Project management method, influence factor prediction model training method and related equipment

Abstract

The invention discloses a project management method, and the method comprises the steps: obtaining a to-be-adjusted project which corresponds to a change demand; performing sweep prediction processing based on the change demand to obtain a sweep factor corresponding to the change demand; performing classification prediction processing on the change demand based on a preset demand classification model to obtain a user demand type corresponding to the change demand; based on the sweep factor and the user demand type, determining an item adjustment strategy corresponding to the change demand; and performing item adjustment on the to-be-adjusted item based on the item adjustment strategy to obtain an adjusted item meeting the change requirement. According to the project management method and device, the corresponding sweep factor and the user demand type can be accurately predicted by changing the demand, then the more accurate project adjustment strategy is determined through the sweep factor and the user demand type, and the management accuracy of project management can be effectively improved through the project adjustment strategy.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and particularly to a project management method, apparatus, computer device, and storage medium. Background Art

[0002] In the engineering field, requirement change is a common and complex issue. As the project progresses, changes in customer requirements, market environment, or technological implementation may all lead to requirement changes. Traditional requirement management models (such as the waterfall model) adopt a linear and phased approach, dividing the requirement management process into strict sequential phases (such as requirement analysis, design, development, testing, and maintenance). Each phase requires the completion of clear deliverables and passing reviews before proceeding to the next phase. However, the characteristic of this method is that requirements are determined upfront. That is, at the initial stage of the project, a fixed specification document is formed through detailed requirement collection and analysis to ensure the traceability and stability of requirements. However, limited by the detailed requirement collection and analysis at the initial stage of the project, the accuracy of project management during requirement changes depends on the initial requirement collection and analysis, and it is often difficult to accurately predict the impact of requirement changes on project schedule, cost, and quality, resulting in project delays, cost overruns, or substandard quality. Existing project management methods have problems with low management accuracy during requirement changes.

[0003] Therefore, how to provide a project management method that can improve management accuracy during requirement changes has become an urgent problem to be solved. Summary of the Invention

[0004] Based on this, it is necessary to provide a project management method, apparatus, computer device, and storage medium for the above technical problems to solve the problem of low management accuracy of traditional methods.

[0005] A project management method, the method includes: Obtain a project to be adjusted, and the project to be adjusted corresponds to a change requirement; Perform a ripple prediction process based on the change requirement to obtain a ripple factor corresponding to the change requirement; Perform a classification prediction process on the change requirement based on a preset requirement classification model to obtain a user requirement type corresponding to the change requirement; Determine a project adjustment strategy corresponding to the change requirement based on the ripple factor and the user requirement type; Perform project adjustment on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirement.

[0006] Optionally, the performing a ripple prediction process based on the change requirement to obtain a ripple factor corresponding to the change requirement includes: Perform predictive processing based on the change requirements to obtain the code modification amount, project completion time, and project completion quality for implementing the change requirements; Determine the impact factor based on the code modification amount, project completion time, and project completion quality.

[0007] Optionally, the determining the impact factor based on the code modification amount, project completion time, and project completion quality includes: Calculate the proportion of code modification amount based on the code modification amount and the total code amount of the project to be adjusted; Calculate the project schedule delay rate based on the project completion time and the planned completion time of the project to be adjusted; Calculate the project quality degradation rate based on the project completion quality and the planned completion quality of the project to be adjusted; Calculate the impact factor based on the proportion of code modification amount, the project schedule delay rate, and the project quality degradation rate.

[0008] Optionally, the calculating the impact factor based on the proportion of code modification amount, the project schedule delay rate, and the project quality degradation rate includes: Obtain the project type of the project to be adjusted; Set a first coefficient for the proportion of code modification amount, a second coefficient for the project schedule delay rate, and a third coefficient for the project quality degradation rate based on the project type; Perform normalization processing on the proportion of code modification amount based on the first coefficient to obtain a first factor; Perform normalization processing on the project schedule delay rate based on the second coefficient to obtain a second factor; Perform normalization processing on the project quality degradation rate based on the third coefficient to obtain a third factor; Calculate the impact factor based on the first factor, the second factor, and the third factor.

[0009] Optionally, the project to be adjusted includes project function modules and project tasks, and the method further includes: Take the change requirements, project function modules, and project tasks as nodes, and use the dependency relationships between the nodes as connection edges to construct a requirements impact graph; Determine the requirements change propagation path corresponding to the project to be adjusted based on the requirements impact graph; Provide the requirements change propagation path to the manager of the project to be adjusted so that the manager can perform the analysis of the change requirements.

