Piece counting task performance dynamic evaluation system based on variable weight
Through the dynamic performance evaluation system of piece-by-piece task with variable weights, T-test and standardized algorithms are used to dynamically adjust the weights, which solves the problems of distortion of the evaluation results and frequent manual adjustments of the traditional performance evaluation system, and achieves a more accurate and efficient performance evaluation.
Patent Information
- Application Number
- CN202510355797.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-25
- Publication Date
- 2025-07-11
AI Technical Summary
The existing performance evaluation system relies on fixed scoring rules and cannot dynamically adapt to diverse task categories and complex statistical needs, resulting in distortion of evaluation results, susceptible to differences in extreme values and data distribution, and requires frequent manual adjustments, which increases costs.
The performance dynamic evaluation system of piece-by-piece task based on variable weights is adopted. By evaluating the statistical dimension framework and backtracking the historical data to perform the weight dynamic evaluation subsystem, the T-test method is used to dynamically adjust the weight parameters, and combined with mapping functions and standardized algorithms, the flexibility and accuracy of performance evaluation are achieved.
It improves the accuracy and comparability of performance evaluation, reduces manual intervention, and improves data processing efficiency and timeliness of parameter configuration.
Smart Images
Figure CN120297790A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of data processing, and in particular to a dynamic performance evaluation system for piecework tasks based on variable weights. Background Art
[0002] Existing performance evaluation systems rely on fixed scoring rules. In the face of diverse task categories and complex statistical requirements, the performance evaluation method with fixed rules is not flexible enough to dynamically adapt to the diversity of various types of tasks. The evaluation results are distorted, vulnerable to extreme values and data distribution differences, and the data processing accuracy is low. With the changes in business requirements and scenarios, the pre-configured fixed weights or parameters are no longer applicable, and manual adjustment by administrators is required frequently, making the evaluation results lag behind the actual situation, affecting both the accuracy of the data and increasing the labor cost. Summary of the Invention
[0003] In order to help solve the above technical problems, this application provides a dynamic performance evaluation system for piecework tasks based on variable weights, adopting the following technical solutions:
[0004] A dynamic performance evaluation system for piecework tasks based on variable weights, wherein the dynamic performance evaluation system for piecework tasks based on variable weights includes a piecework task performance evaluation subsystem and a retrospective historical data execution weight dynamic evaluation subsystem. The piecework task performance evaluation subsystem includes the following connected in sequence:
[0005] An evaluation statistical dimension framework module for setting at least one evaluation dimension, respectively setting an evaluation dimension weight coefficient for each evaluation dimension, and calculating the original score of the original single-piece task on each evaluation dimension;
[0006] A dimension data calculation module includes a sample data unit, a normalization processing unit, a mean and standard deviation unit, and a standard score unit connected in sequence. The sample data unit is used to calculate the subsequent score of each remaining single-piece task on each evaluation dimension. The original score and the subsequent score form a sample set. The mean and standard deviation unit is used to calculate the mean and standard deviation of all scores on each evaluation dimension according to the sample set. The normalization processing unit is used to perform normalization processing on the original scores on each evaluation dimension respectively through the normal distribution algorithm. The standard score unit is used to obtain the standard score on each evaluation dimension based on the normalization processing unit;
[0007] The task-level module includes a mapping function processing unit, a weight parameter reading unit, and a single-task weighted score unit connected in sequence. The mapping function processing unit is used to perform mapping function processing on the standard score on each evaluation dimension to obtain the final dimension score on each evaluation dimension. The weight parameter reading unit is used to calculate the weighted score of each single task on each evaluation dimension according to the final difficulty score and the evaluation dimension weight coefficient. The single-task weighted score unit is used to add the weighted scores on each evaluation dimension to obtain the total performance score.
