0061] If the location is within the restaurant area, the location of this transaction is normal, but the amount of the transaction is suspicious.

[0062] Normalize the amount anomaly and price matching anomaly and update them to normalized values;

[0063] Calculation of authenticity ,in, This is used to indicate a reduction in reasonableness when both the amount and price are abnormal. This indicates that the denominator increases with the degree of outlier, and is used to reduce the likelihood of a true value. The weighting coefficients are set.

[0064] Set a truth threshold ,like If so, then the second-class employees will be marked as third-class employees.

[0065] Optionally, the implementation of the pseudo-overtime behavior scoring strategy, which identifies pseudo-overtime behavior based on the location trajectory of leaving the restaurant, and obtains the overtime work level again within a set detection period for calculating the behavior score, includes:

[0066] Obtain the location trajectories of three types of employees leaving the restaurant and detect the destination of the location trajectories;

[0067] If the destination is not the office and the clock-out message is displayed, then the three types of employees will be marked as pseudo-overtime.

[0068] If the destination is the office, the period from arrival at the office to clocking out is considered a new detection period, during which the overtime work rate of employees who are falsely claiming to be working overtime is recalculated. ;

[0069] Calculate behavioral scores , For mixing ratio;

[0070] Used to represent the impact of the first overtime work on the behavior score. This indicates the impact of the second overtime work on the behavior score.

[0071] Optionally, the execution task verification overtime identification strategy obtains pending tasks and project durations for calculating task requirement scores, and identifies pseudo-overtime behavior based on task requirement scores and behavior scores, including:

[0072] The tasks are categorized into completed tasks. and tasks to be done ;

[0073] Get the project duration for this task, divided into completed duration. and time to completion ;

[0074] Calculate the completion rate of completed and pending tasks. and The calculation formula is as follows:

[0075] , ;

[0076] Set rate ratio ,like Calculate task requirement score ;

[0077] Calculate the probability of pseudo-overtime , As weight;

[0078] Calculate the synergistic amplification effect (1- This indicates that if both the behavior score and the task requirement score decrease simultaneously, the probability of fake overtime increases.

[0079] Set a threshold for the probability of fake overtime ;

[0080] like If so, then the three types of employees are identified as those who are falsely working overtime;

[0081] Employees marked as "fake overtime" will be prompted to upload proof of overtime work. If employees who are "fake overtime" fail to upload proof, they will not receive overtime meal allowances.

[0082] The present invention has the following beneficial effects:

[0083] 1. This enterprise system integration method based on a microservice architecture analyzes multi-dimensional data, including employee location, work behavior, overtime hours, and meal expenses, during monitoring periods after employees leave work. This enables a comprehensive assessment of employee overtime behavior. By locating employee positions and monitoring computer operations (such as keyboard and mouse activity and the work interface), it accurately distinguishes whether employees are truly engaged in work, avoiding the one-sided judgments based solely on clocking in or self-reporting overtime hours. Combining work engagement thresholds and behavior rationality thresholds, employees are categorized into different levels, constructing a hierarchical identification system to effectively identify potential fake overtime behavior. The introduction of meal expense records and consumption authenticity assessment enables linked management of work behavior and subsidy distribution, ensuring fair and reasonable resource allocation. Overall, this method not only improves the transparency and management efficiency of overtime behavior within enterprises but also reduces resource waste caused by insufficient manual review or information silos. It provides a scientific basis and quantitative tools for enterprise overtime management, reducing unnecessary expenditures and management costs while ensuring normal overtime benefits for employees.

[0084] 2. This enterprise system integration method based on a microservice architecture quantifies employees' keyboard and mouse operations and office interface access behavior during the monitoring period to generate terminal indicators. These indicators are then weighted and calculated based on the time employees spend at and away from their workstations to determine overtime work intensity, enabling dynamic evaluation of employees' actual work status. Whitelist and blacklist mechanisms distinguish between valid office interfaces and non-office behaviors, while synergistic gain indicators further improve the accuracy of identifying high-intensity work behaviors. By combining weighted calculations with basic indicators, the method comprehensively reflects employee work engagement, avoiding biases caused by solely relying on login time or attendance records. The advantage of this method is that it transforms office behavior into quantifiable indicators, allowing enterprise managers to objectively and precisely grasp employees' overtime status, thereby scientifically allocating overtime pay, improving management fairness, and effectively reducing resource waste caused by fake overtime or low-intensity operations, achieving quantitative and traceable management of employee work behavior.

