Task solution strategy generation method and device, equipment and medium

Abstract

The invention relates to the technical field of artificial intelligence and the field of financial science and technology, and discloses a task solution strategy generation method, which comprises the following steps: acquiring a task demand, constructing preliminary prompt information containing roles, task requirements and scheme generation guidance based on the task demand, inputting the preliminary prompt information into a preset language model to generate a preliminary solution, and generating a task solution strategy; analyzing the preliminary solution to extract a plurality of solution strategies, constructing review prompt information for each solution strategy and inputting the review prompt information into a preset language model to generate a score, determining a target solution strategy of which the score exceeds a preset score threshold, generating specific execution content for each target solution strategy, analyzing the execution content and determining an execution score, and performing priority ranking on the target solution strategies based on the execution scores to generate a priority list. According to the method, a better solution is generated through analysis of automatic processing task requirements, generation of solution strategies and priority ranking, customer requirements are quickly responded, and the development cost is reduced.

Description

Technical Field

[0001] The present invention relates to the fields of artificial intelligence technology and financial technology, and in particular to a task solution strategy generation method, device, equipment and storage medium. Background Art

[0002] In the process of implementing the corporate business system, the needs of corporate customers in the financial industry are often personalized and customized. Each customer need usually requires the collaboration of product managers, business needs analysts, system development leaders, and interaction designers to complete steps such as demand analysis, solution design, and process design. After that, the development team will develop and cooperate to complete the testing and launch process. Although this traditional demand response process can ensure a high degree of customization, it gradually reveals many deficiencies when faced with a large number of rapidly growing customer needs.

[0003] First, high cost is one of the main problems of existing technologies. In order to meet a large number of personalized needs, financial institutions need to invest more professional team members such as product managers, demand analysts and developers, which greatly increases labor costs. At the same time, each customer demand must go through a complete analysis and design process, resulting in a lengthy project development cycle, which further increases operating costs.

[0004] Secondly, the increase in management complexity is also a major problem in existing technologies. Since the needs of different customers often involve complex business logic and process design, each need to be implemented according to the standardized development process. However, in the face of such a large number of needs, some needs cannot be handled in a timely manner in the strict and standardized process, which may lead to unmet customer expectations. Although streamlining or optimizing the process may improve efficiency, it also increases the risk of implementation, which significantly increases management complexity.

[0005] Reduced customer satisfaction is a pain point that financial companies often encounter in the face of large demands. With the rapid growth of customer demand, the response capacity of the existing development team is limited, and customer needs cannot be met quickly, resulting in a slow launch of demand. This not only affects the business development of customers, but may also lead to complaints and a decline in satisfaction.

[0006] In addition, the long-term focus on solving specific needs has reduced the innovation ability of the product team. Under the pressure of highly customized needs, the product team has invested a lot of energy in solving specific problems, resulting in insufficient investment in new function development and system innovation, and limiting the overall development of the system. Summary of the invention

[0007] The main purpose of the present invention is to provide a task solution strategy generation method, device, equipment and storage medium, aiming to solve the technical problems in the prior art that the task solution design process is highly dependent on manual operation, cannot effectively respond to a large number of personalized needs, resulting in low response efficiency and high development cost.

[0008] To achieve the above object, the present invention provides a method for generating a task solution strategy, comprising:

[0009] Obtain task requirements;

[0010] Based on the task requirements, construct preliminary prompt information including roles, task requirements and solution generation guidance;

[0011] Inputting the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0012] parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution;

[0013] Constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0014] Identifying multiple target resolution strategies with scores exceeding a preset score threshold;

[0015] Constructing execution prompt information for each target solution strategy, inputting the execution prompt information into the preset language model, and generating specific execution content corresponding to each target solution strategy;

[0016] Analyze the specific execution contents and determine the execution score of each specific execution content;

[0017] Based on the execution score of each specific execution content, the corresponding target solution strategies are prioritized to generate a priority list of target solution strategies.

[0018] Furthermore, to achieve the above-mentioned purpose, the present invention provides a task solution strategy generation device, comprising:

[0019] Task requirement acquisition module, used to obtain task requirements;

[0020] A preliminary prompt information generation module, used to construct preliminary prompt information including roles, task requirements and solution generation guidance based on the task requirements;

[0021] A preliminary solution generation module, used to input the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0022] A solution strategy extraction module, used for parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution;

[0023] An evaluation prompt information generation module, used to construct evaluation prompt information for each solution strategy, and input the evaluation prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0024] A solution strategy screening module, for determining multiple target solution strategies with scores exceeding a preset score threshold;

[0025] An execution content generation module is used to construct execution prompt information for each target solution strategy, input the execution prompt information into the preset language model, and generate specific execution content corresponding to each target solution strategy;

[0026] An execution content analysis module is used to analyze the specific execution content and determine the execution score of each specific execution content;

[0027] The priority sorting module is used to prioritize the corresponding target solution strategies based on the execution score of each specific execution content and generate a priority list of the target solution strategies.

[0028] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer device, which includes a memory, a processor, and a task solution strategy generation program stored in the memory and executable on the processor, and the task solution strategy generation program, when executed by the processor, implements the steps of the task solution strategy generation method as described above.

[0029] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, on which a task solution strategy generation program is stored, and when the task solution strategy generation program is executed by a processor, the steps of the task solution strategy generation method described above are implemented.