[0010] A training method for a ripple factor prediction model, the method comprising: Obtain a historical project dataset and a ripple factor prediction model to be trained, the historical project dataset including historical change requirements and the actual ripple factors corresponding to the historical change requirements, the actual ripple factors being calculated based on the proportion of actual code modification amount, actual project schedule delay rate, and actual project quality degradation rate obtained after implementing the historical change requirements in the historical project dataset; Input the historical change requirements into the ripple factor prediction model to be trained for ripple prediction processing to obtain a prediction result; Calculate the loss value between the prediction result and the corresponding actual ripple factor; Taking minimizing the loss value as an optimization objective, adjust the parameters of the ripple factor prediction model to be trained until the loss value converges at the minimum or the number of adjustment times reaches a preset number of times, then stop the adjustment to obtain a trained ripple factor prediction model, and the trained ripple factor prediction model is used to predict ripple factors.

[0011] A project management device, the device comprising: A first acquisition module, configured to acquire a project to be adjusted, and the project to be adjusted corresponds to a change requirement; A first prediction module, configured to perform ripple prediction processing based on the change requirement to obtain the ripple factor corresponding to the change requirement; A second prediction module, configured to perform classification prediction processing on the change requirement based on a preset requirement classification model to obtain the user requirement type corresponding to the change requirement; A first determination module, configured to determine a project adjustment strategy corresponding to the change requirement based on the ripple factor and the user requirement type; A first adjustment module, configured to perform project adjustment on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirement.

[0012] A training device for a ripple factor prediction model, the device comprising: A second acquisition module, configured to acquire a historical project dataset and a ripple factor prediction model to be trained, the historical project dataset including historical change requirements and the actual ripple factors corresponding to the historical change requirements, the actual ripple factors being calculated based on the proportion of actual code modification amount, actual project schedule delay rate, and actual project quality degradation rate obtained after implementing the historical change requirements in the historical project dataset; A first input module, configured to input the historical change requirements into the ripple factor prediction model to be trained for ripple prediction processing to obtain a prediction result; A first calculation module for calculating a loss value between the predicted result and the corresponding actual influence factor. A first adjustment module for adjusting the parameters of the influence factor prediction model to be trained with the objective of minimizing the loss value until the loss value converges at the minimum or the number of adjustment times reaches a preset number, and then stopping the adjustment to obtain a trained influence factor prediction model, which is used to predict the influence factor.

[0013] A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the above project management method or the training method of the influence factor prediction model is implemented.

[0014] A readable storage medium stores computer-readable instructions, which implement the above project management method or the training method of the influence factor prediction model when executed by a processor.

[0015] The above project management method, device, computer device, and storage medium include: obtaining a project to be adjusted, where the project to be adjusted corresponds to a change requirement; performing an influence prediction process based on the change requirement to obtain an influence factor corresponding to the change requirement; performing a classification prediction process on the change requirement based on a preset requirement classification model to obtain a user requirement type corresponding to the change requirement; determining a project adjustment strategy corresponding to the change requirement based on the influence factor and the user requirement type; and performing a project adjustment on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirement. Through the change requirement, the corresponding influence factor and user requirement type can be accurately predicted, and then a more accurate project adjustment strategy can be determined through the influence factor and the user requirement type. The project adjustment strategy can effectively improve the management accuracy of project management. Description of the Drawings

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of a project management method provided by an embodiment of the present invention; Figure 2 It is a flowchart of a training method of an influence factor prediction model provided by an embodiment of the present invention; Figure 3It is a schematic structural diagram of a project management device provided by an embodiment of the present invention; Figure 4 It is a schematic structural diagram of a training device for a ripple factor prediction model provided by an embodiment of the present invention; Figure 5 It is a schematic diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] In one embodiment, as Figure 1 shown, a project management method is provided, including the following steps: 101. Obtain the project to be adjusted.

[0020] In the embodiment of the present invention, the above project management method can be applied to a project management platform, and the above project management platform can be constructed by a server or a server cluster. The above server or server cluster can be any electronic device with functions such as data processing, data analysis, data transmission, and data storage.

[0021] The above project to be adjusted corresponds to a change requirement. The project to be adjusted refers to a system project that has entered the development stage or has been deployed and is being operated or iterated. Its structure and function have not been finally frozen and may still require re-scheduling or technical adjustment due to external change requirements. This project often includes multiple functional modules, subtasks, personnel division of labor, version plans, and other elements.

[0022] After obtaining the project to be adjusted, the change requirements associated with the project can be further extracted. The so-called change requirements refer to new requirements or modification opinions put forward by users, customers, or business parties for the existing system functions, which may be manifested as adding modules, modifying logic, or adjusting interfaces, etc. Usually, they are recorded in the requirements management system and are uniquely associated with the project.

