Project aid decision-making and risk prediction method and device based on large model

Abstract

The invention relates to a project aid decision-making and risk prediction method based on a large model. The method comprises the following steps: acquiring project information related to a project; performing classified management and storage on the project information; before project implementation, according to project information, generating a project implementation plan, drawing a resource input trend chart in a project period, listing a risk problem list, and in the project implementation process, according to the project information, dynamically optimizing the project implementation plan, providing risk early warning, providing decision information for project implementation personnel, and generating a project implementation document; distributing project implementation tasks, pushing risk early warning and decision information, synchronizing project progress information, and updating project implementation documents; and evaluating the decision information and risk early warning effect, and carrying out model training and optimization. Dependence on a single expert is reduced, the management efficiency is improved, the consistency and scientificity of project planning are improved, the review decision process is accelerated, the project period is shortened, and precipitation, reuse and continuous optimization of project management knowledge are achieved.

Description

Methods and apparatus for project-aided decision-making and risk prediction based on large models Technical Field

[0001] This invention relates to the field of project management technology, and in particular to a method and apparatus for project decision support and risk prediction based on a large model. Background Technology

[0002] In the software development lifecycle, projects are typically characterized by high complexity, frequent cross-disciplinary collaboration, long durations, and stringent delivery deadlines. To ensure the orderly progress of projects, companies often employ mature project management systems and process specifications to rigorously control each stage of the project. These stages encompass multiple phases, including project initiation, requirements analysis, architecture design, development and implementation, testing and verification, and final mass production delivery.

[0003] At each stage, the project management team needs to identify and formulate phased goals based on the specific characteristics of the project (such as complexity, time frame, resource allocation, and delivery requirements), and define the corresponding inputs and outputs, such as requirements documents, design schemes, review records, test cases, and reports. These processes typically require repeated communication and collaboration between project managers and technical experts from multiple fields, using multiple rounds of reviews to ensure that the outputs at each stage meet quality and time requirements.

[0004] Traditional processes heavily rely on experienced expert teams, whose judgments and decisions significantly impact project success or failure. However, this also brings several significant problems: High resource investment: Expert resources are scarce, and frequent meetings and reviews require substantial manpower, time, and energy, leading to high costs; High time costs: Long review cycles and complex decision-making processes often affect project schedules and increase the risk of delays; High subjectivity of conclusions: Because the review process depends on the individual knowledge and experience of experts, it is easily influenced by human factors, lacking consistency and traceability. Experience and judgment mainly rely on individual tacit knowledge, making it difficult to standardize and systematize, thus limiting the overall improvement of team capabilities. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned problems in the prior art and provide a method for project-assisted decision-making and risk prediction based on a large model, as well as an apparatus for project-assisted decision-making and risk prediction based on a large model.

[0006] To achieve the above technical objectives and effects, the present invention provides the following technical solution: a method for project-assisted decision-making and risk prediction based on a large model, comprising the following steps: S1, acquiring project information related to the project; S2, classifying, managing, and storing the project information; S3, before project implementation, generating a project implementation plan based on the project information, drawing a resource input trend chart during the project cycle, and listing a list of risk issues; during project implementation, dynamically optimizing the project implementation plan based on the project information, providing risk warnings, providing decision-making information for project implementers, and generating project implementation documents; S4, allocating project implementation tasks according to the project implementation plan, pushing the risk warnings and decision-making information, synchronizing project progress information, and updating project implementation documents according to project progress; S5, evaluating the effectiveness of the decision-making information and risk warnings, and training and optimizing the model.

[0007] The project information comprises four main categories: project documents, project data, team communication information, and external environment information. Project documents include unstructured text data such as project plans, requirements documents, design documents, and meeting minutes. Project data includes structured data such as project progress, resource usage, cost consumption, task status, code commit records, and defect reports. Team communication information includes potential problems and sentiments identified from communication records among team members. External environment information includes collected market trends, industry policies, and competitor activities.

[0008] The project information is categorized and stored as short-term memory, long-term memory, domain knowledge, and a project knowledge graph. The short-term memory contains temporary data and states in the current project session. The long-term memory contains historical project data, risk patterns, decision templates, and best practices in project management. The domain knowledge contains professional knowledge and business rules related to the project management domain. The project knowledge graph is used to construct the relationships between project tasks, resources, risks, and stakeholders.

[0009] The project implementation plan is generated based on a combination of a large language model and a machine learning model, and includes task breakdown, time arrangement, and resource allocation.

[0010] The risk issue list is generated by analyzing project data and external environmental information using machine learning models, anomaly detection technology, and natural language processing technology to identify potential project risks, and then quantitatively assessing the likelihood and impact of these risks.

