Method, medium and device for constructing dynamic saturation evaluation model based on time and resource allocation for language service project management

Abstract

The invention belongs to the technical field of machine translation, and provides a method, medium and device for constructing a dynamic saturation evaluation model based on time and resource allocation for language service item management. The method comprises the following steps: S1, collecting historical data; step S2, preprocessing the historical data; s3, fitting each parameter in the model formula by using a multiple linear regression model, and determining the weight of each variable; s4, on the basis of the model formula in the step S3, part of the historical data preprocessed in the step S2 is used for model training, the other part of the historical data is used for testing and evaluating the prediction performance of the model, and model optimization is conducted according to the evaluation result; and S5, optimizing the model formula according to a model training result, and removing variables with relatively small weights to obtain a final model formula. The method has more accurate resource allocation: more accurate resource and time management is ensured by considering a plurality of dynamic factors, such as language-to-difficulty and project urgency.

Description

Technical Field

[0001] The present invention belongs to the technical field of machine translation, and in particular, relates to a method, medium and device for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management. Background Art

[0002] In the existing field of language service project management, most project management methods rely on static workload estimation models. These models usually determine the time frame and resource requirements of the project based on fixed parameters, such as the total number of words in the project and the estimated translation speed. However, these methods ignore the impact of dynamic factors such as the difficulty of language pairs, the complexity of professional fields, project urgency and personal work efficiency. This leads to inaccurate resource allocation, project delays and cost overruns. Summary of the invention

[0003] The purpose of the present invention is to provide a method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management, so as to solve the technical problems existing in the prior art.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows: A method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management, comprising: Step S1: Collect historical data, including total translated words, language pair difficulty coefficient, field difficulty coefficient, total effective working hours, efficiency coefficient, project management time, reserved time, urgency coefficient, actual completion time, and project results; Step S2: preprocessing the historical data; Step S3: Use the multivariate linear regression model to fit the parameters in the model formula and determine the weight of each variable. The model formula is as follows: SAT = ((T * LDC * FDC) / ((TWH * EC) - PMT -BT)) * UC) Among them, SAT is saturation, T is the total number of words, LDC is the language pair difficulty coefficient, FDC is the field difficulty coefficient, TWH is the total time available for actual translation work in the project, EC is the efficiency coefficient, PMT is the project management time, BT is the buffer time, and UC is the urgency coefficient; Step S4: Based on the model formula of step S3, use part of the historical data preprocessed in step S2 for model training and the other part for testing to evaluate the prediction performance of the model, including the mean square error. , mean absolute error , R-squared value: , where n is the number of samples, is the true value of the i-th sample, is the predicted value of the ith sample, is the average of all true values, and the model is tuned based on the evaluation results; Step S5: According to the model training results, optimize the model formula, remove variables with smaller weights, and obtain the final model formula: (1) (2) In the above formula, the meanings of each variable are as follows: Saturation SAT: workload pressure of the project; Total word count T: the total number of translated words in the project; Language Pair Difficulty Coefficient (LDC): A weighting factor based on the difficulty of the language pair; Field Difficulty Coefficient (FDC): A weighted coefficient based on the professionalism and complexity of the translation field; Project time window TW: effective working hours from the start to the end of the project; Urgency coefficient UC: a weighted coefficient based on the urgency of the project; Efficiency coefficient EC: a weighted coefficient that takes into account individual work efficiency; Project Management Time PMT: The time required for project management and coordination; Buffer time BT: The time reserved for dealing with unforeseen events.

[0005] In a preferred embodiment, in step S2, the preprocessing includes filling or deleting missing data, scaling the data to a uniform range, detecting and processing outliers in the data (to ensure that the data ranges of different variables are consistent, and in addition, it is necessary to detect and process outliers in the data to ensure the accuracy of model fitting).