[0023] The above project can specifically be a software engineering project, a hardware engineering project, or an IT operation and maintenance project, etc. The above project to be adjusted can be specifically understood as a project with change requirements, and the above change requirements can be obtained by the user uploading them to the above project management platform through the above data transmission function.

[0024] 102. Perform ripple prediction processing based on the change requirements to obtain the ripple factor corresponding to the change requirements.

[0025] In the embodiments of the present invention, for each change requirement, impact prediction processing is performed to evaluate its impact intensity on the current project scope, structure, and schedule. The goal of this processing is to output a quantitative value, that is, the impact factor, which is used to measure the development workload, cycle delay, and quality fluctuations that may be caused by this change.

[0026] The impact prediction processing can be implemented based on a pre-trained impact factor prediction model. This model uses historical project change data as training samples to learn the mapping relationship between the characteristics of change requirements (such as requirement type, number of involved modules, module coupling relationship, etc.) and their actual impacts. During prediction, the system inputs the structured information of the current change requirement, and the model outputs a real value as the impact factor of this requirement.

[0027] The impact factor is essentially a composite metric, usually calculated through the following factors: ① the proportion of code modifications caused by the change; ② the project schedule delay rate; ③ the project quality decline rate, and combined with weighted coefficients according to the project type. This metric reflects the "impact scope" of this change in the technical, time, and quality dimensions, and is the basis for formulating subsequent adjustment strategies.

[0028] 103. Perform classification prediction processing on the change requirements based on a preset requirement classification model to obtain the user requirement type corresponding to the change requirements.

[0029] In the embodiments of the present invention, the user requirement type refers to the subjective expectation degree and satisfaction fluctuation characteristics of users for a certain function change. It can include basic type, expected type, and exciting type. Basic requirements are the basic requirements that must be met, and their absence will cause strong dissatisfaction, but their satisfaction will not lead to an increase in satisfaction; expected requirements refer to the functions that users explicitly hope the system to achieve; exciting requirements are innovative features that users do not express but bring a surprise experience after implementation.

[0030] The requirement classification model can be constructed based on implementation methods such as rule judgment, decision tree, and machine learning. The input parameters can include: change requirements, user feedback keywords corresponding to the change requirements, change origin, whether it is a high-frequency / high-weight business module, user portrait, etc. Through this model, it can assist project managers in identifying requirement priorities and value impacts, and provide user-oriented information support for project adjustment. The above requirement classification model can specifically be the Kano model.

[0031] By calling the preset requirement classification model, perform user-oriented type determination on the change requirements, so as to clarify the position of this requirement in the user value system.

[0032] 104. Based on the impact factor and user requirement type, determine the project adjustment strategy corresponding to the change requirements.

[0033] In an embodiment of the present invention, the project adjustment strategy refers to a regulation plan adopted for the current project development plan to implement the change requirements while controlling its impact scope.

[0034] Specifically, when the impact factor is relatively high and the user requirement type belongs to the exciting type, it may be determined that the change is not a core essential requirement, and the priority can be postponed for processing; on the contrary, if the impact factor is relatively low but the requirement type is the basic type, it should be promoted and implemented preferentially. The strategies may include but are not limited to: ① adjusting task priorities; ② re-planning the project stage division; ③ adjusting the personnel resource allocation; ④ introducing a risk buffer period; ⑤ rejecting the implementation of requirements with high costs and low benefits.

[0035] For example, when the requirement impact factor > 0.5, if the requirement is of the expected type, the "Minimum Viable Change" (MVC) strategy is preferentially adopted to implement only the core functions; When the requirement impact factor ≤ 0.5, if the requirement is of the basic type, it is directly incorporated into the current iteration cycle; When the requirement impact factor > 0.7, if the requirement is of the exciting type, a requirement review meeting is automatically triggered.

[0036] 105. Perform project adjustment on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirements.

[0037] In an embodiment of the present invention, based on the project adjustment strategy, scheduling optimization and resource reconstruction are performed on the original project to be adjusted, thereby generating an adjusted project that meets the change requirements. This process may involve multiple levels, including task rearrangement, module replacement, modification of the version release strategy, etc.

[0038] The specific manifestations of project adjustment in the project management platform may be operations such as updating the task status table, modifying the module delivery node, changing the developer assignment plan, increasing or decreasing the testing link, etc. The system can automatically write the updated data into the database or prompt relevant personnel for manual confirmation.

[0039] An adjusted project refers to a project version that has been adapted to the change requirements at the project logic, progress, or architecture level and can be used for subsequent development scheduling, test planning, and version release.