[0085] 3. This enterprise system integration method based on a microservice architecture collects employee location trajectories and compares them with reference paths. Combining trajectory deviation and speed deviation, it calculates the rationality of employee behavior, achieving a dynamic quantitative evaluation of the authenticity of overtime work. It can identify abnormal paths, sudden speed changes, and return journeys after employees leave their workstations, thus distinguishing between genuine overtime and potential pseudo-overtime. By introducing a behavior rationality mapping function, the degree of deviation is transformed into the probability of pseudo-overtime, achieving a quantitative evaluation from 0 to 1, enabling managers to clearly understand the rationality of each employee's overtime. Combining spatial and temporal behavior provides multi-dimensional judgment criteria, significantly improving the accuracy of pseudo-overtime identification while reducing the possibility of manual intervention and misjudgment. It provides enterprises with a scientific, quantifiable, and operable overtime management tool, optimizing overtime supervision processes and reducing the risk of unreasonable compensation payments.

[0086] 4. This enterprise system integration method based on a microservice architecture determines whether consumption behavior is abnormal or involves proxy spending by comparing average meal costs per person, matching menu prices, and verifying location. By calculating the degree of consumption anomaly and identifying proxy spending, it can detect violations such as employees colluding with the restaurant to swipe cards and unusually small transactions, ensuring the fairness and authenticity of overtime meal allowances. Furthermore, through normalization and a comprehensive scoring method, it combines the degree of amount anomaly with the degree of price anomaly to generate an overtime meal allowance authenticity score, providing managers with an intuitive and quantifiable reference. Through the linked analysis of behavioral and consumption data, it comprehensively improves the scientific nature and controllability of meal allowance management, reduces false reimbursements and resource waste, enhances the fairness and transparency of corporate welfare distribution, and strengthens the binding force of the system's implementation.

[0087] 5. This enterprise system integration method based on a microservice architecture analyzes the behavior of employees in three stages. First, it assesses the reasonableness of the behavior based on the employee's trajectory after leaving the restaurant and their return journey, advancing eligible employees from Category I to Category II. Then, it further determines the category by combining assessments of the authenticity of consumption and recalculation of overtime work, progressively identifying Category II employees as Category III employees. This step-by-step approach enables phased and dynamic analysis of employee overtime behavior, achieving high-precision identification of pseudo-overtime behavior. By acquiring completed tasks, pending tasks, and project duration, it calculates task completion rate and task requirement scores, combining these with behavior scores to construct a pseudo-overtime probability assessment model, achieving a comprehensive judgment on the reasonableness of employee overtime. Through rate ratio and pseudo-overtime probability threshold settings, it can dynamically identify employees with abnormal behavior and low task completion rates, prompting them to upload overtime proof to ensure fair compensation distribution. Attached Figure Description

[0088] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0089] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0090] Example 1, refer to Figure 1 A method for enterprise system integration based on microservice architecture, comprising:

[0091] Under existing corporate overtime pay policies, employees who haven't clocked out by 7 PM and have a meal expense record at the company cafeteria are eligible for an overtime meal allowance. The intention is to encourage overtime and reasonably compensate employees for their evening work. However, due to the relatively simplistic rules, some employees exploit loopholes to obtain the allowance by making only small purchases (e.g., 0.1 yuan) at the cafeteria or having someone else swipe their card to meet the eligibility requirements, without actually working overtime. Employees are misclassified as working overtime by delaying clocking in, staying for short periods, or performing low-intensity tasks, thus receiving undue allowances.

[0092] Employees using their cards to pay for meals in the company cafeteria is an existing technology. Companies use card swipe records to track employee meal expenses and distribute overtime pay.

[0093] For any given employee, locate the employee's position.

[0094] Record the number of keystrokes and mouse clicks on employees' office computers and identify the office interface;

[0095] Implement an overtime work evaluation strategy, combine click count and office interface to calculate terminal indicators, identify two types of employee spatiotemporal behaviors for weighting, and calculate the overtime work score by weighted average of terminal indicators.