[0008] The weight dynamic evaluation subsystem for backtracking historical data is used to pull all weighted scores and all final difficulty scores of one period at preset time intervals. The weight dynamic evaluation subsystem for backtracking historical data includes a T test module, which is used to test the merged data of two adjacent periods by the T test method:
[0009]
[0010] Among them, t is the statistic, and are the means of all weighted scores of the first and second periods, respectively. are the variances of all weighted scores of the first and second periods respectively, n1 and n2 are the number of evaluation dimensions of the first and second periods respectively, and the corresponding P value results are found according to the t statistic and t distribution table in the T test to determine whether there is a significant change in the data of the first and second periods. If p<0.05, it is considered that at a 95% confidence level, there is a significant change in the weighted score of the second period compared with the first period. At this time, if the mean of all scores on a dimension in the second period increases by more than the preset rising threshold compared with the first period, the evaluation dimension weight coefficient of the evaluation dimension is reduced to the preset evaluation dimension weight coefficient value.
[0011] The backtracking historical data execution weight dynamic evaluation subsystem also includes:
[0012] A database module, used for storing all weighted scores and all final difficulty scores of each period;
[0013] A scheduler module, the scheduler module is connected to the database module and is used to set the time interval for pulling data and define the granularity of the time period;
[0014] A data cleaning module, the input end of which is connected to the scheduler module, and the output end of which is connected to the T-test module, for automatically checking the pulled data, and if there are missing values, removing the corresponding task records;
[0015] A weight adjustment module, which is connected to the output end of the T-test module. If the increase in the mean value of all scores in a certain dimension in the second period exceeds the preset increase threshold compared with the first period, the administrator is reminded. Furthermore, the weight adjustment module reduces the evaluation dimension weight coefficient of this evaluation dimension to the preset evaluation dimension weight coefficient value. If the increase in the mean value of all scores in a certain dimension in the second period does not exceed the preset increase threshold compared with the first period, the weight adjustment module reduces the evaluation dimension weight coefficient of this evaluation dimension to the preset evaluation dimension weight coefficient value.
[0016] The evaluation dimensions include basic difficulty, completion degree, manual workload, and time efficiency. The basic difficulty is used to measure the complexity and difficulty of the task, the completion degree is used to measure the quality and integrity of the task completion, the manual workload is used to measure the amount of manual input required for the task, and the time efficiency is used to measure the time efficiency of the task completion.
[0017] The mean and standard deviation unit is used to calculate the mean and standard deviation of all scores on each evaluation dimension according to the sample set:
[0018]
[0019] where, μ i is the sample mean, σ i is the sample standard deviation, S ij represents the score of the jth single task in the ith dimension, and N is the number of single tasks.
[0020] The standardization processing unit is used to perform standardization processing on the original scores on each evaluation dimension respectively through the normal distribution algorithm to obtain the standard scores on each evaluation dimension:
[0021]
[0022] where, Z i is the standard score of the ith dimension, and S i is the original score of the ith dimension.
[0023] The mapping function processing unit is used to perform mapping function processing on the standard scores on each evaluation dimension to obtain the final dimension scores on each evaluation dimension:
[0024] The dimension score S i ′ can be expressed as:
[0025] S i ′ = f(Z i );
[0026] On the time efficiency dimension S4, the mapping function is:
[0027]
[0028] The weight parameter reading unit and the single-task weighted score unit are used to calculate the weighted score of each single task in each evaluation dimension according to the final score and the evaluation dimension weight coefficient, and add up the weighted scores in each evaluation dimension to obtain the total performance score:
[0029]
[0030] Among them, W i is the weight coefficient of the i-th dimension, S i ′ is the score of the i-th dimension, n is the number of evaluation dimensions, and W i ×S i ′ is the weighted score.