[0011] The risk warning is based on the combination of time series analysis technology and deep learning technology with project information to predict the development trend of project risks, identify the time stages in which risks may occur, and provide risk warnings in advance according to the risk level.

[0012] The decision information mentioned above utilizes multi-objective optimization technology and expert system technology, combined with project information, to provide multi-dimensional analysis and intelligent suggestions for decision points in the project.

[0013] The project implementation documents are generated using a large language model based on project information, and include project implementation reports, risk reports, and meeting minutes.

[0014] A device for project-assisted decision-making and risk prediction based on a large model, used in the aforementioned method, includes: a data collection module for acquiring project information; a data storage module for classifying and storing project information; a planning and decision-making module for generating project implementation plans, drawing resource input trend charts during the project cycle, listing risk issues, dynamically optimizing project implementation plans, providing risk warnings, providing decision-making information to project implementers, and generating project implementation documents; an execution module for allocating project implementation tasks, pushing risk warnings, pushing decision-making information, synchronizing project progress information, and updating project implementation documents; and a feedback learning module for evaluating the effectiveness of decision-making information and risk warnings, and for training and optimizing the model.

[0015] The beneficial effects of this invention are: by using artificial intelligence and knowledge engineering to perform structured modeling and rule-based reasoning on key knowledge in the project management process, simulating expert judgment behavior, and providing project implementation plans, process node input and output suggestions, risk warnings, and process optimization suggestions, thereby reducing reliance on a single expert, improving management efficiency, enhancing the consistency and scientific nature of project planning, accelerating the review and decision-making process, shortening the project cycle, and realizing the accumulation, reuse, and continuous optimization of project management knowledge. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their descriptions, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: FIG1 is a framework diagram of the apparatus and method of the present invention. Detailed Implementation

[0017] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0018] As shown in Figure 1, the method for project-assisted decision-making and risk prediction based on a large model includes the following steps: S1, acquiring project-related information, which includes four main categories: project documents, project data, team communication information, and external environment information; the project documents include unstructured text data such as project plans, requirement documents, design documents, and meeting minutes; the project data includes structured data such as project progress, resource usage, cost consumption, task status, code commit records, and defect reports; the team communication information includes potential problems and sentiments identified from communication records among team members; and the external environment information includes collected market trends, industry policies, and competitor dynamics.

[0019] S2. Classify and store the project information, which is categorized into short-term memory, long-term memory, domain knowledge, and project knowledge graph. The short-term memory contains temporary data and status in the current project session. The long-term memory contains historical project data, risk patterns, decision templates, and best practices in project management. The domain knowledge contains professional knowledge and business rules related to the project management domain. The project knowledge graph is used to construct the relationships between project tasks, resources, risks, and stakeholders.

[0020] S3. Before project implementation, a project implementation plan is generated based on the project information, a resource input trend chart is drawn during the project cycle, and a risk issue list is compiled. During project implementation, the project implementation plan is dynamically optimized based on the project information, risk warnings are provided, decision-making information is provided to project implementers, and project implementation documents are generated. The project implementation plan is generated based on a large language model and machine learning model combined with project information, and includes task decomposition, time scheduling, and resource allocation. The risk issue list is generated by analyzing project data and external environmental information using machine learning models, anomaly detection technology, and natural language processing technology to identify potential project risks, and then quantitatively assessing the likelihood and impact of these risks. The risk warning is based on time series analysis technology and deep learning technology combined with project information to predict the development trend of project risks, identify potential risk stages, and provide early warnings based on risk levels. The decision-making information uses multi-objective optimization technology and expert system technology combined with project information to provide multi-dimensional analysis and intelligent suggestions for decision points in the project. The project implementation documents are generated using a large language model based on project information and include a project implementation report, risk report, and meeting minutes.

[0021] S4. Assign project implementation tasks according to the project implementation plan, push the risk warning and decision information, synchronize project progress information, update project implementation documents according to project progress; send project implementation task instructions to the project management system to update task status; push risk warnings, decision information, and project progress to relevant project members or managers via email or instant messaging tools.

[0022] Instant messaging tools such as WeChat, QQ, DingTalk, telephone, and SMS.

[0023] S5. Evaluate the effectiveness of the decision information and risk warning, and train and optimize the model. Evaluating the decision information involves automatically assessing the actual effectiveness of the provided decision information, such as whether the decision information effectively reduces risk or improves efficiency. Evaluating the effectiveness of the risk warning involves assessing the accuracy of the provided risk warning and calibrating it based on actual risk events. Model training and optimization involves using new project information and feedback information generated during project implementation to continuously train and optimize the machine learning model, thereby improving the intelligence level and predictive ability of the machine learning model.