[0096] Set up whitelists and blacklists for office-related websites;

[0097] If the office interface is on the whitelist, then set the interface relevance. If the office interface is on the blacklist, then If the office interface is not on the whitelist or blacklist, then ;

[0098] Obtain the relevance of all interfaces viewed by employees during the testing period, and calculate the mean, denoted as . ;

[0099] Get the number of keyboard and mouse clicks Set a maximum number of clicks. Map click counts to , , , This is the sensitivity coefficient;

[0100] The average number of interface relevance, keyboard clicks, and mouse clicks is recorded as the terminal metric. ;

[0101] Calculate collaborative terminal indicators This indicates an additional terminal metric added when the office interface is on the whitelist and the number of clicks increases. For the synergistic gain coefficient;

[0102] Correcting terminal indicators ;

[0103] The two types of spatiotemporal behaviors are being in the current employee position and leaving the current employee position, respectively.

[0104] Record the duration of each employee's time at and away from their workstation, and then divide each instance by the monitoring period to obtain a weight. Calculate employees' overtime work hours , These are basic terminal indicators.

[0105] Set an office activity threshold;

[0106] Obtain employee attendance records for the testing period, filter out employees who did not attend during the testing period, and classify employees whose overtime work level is lower than the work level threshold as a category of employees.

[0107] The purpose of allocating a monitoring period is to dynamically observe employees' behavior after get off work, reflecting their actual overtime commitment by recording keyboard and mouse operations, office interface activity, etc. This reveals employees' true work status and spatial behavior patterns, calculates overtime work intensity, and identifies abnormal behaviors, such as prolonged absence from their posts or low operational intensity, thereby distinguishing between genuine and pseudo-overtime.

[0108] For one type of employee, location is detected in real time to generate location trajectories;

[0109] The implementation of a behavior rationality assessment strategy involves comparing the location trajectory with a reference path to assess the degree of trajectory deviation, and calculating the speed of adjacent locations based on the location trajectory to assess the degree of speed deviation.

[0110] A sampling interval is set, and the location points of employees are collected at each sampling interval. The reference path is obtained by fitting the location trajectories of multiple employees from the same office to the restaurant.

[0111] When the location trajectory does not pass through the designated restaurant area, set ;

[0112] When the location trajectory passes through the designated restaurant area, the degree of trajectory deviation is calculated, specifically as follows:

[0113] Get the length of the location trajectory of a type of employee Get the length of the reference path Calculate the distance deviation term , To prevent constants with a denominator of 0;

[0114] Calculate the position trajectory and the reference path at two position points at the same sampling time, calculate the interval, then calculate the standard deviation of the interval, and normalize the standard deviation to obtain the shape deviation term. ;

[0115] Assess the degree of trajectory deviation , As weight;

[0116] Obtain the displacement of adjacent points, and calculate the velocity of adjacent points by removing bits and using the sampling interval;

[0117] Obtain all speeds along the reference path and calculate the average speed. and variance ;

[0118] The mean velocity value at the position of the track is obtained. Consistent computation speed ;

[0119] Set a speed limit, determine if there is a return trajectory in the location trajectory after leaving the company, and if there is a return trajectory, count the number of speeds exceeding the speed limit in the return trajectory;

[0120] The proportion of burst speed is obtained by statistically analyzing and dividing the number by the total number of location points. ;

[0121] Assess the degree of speed deviation , is the scale factor.

[0122] The rationality of the behavior is calculated based on the degree of trajectory deviation and the degree of speed deviation.

[0123] Calculate the reasonableness of behavior ,in, These are the weighting coefficients. The interaction coefficient represents the degree of behavioral rationality, indicating that the greater the deviation in trajectory and speed, the higher the probability of employees faking overtime.

[0124] Set half-death point Mapping the rationality of behavior to of Specifically:

[0125] , Used to control the steepness of the curve, when hour, The probability of employees faking overtime is low. hour, The probability of employees faking overtime is high.

[0126] If the reasonableness is less than the set reasonableness threshold, the employees in category one will be marked as employees in category two.