[0031] In summary, the piecework task performance dynamic evaluation system constructed in this application includes a comprehensive evaluation and statistical framework with multiple dimensions such as basic difficulty, completion degree, manual workload, and time efficiency. It supports embedding the mapping function of the original scores of each dimension and adopts the standard score method, eliminating the influence of data distribution differences on the evaluation results, solving the evaluation deviation problem caused by uneven data distribution in traditional methods, making the evaluation results more accurate, and making the performance of different types of tasks more comparable. The weight dynamic evaluation subsystem for backtracking historical data in this system, through modules such as a timed task scheduler, version management, and data missing detection module, realizes the dynamic configuration and update of weight parameters. Using the method of system backtracking, it provides the ability to dynamically adjust parameters more quantitatively based on statistical tests. The system uses the timed task scheduler to analyze data changes. When the performance of a certain dimension is significantly different from the previous evaluation results, it automatically generates weight adjustment suggestions, and can more flexibly adapt to the piecework task performance statistical scenarios under different requirements in cooperation with the version management of parameters, and better analyze the change degree of single tasks.
[0032] This application has the following beneficial effects:
[0033] Improved data processing efficiency: Through variable weight configuration and data automation process design, the overall system process runs fully automatically, ensuring the high efficiency and stability of system operation;
[0034] Improved evaluation accuracy: The mapping function and standardization algorithm design eliminate the influence of extreme values, making the evaluation results closer to the true performance distribution;
[0035] Self-optimization of weight parameters: The weight dynamic adjustment mechanism reduces the workload of manual intervention, ensuring the fairness and timeliness of parameter configuration. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 It is a schematic block diagram of an embodiment of the piecework task performance evaluation subsystem of this application;
[0037] Figure 2 A schematic block diagram of an embodiment of the weight dynamic evaluation subsystem for the backtracking historical data of the present application. Detailed implementation manners
[0038] The present application will be further described below with reference to the accompanying drawings. The structure and principle of the present application are very clear to those skilled in the art. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0039] Figure 1 A schematic block diagram of an embodiment of the piecework task performance evaluation subsystem of the present application, Figure 2 A schematic block diagram of an embodiment of the weight dynamic evaluation subsystem for the backtracking historical data of the present application.
[0040] Combined with Figure 1 and Figure 2 It can be understood that the piecework task performance dynamic evaluation system based on variable weights of the present application includes a piecework task performance evaluation subsystem and a weight dynamic evaluation subsystem for backtracking historical data. The piecework task performance evaluation subsystem includes, connected in sequence:
[0041] An evaluation statistical dimension framework module for setting at least one evaluation dimension, respectively setting an evaluation dimension weight coefficient for each evaluation dimension, calculating the original score of the original single task on each evaluation dimension. In this embodiment, the evaluation dimensions include basic difficulty, completion degree, manual workload, and time efficiency. The basic difficulty is used to measure the complexity and difficulty of the task, the completion degree is used to measure the quality and integrity of the task completion, the manual workload is used to measure the amount of manual input required for the task, and the time efficiency is used to measure the time efficiency of the task completion.
[0042] A dimension data calculation module includes a sample data unit, a normalization processing unit, a mean and standard deviation unit, and a standard score unit, which are connected in sequence. The sample data unit is used to calculate the subsequent score of each remaining single task on each evaluation dimension. The original score and the subsequent score form a sample set. The mean and standard deviation unit is used to calculate the mean and standard deviation of all scores on each evaluation dimension according to the sample set. The normalization processing unit is used to perform normalization processing on the original score on each evaluation dimension respectively through a normal distribution algorithm. The standard score unit is used to obtain the standard score on each evaluation dimension based on the normalization processing unit;
[0043] The task-level module includes a mapping function processing unit, a weight parameter reading unit, and a single-task weighted score unit connected in sequence. The mapping function processing unit is used to perform mapping function processing on the standard scores for each evaluation dimension to obtain the final dimension scores for each evaluation dimension. The weight parameter reading unit is used to calculate the weighted scores of each single task in each evaluation dimension according to the final difficulty score and the evaluation dimension weight coefficient. The single-task weighted score unit is used to add up the weighted scores for each evaluation dimension to obtain the total performance score;
[0044] The weight dynamic evaluation subsystem for backtracking historical data is used to pull all weighted scores and all final difficulty scores for one period at preset time intervals. The weight dynamic evaluation subsystem for backtracking historical data includes a T-test module, which is used to test the combined data of two adjacent periods by the T-test method:
[0045]
[0046] where t is the statistic, and are the means of all weighted scores for the first and second periods respectively, are the variances of all weighted scores for the first and second periods respectively. n1 and n2 are the numbers of evaluation dimensions for the first and second periods respectively. According to the t statistic and the t distribution table in the T-test, the corresponding P-value result is found to judge whether there are significant changes between the data of the first and second periods. If p < 0.05, it is considered that at the 95% confidence level, there are significant changes in the weighted scores of the second period compared with the first period. At this time, if the increase in the mean of all scores in a certain dimension in the second period exceeds the preset increase threshold compared with the first period, the evaluation dimension weight coefficient of this evaluation dimension is reduced to the preset evaluation dimension weight coefficient value.
