Employee assessment management method and system based on cloud platform

Abstract

The invention discloses an employee assessment management method and system based on a cloud platform, and particularly relates to the technical field of human resource management.The method comprises the steps that attendance coordinate data, customer evaluation star-level scores and sales order completion quantity are collected in real time through a real-name account system, and a multi-source work data set is generated through abnormal value filtering and sliding window mean value processing; dynamically adjusting the ratio of the customer evaluation weight to the attendance weight, and constructing a five-level achievement step based on extreme value difference equal-ratio division; extracting historical threshold data according to the number of times of achievement unlocking to generate a calibration evaluation benchmark, and verifying and triggering a cross-end game reward instruction in combination with a double-end timestamp difference value; and aligning a re-judgment achievement condition through a timestamp, and sending a virtual asset increment update packet to realize cross-end state synchronization. According to the method, the dynamic adaptive capacity of an evaluation system to service fluctuation is improved, the association accuracy of achievement excitation and work performance is enhanced, the problem of misjudgment caused by asynchronous transmission of data is solved, and the objectivity of an evaluation result and the fault tolerance of the system are ensured.

Description

Technical Field

[0001] This invention relates to the field of human resource management technology, and in particular to an employee performance evaluation management method and system based on a cloud platform. Background Technology

[0002] The integrated system of work performance and gamified incentives refers to establishing digital identities for employees through a real-name account system, using attendance recorders, customer evaluation systems, and sales data interfaces as data collection points, and transmitting standardized data to a cloud storage cluster via HTTPS protocol. A cloud-deployed rule engine performs weighted calculations on indicators such as attendance compliance rate, evaluation scores, and sales completion rate. The calculation results are pushed to a visual interface on mobile terminals via WebSocket protocol, simultaneously triggering the virtual reward distribution mechanism in the game client.

[0003] Existing technologies employ a fixed-weight calculation model, failing to dynamically adjust indicator weights based on business fluctuations. When sales orders experience sustained growth, the evaluation system struggles to reflect changes in actual work quality, leading to delayed assessment results. The rule engine relies on static thresholds to determine achievement, lacking a dynamic interval division mechanism based on historical data fluctuations. This can cause imbalances in achievement acquisition difficulty when overall departmental performance improves, weakening incentive effects. The one-way data push mechanism from the cloud to the client lacks dual-end timestamp verification. Clock discrepancies between mobile and game terminals can cause misjudgments in achievement triggering, resulting in incorrect virtual asset distribution. Abnormal data handling relies solely on simple threshold filtering, without incorporating sliding window mean calculations; occasional outliers can still affect the stability of assessment results. Historical data is used only for storage and lacks a linkage calibration mechanism with the current assessment benchmark, potentially leading to accumulated assessment biases over time. Existing data synchronization mechanisms lack a game reward re-judgment process; when network latency causes data version inconsistencies, directly using the initial judgment result can cause cross-platform inconsistency issues. Summary of the Invention

[0004] The main objective of this invention is to provide a cloud-based employee performance evaluation management method and system. By integrating multi-source data, calculating dynamic performance coefficients, constructing tiered achievement ladders, verifying cross-platform game rewards, and providing virtual incentive feedback, the system improves the fairness, real-time performance, and incentive effectiveness of performance evaluation management.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A cloud-based employee performance evaluation management method and system, with the following specific steps:

[0007] Step 1: Collect employee attendance and clock-in coordinate data, customer rating star ratings, and original records of sales order completion volume using real names. Calculate the standard deviation of each indicator within a set period. Mark data that deviates from the department average by a set multiple as outliers. Take the sliding window mean of the valid data to generate a multi-source working dataset.