[0006] In a preferred embodiment, the specific method of using the multiple linear regression model to fit the parameters in the model formula in step S3 to determine the weight of each variable is as follows: Step S3.1: Divide the historical data into independent variables and dependent variables, the independent variables include T, LDC, FDC, TWH, EC, PMT, BT, UC, and the dependent variable is the actual completion time or project results; Step S3.2: Use the multiple linear regression model to estimate the coefficients of the independent and dependent variables to obtain the initial form of the model; Step S3.3: Use statistical significance tests to determine whether the weight of each variable is significant, remove variables with insignificant weights, and retain variables with significant weights; Step S3.4: Check the fitting effect of the model formula based on residual analysis. If the fitting effect of the model formula is poor, re-execute steps S3.2 to 3.3. If the fitting effect of the model formula is good, the variables with significant weights retained in step S3.3 are used as the determined variables and their weights.

[0007] In a preferred embodiment, in step S4, the model tuning method is as follows: if the evaluation indicators of the model do not meet expectations, the model is optimized by cross-validation, feature engineering or adjusting regularization parameters.

[0008] In a preferred embodiment, in step S5, the optimization method is as follows: simplify the model formula by a stepwise regression method or a Lasso regression method, and remove variables with smaller weights.

[0009] In a preferred embodiment, the method for determining the variable with a smaller weight is as follows: by comparing the standardized regression coefficients of various variables on the target variable, it is determined which variables have a smaller impact.

[0010] In order to achieve the above objectives, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management as described above.

[0011] In order to achieve the above-mentioned purpose, the present invention also provides a device for implementing a method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management, comprising: a processor and a memory, the memory being used to store a computer program; the processor is connected to the memory, and is used to execute the computer program stored in the memory, so that the device for implementing a method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management executes the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management as described above.

[0012] Compared with the prior art, the present invention has the following beneficial effects: (1) More precise resource allocation according to the present invention: by considering multiple dynamic factors such as language pair difficulty and project urgency, more precise resource and time management is ensured.

[0013] (2) According to the present invention, it is flexible: it can adapt to different types of project needs, including urgent and difficult projects.

[0014] (3) According to the present invention, efficiency is improved: by optimizing resource allocation, project delays and cost overruns are reduced, thereby improving overall work efficiency.

[0015] (4) According to the present invention, risk reduction: Buffer time and project management time are introduced to reduce the impact of unforeseen events on project progress. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a schematic diagram of the process of Example 1 of the present invention.

[0017] Figure 2 Schematic diagram of the impact of different parameters on project saturation assessment. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to have a clearer understanding and knowledge of the present invention, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described below are only used to explain the present invention, and are convenient for understanding. The technical solutions provided by the present invention are not limited to the technical solutions provided by the following embodiments, and the technical solutions provided by the embodiments should not limit the protection scope of the present invention.

[0019] It should be noted that the illustrations provided in the following embodiments are only used to illustrate the basic concept of the present invention in a schematic manner. Therefore, the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the form, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0020] Example 1 like Figure 1 , 2 As shown, this embodiment provides a method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management, which includes the following steps: 1. Step S1: Collect historical data In this step, historical data includes total translated words, language pair difficulty coefficient, field difficulty coefficient, total effective working hours, efficiency coefficient, project management time, reserved time, urgency coefficient, actual completion time, and project results.

[0021] 2. Step S2: Preprocessing historical data Preprocessing methods include filling or deleting missing data, scaling data to a uniform range, detecting and processing outliers in the data to ensure that the data ranges of different variables are consistent. In addition, it is necessary to detect and process outliers in the data to ensure the accuracy of model fitting.

[0022] 3. Step S3: Use the multivariate linear regression model to fit the parameters in the model formula and determine the weights of each variable The model formula is expressed as follows: SAT = ((T * LDC * FDC) / ((TWH * EC) - PMT -BT)) * UC) Among them, SAT is saturation, T is the total number of words, LDC is the language pair difficulty coefficient, FDC is the field difficulty coefficient, TWH is the total time available for actual translation work in the project (the result after deducting non-working time between the start and deadline of the project), EC is the efficiency coefficient, PMT is the project management time, BT is the buffer time, and UC is the urgency coefficient.