[0047] Specifically, the mean and standard deviation unit is used to calculate the mean and standard deviation of all scores for each evaluation dimension according to the sample set:
[0048]
[0049] where μ i is the sample mean, σ i is the sample standard deviation, S ij represents the score of the jth single task in the ith dimension, and N is the number of single tasks.
[0050] The standardization processing unit is used to perform standardization processing on the original scores for each evaluation dimension respectively through the normal distribution algorithm to obtain the standard scores for each evaluation dimension:
[0051]
[0052] where Zi is the standard score for the i-th dimension, S i is the original score for the i-th dimension.
[0053] The mapping function processing unit is used to perform mapping function processing on the standard scores for each evaluation dimension to obtain the final dimension scores for each evaluation dimension:
[0054] The final dimension score S i ' can be expressed as:
[0055] S i ' = f(Z i );
[0056] In the time efficiency dimension S4, the mapping function is:
[0057]
[0058] The read weight parameter unit and the single-piece task weighted score unit are used to calculate the weighted scores of each single-piece task in each evaluation dimension according to the final difficulty score and the evaluation dimension weight coefficient, and add up the weighted scores in each evaluation dimension to obtain the total performance score:
[0059]
[0060] where, W i is the weight coefficient for the i-th dimension, S i ' is the final difficulty score for the i-th dimension, n is the number of evaluation dimensions, W i ×S i ' is the weighted score.
[0061] In actual work, the piecework task performance evaluation subsystem can work in the following ways:
[0062] 1: Construct an evaluation statistical dimension framework. Combining the characteristics of piecework tasks in the data verification scenario, set the evaluation dimensions as basic difficulty, completion degree, manual workload, and time efficiency.
[0063] 2: Collect sample data and perform standardization processing. Collect the original data of all piecework tasks in these four dimensions for the past six months to form a sample set. Perform standardization processing on the sample data through the scipy.stats normal distribution standard score method.
[0064] 3: Process the standard scores according to the scoring mapping function set by the business to obtain the final dimension scores.
[0065] 4: Calculate the performance scores of single-piece tasks by weighting the standard scores of each dimension according to the currently set weight parameters.
[0066]
[0067] 2. The weight dynamic evaluation subsystem for backtracking historical data in the piece-rate task performance dynamic evaluation system according to claim 1 further includes:
[0068] A database module for storing all weighted scores and all final difficulty scores in a certain period.
[0069] A scheduler module, which is connected to the database module, for setting the time interval for pulling data and defining the time period granularity.
[0070] A data cleaning module, whose input end is connected to the scheduler module and output end is connected to the T-test module, for automatically checking the pulled data. If there are missing values, the corresponding task records are excluded.
[0071] A weight adjustment module, which is connected to the output end of the T-test module. If the increase in the mean value of all scores in a certain dimension in the second period compared to the first period exceeds the preset increase threshold, it reminds the administrator. Then the weight adjustment module reduces the evaluation dimension weight coefficient of this evaluation dimension to the preset evaluation dimension weight coefficient value. If the increase in the mean value of all scores in a certain dimension in the second period compared to the first period does not exceed the preset increase threshold, the weight adjustment module reduces the evaluation dimension weight coefficient of this evaluation dimension to the preset evaluation dimension weight coefficient value.