[0040] Perform project adjustment on the project to be adjusted through the project adjustment strategy to obtain an adjusted project that meets the change requirements, realizing a closed-loop response from requirement identification to structural adjustment, and ensuring that the change requirements are quickly incorporated into the existing project process under controlled costs and a clear path.

[0041] In an embodiment of the present invention, a project to be adjusted is obtained, and there is a change requirement corresponding to the project to be adjusted; a ripple prediction process is performed based on the change requirement to obtain a ripple factor corresponding to the change requirement; a classification prediction process is performed on the change requirement based on a preset requirement classification model to obtain a user requirement type corresponding to the change requirement; based on the ripple factor and the user requirement type, a project adjustment strategy corresponding to the change requirement is determined; and a project adjustment is performed on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirement. Through the change requirement, the corresponding ripple factor and user requirement type can be accurately predicted, and then a more accurate project adjustment strategy can be determined through the ripple factor and the user requirement type. The project management accuracy can be effectively improved through the project adjustment strategy.

[0042] It can be understood that in the specific implementation of the present application, data related to the project to be adjusted, change requirements, project types, historical project data sets, etc. are involved. When the embodiments in the present application are applied to specific products or technologies, user permission or consent needs to be obtained, and the collection, use, and processing of relevant data, as well as the construction, training, and use of the ripple factor prediction model, need to comply with relevant laws, regulations, and standards in relevant countries and regions.

[0043] Optionally, in the step of performing a ripple prediction process based on the change requirement to obtain a ripple factor corresponding to the change requirement, a prediction process can also be performed based on the change requirement to obtain the code modification amount, project completion time, and project completion quality required to implement the change requirement; and the ripple factor is determined based on the code modification amount, project completion time, and project completion quality.

[0044] In an embodiment of the present invention, the prediction process of the above change requirement can be implemented by any one of the requirement change prediction models, and the above requirement change prediction module can be constructed by one or more algorithms such as Random Forest Regressor, gradient boosting trees (such as XGBoost, LightGBM), and Multi-Layer Perceptron (MLP) regression models.

[0045] A prediction process is performed based on the change requirement to obtain the code modification amount, project completion time, and project completion quality required to implement the change requirement. Among them, the code modification amount can be expressed as the number of lines of code added, deleted, or replaced; the project completion time represents the total construction period estimated from the start of development to online delivery; and the project completion quality can be estimated by indicators such as the expected number of defects, defect density, or user acceptance pass rate.

[0046] After obtaining the above-mentioned code modification volume, project completion time, and project completion quality, parameter normalization can be performed on the above-mentioned code modification volume, project completion time, and project completion quality to obtain a code modification volume factor, a project completion time factor, and a project completion quality factor, so as to eliminate the dimensional difference and make the parameters comparable. The above parameter normalization can specifically be methods such as Min-Max Normalization, Z-Score normalization (standard deviation normalization), etc.

[0047] After obtaining the above-mentioned code modification volume factor, project completion time factor, and project completion quality factor, the above-mentioned code modification volume factor, project completion time factor, and project completion quality factor can be weighted and averaged to obtain the above-mentioned impact factor.

[0048] Optionally, in the step of determining the impact factor based on the code modification volume, project completion time, and project completion quality, the proportion of code modification can also be calculated based on the code modification volume and the total amount of code of the project to be adjusted; the project schedule delay rate can be calculated based on the project completion time and the planned completion time of the project to be adjusted; the project quality decline rate can be calculated based on the project completion quality and the planned completion quality of the project to be adjusted; the impact factor can be calculated based on the proportion of code modification, the project schedule delay rate, and the project quality decline rate.

[0049] In the embodiment of the present invention, the total amount of code refers to the total number of source code lines under the current version of the project to be adjusted, including the sum of the effective codes in the business logic, configuration files, interface definitions, scripts, etc. of each module. This indicator can be statistically obtained through the main branch in the code repository (such as Git) and used as the denominator benchmark for measuring the proportion of code modification. The planned completion time refers to the original scheduling target set for the project in the project management system, that is, the total number of days or time period expected from the start of development to the completion of delivery. This time generally comes from project approval documents, Gantt chart scheduling, or iteration schedules, representing the original progress baseline of the project. The planned completion quality refers to the delivery quality target set for the project under the original plan, which can be measured by multiple quantitative indicators, such as: The number of defects does not exceed N; The defect density is less than M per thousand lines of code; The first-pass acceptance rate of users ≥ X%; The score of the satisfaction survey ≥ Y points; The above indicators can be combined or weighted as the evaluation benchmark of "planned quality", which is usually stored in the quality management platform or project acceptance documents.