[0008] Step 2: Call the multi-source working dataset, detect the period when the number of completed sales orders continuously exceeds the department's average for a set duration, increase the weight of customer rating star rating to a set multiple of the base value, reduce the weight of attendance coordinate data to a set ratio, calculate the weighted comprehensive value, and generate a dynamic performance coefficient;

[0009] Step 3: Based on the dynamic performance coefficient, extract the extreme value difference of attendance achievement rate within the most recent set period, divide the extreme value difference by the set level to generate the base of the equal ratio interval, and build a five-level achievement ladder by superimposing the base with the minimum achievement rate as the starting point. When the sales order volume fluctuates beyond the standard deviation set multiple, recalculate the extreme value to generate the dynamic achievement threshold interval.

[0010] Step 4: Count the number of times employees unlock game achievements. When the same type of achievement reaches the set number of times, extract the historical data of the dynamic achievement threshold range associated with the achievement, calculate the standard deviation of sales order volume, and add the current evaluation benchmark to the standard deviation by the set ratio to generate a calibration evaluation benchmark.

[0011] Step 5: Obtain the achievement trigger timestamp and data version identifier from the mobile and game platforms, calculate the absolute value of the timestamp difference, and when the difference exceeds the set fault tolerance threshold or the version identifier is discontinuous, call the calibration evaluation benchmark to re-execute the achievement condition judgment and generate cross-platform game reward instructions.

[0012] Step 6: Detect cross-platform game reward instructions, retrieve the latest multi-source working dataset, re-execute achievement condition judgment, and if the judgment result differs from the initial record, send a virtual asset incremental update package to the game client and push a status change notification to the mobile client to generate a cross-platform status synchronization record.

[0013] Preferably, the method for calculating the mean of the sliding window in step 1 is as follows: using a time series alignment method, the attendance coordinate data is discretized by hourly granularity, customer rating star ratings are matched by order completion timestamps, and the sales order completion volume is aggregated by natural day segments.

[0014] Preferably, the method for increasing the customer evaluation weight in step 2 is as follows: when the sales order exceeds the average for a set number of consecutive days, the weight is increased by 0.1 times daily, with a maximum of 1.5 times the base value.

[0015] Preferably, the method for increasing the customer evaluation weight in step 2 is as follows: when the sales order exceeds the average for a set number of consecutive days, the weight is increased by 0.1 times daily, with a maximum of 1.5 times the base value.

[0016] Preferably, a data verification step is added before step 3: the dynamic performance coefficient is hashed and compared with the latest node of the pre-stored hash chain in the cloud. If the verification fails, the coefficient is reverted to the previous two valid versions of the dataset and regenerated.

[0017] Preferably, the generation of the calibration evaluation benchmark in step 4 is specifically as follows: when the sales standard deviation exceeds the average historical standard deviation of the department, a weighted fusion method is used to replace 0.5 times the standard deviation with the algebraic sum of 0.3 times the standard deviation and 0.2 times the historical mean.

[0018] Preferably, the timestamp difference calculation in step 5 uses the NTP protocol to synchronize dual-end clock sources, with the mobile terminal using the timestamp of the last heartbeat packet and the game terminal using the timestamp of the server receiving confirmation.

[0019] Preferably, the virtual asset incremental update in step 6 includes the virtual currency game reward amount corresponding to the achievement difference value, which is calculated as: difference achievement level number × base currency unit × time decay coefficient, where the time decay coefficient = 1 / (1+0.1×delay days).

[0020] Preferably, the status change notification in step 6 adopts a dual-channel redundant push mechanism, simultaneously sending a Base64 encoded data packet containing an achievement change summary to the mobile message queue and the SMS gateway.

[0021] A cloud-based employee performance evaluation management system, used to execute the above-mentioned employee performance evaluation management method, includes:

[0022] The multi-source data acquisition module is used to collect attendance coordinate data, customer rating star ratings, and original records of sales order completion in real time through the real-name account system, perform standard deviation calculation and outlier marking, and output multi-source working datasets.

[0023] The dynamic weight calculation module calls a multi-source working dataset, detects periods when sales orders continuously exceed the average value, performs operations to increase customer evaluation weight and decrease attendance weight, and outputs dynamic performance coefficients.

