Employee efficiency evaluation algorithm based on linear model non-burying point
Through the non-buried point algorithm based on the linear model, the problem of inaccurate employee efficiency evaluation in complex scenarios in the existing technology is solved, and fast and accurate efficiency evaluation is achieved, which facilitates task optimization.
Patent Information
- Application Number
- CN202411722646.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-28
- Publication Date
- 2025-01-21
AI Technical Summary
The existing data point buried technology is difficult to accurately evaluate employee efficiency in complex scenarios, and there are problems of duplicate development and data deviation.
Using a linear model -based non -buried employee efficiency evaluation algorithm, through modeling the Internet online consultation service items, establishing task modularity and memory library, using big data for linear modeling, obtaining analytical solutions of various tasks, so as to obtain the analysis of each task, so as to obtain the analysis of each task. Evaluate employee efficiency.
In the case of avoiding repeated buried work, quickly and accurately obtain real efficiency, facilitate task optimization and adjustment, and bypass the disadvantages of traditional methods.
Smart Images
Figure CN119338325A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data embedding and efficiency evaluation technology, and more specifically, to an employee efficiency evaluation algorithm based on linear model non-embedding. Background Art
[0002] Currently in the Internet industry, the measurement of a thing or the change of a trend is often reflected through a series of complex and diverse indicators, but obtaining the results of these indicators requires data collection. A very common method in the industry is data embedding. The specific operation is often to click a button, stay on a page, or scroll through web page information, and record the number of clicks, browsing time, cancellation and rollback data, and then store them in the relevant database for subsequent use.
[0003] Conventional data tracking can theoretically obtain all the desired data, but it is also accompanied by great difficulties. The main reason is that data tracking will involve increasing the development of front-end and back-end engineers, and it is not flexible and changeable. With changes in business or page UI design, these tracking points often need to be carried out again. Secondly, there are also some difficulties in the actual production process of data tracking, which causes the statistical data to deviate from the actual and fail to achieve the desired purpose. For example, currently: the computer of hospital doctors generally stores a large amount of patient information, such as medical records, examination reports, etc. When consulting online on the Internet, doctors may jump to different pages many times during a service to view patient information. Even considering emergencies, doctors stay on a certain page for a long time. Therefore, the evaluation of the doctor's time spent on this service through tracking is inaccurate, and the doctor jumps to the page many times, in fact, for this service. The efficiency of this service cannot be well measured by clicking buttons or page time. For this reason, this solution proposes an employee efficiency evaluation algorithm based on linear model non-tracking. Summary of the invention
[0004] 1. Technical problems to be solved: In view of the problems existing in the prior art, the purpose of the present invention is to provide an employee efficiency evaluation algorithm based on a linear model and non-buried points. The algorithm is based on a large number of data samples and uses linear modeling to obtain analytical solutions for each task. When the data samples are rich, it will be infinitely close to the actual situation. At the same time, each task will be processed by staff every day, and the accumulation of data will quickly become rich, which can reflect the actual situation relatively quickly and facilitate the optimization and adjustment of different tasks.
[0005] 2. Technical solution: To solve the above problems, the present invention adopts the following technical solutions.
[0006] An employee efficiency evaluation algorithm based on a linear model and non-embedded point, comprising: pre-modeling the service items of online medical consultation on the Internet to form a database, wherein the database has divided task modules, and based on the evaluation items corresponding to the processes in the service item library, an evaluation item record model is formed to form a memory library; Call the database, task module and memory according to the project task to be evaluated, process under the task module divided by the target, encapsulate the work content, and obtain the efficiency evaluation of a certain task project; The obtained efficiency evaluation data of a certain task project is recorded according to the data in the memory bank, and the evaluation time of the task project can be obtained through the established equation.
[0007] A further improvement is that the database of online consultation services on the Internet includes: The average time required to complete the existing standardized (young people) daily tasks and the average time required to complete the existing specialized (elderly people) tasks are taken and entered into the database to form a database for completing tasks with specific standard time.
[0008] Further improvements are: the efficiency assessment of a task project includes: Based on the database and memory, a time comparison library is established in the form of comparing the existing task completion time with the ongoing task completion time; Set up a separate time point for a task item; Using the corresponding divided task modules, the completion time of any task item and the task items in the comparison library are set to the same module; Repeat setting the time point for the next task item to obtain the end time under the same task; When multiple tasks are running in parallel, the time of other tasks is recorded synchronously, and the start time is ignored, and the end time is recorded uniformly.