[0050] Specifically, the calculation of the proportion of code modification volume can be: code modification volume / total code volume of the items to be adjusted. The calculation of the project progress delay rate can be: (project completion time - planned completion time) / planned completion time. The calculation of the project quality decline rate can be: (project completion quality - planned completion quality) / planned completion quality.

[0051] After calculating the above three indicators, namely the proportion of code modification volume, project progress delay rate, and project quality decline rate, parameter normalization processing (such as min-max normalization) can be further performed, and then the final "ripple factor RPF" is formed through weighted combination, providing a quantitative basis for subsequent project strategy adjustment.

[0052] Optionally, in the step of calculating the ripple factor based on the proportion of code modification volume, project progress delay rate, and project quality decline rate, the project type of the item to be adjusted can also be obtained; based on the project type, the first coefficient of the proportion of code modification volume, the second coefficient of the project progress delay rate, and the third coefficient of the project quality decline rate are set; the proportion of code modification volume is normalized based on the first coefficient to obtain the first factor; the project progress delay rate is normalized based on the second coefficient to obtain the second factor; the project quality decline rate is normalized based on the third coefficient to obtain the third factor; and the ripple factor is calculated based on the first factor, the second factor, and the third factor.

[0053] In the embodiments of the present invention, the above project type can be agile development or waterfall model. Agile development projects have the characteristics of rapid iteration, small version updates, and frequent deliveries, and are often more sensitive to time, with a higher tolerance for quality control and scope expansion. The waterfall model projects have strict phase divisions (such as requirements, design, development, testing, go-live, etc.), usually with fine upfront planning, being more conservative about scope changes, and emphasizing quality and scope control.

[0054] Different project types have different tolerances for "code changes, progress delays, and quality declines". Therefore, when calculating the ripple factor, different weight coefficients should be assigned according to the project type.

[0055] Then, based on the project type, weight parameters are selected from a preset coefficient template, including: The first coefficient (α): used to measure the impact of the proportion of code modification volume on the ripple factor; The second coefficient (β): used to measure the impact of the project progress delay rate on the ripple factor; The third coefficient (γ): used to measure the impact of the project quality decline rate on the ripple factor; The sum of the coefficients should satisfy the normalization constraint: α + β + γ = 1; Then, based on these three coefficients respectively, weighted normalization processing is performed on the corresponding ratio indicators: Normalize the proportion of code modification volume based on the first coefficient to obtain the first factor; Normalize the project schedule delay rate based on the second coefficient to obtain the second factor; Normalize the project quality degradation rate based on the third coefficient to obtain the third factor; The normalization method can adopt linear normalization, Z-Score standardization, or processing with upper and lower limit templates defined according to the project scale to ensure the comparability of each indicator.

[0056] Finally, based on the above three factors, calculate the ripple factor (RPF) of the change requirement by using the weighted summation method.

[0057] The calculation of the above ripple factor can be further illustrated by the following formula:

[0058] Among them, RPF represents the ripple factor, α represents the first coefficient used to measure the impact of the proportion of code modification volume on the ripple factor; β represents the second coefficient used to measure the impact of the project schedule delay rate on the ripple factor; γ represents the third coefficient used to measure the impact of the project quality degradation rate on the ripple factor; ΔC / C represents the proportion of code modification volume, ΔT / T represents the project schedule delay rate, and ΔQ / Q represents the project quality degradation rate.

[0059] Optionally, the project management method can also use the change requirement, project function module, and project task as nodes, and use the dependency relationship between each node as the connection edge to construct a requirement ripple map; based on the requirement ripple map, determine the requirement change propagation path corresponding to the project to be adjusted; provide the requirement change propagation path to the manager of the project to be adjusted so that the manager can analyze the change requirement.

[0060] In the embodiment of the present invention, the project to be adjusted includes a project function module and a project task. The project function module refers to a functional unit divided according to the business logic in the project, such as a user management module, an order processing module, a data synchronization module, etc., which is the basic composition of the code structure and business logic. The project task refers to executable work items such as development, testing, and deployment, which are usually scheduled through a task management tool and are closely associated with specific modules.

[0061] Specifically, the dependency edges between nodes can be established by analyzing the direct association between the change requirement and the module, the call relationship between the modules, the mapping relationship between the module and the task, etc., for example: Change requirement → Affected module (function implementation dependency); Module A → Module B (call or inheritance dependency); Module → Task (realize task association).

[0062] The graph structure formed by the above dependencies is the requirements impact spectrum.

[0063] Next, based on this spectrum, propagation path analysis can be performed, that is, identifying all modules and task paths affected by the downward propagation through the dependency chain starting from the changed requirement node, forming a requirements change propagation path. This path can clearly mark the depth and breadth of the change impact, revealing the core risk points and node combinations that may cause chain effects.