[0030] Beneficial effects: The present invention relates to the fields of artificial intelligence technology and financial technology, and discloses a method for generating a task solution strategy, which obtains task requirements, constructs preliminary prompt information including roles, task requirements and solution generation guidelines based on the task requirements, inputs the preliminary prompt information into a preset language model to generate a preliminary solution, parses the preliminary solution to extract multiple solution strategies, constructs review prompt information for each solution strategy and inputs the preset language model to generate a score, determines a target solution strategy whose score exceeds a preset score threshold, generates specific execution content for each target solution strategy, analyzes the execution content and determines the execution score, prioritizes the target solution strategy based on the execution score, and generates a priority list. The present invention generates a better solution by automating the analysis of task requirements, the generation and priority sorting of solution strategies, reduces dependence on manual operations, achieves rapid response to customer needs and reduces development costs, and improves the flexibility and innovation capabilities of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0032] Figure 1 A schematic diagram of an application environment of a task solution strategy generation method in an embodiment of the present invention;

[0033] Figure 2 A schematic diagram of a flow chart of an embodiment of a method for generating a task solution strategy of the present invention;

[0034] Figure 3 A schematic diagram of functional modules of a preferred embodiment of a task solving strategy generating device of the present invention;

[0035] Figure 4 A schematic diagram of the structure of a computer device in one embodiment of the present invention;

[0036] Figure 5 FIG. 4 is another schematic diagram of the structure of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0037] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0038] The task solution strategy generation method provided by the embodiment of the present invention can be applied in Figure 1In an application environment, the user terminal communicates with the server terminal through a network. The server terminal can obtain task requirements through the user terminal, construct preliminary prompt information including roles, task requirements and solution generation guidance based on the task requirements, input the preliminary prompt information into a preset language model to generate a preliminary solution, parse the preliminary solution to extract multiple solution strategies, construct review prompt information for each solution strategy and input the preset language model to generate a score, determine the target solution strategy whose score exceeds the preset score threshold, generate specific execution content for each target solution strategy, analyze the execution content and determine the execution score, prioritize the target solution strategy based on the execution score, and generate a priority list. The present invention generates a better solution by automating the analysis of task requirements, the generation and priority sorting of solution strategies, reduces the dependence on manual operations, realizes rapid response to customer needs and reduces development costs, and improves the flexibility and innovation ability of the system. Among them, the user terminal can be but is not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices. The server terminal can be implemented with an independent server or a server cluster composed of multiple servers. The present invention is described in detail below through specific embodiments.

[0039] See also Figure 2 , Figure 2 This is a flow chart of an embodiment of a method for generating a task solution strategy provided by the present invention. It should be noted that although a logical sequence is shown in the flow chart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0040] like Figure 2 As shown, the task solution strategy generation method proposed in the present invention includes the following steps:

[0041] S10, obtaining task requirements;

[0042] In this embodiment, the system collects and organizes detailed requirements of tasks based on the input of users or customers. These task requirements are specific requirements for the functions, performance or improvements of the target system, and may include various types of information, such as operating procedures, business rules, system configuration, etc.

[0043] Task requirements can come from multiple sources, including user input, feedback from business departments, system monitoring data, etc. It usually involves the following key aspects:

[0044] Goal: The ultimate goal of the requirement, such as improving system processing efficiency or adding a certain function.

[0045] Constraints: Technical limitations or business rules in performing a task.

[0046] Priority: The urgency of the requirement, which may affect subsequent task scheduling and resource allocation.

[0047] Interaction requirements: User interaction requirements for the system interface.

[0048] Business processes: specific business processes involved, such as capital flow, customer management, transaction process, etc.

[0049] The acquisition of task requirements is the core foundation of the technical process. Subsequent operations such as solution generation and priority sorting all rely on accurate and complete requirements collection.

[0050] Users can manually input task requirement information through the system interface. The system provides an interactive interface (UI) where users can fill in the requirements in the input box and classify and prioritize them. The system performs a preliminary analysis based on the information entered by the user and stores the task requirements in the database for subsequent steps. For example, in the financial field, business analysts can input customers' loan approval requirements, including approval standards, risk assessment models, approval processes, etc., through the requirements management interface.

[0051] The system can also automatically obtain task requirement information from other systems or data sources through API interfaces. This method is suitable for financial data management systems. The system can automatically collect historical transaction data and customer behavior data and automatically generate task requirements. This method improves the efficiency of data collection and reduces manual input errors. For example, in a financial system, credit risk scores or historical transaction records automatically extracted from the customer management system are used as task requirements for risk management or loan approval.

[0052] Through the preset form format, users can select and fill in standardized task requirements. The fields in the form will limit the type, scope and details of the requirements, reducing the complexity of input. The system parses and reads the requirement information according to the fixed format in the form. For example, financial companies can provide standardized forms, and users only need to select the required product category (such as corporate loans, personal loans), and enter information such as the amount and interest rate, and the system can input these requirements as task requirements.

[0053] By acquiring task requirements through automated and standardized means, an efficient demand collection process is achieved, the complexity of manual operations is reduced, and the system response speed is improved. Especially in large-scale financial business scenarios, the accuracy and efficiency of demand management are significantly improved.

[0054] S20, constructing preliminary prompt information including roles, task requirements and solution generation guidance based on the task requirements;

[0055] In this embodiment, a preliminary prompt information is generated according to the task requirements so that the subsequent language model can generate a solution based on these prompts. The prompt information includes three main parts:

[0056] Role: The person or system role involved in the execution of a task. These roles are usually defined based on the complexity and specific needs of the task. For example, in the financial field, roles may include account managers, risk control experts, financial personnel, etc.

[0057] Task requirements: define the specific responsibilities and goals of each role in performing the task. Task requirements usually include the operations and goals that the role needs to complete. For example, an account manager is responsible for collecting customer information and conducting preliminary review, while a risk control expert is responsible for assessing risks.

[0058] Solution generation guidance: Provide guidance information for the language model to generate solutions. The guidance includes the task objectives, business rules, priorities, and constraints that need to be followed.

[0059] The system automatically or semi-automatically generates the corresponding roles according to the task requirements. For example, in a financial system, the task requirements may involve loan approval or fund management. The system will generate roles such as account manager, risk control personnel, and approval personnel according to the requirements. Roles can be selected from the preset role library or customized according to requirements. Example: In the task requirements of loan approval, the roles may include account manager, financial manager, risk control expert, legal advisor, etc. The system automatically determines the roles involved in the approval based on the loan type and amount.

[0060] The system generates corresponding task requirements based on the responsibilities of each role. For example, in the fund management task, the task requirement of the account manager may be to collect the financial information of the customer, the task of the risk control personnel is to assess the risk of fund flow, and the approval personnel need to review and approve the fund plan. Example: For a corporate fund management demand involving multiple departments, the system defines the task requirement for the account manager to collect customer financial information, the task requirement for the risk control personnel to review the potential risks of fund flow, and the task requirement for the financial personnel to approve and allocate funds.