[0024] The achievement threshold generation module, based on dynamic performance coefficients, extracts the extreme difference of attendance compliance rate, executes the logic of constructing a five-level achievement ladder and dynamic recalculation, and outputs the dynamic achievement threshold range.

[0025] The benchmark calibration module counts the number of times game achievements are unlocked, extracts historical data of dynamic achievement threshold range, performs sales standard deviation calculation and benchmark update, and outputs calibration evaluation benchmark.

[0026] The cross-platform verification module obtains the timestamps and version identifiers of the mobile and game platforms, performs difference limit judgment and achievement condition re-judgment, and outputs cross-platform game reward instructions.

[0027] The status synchronization execution module detects cross-platform game reward commands, retrieves the latest three-dimensional data to perform timestamp alignment and achievement re-judgment, sends virtual asset update packages and notifications when discrepancies are found, and outputs cross-platform status synchronization records.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] This invention eliminates occasional outlier interference and improves the quality of the basic dataset by using multi-source data sliding window mean processing. It employs a dynamic weight adjustment mechanism to intelligently adjust the weighting of evaluation and attendance based on periods of sustained sales order outperformance, enhancing the adaptability of performance assessment scenarios. When constructing a five-level dynamic achievement ladder, it introduces an extreme value difference proportional division method, combined with a threshold recalculation mechanism triggered by the standard deviation of sales fluctuations, to achieve dynamic matching between the achievement system and work status. It establishes a calibration evaluation benchmark based on the number of achievement unlocks, updating the benchmark value through historical data standard deviation fusion, reducing the cumulative deviation of evaluation results. It designs a cross-platform timestamp difference verification and game reward instruction generation mechanism, combined with three-dimensional data timestamp alignment and re-judgment logic, effectively solving the problem of achievement misjudgment caused by asynchronous data between the two platforms. Virtual asset incremental update packages and dual-channel notifications are executed collaboratively to ensure the real-time nature of status changes and the accuracy of game rewards. The entire processing logic forms a closed-loop feedback system through dynamic parameter adjustment, multi-condition trigger recalculation, and cross-platform data collaborative verification, improving system fault tolerance and response accuracy while ensuring the objectivity of the evaluation. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the workflow of the present invention. Detailed Implementation

[0031] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0032] As shown in the figure, a cloud-based employee performance management method includes the following steps:

[0033] Step 1: Collect employee attendance and clock-in coordinate data, customer rating star ratings, and original records of sales order completion through the real-name account system. Calculate the standard deviation of each indicator within the set period, mark data that deviates from the department average by a set multiple as outliers, and take the sliding window mean of the valid data to generate a multi-source working dataset.

[0034] Step 2: Call the multi-source working dataset, detect the period when the number of completed sales orders continuously exceeds the department's average for a set duration, increase the weight of customer rating star rating to a set multiple of the base value, reduce the weight of attendance coordinate data to a set ratio, calculate the weighted comprehensive value, and generate a dynamic performance coefficient;

[0035] Step 3: Based on the dynamic performance coefficient, extract the extreme value difference of attendance achievement rate within the most recent set period, divide the extreme value difference by the set level to generate the base of the equal ratio interval, and build a five-level achievement ladder by superimposing the base with the minimum achievement rate as the starting point. When the sales order volume fluctuates beyond the standard deviation set multiple, recalculate the extreme value to generate the dynamic achievement threshold interval.

[0036] Step 4: Count the number of times employees unlock game achievements. When the same type of achievement reaches the set number of times, extract the historical data of the dynamic achievement threshold range associated with the achievement, calculate the standard deviation of sales order volume, and add the current evaluation benchmark to the standard deviation by the set ratio to generate a calibration evaluation benchmark.

[0037] Step 5: Obtain the achievement trigger timestamp and data version identifier from the mobile and game platforms, calculate the absolute value of the timestamp difference, and when the difference exceeds the set fault tolerance threshold or the version identifier is discontinuous, call the calibration evaluation benchmark to re-execute the achievement condition judgment and generate cross-platform game reward instructions.