[0064] Finally, the generated requirements change propagation path can be visualized or structured and output to the managers of the projects to be adjusted. Based on this, project managers can quickly understand issues such as secondary development, associated testing, version conflicts, or resource rearrangement that may be triggered by the changed requirements, thereby improving the accuracy of change analysis and decision-making efficiency. For example, when a user proposes a requirement of "adding a payment method", the system automatically marks the affected modules (such as the order system, financial system) and potential risks (such as data consistency verification).

[0065] This spectrum analysis mechanism not only improves the visibility and transparency of project management but also serves as the source of the structural features of the impact factor prediction model, providing graph neural structure support for subsequent machine learning.

[0066] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0067] In one embodiment, as Figure 2 shown, a training method for an impact factor prediction model is provided, including the following steps: 201. Obtain a historical project dataset and an impact factor prediction model to be trained.

[0068] In the embodiments of the present invention, the historical project dataset includes historical change requirements and the actual impact factors corresponding to the historical change requirements. The actual impact factors are calculated based on the proportion of actual code modification amount, actual project schedule delay rate, and actual project quality degradation rate obtained after implementing the historical change requirements in the historical project dataset.

[0069] Obtain a project dataset containing rich historical change samples and an impact factor prediction model to be trained.

[0070] The historical project dataset includes historical change requirements recorded in multiple completed projects and their corresponding actual impact factors.

[0071] The actual impact factor is the true impact degree generated by each historical change requirement during the actual implementation process, and its calculation is based on the following three key indicators: Ratio of actual code modification amount (ΔC / C): The ratio of the actual code change amount generated by the change to the total amount of the original project code; Actual project schedule delay rate (ΔT / T): The percentage of the actual construction period delayed relative to the original planned construction period; Actual project quality degradation rate (ΔQ / Q): The decline amplitude of the number of defects or other quality indicators relative to the original plan.

[0072] The above three indicators can be obtained through the project management system, code repository analysis tool, and test / defect tracking system. The actual impact factor of the historical sample is calculated through weighted combination and used as the input of the supervision signal (label). Specifically, the ratio of the actual code modification amount can be calculated by counting the number of newly added / modified code lines through the version control system (such as Git); the actual project schedule delay rate can be calculated by comparing the planned construction period with the actual construction period and calculating the number of delayed days or percentage; the actual project quality degradation rate can be calculated by counting the increase in the number of defects in the test report or evaluating the change in the quality score through the code review tool.

[0073] 202. Input the historical change requirement into the to-be-trained impact factor prediction model for impact prediction processing to obtain the prediction result.

[0074] In the embodiment of the present invention, the above historical project dataset may further include input features corresponding to each historical change requirement. When inputting the historical change requirement, the input features corresponding to each historical change requirement can be input into the to-be-trained impact factor prediction model for impact prediction.

[0075] The input features may include but are not limited to: Change requirement type (functional / non-functional); Number of involved modules; Module historical change frequency; Requirement importance level (such as Kano classification); Graph structure features (such as module dependency centrality, path depth, that is, the above-mentioned requirement impact graph).

[0076] The above to-be-trained model can adopt structures such as random forest, gradient boosting tree, deep neural network, graph neural network, etc., and support multi-dimensional feature input and continuous variable regression output.

[0077] The prediction result of the above model is the predicted impact factor value of this requirement, that is, the speculation on the impact degree of this requirement when the execution result is not observed.

[0078] 203. Calculate the loss value between the predicted result and the corresponding actual impact factor.

[0079] In the embodiment of the present invention, calculate the error between the prediction result of the calculation model and the true impact factor in the historical record, and generate a loss value. The loss value can be in the form of mean square error (MSE), mean absolute error (MAE), etc.

[0080] 204. Taking the minimization of the loss value as the optimization goal, adjust the parameters of the impact factor prediction model to be trained until the loss value converges at the minimum or the number of adjustment times reaches the preset number of times, then stop the adjustment to obtain the trained impact factor prediction model.

[0081] In the embodiment of the present invention, the trained impact factor prediction model is used to predict the impact factor of the project management method. The parameters of the model to be trained can be iteratively updated by taking the minimization of the loss value as the optimization goal and using methods such as gradient descent, stochastic optimization, and tree structure pruning. When the loss value reaches near the preset minimum value, or the model training reaches the maximum number of epochs / convergence threshold, the training process ends, and the trained impact factor prediction model is output. This model can be deployed in the project management platform. When a newly proposed change requirement is encountered in the future, it can predict the value of its impact factor in real time to assist managers in formulating reasonable project adjustment strategies and realizing the automation of change impact assessment.