[0023] In this step, the multiple linear regression model is used to fit the parameters in the model formula. The specific method for determining the weight of each variable is as follows: Step S3.1: Divide the historical data into independent variables and dependent variables, the independent variables include T, LDC, FDC, TWH, EC, PMT, BT, UC, and the dependent variable is the actual completion time or project results; Step S3.2: Use the multiple linear regression model to estimate the coefficients of the independent and dependent variables to obtain the initial form of the model; Step S3.3: Use statistical significance tests to determine whether the weight of each variable is significant, remove variables with insignificant weights, and retain variables with significant weights; Step S3.3 is the first variable screening, focusing on significance tests to ensure that each variable in the model has a statistically significant effect on the target variable and remove insignificant variables. This step makes the model more interpretable; Step S3.4: Check the fitting effect of the model formula based on residual analysis. If the fitting effect of the model formula is poor, re-execute steps S3.2 to S3.3 until it meets the conditions for a good fitting effect of the model formula. If the fitting effect of the model formula is good, the variables with significant weights retained in step S3.3 are used as the determined variables and their weights.

[0024] Step S4: Evaluate the prediction performance of the model and tune the model based on the evaluation results Based on the model formula of step S3, use part of the historical data preprocessed in step S2 for model training and the other part for testing to evaluate the prediction performance of the model, including the mean square error. , mean absolute error , R-squared value: , where n is the number of samples, is the true value of the i-th sample, is the predicted value of the ith sample, It is the average of all true values, and the model is tuned according to the evaluation results.

[0025] In this step, the model tuning method is as follows: If the evaluation indicators of the model do not meet expectations, the model is optimized through cross-validation, feature engineering (such as removing variables with strong collinearity or adding polynomial terms), or adjusting regularization parameters.

[0026] 5. Step S5: Optimize the model formula based on the model training results According to the model training results, the model formula is optimized and variables with smaller weights are removed. The optimization method is as follows: Simplify the model formula through stepwise regression method or Lasso regression method, and remove variables with smaller weights. The method for determining variables with smaller weights is as follows: Determine which variables have smaller effects by comparing the standardized regression coefficients of each variable on the target variable.

[0027] Through the above, the final model formula is obtained, which is expressed as follows: (1) (2) In the above formula, the meanings of each variable are as follows: Saturation SAT: workload pressure of the project; Total word count T: the total number of translated words in the project; Language Pair Difficulty Coefficient (LDC): A weighting factor based on the difficulty of the language pair; Field Difficulty Coefficient (FDC): A weighted coefficient based on the professionalism and complexity of the translation field; Project time window TW: effective working hours from the start to the end of the project; Urgency coefficient UC: a weighted coefficient based on the urgency of the project; Efficiency coefficient EC: a weighted coefficient that takes into account individual work efficiency; Project Management Time PMT: The time required for project management and coordination; Buffer time BT: The time reserved for dealing with unforeseen events.

[0028] Step S5 is to optimize the model again based on the weight size after the model has passed the significance screening. Here, the model is further simplified by eliminating variables with smaller weights. Different from the significance-based screening in S3.3, this is based on the comparison of standardized weights to reduce variables that contribute less to the target variable, making the model simpler and more generalizable. The purpose of the two weight optimizations is to gradually streamline the model to ensure that it has higher interpretability and convenience in practical application while meeting the prediction accuracy. The screening in step S3.3 reduces irrelevant or weakly correlated variables, while the screening in step S5 further removes variables that have little effect on saturation while ensuring model performance.

[0029] Example 2 This embodiment provides a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to implement the method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management provided in Example 1. A person of ordinary skill in the art can understand that all or part of the steps of implementing the method provided in Example 1 can be completed by hardware related to the computer program, and the above-mentioned computer program can be stored in a computer-readable storage medium, and when the program is executed, the steps of the method provided in Example 1 are executed; and the above-mentioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

[0030] Example 3 This embodiment provides an apparatus for implementing a method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management, comprising: a processor and a memory; the memory is used to store a computer program; the processor is connected to the memory, and is used to execute the computer program stored in the memory, so that the apparatus for implementing a method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management executes the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management provided in Example 1.

[0031] Specifically, the memory includes: ROM, RAM, disk, USB flash drive, memory card or CD and other media that can store program codes.

[0032] Preferably, the processor can be a general-purpose processor, including a central processing unit, a network processor, etc.; it can also be a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component.