[0061] The system generates guidance information for the language model based on the task requirements. These guidance are used to clarify the boundary conditions and goals of the language model when generating solutions. The guidance may include business rules (such as loan interest rates, approval time limits), risk control requirements, customer preferences, etc. The guidance is usually expressed in natural language to provide clear contextual information for the model. Example: In a loan approval task in the financial field, the solution generation guidance may include business rules (such as the risk level of the loan, interest rate range), approval time limits, risk control requirements (such as credit score requirements for the loan amount), and special preferences of customers (such as a preference for a shorter approval cycle).

[0062] By automatically building roles, task requirements, and solution generation guidelines, the workload of manually defining and managing complex role tasks can be effectively reduced. By combining natural language processing technology, the system can quickly provide accurate contextual information for the language model, thereby generating solutions that better meet actual needs. In the financial field, it can speed up the business approval process, reduce the error rate of manual operations, and improve business response efficiency and accuracy.

[0063] S30, inputting the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0064] In this embodiment, the constructed preliminary prompt information is input into the preset language model to generate a preliminary solution related to the task requirements. The preset language model is a system based on natural language processing (NLP) and machine learning, which can generate preliminary solutions that meet business needs by parsing and learning the input information. The preliminary prompt information usually includes roles, task requirements and solution generation guidelines, which provide the language model with context and guidance for generating solutions. Based on the prompt information, the language model combines its training data and language understanding capabilities to generate possible task execution solutions or business strategies, which are then further parsed and optimized.

[0065] The system first passes the constructed preliminary prompt information to the preset language model. The input data structure is usually in natural language format. The language model understands the key elements in the task requirements by parsing these natural language descriptions. Example: In the financial field, if the task requirement is to process a customer's loan approval, the preliminary prompt information may include the loan amount, repayment period, risk assessment criteria, etc. The preset language model will generate a preliminary loan approval plan based on this information, including the recommended interest rate, repayment plan, risk control measures, etc.

[0066] The preset language model analyzes the input prompt information through a computational model (such as GPT, BERT, etc.), and generates multiple potential solutions in combination with the model's knowledge base. The solution is usually output in natural language, including the specific steps of business operations, precautions, and the allocation of responsibilities of relevant roles. Example: In a financial system, a language model may generate a set of solutions that propose different interest rates and repayment periods for different risk levels. The model outputs a preliminary loan approval plan based on the customer's credit score, historical transaction data, and other information.

[0067] The preliminary solution generated by the language model may be further optimized through internal logic. The system can adjust the generated solution as needed, such as optimizing the model output results according to financial regulatory requirements and internal business rules to ensure the feasibility and compliance of the solution. Example: In financial business approval, the generated solution may need to be further adjusted according to the bank's loan approval standards. For example, if the interest rate in the preliminary solution exceeds the bank's regulatory limit, the system will automatically adjust and optimize the solution.

[0068] By inputting the preliminary prompt information into the language model, multiple preliminary solutions can be automatically generated, greatly reducing the time of manual operation and improving the efficiency of the system in processing task requirements. Through the natural language processing and learning capabilities of the language model, the generated solutions can better meet the task requirements, further improving the intelligent level of task solving.

[0069] S40, analyzing the preliminary solution, and extracting multiple solution strategies from the preliminary solution;

[0070] In this embodiment, the preliminary solution generated from the preset language model is parsed, its internal structure and key elements are analyzed, and multiple feasible solution strategies are extracted from it. The preliminary solution is usually an overall response to the task requirements, while the solution strategy is a specific implementation plan for a specific task goal or problem. The process of extracting the solution strategy usually includes operations such as identification, decomposition, and classification of the key content in the solution, so that these strategies can be evaluated and optimized in subsequent steps. The key operations of parsing the preliminary solution include:

[0071] Identify the core elements of the solution: such as task objectives, problem description, role allocation, resource allocation, etc.

[0072] Decompose the preliminary solution: Decompose the overall solution into multiple logically independent parts to facilitate the extraction of different solution strategies.

[0073] Strategy extraction and classification: According to each logical part, the corresponding solution strategy is extracted and the strategies are classified according to task requirements.

[0074] The system receives and parses the preliminary solution generated by the language model and identifies the core elements. The core elements may include task objectives, operation steps, problem description, resource requirements, and execution conditions. Natural language processing (NLP) technology is used to automatically analyze text, identify these elements and convert them into structured data. Example: In the financial field, if the preliminary solution is about the loan approval process, the system will identify the key parts of the solution, such as the loan amount, customer risk assessment, approval process, etc., and use them as the basis for subsequent strategy extraction.

[0075] The system breaks down the preliminary solution into multiple independent parts, which can be different logical modules in the solution. For example, for the loan approval process, it may include risk assessment, repayment plan, approval conditions, etc. These independent parts can be extracted and processed separately. Example: The system may break down the loan approval process into risk assessment module, fund allocation module, approval decision module, etc., and the information in each module is extracted independently.

[0076] The system extracts solution strategies from the decomposed preliminary solutions. Solution strategies are specific execution steps or methods that provide detailed implementation plans for different task objectives. For example, solution strategies may be how to manage loan risks, how to adjust repayment periods, or how to determine the approval process. Example: In the scenario of loan approval, the system may extract different loan risk management strategies, such as setting different loan interest rates based on customer credit scores, or reducing default risks by extending repayment periods.

[0077] The system classifies each strategy based on its mission objective. Classification can be based on the priority of the task, the applicable scenario of the strategy, or the complexity of the strategy. For example, high-risk loan approval strategies may be classified separately, while standardized low-risk approval strategies are classified in another category. Example: In a financial scenario, the system classifies loan strategies based on risk assessment and interest rate adjustment strategies based on customer credit scores separately to form different categories of solution strategies.

[0078] By analyzing the preliminary solutions generated by the language model, multiple feasible solution strategies can be effectively extracted and classified according to task requirements. Automated strategy extraction and classification reduces manual intervention and improves the accuracy and efficiency of task execution.