[0009] When tasks opened on multiple terminals are performed synchronously, the task start time is ignored and only the end time is recognized.
[0010] Further improvements are: the efficiency assessment of a task project also includes: Define the working hours within a certain period of time, generally selected as the working hours of a normal working day, and determine the work that needs to be done in a certain time interval in order to establish a relationship with the number of completed tasks; The designated working time periods can be modified at will according to actual conditions.
[0011] Further improvements are: the efficiency assessment of a task project also includes: The number of task items completed within a specific time period each day is counted to form a single-day project measurement library, which can be compared with the existing standardized number of projects per day to obtain efficiency evaluation results.
[0012] A further improvement is that the task project evaluation time is determined based on 8 working hours per day, in minutes, for a total of 480 minutes, as follows: ; The task items can be represented by X, and different task representation methods are X i ; Today X i The number of tasks completed is K i ; The completion time of each task is t i ; i is the corresponding project number.
[0013] Further improvement is: according to different task items obtained by different pipelines, the completion is also different, and the same equation is established for the data under each account: .
[0014] A further improvement is to include the step of solving the equation: When the amount of data in the equation exceeds the number of tasks, the parsing condition is met; When a linear relationship appears between two equations, the amount of data continues to be accumulated. When the row rank of the linear system is greater than or equal to the column rank, the linear system can be solved analytically, and the result is the time evaluation of each task.
[0015] A further improvement is that the evaluation results after the equation is resolved can be compared according to actual scenarios, or solved using the same account data alone, compared between different accounts, and then used for data analysis and optimization.
[0016] 3. Beneficial effects: Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: The invention is reasonably designed, and uses mathematical modeling to achieve an algorithm that can estimate the real efficiency even when it is difficult to place points under complex conditions. It provides a solution to the efficiency evaluation under complex conditions while avoiding a large amount of repeated point placement work, bypassing the drawbacks of traditional methods. At the same time, based on big data, it can quickly and accurately obtain the real efficiency situation, which is convenient for the optimization and adjustment of difficult tasks. The present invention can also evaluate the efficiency of a task by modularizing the task and encapsulating the work content. By recording the past situation of the daily task completion in a digital way and establishing a relationship, a piece of data can be collected from each account every day to establish an equation. Since there are generally limited tasks, a task can be regarded as a feature. Therefore, the evaluation time for each project can be obtained as early as when the data equation is consistent with the number of tasks.
[0017] It should be noted that the structures not introduced in the present invention are the same as the prior art or can be implemented by using the prior art, and are not described in detail here because they do not involve the design points and improvement directions of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 A flow chart for the overall evaluation of the present invention. DETAILED DESCRIPTION
[0019] To facilitate understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Example
[0020] The present invention aims to solve the problem that data embedding is too complicated and there are deviations between the embedded data and the actual data.
[0021] In order to evaluate the time taken for a certain event of the platform business, the conventional solution in the Internet industry is to record the start and end time of the operation. However, if a certain business requires close cooperation between operators at both ends to complete, and the operator needs to jump out of the page to find other information, or is temporarily delayed by an emergency and completed later, or when multiple projects are processed at the same time, it will cause a huge error in the middle time of the tracking solution. Therefore, in order to solve such problems, an algorithm for evaluating the efficiency of medical assistants based on the analytical solution of linear systems is proposed. The algorithm is based on a large number of data samples and uses linear modeling to obtain the analytical solution of each task. When the data samples are rich, it will be infinitely close to the real situation. At the same time, each task will be processed by staff every day, and the accumulation of data will quickly become rich, which can reflect the real situation relatively quickly.
[0022] The present invention utilizes mathematical models to provide a solution to efficiency evaluation in complex situations while avoiding a large amount of repetitive tracking work, bypassing the drawbacks of traditional methods. At the same time, based on big data, the real efficiency situation can be quickly and accurately obtained, which facilitates the optimization and adjustment of difficult tasks.