[0072] In actual work, the weight dynamic evaluation subsystem for backtracking historical data works in the following way:
[0073] 1: Set the backtracking period to one month, and the system scheduler starts at the end of each month. The system pulls the sum of weighted scores of each evaluation dimension in this month and last month from the database according to the time granularity through the query interface.
[0074] 2: Automatically check the pulled data. If there are missing values, exclude the corresponding task records.
[0075] 3: Calculate the indicators. Assume that the mean value of the weighted score of the manual workload dimension in this month is Y t = 1.2, and the mean value in the last month is Y t-1 = 1, the sample sizes are n t = 980, n t-1 = 785, and the variances are respectively
[0076] 4: The T-test method is adopted, and the stats.ttest_ind function is called to test the data of the current period and the previous period to obtain the P-value. After calculation, the P-value is less than 0.05, and the original mean of the time efficiency dimension has increased by more than the threshold of 0.1 (assuming the threshold is 0.1), indicating that the overall efficiency of all employees in the current period has improved under this task category.
[0077] 5: After completing the previous step, the system prompts the administrator. The administrator combines the actual business situation to judge and confirm that the weight of the time efficiency dimension is adjusted from 0.2 to 0.15, completing the dynamic optimization of the weight parameters and making the subsequent performance evaluation more in line with the actual business situation.
[0078] In summary, the dynamic performance evaluation system for piecework tasks based on variable weights in this application builds a comprehensive evaluation framework integrating multiple dimensions such as basic difficulty, completion degree, manual workload, and time efficiency. By embedding the original score mapping function and standard score method for each dimension, it effectively eliminates the interference of data distribution differences on the evaluation results, solves the deviation problem caused by uneven data distribution in traditional evaluations, and enhances the comparability of cross-category task performance scores.
[0079] The weight dynamic evaluation subsystem for backtracking historical data in the system adopts a timed task scheduler, version management, and data missing detection module to achieve dynamic configuration and update of weight parameters. By combining system backtracking with statistical tests, it quantifies the ability to dynamically adjust parameters, flexibly adapts to various piecework task performance statistical scenarios, accurately analyzes the changes in single-piece tasks, and significantly reduces the cost of manual intervention.
Claims
1. A dynamic performance evaluation system for piecework tasks based on variable weights, characterized in that, The piece-rate task performance dynamic evaluation system based on variable weights includes a piece-rate task performance evaluation subsystem and a back-tracing historical data execution weight dynamic evaluation subsystem, wherein the piece-rate task performance evaluation subsystem includes: an evaluation statistical dimension framework module connected in sequence, for setting at least one evaluation dimension, and setting an evaluation dimension weight coefficient for each evaluation dimension, and calculating the original score of the original single-piece task on each evaluation dimension; A dimensional data calculation module, including a sample data unit, a standardization processing unit, a mean and standard deviation unit, and a standard score unit connected in sequence, wherein the sample data unit is used to calculate the subsequent score of each remaining single task on each evaluation dimension, and the original score and the subsequent score form a sample set, the mean and standard deviation unit is used to calculate the mean and standard deviation of all scores on each evaluation dimension according to the sample set, the standardization processing unit is used to perform standardization processing on the original score on each evaluation dimension respectively by a normal distribution algorithm, and the standard score unit is used to obtain the standard score on each evaluation dimension based on the standardization processing unit; a task-level module, including a mapping function processing unit, a weight parameter reading unit, and a single-task weighted score unit connected in sequence, the mapping function processing unit is used to perform mapping function processing on the standard score on each evaluation dimension to obtain the final dimensional score on each evaluation dimension, the weight parameter reading unit is used to calculate the weighted score of each single task on each evaluation dimension according to the final difficulty score and the evaluation dimension weight coefficient, and the single-task weighted score unit is used to add the weighted scores on each evaluation dimension to obtain the total performance score; The weight dynamic evaluation subsystem for backtracking historical data is used to pull all weighted scores and all final difficulty scores of one period at a preset time interval. The weight dynamic evaluation subsystem for backtracking historical data includes a T test module, which is used to test the merged data of two adjacent periods by the T test method: where t is the statistic, and are the means of all weighted scores in the first and second periods respectively, are the variances of all weighted scores in the first and second periods respectively. n1 and n2 are the numbers of evaluation dimensions in the first and second periods respectively. The corresponding P - value result is found according to the t - statistic and t - distribution table in the T - test to determine whether there are significant changes in the data of the first and second periods. If p < 0.05, it is considered that at the 95% confidence level, there are significant changes in the weighted scores of the second period compared with the first period. At this time, if the increase in the mean of all scores on a certain dimension in the second period compared with the first period exceeds the preset increase threshold, the evaluation dimension weight coefficient of this evaluation dimension is reduced to the preset evaluation dimension weight coefficient value.