[0079] S50, constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0080] In this embodiment, a prompt information for review is constructed for each solution strategy extracted from the preliminary solution, and the prompt information is input into a preset language model to generate a score for the solution strategy. The construction of the review prompt information is based on the core content of each solution strategy and multiple review dimensions, with the purpose of providing the model with clear review criteria so that the model can give an accurate score. The construction of the review prompt information needs to consider the following elements:

[0081] Core content identification: Extract key content for each solution strategy, such as task objectives, execution steps, resource requirements, potential risks, etc.

[0082] Review dimensions: Select multiple dimensions to evaluate each solution strategy based on the needs. These dimensions may include the feasibility of the strategy, the cost of implementation, the expected effect, the risk, etc.

[0083] Model scoring: The constructed review prompt information is input into the preset language model, and the model generates a score for each solution strategy based on the prompt information.

[0084] The system first identifies and extracts the core content of each solution strategy. These core contents may be task objectives, specific operation steps, resource requirements (such as manpower, time, funds) or potential risks. The purpose of identifying the core content is to provide accurate context information for the review prompt information so that the model can better understand the essence of the solution strategy. Example: In a financial system, the core content of a loan approval strategy may include approval standards, loan amount, repayment period, risk control measures, etc. The system identifies these contents through natural language processing technology and marks them as core information.

[0085] The system selects multiple review dimensions based on the characteristics of each solution strategy. Common review dimensions include:

[0086] Feasibility: Evaluate the practical operability of the strategy and whether it can be implemented smoothly.

[0087] Execution cost: includes the consumption of resources such as time, money and manpower.

[0088] Expected results: Whether the solution strategy can achieve the task objectives and whether it has significant business effects.

[0089] Risk assessment: The risks that may arise from executing the strategy. The system automatically selects appropriate review dimensions based on different task requirements.

[0090] Example: For a fund allocation strategy in the financial field, the system may select feasibility, execution cost, and risk assessment as the main review dimensions.

[0091] The system constructs review prompt information based on the extracted core content and the selected review dimensions. The prompt information is organized in natural language and provides a clear review framework for the model. Review prompt information usually includes task background, review criteria, target requirements, etc. Example: For a loan approval strategy, the review prompt information may include: "Evaluate the feasibility of the loan approval strategy, analyze the execution cost and potential risks of the strategy based on the customer's credit score, risk assessment and repayment plan."

[0092] The system inputs the constructed review prompt information into the preset language model. The language model generates a score for each solution strategy based on the prompt information. The score result is usually a quantitative score used to represent the performance of the strategy in each review dimension. Example: For a loan approval strategy, the language model generates a comprehensive score based on the prompt information, such as 8 points (out of 10 points), indicating that the strategy has high feasibility and low risk.

[0093] Through the above steps, the system can automatically construct clear review prompt information for each solution strategy, and input it into the preset language model to generate a score, which improves the review efficiency and accuracy. Based on multi-dimensional review prompt information, the system can comprehensively evaluate the feasibility, cost, effect and risk of each solution strategy to ensure that the optimal strategy is selected.

[0094] S60, determining a plurality of target solution strategies whose scores exceed a preset score threshold;

[0095] In this embodiment, the system selects multiple target solution strategies with scores exceeding a preset threshold based on the previously generated solution strategy scores. The purpose of this process is to automatically select strategies with higher feasibility through quantitative criteria for subsequent prioritization and execution. During the scoring process, different solution strategies are evaluated based on multiple dimensions and receive a comprehensive score. The preset scoring threshold is a minimum standard set by the system, and solution strategies below this threshold will be deemed to be insufficiently feasible.

[0096] The system sets an initial scoring threshold for the task requirements, which may be based on historical project data, industry standards, or business requirements. The system screens the scores of all solution strategies, retains strategies with scores above the threshold, and marks these strategies as target solution strategies for subsequent optimization and execution steps.

[0097] The system first sets an initial scoring threshold for the task requirement. This threshold can be set in a variety of ways, including statistical analysis based on historical data, business rule setting, or user-defined thresholds. The choice of scoring threshold needs to be adjusted according to the importance of the task and the business risk. Example: In the loan approval process in the financial field, the system may set a threshold that requires the solution strategy to score at least 7 points (out of 10 points) to ensure that the approval strategy has sufficient feasibility and risk control capabilities.

[0098] The system adjusts the initial scoring threshold based on the specific characteristics of the task requirements or historical project data. For example, by analyzing the score distribution of successfully executed solution strategies in historical projects, the system can dynamically adjust the scoring threshold to meet the needs of different scenarios. For example, for low-risk loans, the threshold can be slightly lower, while for high-risk projects, the threshold may be higher. Example: If historical data shows that most successful loan approval strategies score above 8 points, the system may raise the scoring threshold to 8 points.

[0099] The system will filter out all solution strategies with scores higher than the threshold according to the set score threshold, and mark these strategies as target solution strategies. The filtering operation is usually implemented based on database query or score sorting. The selected target solution strategies will enter the next execution plan or further evaluation. Example: In the loan approval process, the system will filter out approval strategies with scores higher than 8 points. These strategies may include different interest rates, repayment plans and risk control measures. Strategies with scores lower than 8 points will be excluded.

[0100] The system marks the selected solution strategies as target solution strategies for further processing. The tag can contain information such as the strategy ID, score, priority, etc., to facilitate the execution and optimization of subsequent steps. Example: The system marks all solution strategies with a score higher than 8 as target strategies, stores them in the task management module, and prepares for subsequent priority sorting and execution.

[0101] By setting a scoring threshold, the target solution strategies with scores above the threshold are automatically screened out, significantly improving the automation and efficiency of the task execution process. By dynamically analyzing historical data to adjust the scoring threshold, the system can ensure that the selected solution strategies are suitable for different business scenarios and task requirements, reducing the complexity of manual screening strategies.