[0038] Step 6: Detect game reward instructions, retrieve the latest multi-source working dataset, re-execute achievement condition judgment, if the judgment result differs from the initial record, send a virtual asset incremental update package to the game client, and push a status change notification to the mobile client to generate a cross-platform status synchronization record.

[0039] This invention also supports superiors or leaders in reviewing or evaluating attendance and work performance at regular intervals.

[0040] For example, the software sends attendance and performance questionnaires to company leaders, superiors, or department heads at regular intervals, such as daily, weekly, or monthly. The software includes a "normal" option; if everything is normal, superiors and leaders can select this option. If an employee's performance is substandard, the software selects or searches for that employee and assigns a negative review. Once the superiors and leaders have completed their evaluations, the instructions are sent to the software and its server or cloud server. The resulting evaluations and reviews then affect the employee's game rewards.

[0041] Games and work software can be transferred between different companies. Work software and various functions of games can also be broken down and completed by multiple companies. For example, the original account, the original mobile phone number and the customer's registration information can be shared with another company to create an account, or the original account and account data can be used to log in to another company's software. The game software company can also be a variety of different companies.

[0042] The present invention will be further disclosed below with reference to specific examples:

[0043] Step 1: Multi-source data acquisition and outlier labeling

[0044] The system collects employee attendance and clock-in coordinates in real time through a real-name account system, using GPS positioning to record precise time and location; it also collects customer ratings based on the order after-sales evaluation system, with the data structure including order number, rating, and timestamp; and it collects sales order completion volume to count the number of orders actually completed by employees each day.

[0045] Among them, attendance coordinate data is discretized at the hourly granularity to form a time series; customer rating star ratings are matched based on order completion timestamps; and sales order completion volume is aggregated by natural day.

[0046] The system calculates the standard deviation of each indicator within a set assessment period (e.g., 7 days, 30 days), uses a multiple of the departmental mean (e.g., 2 times the standard deviation) to identify and mark outliers. Valid data is smoothed using a sliding window mean to form a multi-source working dataset.

[0047] For example, Zhang, an employee in a sales department, had the following daily order completion counts over 7 days: [12, 14, 11, 19, 9, 10, 50]. The department's daily average was 13, and the standard deviation was 4.5. The order volume "50" on the 7th day deviated significantly. Based on the set outlier detection factor of 2, the upper limit for outliers is 13 + 2 × 4.5 = 22. Since order volume 50 > 22, it was marked as an outlier. The remaining data were calculated using a 3-day sliding window mean to generate a smoothed sales series for subsequent analysis.

[0048] Step 2: Dynamic Weight Adjustment and Performance Coefficient Generation

[0049] The system calls a multi-source working dataset and checks the following: If an employee's sales order completion volume continuously exceeds the department average for a cumulative period of a set number of days (e.g., exceeding the average for 3 consecutive days), the system initiates dynamic weight adjustment:

[0050] Specific rules: The weight of customer rating star rating increases by 0.1 times daily, up to a maximum of 1.5 times the base value; the weight of attendance coordinate data is simultaneously reduced to a set ratio (e.g., reduced to 0.7 times the base value); a comprehensive weighted value is calculated to generate a dynamic performance coefficient, which serves as the core reference indicator for subsequent achievements and assessments.

[0051] For example: Xiao Zhang's order completion volume for days 1-3 was 14, 15, and 16 respectively, exceeding the department's daily average of 13 for three consecutive days. The system adjusts the weights as follows: Day 1: Customer rating weight = base value × 1.1; Day 2: Customer rating weight = base value × 1.2; Day 3: Customer rating weight = base value × 1.3 (if the base value is 0.4, then the customer rating weight for day 3 is 0.52); at the same time, the attendance weight decreases from the base value of 0.3 to 0.7 × 0.3 = 0.21; based on the above dynamic weights, the corresponding period's comprehensive performance coefficient is calculated using the following formula: Performance coefficient = Attendance score × Attendance weight + Sales score × Sales weight + Customer rating × Customer weight

[0052] Step 3: Construction of a five-level achievement ladder and generation of dynamic threshold ranges

[0053] Based on the dynamic performance coefficient, the system calculates the extreme difference (the difference between the maximum and minimum values) of the attendance achievement rate within the most recent set period; the extreme difference is divided by the 5-level interval series to generate the base of the equal interval; the construction of the five-level achievement ladder adopts a decreasing interval span pattern for the first three levels and an increasing interval span pattern for the last two levels, where the first level span is 28% of the extreme difference, the second level is 22%, the third level is 18%, the fourth level is 20%, and the fifth level is 25%.