[0024] This method utilizes artificial intelligence and knowledge engineering to perform structured modeling and rule-based reasoning of key knowledge in the project management process, simulates expert judgment behavior, and provides suggestions for project implementation plans, process node inputs and outputs, risk warnings, and process optimization. This reduces reliance on a single expert, improves management efficiency, enhances the consistency and scientific nature of project planning, accelerates the review and decision-making process, shortens the project cycle, and enables the accumulation, reuse, and continuous optimization of project management knowledge.

[0025] A device for project-assisted decision-making and risk prediction based on a large model, used in the aforementioned method, includes: a data collection module for acquiring project information; a data storage module for classifying and storing project information; a planning and decision-making module for generating project implementation plans, drawing resource input trend charts during the project cycle, listing risk issues, dynamically optimizing project implementation plans, providing risk warnings, providing decision-making information to project implementers, and generating project implementation documents; an execution module for allocating project implementation tasks, pushing risk warnings, pushing decision-making information, synchronizing project progress information, and updating project implementation documents; and a feedback learning module for evaluating the effectiveness of decision-making information and risk warnings, and training and optimizing the model. The data storage module is communicatively connected to the data collection module, the planning and decision-making module, and the feedback learning module, respectively. The planning and decision-making module is also communicatively connected to the data collection module, the execution module, and the feedback learning module, respectively. The execution module is also communicatively connected to the data collection module and the feedback learning module, respectively.

[0026] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0027] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for project-assisted decision-making and risk prediction based on a large model, characterized in that, Includes the following steps: S1. Obtain project information related to the project; S2. Classify, manage, and store the project information; S3. Before project implementation, generate a project implementation plan based on the project information, draw a resource input trend chart during the project cycle, and list a list of risk issues. During project implementation, dynamically optimize the project implementation plan based on the project information, provide risk warnings, provide decision-making information for project implementers, and generate project implementation documents. S4. Assign project implementation tasks according to the project implementation plan, push the risk warning and decision information, synchronize project progress information, and update project implementation documents according to the project progress; S5. Evaluate the effectiveness of the decision information and risk warning, and perform model training and optimization.

2. The method according to claim 1, characterized in that: The project information comprises four main categories: project documents, project data, team communication information, and external environment information. Project documents include unstructured text data such as project plans, requirements documents, design documents, and meeting minutes. Project data includes structured data such as project progress, resource usage, cost consumption, task status, code commit records, and defect reports. Team communication information includes potential problems and sentiments identified from communication records among team members. External environment information includes collected market trends, industry policies, and competitor activities.

3. The method according to claim 2, characterized in that: The project information is categorized and stored as short-term memory, long-term memory, domain knowledge, and a project knowledge graph. The short-term memory contains temporary data and states in the current project session. The long-term memory contains historical project data, risk patterns, decision templates, and best practices in project management. The domain knowledge contains professional knowledge and business rules related to the project management domain. The project knowledge graph is used to construct the relationships between project tasks, resources, risks, and stakeholders.

4. The method according to claim 3, characterized in that: The project implementation plan is generated based on a combination of large language models and machine learning models and project information. The project implementation plan includes task breakdown, time arrangement, and resource allocation.

5. The method according to claim 3, characterized in that: The risk list is generated by analyzing project data and external environmental information using machine learning models, anomaly detection technology, and natural language processing technology to identify potential project risks, and then quantitatively assessing the likelihood and impact of these risks.

6. The method according to claim 3, characterized in that: The risk warning is based on the combination of time series analysis and deep learning technologies with project information to predict the development trend of project risks, identify the time stages in which risks may occur, and provide risk warnings in advance according to the risk level.

7. The method according to claim 3, characterized in that: The decision information is provided by combining multi-objective optimization technology and expert system technology with project information to provide multi-dimensional analysis and intelligent suggestions for decision points in the project.

8. The method according to claim 3, characterized in that: The project implementation documents are generated using a large language model based on project information, and include project implementation reports, risk reports, and meeting minutes.

9. An apparatus for project-assisted decision-making and risk prediction based on a large model, used in the method described in any one of claims 1 to 8, characterized in that, include: The data collection module is used to obtain project information; The data storage module is used to categorize and store project information. The planning and decision-making module is used to generate project implementation plans, draw resource input trend charts during the project cycle, list risk issues, dynamically optimize project implementation plans, provide risk warnings, provide decision-making information for project implementers, and generate project implementation documents; the execution module is used to allocate project implementation tasks, push risk warnings, push decision-making information, synchronize project progress information, and update project implementation documents. The feedback learning module is used to evaluate the effectiveness of decision-making information and risk warnings, and to train and optimize models.

Images