[0082] Through the training method in this embodiment, the real feedback data in historical projects can be fully utilized to automatically construct a change impact model with prediction ability, enabling future change responses to have a data-driven foundation, significantly improving the prediction accuracy and the rationality of project resource allocation, and adapting to the management requirements of different types of projects.

[0083] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0084] In one embodiment, a project management device is provided, and this project management device corresponds one-to-one with the project management method in the above embodiment. As Figure 3 shown, this project management device includes a first acquisition module 301, a first prediction module 302, a second prediction module 303, a first determination module 304, and a first adjustment module 305. The detailed descriptions of each functional module are as follows: The first acquisition module 301 is used to acquire the project to be adjusted, and the project to be adjusted corresponds to a change requirement; The first prediction module 302 is used to perform impact prediction processing based on the change requirement to obtain the impact factor corresponding to the change requirement; The second prediction module 303 is configured to perform classification prediction processing on the change requirements based on a preset requirement classification model to obtain the user requirement type corresponding to the change requirements; The first determination module 304 is configured to determine a project adjustment strategy corresponding to the change requirements based on the impact factor and the user requirement type; The first adjustment module 305 is configured to perform project adjustment on the project to be adjusted based on the project adjustment strategy to obtain an adjusted project that meets the change requirements.

[0085] For the specific limitations of the project management device, reference may be made to the limitations of the project management method in the foregoing text, which will not be elaborated herein. Each module in the above project management device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor of the computer device in the form of hardware or be independent of it, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0086] In one embodiment, a training device for an impact factor prediction model is provided. The training device for the impact factor prediction model corresponds one-to-one with the training method for the impact factor prediction model in the foregoing embodiment. As Figure 4 shown, the training device for the impact factor prediction model includes a second acquisition module 401, a first input module 402, a first calculation module 403, and a first adjustment module 404. The detailed description of each functional module is as follows: The second acquisition module 401 is configured to acquire a historical project data set and an impact factor prediction model to be trained. The historical project data set includes historical change requirements and the actual impact factors corresponding to the historical change requirements. The actual impact factors are calculated based on the proportion of the actual code modification amount, the actual project schedule delay rate, and the actual project quality degradation rate obtained after the historical project data set implements the historical change requirements; The first input module 402 is configured to input the historical change requirements into the impact factor prediction model to be trained for impact prediction processing to obtain a prediction result; The first calculation module 403 is configured to calculate a loss value between the prediction result and the corresponding actual impact factor; The first adjustment module 404 is configured to take minimizing the loss value as an optimization objective to adjust the parameters of the impact factor prediction model to be trained until the loss value converges at the minimum or the number of adjustment times reaches a preset number of times, and then stop the adjustment to obtain a trained impact factor prediction model. The trained impact factor prediction model is used to predict the impact factor described in the project management method.

[0087] For the specific limitations on the training device of the sweep factor prediction model, reference can be made to the limitations on the training method of the sweep factor prediction model in the foregoing text, which will not be elaborated here. Each module in the above-mentioned training device of the sweep factor prediction model can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0088] In one embodiment, a computer device is provided. This computer device can be a terminal device, and its internal structure diagram can be as Figure 5 shown. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium. The readable storage medium stores computer-readable instructions. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instructions are executed by the processor, a project management method or a training method of the sweep factor prediction model is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0089] In an embodiment of the present application, a computer device is provided, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above-mentioned project management method or the training method of the sweep factor prediction model are implemented.

[0090] In an embodiment of the application, a readable storage medium is provided. The readable storage medium stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps of the above-mentioned project management method or the training method of the sweep factor prediction model are implemented.

[0091] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through computer-readable instructions. The computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0092] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0093] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A project management method, characterized in that, The method includes: Obtain the item to be adjusted, and there is a change requirement corresponding to the item to be adjusted; Perform a ripple prediction process based on the change requirement to obtain a ripple factor corresponding to the change requirement; Perform a classification prediction process on the change requirement based on a preset requirement classification model to obtain a user requirement type corresponding to the change requirement; Determine a project adjustment strategy corresponding to the change requirement based on the ripple factor and the user requirement type; Perform project adjustment on the item to be adjusted based on the project adjustment strategy to obtain an adjusted item that meets the change requirement.

2. The project management method according to claim 1, characterized in that The performing a ripple prediction process based on the change requirement to obtain a ripple factor corresponding to the change requirement includes: Perform a prediction process based on the change requirement to obtain the code modification amount, project completion time, and project completion quality for implementing the change requirement; Determine the ripple factor based on the code modification amount, project completion time, and project completion quality.