[0033] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management, characterized in that: include: Step S1: Collect historical data, including total translated words, language pair difficulty coefficient, field difficulty coefficient, total effective working hours, efficiency coefficient, project management time, reserved time, urgency coefficient, actual completion time, and project results; Step S2: preprocessing the historical data; Step S3: Use the multivariate linear regression model to fit the parameters in the model formula and determine the weight of each variable. The model formula is as follows: SAT = ((T * LDC * FDC) / ((TWH * EC) - PMT -BT)) * UC) Among them, SAT is saturation, T is the total number of words, LDC is the language pair difficulty coefficient, FDC is the field difficulty coefficient, TWH is the total time available for actual translation work in the project, EC is the efficiency coefficient, PMT is the project management time, BT is the buffer time, and UC is the urgency coefficient; Step S4: Based on the model formula of step S3, use part of the historical data preprocessed in step S2 for model training and the other part for testing to evaluate the prediction performance of the model, including the mean square error. , mean absolute error , R-squared value: , where n is the number of samples, is the true value of the i-th sample, is the predicted value of the ith sample, is the average of all true values, and the model is tuned based on the evaluation results; Step S5: According to the model training results, optimize the model formula, remove variables with smaller weights, and obtain the final model formula: （1） （2） In the above formula, the meanings of each variable are as follows: Saturation SAT: workload pressure of the project; Total word count T: the total number of translated words in the project; Language Pair Difficulty Coefficient (LDC): A weighting factor based on the difficulty of the language pair; Field Difficulty Coefficient (FDC): A weighted coefficient based on the professionalism and complexity of the translation field; Project time window TW: effective working hours from the start to the end of the project; Urgency coefficient UC: a weighted coefficient based on the urgency of the project; Efficiency coefficient EC: a weighted coefficient that takes into account individual work efficiency; Project Management Time PMT: The time required for project management and coordination; Buffer time BT: The time reserved for dealing with unforeseen events.

2. The method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management according to claim 1, characterized in that: In step S2, the preprocessing includes filling or deleting missing data, scaling the data to a uniform range, and detecting and processing outliers in the data.

3. The method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management according to claim 2, characterized in that: In step S3, the specific method of using the multiple linear regression model to fit the parameters in the model formula and determining the weight of each variable is as follows: Step S3.1: Divide the historical data into independent variables and dependent variables, the independent variables include T, LDC, FDC, TWH, EC, PMT, BT, UC, and the dependent variable is the actual completion time or project results; Step S3.2: Use the multiple linear regression model to estimate the coefficients of the independent and dependent variables to obtain the initial form of the model; Step S3.3: Use statistical significance tests to determine whether the weight of each variable is significant, remove variables with insignificant weights, and retain variables with significant weights; Step S3.4: Check the fitting effect of the model formula based on residual analysis. If the fitting effect of the model formula is poor, re-execute steps S3.2 to S3.

3. If the fitting effect of the model formula is good, the variables with significant weights retained in step S3.3 are used as the determined variables and their weights.

4. The method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management according to claim 3, characterized in that: In step S4, the model tuning method is as follows: if the evaluation index of the model does not meet expectations, the model is optimized by cross-validation, feature engineering or adjusting regularization parameters.

5. The method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management according to claim 4, characterized in that: In step S5, the optimization method is as follows: simplify the model formula by using a stepwise regression method or a Lasso regression method, and remove variables with smaller weights.

6. The method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management according to claim 5, characterized in that: The method for determining the variable with a smaller weight is as follows: by comparing the standardized regression coefficients of various variables on the target variable, determine which variables have a smaller impact.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program is executed by a processor to implement the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management according to any one of claims 1 to 6.

8. A device for implementing a method for constructing a dynamic saturation evaluation model based on time and resource allocation for language service project management, comprising: A processor and a memory, characterized in that the memory is used to store a computer program; the processor is connected to the memory and is used to execute the computer program stored in the memory, so that the device for implementing the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management executes the method for constructing a dynamic saturation assessment model based on time and resource allocation for language service project management as described in any one of claims 1 to 6.

Images