2. The dynamic evaluation system for piecework task performance based on variable weights according to claim 1, characterized in that, The backtracking historical data execution weight dynamic evaluation subsystem also includes: A database module, used for storing all weighted scores and all final difficulty scores of the said one period; A scheduler module, the scheduler module is connected to the database module and is used to set the time interval for pulling data and define the granularity of the time period; A data cleaning module, the input end of which is connected to the scheduler module, and the output end of which is connected to the T-test module, for automatically checking the pulled data, and if there are missing values, removing the corresponding task records; A weight adjustment module, wherein the weight adjustment module is connected to the output end of the T-test module. If the increase in the mean of all scores on a certain dimension in the second period exceeds the preset increase threshold compared with the first period, the administrator is reminded, and then the weight adjustment module reduces the evaluation dimension weight coefficient of the evaluation dimension to the preset evaluation dimension weight coefficient value. If the increase in the mean of all scores on a certain dimension in the second period does not exceed the preset increase threshold compared with the first period, the weight adjustment module reduces the evaluation dimension weight coefficient of the evaluation dimension to the preset evaluation dimension weight coefficient value.
3. The dynamic evaluation system for piecework task performance based on variable weights according to claim 1, wherein, The evaluation dimensions include basic difficulty, completion rate, manual workload, and time efficiency. The basic difficulty is used to measure the complexity and difficulty of the task. The completion rate is used to measure the quality and integrity of the task completion. The manual workload is used to measure the amount of manual input required for the task. The time efficiency is used to measure the time efficiency of the task completion.
4. The dynamic performance evaluation system for piecework tasks based on variable weights according to claim 1, characterized in that The mean and standard deviation unit is used to calculate the mean and standard deviation of all scores on each evaluation dimension according to the sample set: Among them, μ i is the sample mean, σ i is the sample standard deviation, S ij represents the score of the j-th single task in the i-th dimension, and N is the number of single tasks.
5. The dynamic evaluation system for piecework task performance based on variable weights according to claim 4, wherein The standardization processing unit is used to perform standardization processing on the original scores on each evaluation dimension respectively through the normal distribution algorithm to obtain the standard scores on each evaluation dimension: Among them, Z i is the standard score of the i-th dimension, and S i is the original score of the i-th dimension.
6. The dynamic evaluation system for piecework task performance based on variable weights according to claim 5, wherein The mapping function processing unit is used to perform mapping function processing on the standard scores on each evaluation dimension to obtain the final dimension scores on each evaluation dimension: Final dimension score S i′ Can be expressed as: S i′ = f(Z i ); On the time efficiency dimension S4, the mapping function is:
7. The dynamic performance evaluation system for piecework tasks based on variable weights according to claim 6, characterized in that, The weight parameter reading unit and the single-task weighted score unit are used to calculate the weighted scores of each single task on each evaluation dimension according to the final difficulty score and the evaluation dimension weight coefficient, and add up the weighted scores on each evaluation dimension to obtain the total performance score: Among them, W i is the weight coefficient of the i-th dimension, S i′ is the final difficulty score of the i-th dimension, n is the number of evaluation dimensions, W i ×S i′ is the weighted score.