[0102] S70, constructing execution prompt information for each target solution strategy, inputting the execution prompt information into the preset language model, and generating specific execution content corresponding to each target solution strategy;

[0103] In this embodiment, the system constructs corresponding execution prompt information for each screened target solution strategy, and inputs these prompt information into the preset language model to generate specific execution content. The construction of execution prompt information is to provide sufficient context information to the language model to generate detailed, executable steps or operation plans. The execution prompt information needs to cover the execution goals of the solution strategy, the resources involved, the operation steps and related constraints, etc. The preset language model generates specific execution content corresponding to the task requirements based on the input prompt information. These execution contents can be specific operation processes, resource allocation methods, specific steps for task completion, etc.

[0104] The system constructs execution prompts based on the characteristics of each target solution strategy. The prompts usually contain the following elements:

[0105] Execution Goal: What is the ultimate mission goal of each goal-solving strategy.

[0106] Resource requirements: the resources needed to implement the strategy (such as manpower, money, time, etc.).

[0107] Operational steps: The specific operational steps required to complete the task, including timetable, task allocation, etc.

[0108] Constraints: business rules, laws and regulations, technical constraints, etc. that may be encountered in the execution strategy. The system will automatically generate prompts based on this information to ensure that the language model can follow these guidelines when generating execution content.

[0109] Example: In a loan approval scenario in the financial sector, the execution prompt may include: "Develop a repayment plan based on the customer's risk assessment results to ensure that the loan approval process complies with risk control requirements and repayment capacity assessment."

[0110] The system inputs the constructed execution prompt information into the preset language model, and the language model generates the specific execution content of each target solution strategy based on these prompt information. The model understands the input information through natural language processing technology, and generates specific operation steps related to the task in combination with its knowledge base. Example: After the system inputs the execution prompt information into the language model, the model generates a detailed loan approval process based on the prompt, including specific execution content such as loan interest rate calculation, repayment period, and customer follow-up.

[0111] The language model generates specific execution content corresponding to the target solution strategy based on the prompt information. These execution contents are usually operational task instructions or processes, which are convenient for business personnel or systems to perform subsequent operations. The generated execution content may include specific operation plans, task allocation tables, time schedules, resource allocation plans, etc. Example: For the loan approval process, the language model may generate the following specific execution content: "After receiving the materials submitted by the customer, the risk control personnel must complete the risk assessment within 2 working days; the approval department confirms the loan conditions within 3 working days and notifies the customer."

[0112] The generated execution content may be optimized and adjusted according to business rules or actual operation conditions. For example, the system can adjust the steps in the execution content according to different business conditions to ensure the feasibility and compliance of its execution. Example: If the loan approval strategy contains a complex risk assessment process, the system may adjust the timing or task allocation of the execution content to ensure that the approval process is completed within the scheduled time.

[0113] Through the above steps, the system can automatically generate execution prompt information for each target solution strategy, and generate detailed execution content through the preset language model. This automated process reduces the complexity of manual intervention and improves the efficiency and accuracy of task execution. Through natural language processing technology, the system can generate specific operation plans that meet task requirements and can be adjusted and optimized according to business rules or constraints.

[0114] S80, analyzing the specific execution contents to determine the execution score of each specific execution content;

[0115] In this embodiment, the generated specific execution contents are analyzed in depth and scored for each execution content to quantify its feasibility, cost, expected effect and potential risk. The execution score is a key factor in the subsequent priority sorting. It ensures that the selected strategy can meet the task objectives and have a high success rate during execution through multi-dimensional analysis. The main dimensions of the score include:

[0116] Execution feasibility: Analyze the feasibility of the strategy in terms of technology and resources encountered during actual implementation.

[0117] Implementation cost: Assess the resources required to implement the strategy, including time, manpower, money, etc.

[0118] Expected results: Predict whether the implementation of the strategy can achieve the expected goals, including business benefits and problem-solving effects.

[0119] Potential risks: Analyze the risks that may be faced in executing the strategy, especially in terms of business compliance and market changes.

[0120] The system evaluates the feasibility of the execution content in the real environment by analyzing each operation step of the execution content. The feasibility analysis mainly includes technical resources, skills required for execution, system capabilities and external conditions. The system scores the feasibility of each execution content based on the availability of resources and the complexity of execution. Example: In the financial system, for the loan approval process, the system will analyze the technical conditions in the approval process, such as whether the automated approval system can handle complex credit assessments. A more feasible strategy may include using automated tools to complete loan assessments, while a manual approval strategy may have a lower feasibility.

[0121] The system conducts cost analysis based on the resources (including time, manpower, and funds) required to execute the content. The system uses historical data or project estimates to evaluate the specific resources required to execute the content and generate an execution cost score. The higher the cost score, the lower the difficulty of execution and the higher the cost-effectiveness. Example: For the loan approval process, the system will evaluate the manual review time required for the approval process, the workload of risk control personnel, and the possible customer communication costs involved. If a certain execution content requires a lot of manual intervention, its cost score may be lower.

[0122] The system scores the expected effect of the execution content by analyzing whether it can achieve the expected business goals. The system generates an expected effect score based on the degree to which the execution content solves the task requirements and the potential business benefits. The higher the expected effect, the better the strategy can meet business needs. Example: If a loan approval strategy can effectively reduce approval time and reduce risks, its expected effect score will be high. On the contrary, if the strategy reduces risks while seriously affecting customer experience, the expected effect score will be low.

[0123] The system analyzes the potential risks that may be brought about by executing the strategy, including technical failures, market changes, legal compliance, etc. Based on the risk assessment results, the system generates a risk score for each execution content. The lower the risk score, the lower the risk of the strategy, and the more suitable it is for execution. Example: In the loan approval process, if a certain execution strategy needs to be implemented in a high-risk customer group and there is a lack of effective risk control measures, the risk score of the strategy may be high, while the approval strategy for low-risk customers may have a lower risk score.

[0124] The system weights the scores of feasibility, cost, expected effect and potential risk to generate a comprehensive execution score for each specific execution content. The comprehensive score can be used for subsequent priority sorting to ensure that the best strategy is executed first. Example: The system may generate a comprehensive execution score for a loan approval strategy, such as 7.8 points (out of 10 points), indicating that the strategy has good feasibility and expected effect, as well as low execution cost and risk.