3. The project management method according to claim 2, wherein The determining the ripple factor based on the code modification amount, project completion time, and project completion quality includes: Calculate a code modification amount ratio based on the code modification amount and the total code amount of the item to be adjusted; Calculate a project schedule delay rate based on the project completion time and the planned completion time of the item to be adjusted; Calculate a project quality degradation rate based on the project completion quality and the planned completion quality of the item to be adjusted; Calculate the ripple factor based on the code modification amount ratio, the project schedule delay rate, and the project quality degradation rate.

4. The project management method according to claim 3, wherein The calculating the ripple factor based on the code modification amount ratio, the project schedule delay rate, and the project quality degradation rate includes: Obtain the project type of the item to be adjusted; Set a first coefficient for the code modification amount ratio, a second coefficient for the project schedule delay rate, and a third coefficient for the project quality degradation rate based on the project type; Perform normalization processing on the code modification amount ratio based on the first coefficient to obtain a first factor; Perform normalization processing on the project schedule delay rate based on the second coefficient to obtain a second factor; Perform normalization processing on the project quality degradation rate based on the third coefficient to obtain a third factor; Calculate the ripple factor based on the first factor, the second factor, and the third factor.

5. The project management method according to claim 1, characterized in that The item to be adjusted includes a project function module and project tasks, and the method further includes: Use the change requirement, the project function module, and the project tasks as nodes, and use the dependency relationship between each node as a connection edge to construct a requirement ripple map; Determine a requirement change propagation path corresponding to the item to be adjusted based on the requirement ripple map; Provide the requirement change propagation path to the manager of the item to be adjusted so that the manager can perform analysis of the change requirement.

6. A training method for a spillover factor prediction model, characterized in that, The method includes: Obtain a historical project dataset and a to-be-trained impact factor prediction model, where the historical project dataset includes historical change requirements and the actual impact factors corresponding to the historical change requirements, and the actual impact factors are calculated based on the proportion of actual code modification amount, actual project schedule delay rate, and actual project quality degradation rate obtained after implementing the historical change requirements in the historical project dataset; Input the historical change requirements into the to-be-trained impact factor prediction model for impact prediction processing to obtain a prediction result; Calculate the loss value between the prediction result and the corresponding actual impact factor; Taking the minimization of the loss value as the optimization objective, adjust the parameters of the to-be-trained impact factor prediction model until the loss value converges at the minimum or the number of adjustment times reaches the preset number of times, then stop the adjustment to obtain a trained impact factor prediction model, and the trained impact factor prediction model is used to predict the impact factor as described in any one of claims 1 to 5.

7. A project management device, characterized in that, The device includes: A first acquisition module, configured to acquire a to-be-adjusted project, and the to-be-adjusted project corresponds to a change requirement; A first prediction module, configured to perform impact prediction processing based on the change requirement to obtain the impact factor corresponding to the change requirement; A second prediction module, configured to perform classification prediction processing on the change requirement based on a preset requirement classification model to obtain the user requirement type corresponding to the change requirement; A first determination module, configured to determine the project adjustment strategy corresponding to the change requirement based on the impact factor and the user requirement type; A first adjustment module, configured to adjust the to-be-adjusted project based on the project adjustment strategy to obtain an adjusted project that meets the change requirement.

8. A training device for a spillover factor prediction model, characterized in that Includes: A second acquisition module, configured to acquire a historical project dataset and a to-be-trained impact factor prediction model, where the historical project dataset includes historical change requirements and the actual impact factors corresponding to the historical change requirements, and the actual impact factors are calculated based on the proportion of actual code modification amount, actual project schedule delay rate, and actual project quality degradation rate obtained after implementing the historical change requirements in the historical project dataset; A first input module, configured to input the historical change requirements into the to-be-trained impact factor prediction model for impact prediction processing to obtain a prediction result; A first calculation module, configured to calculate the loss value between the prediction result and the corresponding actual impact factor; A first adjustment module, configured to take the minimization of the loss value as the optimization objective, adjust the parameters of the to-be-trained impact factor prediction model until the loss value converges at the minimum or the number of adjustment times reaches the preset number of times, then stop the adjustment to obtain a trained impact factor prediction model, and the trained impact factor prediction model is used to predict the impact factor as described in any one of claims 1 to 5.

9. A computer device, comprising a memory, a processor, and computer-readable instructions stored on the memory and running on the processor, characterized in that, When the processor executes the computer-readable instructions, it implements the project management method as described in any one of claims 1 to 5 or the training method of the impact factor prediction model as described in claim 6.

10. A readable storage medium having computer-readable instructions stored thereon, characterized in that, When the computer-readable instructions are executed by a processor, the project management method according to any one of claims 1 to 5 or the training method of the impact factor prediction model according to claim 6 is implemented.

Images