[0054] By overlaying interval baselines, a five-level achievement ladder is constructed, starting with the minimum achievement rate. If the sales order volume fluctuates beyond a multiple of the department's historical standard deviation (e.g., 1.5 times), the system recalculates the extreme values ​​and ladder thresholds to ensure dynamic adaptation to the actual performance level of employees.

[0055] To enhance data security, the system calculates the hash value of the dynamic performance coefficients before execution and matches it with the latest node of the pre-stored hash chain in the cloud. If the verification fails, the system automatically reverts to the previous two valid versions of the dataset and regenerates the coefficients.

[0056] For example: Xiao Zhang's 7-day attendance compliance rate is as follows (%): [100, 95, 100, 85, 80, 90, 100], with a maximum value of 100%, a minimum value of 80%, and an extreme value difference of 20%. The five-level tiered range is calculated proportionally as follows: Level 1: 80%–85.6% (20% × 28%), Level 2: 85.6%–90%, Level 3: 90%–93.6%, Level 4: 93.6%–97.6%, Level 5: 97.6%–100%.

[0057] If the standard deviation of subsequent order fluctuations exceeds the set value, the system will recalculate the extreme value range and adjust the above-mentioned step boundaries.

[0058] Step 4: Benchmark Calibration and Achievement Recording

[0059] The system tracks the number of times employees unlock achievements in the game. When the cumulative number of the same type of achievement reaches a set number (e.g., 5 times), the system extracts historical data for the corresponding dynamic achievement threshold range and calculates the standard deviation of sales order volume.

[0060] The benchmark calibration logic is as follows: if the current standard deviation exceeds the department's historical average standard deviation, a weighted fusion is used; the calibration evaluation benchmark = 0.3 times the current standard deviation + 0.2 times the historical average. - Replace the original benchmark and dynamically adjust the evaluation accuracy.

[0061] For example: Xiao Zhang's recent sales order standard deviation is 6, and the historical standard deviation mean is 3. The calibration evaluation benchmark = 0.3×6+0.2×3=1.8+0.6=2.4. The system will update the original evaluation benchmark (such as 2) to 2.4 for subsequent dynamic re-judgment of achievement threshold.

[0062] Step 5: Cross-platform verification and game reward determination

[0063] The system obtains the timestamp of the last heartbeat packet from the mobile device (synchronized via NTP protocol); the game server receives the confirmation timestamp and data version identifier. The absolute value of the timestamp difference is calculated. If it exceeds a set fault tolerance threshold (e.g., 300 seconds) or the versions are discontinuous, the system calls the calibration evaluation benchmark and re-evaluates the achievement conditions; if the evaluation result changes, a cross-platform game reward instruction is generated.

[0064] For example, if the mobile device records a timestamp of 2025-06-01 14:03:00, while the game device records a timestamp of 2025-06-01 14:10:00, the difference of 7 minutes exceeds the tolerance threshold of 300 seconds, triggering the system's calibration mechanism. Additionally, discontinuous version number determinations (such as jumping from v3 to v5) will also trigger the game's reward determination logic.

[0065] Step 6: Cross-device state synchronization and virtual incentive feedback

[0066] The system receives the game reward instruction, retrieves the latest multi-source working dataset, and re-executes timestamp alignment and achievement determination. If a discrepancy is found, it sends an incremental update package of virtual assets to the game client. The game reward content includes: game reward amount = number of difference achievement levels × base currency unit × time decay coefficient, where the time decay coefficient = 1 / (1+0.1× number of delay days).