[0125] Through the above steps, the system can automatically conduct multi-dimensional analysis of each specific execution content and generate execution scores, reducing the complexity of manual participation and improving the evaluation efficiency of execution strategies. The multi-dimensional analysis of execution scores ensures a comprehensive evaluation of strategies in terms of feasibility, cost, effect and risk, allowing the system to prioritize the best strategies with low execution costs, good expected effects and controllable risks.

[0126] S90, based on the execution score of each specific execution content, prioritize the corresponding target solution strategies and generate a priority list of the target solution strategies.

[0127] In this embodiment, the execution score of each specific execution content is sorted according to different priorities to generate a priority list of target solution strategies. By comprehensively sorting multiple target solution strategies, the system can identify the solution strategies that are executed first, thereby optimizing resource allocation and task execution processes. Priority sorting is based on the comprehensive execution score of each strategy, which includes scores in multiple dimensions such as feasibility, execution cost, expected effects, and potential risks. The system prioritizes strategies with higher scores through a sorting algorithm to ensure the efficiency of task execution and the optimization of results.

[0128] The system first collects the execution scores of all specific execution contents. The execution contents corresponding to each target solution strategy have been scored in the previous step, and the system aggregates these scores for priority sorting. Example: In the loan approval scenario in the financial field, the system collects the execution scores of each approval strategy, which include the feasibility of the approval process, execution cost, customer satisfaction prediction, and risk assessment.

[0129] The system sets the priority sorting rules based on the task requirements and business goals. These rules can be adjusted according to specific needs. For example, in some scenarios, execution cost may be more important, while in other scenarios, expected effect may have a higher priority. The system can determine the importance of different dimensions in the sorting by setting weighting rules. Example: In the fund allocation scenario, the system may set execution cost as the dimension with the highest weight, while in the risk control scenario, expected effect and risk control have higher weights.

[0130] The system prioritizes each solution strategy based on its execution score. The sorting algorithm can be a simple high to low sorting, or a complex weighted sorting based on different task requirements. The system ensures that solution strategies with high scores, low execution costs, and controllable risks are executed first. Example: For a loan approval process, the system may give priority to those approval strategies that score high in terms of cost and risk control, and put these strategies at the top of the priority list.

[0131] The system generates a priority list of target solution strategies based on the sorting results. The priority list can be presented in the form of structured data, listing all target solution strategies and their corresponding priority rankings. The list may also include information such as the comprehensive score of the strategy, the execution steps, and the estimated completion time. Example: In a financial loan approval scenario, the priority list may include the priority order of different approval strategies, such as "low-risk customers first" and "automated approval first", etc., to help banks optimize approval resources.

[0132] The system can dynamically adjust the priority list based on changes in task requirements. For example, during the execution of a task, new task requirements or changes in external conditions (such as market fluctuations, regulatory changes) may cause the priority of certain strategies to change. The system will re-evaluate the execution score of the solution strategy based on the new conditions and update the priority ranking. Example: If market conditions suddenly change, causing loan approval for high-risk customers to become more urgent, the system will automatically adjust the priority list and increase the priority of these strategies.

[0133] By prioritizing the execution scores of each target solution strategy, the optimal strategy can be automatically screened and executed in order, improving the efficiency of resource use and the accuracy of task execution. Through weighted sorting and dynamic adjustment, the system can adjust priorities in a timely manner according to changes in business needs to ensure that tasks are completed on time and efficiently.

[0134] The present invention relates to the fields of artificial intelligence technology and financial technology, and discloses a method for generating a task solution strategy, which obtains task requirements, constructs preliminary prompt information including roles, task requirements and solution generation guidance based on the task requirements, inputs the preliminary prompt information into a preset language model to generate a preliminary solution, parses the preliminary solution to extract multiple solution strategies, constructs review prompt information for each solution strategy and inputs the preset language model to generate a score, determines a target solution strategy whose score exceeds a preset score threshold, generates specific execution content for each target solution strategy, analyzes the execution content and determines the execution score, prioritizes the target solution strategy based on the execution score, and generates a priority list. The present invention generates a better solution by automating the analysis of task requirements, the generation and priority sorting of solution strategies, reduces dependence on manual operations, achieves rapid response to customer needs and reduces development costs, and improves the flexibility and innovation capabilities of the system.

[0135] In one embodiment, the above S80 includes:

[0136] S801, analyzing the execution cost of each specific execution content based on the resource consumption, time investment and complexity of each specific execution content, and generating an execution cost score;

[0137] S802, based on the possibility of each specific execution content achieving the expected effect and the stability of the execution process, the expected effect of each specific execution content is scored to generate an expected effect reliability score;

[0138] S803, scoring the applicability of each specific execution content based on the user group or business scenario coverage of each specific execution content to generate an applicability score;

[0139] S804, generating a comprehensive execution score for each specific execution content based on the execution cost score, expected effect reliability score and applicability score of each specific execution content.

[0140] In this embodiment, the system scores each specific execution content through multi-dimensional analysis in order to evaluate its executability, effectiveness and applicability under different conditions. The execution score is mainly used to conduct a scientific and quantitative evaluation of the specific execution content to ensure that the execution plan with the least resource consumption, the highest execution efficiency and the widest applicability can be given priority among multiple strategies. The system will analyze based on the specific characteristics of each specific execution content and generate scores based on multiple dimensions, including resource consumption, time investment, task complexity, the possibility of achieving the expected effect, the stability of the execution process, and the breadth of coverage of user groups or business scenarios. Through these analyses, the system can provide an accurate quantitative basis for subsequent policy priority sorting.

[0141] The system generates an execution cost score based on a detailed analysis of each specific execution content in terms of resource consumption, time investment, and task complexity. Resource consumption includes the manpower, material resources, and technical equipment required to complete the task, time investment is the estimated time required to execute the task, and complexity measures the technical difficulty of the execution content and the complexity of the steps.