[0127] When Category 2 employees dine at the restaurant, record the amount spent.

[0128] Implement a consumption authenticity assessment strategy, assess the authenticity of consumption amount based on the average meal cost per person, and combine location matching verification to analyze card swipe intervals to identify proxy swipe behavior;

[0129] Obtain the restaurant's daily spending records, calculate the average, and record it as the average meal cost per person. ;

[0130] Obtain the consumption amount of Category II employees Set a benchmark consumption amount ;

[0131] like Then set the amount anomaly level. ;

[0132] like , ,otherwise ;

[0133] Get the prices of all dishes in the restaurant Calculate the price matching anomaly. ;

[0134] Get the employee to whom the current consumption record belongs, and get the location of the employee;

[0135] If the location is outside the restaurant area, retrieve the consumption records before this consumption record, calculate the time interval between adjacent consumption records, and set the interval threshold to 20 seconds;

[0136] If the time interval is less than the interval threshold;

[0137] If the transaction is deemed to be a proxy transaction, then multiple subsequent transaction records will be obtained, and the time interval between adjacent transaction records will be calculated.

[0138] If the time interval between consecutive adjacent consumption records is less than the interval threshold, multiple consumption records after this consumption record will be identified as proxy spending.

[0139] In real-world business settings, employees may use other people's meal cards to dine, meaning the employee swiping the card is not the actual diner. Directly linking the transaction record to the employee swiping the card could lead to misjudgments about their overtime work or the authenticity of meal expenses.

[0140] Based on experience: Regular employees typically swipe their cards 5-10 minutes apart during meal service (from picking up their food to paying and leaving); proxy swipes, however, can be completed in quick succession within seconds to tens of seconds. Normal meal intervals are far greater than the threshold and will not be mistakenly identified as proxy swipes. Short, consecutive swipes are sufficient to identify proxy swipes, eliminating the need to link each transaction to an employee or additionally check the employee responsible for each transaction, allowing for rapid identification of potential proxy swipe behavior.

[0141] Employees whose multiple spending records are linked to the proxy spending activity will not be eligible for overtime meal allowances.

[0142] If the location is within the restaurant area, the location of this transaction is normal, but the amount of the transaction is suspicious.

[0143] Normalize the amount anomaly and price matching anomaly and update them to normalized values;

[0144] Calculation of authenticity ,in, This is used to indicate a reduction in reasonableness when both the amount and price are abnormal. This indicates that the denominator increases with the degree of outlier, and is used to reduce the likelihood of a true value. The weighting coefficients are set.

[0145] Set a truth threshold ,like If so, then the second-class employees will be marked as third-class employees.

[0146] Based on the strategy for assessing the authenticity of consumption, employees in category two were reclassified as employees in category three.

[0147] In this embodiment, if an employee leaves work after 7:00 PM, the system records that they are still in the office. However, in reality, they are just sitting at their workstation browsing the web, occasionally typing on the keyboard, or spending only 0.1 yuan at the cafeteria in an attempt to claim overtime pay. Other employees leave the office briefly after get off work, then return and clock in again, seemingly having completed overtime. These fake overtime behaviors are difficult to identify based solely on clock-in or meal expense records.

[0148] This solution dynamically tracks employees' keyboard and mouse operations and office interface usage during monitoring periods to calculate overtime work intensity. It then assesses the reasonableness of these behaviors by comparing location trajectories with reference paths and identifies fraudulent or abnormal spending through consumption authenticity analysis. This includes identifying fake overtime, low-effort overtime, and fraudulent meal allowance payments.

[0149] Implement a pseudo-overtime behavior scoring strategy, identify pseudo-overtime behavior based on the location trajectory of leaving the restaurant, identify secondary overtime behavior, and recalculate the overtime work degree for the purpose of calculating the behavior score;

[0150] Obtain the location trajectories of three types of employees leaving the restaurant and detect the destination of the location trajectories;

[0151] If the destination is not the office and the clock-out message is displayed, then the three types of employees will be marked as pseudo-overtime.