[0023] The specific steps are as follows: 1. By dividing the work according to the actual application scenarios, modular task items are obtained, and finally these task items are evaluated. The evaluation method is determined by time, that is, the longer the time is, the lower the efficiency; 2. Model modular tasks; 3. Build a model and solve it (1) Assumptions By assuming that it is the time taken to complete the task today, all the time in the defined time interval today is constructed. Task items can be represented by X, and different task representation methods are X i ; Today X i The number of completed tasks corresponding to k i ; The completion time of each task is t i ; i is the corresponding project number; (2) Model establishment Based on the assumption, it is easy to get the identity: if we work 8 hours a day, in minutes, the total is 480 minutes; ; The task items obtained by different pipelines are different, and the completion is also different. Therefore, the same equation can be established for the data under each account; .
[0024] When the amount of equation data exceeds the number of tasks, you can try to find an analytical solution. If a linear relationship appears between two equations, continue to accumulate data. When the row rank of the linear system is greater than or equal to the column rank, the linear system can obtain an analytical solution, and the result is the time evaluation of each task.
[0025] After obtaining the usage time, you can compare it according to the actual scenario, or use the same account data alone to solve it, compare it between different accounts, and then use it for data analysis and optimization.
[0026] Example 1: Modularize the work to be done 1. Modularization tasks that need to be evaluated: First, the tasks to be evaluated are divided into modules so that we know which projects will be evaluated.
[0027] 2. Record the end time of the task: This step requires a simple tracking point to record the time when the task is finally completed, so as to determine the day when the task is completed. Because it involves multi-end operations, it may be necessary for the other end to respond to the task before the work is completed, so there will be multiple tasks in parallel. Extending to the actual scenario, it is possible that the task started today will not be completed until tomorrow, so other tasks will be carried out simultaneously. However, in the present invention, by ignoring the start time, it is determined that the task that ends today is completed today.
[0028] 3. Set aside a certain period of time (working time) This step is to determine the work that needs to be done in a certain time interval so as to establish a relationship between the remaining and completed tasks. Generally, the working hours of the working day are selected, such as 9:00-12:00, 14:00-17:00. 4. Count the tasks completed during this period of time every day Get which module tasks have been completed that day and how many of each have been completed. These together constitute the total duration of this period.
[0029] Example 2: Make assumptions and build models based on the divided tasks (1) Assumptions a. Assume that all work content comes from well-divided task projects; b. Assume that all work completion methods need to be operated through the platform; c. Assume that all work is completed within a fixed time period; d. The time to complete all tasks is greater than 0; e. The number of completed tasks is greater than or equal to 0; (2) Mathematical symbol P i Pi a. The task item set is P, i is the corresponding task number, P i corresponds to the i-th task; b. Completion time: t i , t i For task P i For application time; c. Completed quantity: K i , K i For task P i The corresponding completion quantity; d. The time taken to complete the tasks on the day constitutes a set fixed time, the length of the fixed time period is C; (3) Model building According to assumption d, the tasks completed on the day constitute the total working time of the day, so the number of all projects completed on the day and the sum of the corresponding time should be this fixed time; that is, the relationship for each day is: ; As time goes by, a large number of equations will be obtained, and the following equations can be obtained: ; In the matrix, J represents the number of equations, K represents the coefficient matrix, T represents the time matrix corresponding to each task, and C represents the matrix composed of fixed time for each equation. We can further obtain: .
[0030] Example 3: Model Solution (1) Model solution conditions For formula (2), we want to obtain the t of each project i , that is, through formula (2) we can get i Analytical solution. Then the condition that the coefficient matrix K is a non-singular matrix must be met, that is: ; To ensure that K is a non-singular matrix, the number of equations should at least be greater than or equal to the number of tasks, that is, in the equation, j ≥ n must be satisfied, and the nonlinear correlation between the equations must also be ensured.
[0031] After satisfying the above conditions and obtaining the analytical solution of the time required for each project, it is still necessary to ensure that the assumption condition d is satisfied, that is, t i >0.
[0032] (2) Model algorithm From a theoretical perspective, the analytical solution obtained by the model is the desired result, that is, the average time taken for each task, which can be used to evaluate efficiency. However, during the solution process, it was found that the analytical solution would inevitably result in negative values. According to the limitation of assumption d, the result cannot have negative values. Therefore, this problem can be further transformed into a function fitting problem.
[0033] Since the specific task is a fixed item, this linear solution problem is transformed into a linear function fitting problem, in which the characteristic parameter needs to be greater than 0.