[0142] Resource consumption involves estimating the physical resources, software resources, or human resources required for the project. The system evaluates the resource utilization efficiency of the execution plan by analyzing these requirements. The system will evaluate the time required for task execution based on the specific operation steps of the execution content and the external conditions it depends on (such as approval process, system response time, etc.). The system evaluates the difficulty of execution through factors such as the number of task processes, the complexity of the decision path, and whether it involves multi-party cooperation. The system comprehensively considers resource consumption, time investment, and complexity to generate an execution cost score. The higher the score, the lower the cost and the higher the execution efficiency.

[0143] The system will score each specific execution content based on the possibility of achieving the expected effect and the stability of the execution process, and generate an expected effect reliability score. The system analyzes whether the execution of the strategy can achieve its set business goals, as well as potential uncertainties or problems in the execution process.

[0144] The system analyzes historical data, strategy models, and current conditions to evaluate whether the execution strategy can effectively achieve its goals. By evaluating the reliability of system operations, the stability of data processing, and external dependencies, the system analyzes whether the execution process will face uncertainties or potential obstacles to ensure a smooth execution process. By analyzing the stability of strategy execution and the possibility of achieving goals, the system generates an expected effect reliability score. The higher the score, the better the execution plan performs in terms of effect and stability.

[0145] The system evaluates the applicability of each specific execution content in different business scenarios and generates a suitability score. The suitability analysis evaluates whether the execution content has a wide range of user groups or business scenario coverage. The higher the applicability, the more business scenarios the strategy can play a role in. Analyze the range of users that the execution content can cover, such as whether it is applicable to standard users, special user groups, or user groups under specific business conditions. The system analyzes the scope of application of the strategy in various business scenarios and evaluates its versatility or whether it is only applicable to specific scenarios. The system scores based on the diversity and versatility of the execution content. The applicability score is generated by analyzing the breadth of users and business scenarios covered by the strategy. The higher the score, the wider the scope of application of the execution plan.

[0146] The system comprehensively analyzes the three scoring dimensions of execution cost, reliability of expected effects, and applicability to generate a comprehensive execution score for each specific execution content. Based on task requirements or business goals, the system can assign different weights to different dimensions to ensure that the score meets the actual business priority requirements. The system performs weighted scoring based on the importance of each dimension and flexibly adjusts the influence of each dimension. For example, if the task emphasizes resource conservation, the weight of the execution cost score will increase, while in high-risk business scenarios, the reliability of expected effects may have a greater weight. The final comprehensive execution score is generated through weighted calculation and is used to measure the comprehensive performance of the execution plan to ensure that the priority in the subsequent task execution is reasonable and scientific.

[0147] This embodiment can comprehensively evaluate the resource consumption, expected effect and applicability of each execution content through multi-dimensional analysis and weighted calculation, ensuring that the scoring results can accurately reflect the actual feasibility and business benefits of the strategy. By generating a comprehensive execution score, the system can provide a scientific basis for the priority sorting of subsequent tasks, effectively improving resource utilization and decision-making efficiency.

[0148] In one embodiment, the above S20 includes:

[0149] S201, analyzing the task requirements and generating task requirement analysis results;

[0150] S202, based on the task requirement analysis result, setting the roles required for task execution and generating a description of responsibilities for each role;

[0151] S203, generating task requirements for each role based on the job description of each role;

[0152] S204, generating a solution generation guide for guiding a preset language model to generate a preliminary solution based on the task requirement analysis result and the task requirements of each role;

[0153] S205, integrating the task requirement analysis results, the task requirements of each role and the solution generation guide into preliminary prompt information.

[0154] In this embodiment, by analyzing the task requirements, guidance information including roles, task requirements and guidance language model generation solutions is generated. This information is the basis for the language model to generate a preliminary solution. By comprehensively analyzing the task requirements, the system can identify the specific goals of the task, the problems to be solved, the business scenarios, and the role responsibilities and task requirements required for execution. Subsequently, this information is integrated into a complete prompt information as the basis for input into the language model.

[0155] The system first analyzes all aspects of the task requirements and extracts the key elements of the task. The task requirements analysis results include the expected results, goals, existing problems, and specific usage scenarios of the task. The analysis results provide complete context and background information for subsequent steps to ensure that the guidance for task execution is accurate.

[0156] The system determines the final result that needs to be achieved by the task by analyzing customer needs or business goals. It also identifies the specific results or completion standards that the task needs to achieve. It identifies the obstacles or challenges that the task currently faces and identifies the problems that need to be solved through solutions. It determines the business scenario or operating environment for the task to be executed, such as loan approval in the financial field. The system can parse the task requirements in the text through natural language processing technology, automatically identify the key elements of the task and generate analysis results.

[0157] Based on the task requirement analysis results, the system sets the roles required for the task execution process and generates specific job descriptions for each role. The system determines which roles play a key role in the task based on the complexity of the task, the areas involved, and the business needs, and specifies clear responsibilities for each role.

[0158] The system assigns the required roles, such as business analyst, development engineer, financial expert, etc., based on the operational complexity of the task. Generate specific job descriptions for each role to ensure that each role in the task has a clear division of responsibilities during execution. The system can automatically match the appropriate roles through standard task-based templates and generate specific job descriptions based on business needs.

[0159] Based on the job description of each role, the system further generates specific task requirements that the role needs to complete during the task execution process. The task requirements clearly define the operations or goals that each role should complete during execution to ensure the accuracy of task execution. The system sets specific task goals, operation steps and deliverables for each role. The task requirements of each role are based on the job description and are consistent with the overall task goals to ensure seamless collaboration between roles. Task requirements can be automatically generated by process management tools, or they can be matched with corresponding operation steps based on established templates.

[0160] Based on the results of the task requirement analysis and the task requirements of each role, the system generates a complete solution generation guide. This guide is used to guide the language model to generate a preliminary solution, which contains information such as the key elements of the task, role responsibilities, and task requirements, ensuring that the language model can generate an effective preliminary solution based on these prompts. The solution generation guide includes a detailed description of the task objectives, steps to solve the problem, resource requirements, and business scenarios, which are passed to the preset language model as input information. The system ensures that the solution generation guide contains sufficiently detailed content so that the language model can generate accurate and compliant preliminary solutions. The system uses data structuring tools and natural language generation (NLG) technology to convert task requirements and role requirements into model-readable prompt information.