[0152] If the destination is the office, the period from arrival at the office to clocking out is considered a new detection period, during which the overtime work rate of employees who are falsely claiming to be working overtime is recalculated. ;

[0153] Calculate behavioral scores , For mixing ratio;

[0154] Used to represent the impact of the first overtime work on the behavior score. This indicates the impact of the second overtime work on the behavior score.

[0155] The final behavioral score is calculated by integrating all scores because employee overtime behavior is dynamic and multi-stage, and indicators from a single stage cannot fully reflect its true nature. The behavioral score is not an absolute judgment, but rather the result of a comprehensive system evaluation. By quantifying indicators from multiple stages, it reduces the misleading influence of single-point anomalies or occasional events on the judgment.

[0156] In this example, the employee operated a computer in the office from 19:00 to 19:30, but the intensity of operation was very low, and the amount of card spending in the cafeteria was unusually low. He then returned to the office from 20:00 to 20:15, operated the computer briefly, and then clocked out. If only the first instance of overtime work is considered, it might be mistaken for normal overtime work; if only the second short-term operation is considered, the behavioral problems might be underestimated.

[0157] By integrating scores for two overtime work sessions, scores for reasonable behavior, scores for genuine consumption, and scores for the second overtime work session, the system comprehensively quantifies employees' input and abnormal behavior throughout the entire testing period, resulting in a more scientific and objective behavioral score.

[0158] The strategy for verifying overtime work is implemented by obtaining the duration of pending tasks and projects to calculate task requirement scores and identifying pseudo-overtime behavior based on task requirement scores and behavior scores.

[0159] The tasks are categorized into completed tasks. and tasks to be done ;

[0160] Get the project duration for this task, divided into completed duration. and time to completion ;

[0161] Calculate the completion rate of completed and pending tasks. and The calculation formula is as follows:

[0162] , ;

[0163] Set rate ratio ,like Calculate task requirement score ;

[0164] Calculate the probability of pseudo-overtime , As weight;

[0165] Calculate the synergistic amplification effect (1- This indicates that if both the behavior score and the task requirement score decrease simultaneously, the probability of fake overtime increases.

[0166] Set a threshold for the probability of fake overtime ;like If so, then the three types of employees are identified as those who are falsely working overtime;

[0167] Introducing task demand scoring closely links employee overtime behavior with actual work tasks. By calculating the completion rate of completed tasks and pending tasks, the rationality and necessity of employee overtime can be assessed.

[0168] Employees identified as engaging in fraudulent overtime work are prompted to upload proof of their overtime work. If they fail to do so, overtime meal allowances will not be provided, but a remedial opportunity is offered: employees are prompted to upload proof of their overtime work to justify the work. If they can provide valid proof, the determination can be revised, and they will still receive the corresponding overtime meal allowance; if they fail to upload proof or provide insufficient evidence, the allowance will not be issued. This approach not only improves the fairness of identifying fraudulent overtime work but also respects employees' rights to appeal and correct errors, ensuring more reasonable and transparent overtime allowance distribution, while enhancing the flexibility and trustworthiness of corporate management.

[0169] To verify the effectiveness of this solution in identifying overtime work and managing meal allowances, an experimental comparison was conducted. The results showed that, with a fixed daily meal allowance of 14 yuan, relying solely on clock-in time or restaurant card swipe records resulted in approximately 32% of employees falsely claiming overtime or having their meal allowances swiped on their behalf. After adopting this solution's comprehensive scoring system, which considers overtime work frequency, behavioral reasonableness, and the authenticity of consumption, the false overtime work identification rate increased to 94%, the meal allowance distribution accuracy reached 96%, and cases of swiping on behalf of others or abnormal claims were significantly reduced.