[0034] Specifically, a linear regression model is used for fitting. The python pseudo code is as follows: def get_event_time_cost(data_path: str): # -- step 1 load data -- # x, y = get_event_and_time(data_path) # -- step 2 caculate each event time cost -- # reg = LinearRegression(positive=True, fit_intercept=False) reg.fit(x, y) print(reg.score(x, y)) weight = reg.coef_ # -- step 3 display result -- # table = PrettyTable() table.field_names = [f"Average time of event {i} (min)" for i in range(len(weight))] table.add_row([f'{v:.2f}' for v in weight.tolist()]).
[0035] By quickly accumulating data every day and eventually training with large samples, the feature parameters will be continuously updated, making the expression of the real situation more and more accurate. Therefore, the time evaluation of each task will become more and more accurate. After convergence, it will bypass the traditional tracking point and obtain the efficiency evaluation result.
[0036] The above-described embodiments only express a certain implementation mode of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the patent scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.
Claims
1. An employee efficiency evaluation algorithm based on linear model non-embedded point, characterized in that: include: Modeling the service items of online medical consultation on the Internet in advance to form a database, wherein the database has divided task modules, and based on the evaluation items corresponding to the processes in the service item library, an evaluation item record model is formed to form a memory library; Call the database, task module and memory according to the project task to be evaluated, process under the task module divided by the target, encapsulate the work content, and obtain the efficiency evaluation of a certain task project; The obtained efficiency evaluation data of a certain task project is recorded according to the data in the memory bank, and the evaluation time of the task project can be obtained through the established equation.
2. According to claim 1, a linear model-based non-buried point employee efficiency evaluation algorithm is characterized in that: The database of online medical consultation services in advance includes: The average time required to complete the existing standardized (young people) daily tasks and the average time required to complete the existing specialized (elderly people) tasks are taken and entered into the database to form a database for completing tasks with specific standard time.
3. According to claim 1, the employee efficiency evaluation algorithm based on linear model non-buried point is characterized in that: The efficiency assessment of a task project includes: Based on the database and memory, a time comparison library is established in the form of comparing the existing task completion time with the ongoing task completion time; Set up a separate time point for a task item; Using the corresponding divided task modules, the completion time of any task item and the task items in the comparison library are set to the same module; Repeat setting the time point for the next task item to obtain the end time under the same task; When multiple tasks are running in parallel, the time of other tasks is recorded synchronously, and the start time is ignored, and the end time is recorded uniformly; When tasks opened on multiple terminals are performed synchronously, the task start time is ignored and only the end time is recognized.
4. According to claim 3, the employee efficiency evaluation algorithm based on linear model non-buried point is characterized in that: The efficiency assessment of a task project also includes: Define the working hours within a certain period of time, generally selected as the working hours of a normal working day, and determine the work that needs to be done in a certain time interval in order to establish a relationship with the number of completed tasks; The designated working time periods can be modified at will according to actual conditions.
5. According to claim 4, the employee efficiency evaluation algorithm based on linear model non-buried point is characterized in that: The efficiency assessment of a task project also includes: The number of task items completed within a specific time period each day is counted to form a single-day project measurement library, which can be compared with the existing standardized number of projects per day to obtain efficiency evaluation results.
6. According to the employee efficiency evaluation algorithm based on linear model and non-embedded points in claim 1, in general, the total time c per day can be freely determined. The current feature is that the evaluation time of the task project is based on 8 hours of work per day, in minutes, that is, the determination formula under the benchmark of c taking a value of 480 minutes is: ； The task items can be represented by X, and different task representation methods are X i ; Today X i The number of tasks completed is K i ; The completion time of each task is t i ; i is the corresponding project number.
7. The employee efficiency evaluation algorithm based on linear model non-buried point according to claim 6 is characterized in that: Different pipelines have different task items and different completions. The same equation is established for the data under each account: 。 8. The employee efficiency evaluation algorithm based on linear model non-buried points according to claims 6-7 is characterized in that: This involves solving the equation: When the amount of data in the equation exceeds the number of tasks, the parsing condition is met; When a linear relationship appears between two equations, the amount of data continues to be accumulated. When the row rank of the linear system is greater than or equal to the column rank, the linear system can be solved analytically, and the result is the time evaluation of each task.
9. The employee efficiency evaluation algorithm based on linear model non-buried point according to claim 8 is characterized in that: The evaluation results after the equation is analyzed can be compared according to actual scenarios, or solved using the same account data alone, compared between different accounts, and then used for data analysis and optimization.