[0067] Simultaneously, a dual-channel redundant push mechanism is adopted, which sends Base64 encoded data packets containing achievement change summaries to the mobile terminal message queue and SMS gateway respectively, thereby achieving cross-terminal status synchronization.

[0068] For example: Xiao Zhang should have obtained Level 4 achievement, but due to synchronization failure, he was only recognized as Level 2, resulting in a difference of 2 levels. The base currency unit is 100 coins, and the delay is 3 days. Therefore: the time decay coefficient = 1 / (1+0.1×3) = 1 / 1.3 ≈ 0.769, and the game reward amount = 2×100×0.769 ≈ 153.8 coins. The system pushes an update package containing the game reward assets to the game client and notifies employees via a dual-channel approach: message queue and SMS.

[0069] This invention also discloses a cloud-based employee performance evaluation management system for executing the above-described employee performance evaluation management method, comprising:

[0070] The multi-source data acquisition module is used to collect attendance coordinate data, customer rating star ratings, and original records of sales order completion in real time through the real-name account system, perform standard deviation calculation and outlier marking, and output multi-source working datasets.

[0071] The dynamic weight calculation module calls a multi-source working dataset, detects periods when sales orders continuously exceed the average value, performs operations to increase customer evaluation weight and decrease attendance weight, and outputs dynamic performance coefficients.

[0072] The achievement threshold generation module, based on dynamic performance coefficients, extracts the extreme difference of attendance compliance rate, executes the logic of constructing a five-level achievement ladder and dynamic recalculation, and outputs the dynamic achievement threshold range.

[0073] The benchmark calibration module counts the number of times game achievements are unlocked, extracts historical data of dynamic achievement threshold range, performs sales standard deviation calculation and benchmark update, and outputs calibration evaluation benchmark.

[0074] The cross-platform verification module obtains the timestamps and version identifiers of the mobile and game platforms, performs difference limit judgment and achievement condition re-judgment, and outputs cross-platform game reward instructions.

[0075] The status synchronization execution module detects cross-platform game reward commands, retrieves the latest three-dimensional data to perform timestamp alignment and achievement re-judgment, sends virtual asset update packages and notifications when discrepancies are found, and outputs cross-platform status synchronization records.

[0076] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A cloud-based employee performance evaluation and management method, characterized in that, The specific steps are as follows: Step 1: Collect employee attendance and clock-in coordinate data, customer rating star ratings, and original records of sales order completion volume using real names. Calculate the standard deviation of each indicator within a set period. Mark data that deviates from the department average by a set multiple as outliers. Take the sliding window mean of the valid data to generate a multi-source working dataset. Step 2: Call the multi-source working dataset, detect the period when the number of completed sales orders continuously exceeds the department's average for a set duration, increase the weight of customer rating star rating to a set multiple of the base value, reduce the weight of attendance coordinate data to a set ratio, calculate the weighted comprehensive value, and generate a dynamic performance coefficient; Step 3: Based on the dynamic performance coefficient, extract the extreme value difference of attendance achievement rate within the most recent set period, divide the extreme value difference by the set level to generate the base of the equal ratio interval, and build a five-level achievement ladder by superimposing the base with the minimum achievement rate as the starting point. When the sales order volume fluctuates beyond the standard deviation set multiple, recalculate the extreme value to generate the dynamic achievement threshold interval. Step 4: Count the number of times employees unlock game achievements. When the same type of achievement reaches the set number of times, extract the historical data of the dynamic achievement threshold range associated with the achievement, calculate the standard deviation of sales order volume, and add the current evaluation benchmark to the standard deviation by the set ratio to generate a calibration evaluation benchmark. Step 5: Obtain the achievement trigger timestamp and data version identifier from the mobile and game platforms, calculate the absolute value of the timestamp difference, and when the difference exceeds the set fault tolerance threshold or the version identifier is discontinuous, call the calibration evaluation benchmark to re-execute the achievement condition judgment and generate cross-platform game reward instructions. Step 6: Detect cross-platform game reward instructions, retrieve the latest multi-source working dataset, re-execute achievement condition judgment, and if the judgment result differs from the initial record, send a virtual asset incremental update package to the game client and push a status change notification to the mobile client to generate a cross-platform status synchronization record.

2. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The method for calculating the sliding window mean in step 1 is as follows: using time series alignment, the attendance coordinate data is discretized by hourly granularity, customer rating star ratings are matched by order completion timestamps, and the sales order completion volume is aggregated by natural day segments.

3. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The specific method for increasing customer evaluation weight in step 2 is as follows: when a sales order exceeds the average for a set number of consecutive days, the weight is increased by 0.1 times daily, with a maximum increase of no more than 1.5 times the base value.

4. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The construction of the five-level achievement ladder described in step 3 adopts a decreasing interval span pattern for the first three levels and an increasing interval span pattern for the last two levels. The first level has an interval span of 28% of the extreme value difference, the second level 22%, the third level 18%, the fourth level 20%, and the fifth level 25%.

5. The employee performance evaluation and management method based on a cloud platform according to claim 4, characterized in that: Before step 3 is executed, a data verification step is added: the dynamic performance coefficient is hashed and compared with the latest node of the pre-stored hash chain in the cloud. If the verification fails, the coefficient is reverted to the previous two valid versions of the dataset and regenerated.

6. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The specific method for generating the calibration evaluation benchmark in step 4 is as follows: when the sales standard deviation exceeds the average historical standard deviation of the department, a weighted fusion method is used to replace 0.5 times the standard deviation with the algebraic sum of 0.3 times the standard deviation and 0.2 times the historical mean.

7. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The timestamp difference calculation in step 5 uses the NTP protocol to synchronize the dual-end clock sources. For the mobile terminal, the timestamp of the last heartbeat packet is used, and for the game terminal, the timestamp of the server's received confirmation is used.

8. The employee performance evaluation and management method based on a cloud platform according to claim 1, characterized in that: The incremental update of virtual assets mentioned in step 6 includes the amount of virtual currency game rewards corresponding to the achievement difference value. The calculation method is: difference achievement level number × base currency unit × time decay coefficient, where the time decay coefficient = 1 / (1+0.1×delay days).

9. The employee performance evaluation and management method based on a cloud platform according to claim 8, characterized in that: The status change notification in step 6 adopts a dual-channel redundant push mechanism, simultaneously sending a Base64 encoded data packet containing an achievement change summary to both the mobile message queue and the SMS gateway.

10. A cloud-based employee performance evaluation management system, used to execute the employee performance evaluation management method according to any one of claims 1-9, characterized in that, include: The multi-source data acquisition module is used to collect attendance coordinate data, customer rating star ratings, and original records of sales order completion in real time through the real-name account system, perform standard deviation calculation and outlier marking, and output multi-source working datasets. The dynamic weight calculation module calls a multi-source working dataset, detects periods when sales orders continuously exceed the average value, performs operations to increase customer evaluation weight and decrease attendance weight, and outputs dynamic performance coefficients. The achievement threshold generation module, based on dynamic performance coefficients, extracts the extreme difference of attendance compliance rate, executes the logic of constructing a five-level achievement ladder and dynamic recalculation, and outputs the dynamic achievement threshold range. The benchmark calibration module counts the number of times game achievements are unlocked, extracts historical data of dynamic achievement threshold range, performs sales standard deviation calculation and benchmark update, and outputs calibration evaluation benchmark. The cross-platform verification module obtains the timestamps and version identifiers of the mobile and game platforms, performs difference limit judgment and achievement condition re-judgment, and outputs cross-platform game reward instructions. The status synchronization execution module detects cross-platform game reward commands, retrieves the latest three-dimensional data to perform timestamp alignment and achievement re-judgment, sends virtual asset update packages and notifications when discrepancies are found, and outputs cross-platform status synchronization records.

Images