[0170] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0171] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for enterprise system integration based on microservice architecture, characterized in that: include: The testing period is set after the company's closing time, and testing takes place during that period. For any given employee, locate the employee's position. Record the number of keystrokes and mouse clicks on employees' office computers and identify the office interface; Implement an overtime work evaluation strategy, combine click count and office interface to calculate terminal indicators, identify two types of employee spatiotemporal behaviors for weighting, and calculate the overtime work score by weighted average of terminal indicators. Set an office activity threshold; Obtain employee attendance records for the testing period, filter out employees who did not attend during the testing period, and classify employees whose overtime work level is lower than the work level threshold as a category of employees. For one type of employee, real-time location detection generates location trajectories. The implementation of a behavior rationality assessment strategy involves comparing the location trajectory with a reference path to assess the degree of trajectory deviation, and calculating the speed of adjacent locations based on the location trajectory to assess the degree of speed deviation. The reasonableness of the behavior is calculated based on the degree of deviation from the trajectory and the degree of deviation from the speed. If the reasonableness is less than the set reasonableness threshold, the employees of Category I are marked as Category II employees. When Category 2 employees dine at the restaurant, record the amount spent. Implement a consumption authenticity assessment strategy, assess the authenticity of consumption amount based on the average meal cost per person, and combine location matching verification to analyze card swipe intervals to identify proxy swipe behavior; Based on the strategy for assessing the authenticity of consumption, employees in category two were reclassified as employees in category three. Implement a pseudo-overtime behavior scoring strategy, identify pseudo-overtime behavior based on the location trajectory of leaving the restaurant, identify secondary overtime behavior, and recalculate the overtime work degree for the purpose of calculating the behavior score; The system implements a task verification and overtime identification strategy, obtains pending tasks and project durations to calculate task requirement scores, and identifies pseudo-overtime behavior based on task requirement scores and behavior scores.

2. The enterprise system integration method based on microservice architecture according to claim 1, characterized in that: The overtime work evaluation strategy combines click counts and office interface metrics to calculate terminal indicators, identifies two types of employee spatiotemporal behaviors for weighting, and calculates the overtime work level by weighted averaging of terminal indicators, including: Set up whitelists and blacklists for office-related websites; If the office interface is on the whitelist, then set the interface relevance. If the office interface is on the blacklist, then If the office interface is not on the whitelist or blacklist, then ; Obtain the relevance of all interfaces viewed by employees during the testing period, and calculate the mean, denoted as . ; Get the number of keyboard and mouse clicks Set a maximum number of clicks. Map click counts to , , , This is the sensitivity coefficient; The average number of interface relevance, keyboard clicks, and mouse clicks is recorded as the terminal metric. ; Calculate collaborative terminal indicators This indicates an additional terminal metric added when the office interface is on the whitelist and the number of clicks increases. For the synergistic gain coefficient; Correcting terminal indicators ; The two types of spatiotemporal behaviors are being in the current employee position and leaving the current employee position, respectively. Record the duration of each employee's time at and away from their workstation, and then divide each instance by the monitoring period to obtain a weight. Calculate employees' overtime work hours , These are basic terminal indicators.

3. The enterprise system integration method based on microservice architecture according to claim 2, characterized in that: The rationality assessment strategy for the execution behavior compares the location trajectory with a reference path to assess the degree of trajectory deviation, and calculates the speed of adjacent positions based on the location trajectory to assess the degree of speed deviation, including: A sampling interval is set, and the location points of employees are collected at each sampling interval. The reference path is obtained by fitting the location trajectories of multiple employees from the same office to the restaurant. When the location trajectory does not pass through the designated restaurant area, set ; When the location trajectory passes through the designated restaurant area, the degree of trajectory deviation is calculated, specifically as follows: Get the length of the location trajectory of a type of employee Get the length of the reference path Calculate the distance deviation term , To prevent constants with a denominator of 0; Calculate the position trajectory and the reference path at two position points at the same sampling time, calculate the interval between the two position points, calculate the standard deviation of the interval, and normalize the standard deviation to obtain the shape deviation term. ; Assess the degree of trajectory deviation , As weight; Obtain the displacement of adjacent points, and calculate the velocity of adjacent points by removing bits and using the sampling interval; Obtain all speeds along the reference path and calculate the average speed. and variance ; The mean velocity value at the position of the track is obtained. Consistent computation speed ; Set a speed limit, determine if there is a return trajectory in the location trajectory after leaving the company, and if there is a return trajectory, count the number of speeds exceeding the speed limit in the return trajectory; The proportion of burst speed is obtained by statistically analyzing and dividing the number by the total number of location points. ; Assess the degree of speed deviation , is the scale factor.

4. The enterprise system integration method based on microservice architecture according to claim 3, characterized in that: The calculation of behavioral reasonableness based on the degree of trajectory deviation and speed deviation includes: Calculate the reasonableness of behavior ,in, These are the weighting coefficients. The interaction coefficient represents the degree of behavioral rationality, indicating that the greater the deviation in trajectory and speed, the higher the probability of employees faking overtime. Set half-death point Mapping the rationality of behavior to of Specifically: , Used to control the steepness of the curve, when hour, The probability of employees faking overtime is low. hour, The probability of employees faking overtime is high.

5. The enterprise system integration method based on microservice architecture according to claim 4, characterized in that: The aforementioned consumption authenticity assessment strategy evaluates the authenticity of consumption amounts based on the average meal cost per person, and combines location matching verification to analyze card swipe intervals to identify proxy swipe behavior, including: Obtain the restaurant's daily spending records, calculate the average, and record it as the average meal cost per person. ; Obtain the consumption amount of Category II employees Set a benchmark consumption amount ; like Then set the amount anomaly level. ; like , ,otherwise ; Get the prices of all dishes in the restaurant Calculate the price matching anomaly. ; Get the employee to whom the current consumption record belongs, and get the location of the employee; If the location is outside the restaurant area, retrieve the consumption records before this consumption record, calculate the time interval between adjacent consumption records, and set an interval threshold; If the time interval is less than the interval threshold; If the transaction is deemed to be a proxy transaction, then multiple subsequent transaction records will be obtained, and the time interval between adjacent transaction records will be calculated. If the time interval between consecutive adjacent consumption records is less than the interval threshold, multiple consumption records after this consumption record will be identified as proxy spending. Employees whose multiple spending records are linked to the proxy spending activity will not be eligible for overtime meal allowances.

6. The enterprise system integration method based on microservice architecture according to claim 5, characterized in that: The strategy for assessing the authenticity of consumption, which evaluates the authenticity of consumption amount based on the average meal cost per person and verifies it in conjunction with location matching, also includes: If the location is within the restaurant area, the location of this transaction is normal, but the amount of the transaction is suspicious. Normalize the amount anomaly and price matching anomaly and update them to normalized values; Calculation of authenticity ,in, This is used to indicate a reduction in reasonableness when both the amount and price are abnormal. This indicates that the denominator increases with the degree of outlier, and is used to reduce the likelihood of a true value. The weighting coefficients are set. Set a truth threshold ,like If so, then the second-class employees will be marked as third-class employees.

7. The enterprise system integration method based on microservice architecture according to claim 6, characterized in that: The pseudo-overtime behavior scoring strategy identifies pseudo-overtime behavior based on the location trajectory of leaving the restaurant, and obtains the overtime work level again within a set detection period to calculate the behavior score, including: Obtain the location trajectories of three types of employees leaving the restaurant and detect the destination of the location trajectories; If the destination is not the office and the clock-out message is displayed, then the three types of employees will be marked as pseudo-overtime. If the destination is the office, the period from arrival at the office to clocking out is considered a new detection period, during which the overtime work rate of employees who are falsely claiming to be working overtime is recalculated. ; Calculate behavioral scores , For mixing ratio; Used to represent the impact of the first overtime work on the behavior score. This indicates the impact of the second overtime work on the behavior score.

8. The enterprise system integration method based on microservice architecture according to claim 7, characterized in that: The overtime identification strategy for task verification involves obtaining pending tasks and project durations to calculate task requirement scores, and identifying pseudo-overtime behavior based on task requirement scores and behavior scores, including: The tasks are categorized into completed tasks. and tasks to be done ; Get the project duration for this task, divided into completed duration. and time to completion ; Calculate the completion rate of completed and pending tasks. and The calculation formula is as follows: ， ； Set rate ratio ,like Calculate task requirement score ; Calculate the probability of pseudo-overtime , As weight; Calculate the synergistic amplification effect (1- This indicates that if both the behavior score and the task requirement score decrease simultaneously, the probability of fake overtime increases. Set a threshold for the probability of pseudo-overtime ; like If so, then the three types of employees are identified as those who are falsely working overtime; Employees marked as "fake overtime" will be prompted to upload proof of overtime work. If employees who are "fake overtime" fail to upload proof, they will not receive overtime meal allowances.

Images