[0161] The system integrates all generated task requirement analysis results, role task requirements, and solution generation guidelines into a complete prompt information and passes it to the language model as input. This prompt information is the basis for the language model to generate a preliminary solution in the subsequent steps. The system uses the information integration module to merge a variety of different information and generate structured prompt information. The prompt information needs to be formatted to ensure that the language model can correctly parse and understand the prompt content. The system uses input optimization algorithms and natural language processing tools to package and integrate the generated analysis results and guidance information into standardized prompt text.

[0162] This embodiment can automatically construct prompt information for language model generation of preliminary solutions through task requirement analysis, role allocation, and task requirement generation, reducing the complexity of manual intervention and improving the accuracy and efficiency of task execution. Through detailed analysis of task requirements and clear allocation of roles, the system can ensure that the execution plan of each task accurately matches the business requirements, and optimize the process of prompt information generation through natural language processing technology, further improving the effectiveness of language model generation of preliminary solutions.

[0163] In one embodiment, the above S30 includes:

[0164] S301, determining a preset language model based on the task requirement;

[0165] S302, formatting the preliminary prompt information so that the preliminary prompt information complies with the input format requirement of the preset language model;

[0166] S303, setting a priority for specific content in the processed preliminary prompt information based on the task requirement;

[0167] S304, segmenting the preliminary prompt information based on the priority to generate a plurality of priority information segments;

[0168] S305, inputting the plurality of priority information segments into the preset language model in sequence according to the priority order;

[0169] S306, after each priority information segment is input, obtaining a generated intermediate solution;

[0170] S307, after all priority information segments are input, a complete preliminary solution is generated based on the corresponding intermediate solutions.

[0171] In this embodiment, the corresponding preliminary solution is generated by gradually inputting the preliminary prompt information into the preset language model. The process includes selecting a suitable preset language model, ensuring that the prompt information format meets the input requirements of the model, and processing the information in segments according to the task requirements to generate solutions step by step. This method ensures that the multi-level information in complex tasks can be correctly parsed by the model and a complete solution can be gradually generated.

[0172] The system first determines the preset language model suitable for the current task based on the task requirements. Different task requirements may require different language models, and the choice of model will be adjusted according to the complexity, scale and business scenario of the task.

[0173] The system selects a preset language model that is suitable for handling these requirements based on the goal, domain, and complexity of the task. For example, complex risk assessment in the financial field may require the use of a specific model, while general text generation tasks may use other types of models. The system can tune the parameters of the model based on the task requirements so that it can better handle the task requirements. Through the model selection module, the system automatically selects the appropriate language model and configures it before the task is executed.

[0174] The system formats the preliminary prompt information to ensure that the input content meets the input format requirements of the preset language model. Formatting includes text normalization, information structuring, etc., so that the language model can correctly parse and process the input data. The system converts the unstructured text in the prompt information into structured data, including titles, content paragraphs, tags, etc., to ensure that the information is clear and parsable. Different language models may have different requirements for the input format, and the system ensures that the prompt information meets these requirements, such as appropriate sentence length, keywords, etc. Using the formatting module, the system cleans the text, standardizes the content, and optimizes the semantics of the prompt information to ensure that it meets the parsing requirements of the model.

[0175] The system sets priorities for specific content in the prompt information based on task requirements. Priority setting is to ensure that important information is processed first during subsequent segment processing and model input. Prioritization is usually based on the importance, urgency and impact of the information on the task. The system divides the content in the prompt information into high priority, medium priority and low priority according to importance to ensure that key information is input into the model first. Priority setting is based on task goals, such as problems that need to be solved immediately or resource allocation in the task are given a higher priority. Through the priority algorithm, the system automatically assigns the priority of each prompt information according to the specific needs of the task and the importance of the information.

[0176] The system processes the preliminary prompt information in segments according to the set priority and generates multiple priority information segments. Each priority information segment contains different parts of the prompt information, and the processing order of the information segment is determined according to its priority. The system divides the prompt information into multiple small segments according to priority, and each information segment contains the key content, detailed information or auxiliary information for task execution. The information segments with high priority are processed first to ensure that the key steps or core information in the task are input into the model first. Through the information segmentation processing module, the system automatically divides the prompt information into multiple small segments according to the set priority rules, and sets the processing order for each information segment.

[0177] The system inputs the segmented prompt information into the language model one by one according to the set priority order. Each time an information segment is input, the model processes the information step by step and generates the corresponding intermediate solutions. The system inputs information segments of different priorities into the language model in turn to ensure that the processing order of the information meets the task requirements and the parsing logic of the model. After each input, the system records the intermediate solutions generated by the model to provide contextual support for the input of subsequent information segments. Through the information scheduling module, the system controls the input order of information segments to ensure that high-priority content is processed first and low-priority content is supplemented according to task requirements.

[0178] Whenever a priority information segment is input into the language model, the system immediately obtains the intermediate solution generated by the model. The intermediate solution provides the processing results of this part of information, which may include specific operation suggestions, task steps or preliminary solution strategies. After each information segment is input, the system immediately receives the intermediate solution returned by the model to ensure that the information in the task execution process is completed step by step. The system analyzes the feedback of the intermediate solution, evaluates the accuracy and effectiveness of the model processing information, and provides a basis for the subsequent input of information segments. Through the real-time feedback mechanism, the system can obtain the solution generated by the model in a timely manner after each information segment is input, and record and analyze it.

[0179] After all priority information segments are entered, the system integrates and generates a complete preliminary solution based on each intermediate solution. The solution combines all levels of information required by the task and provides a comprehensive task solution. The system integrates each intermediate solution to ensure the logical coherence of the information and generates a complete task execution plan. The system may optimize and adjust the generated preliminary solution to ensure that it meets the task requirements and business goals. Through the solution integration module, the system automatically splices the intermediate solutions generated step by step into a complete preliminary solution and optimizes the solution.

[0180] This embodiment realizes the step-by-step processing and solution generation of complex tasks by gradually inputting priority information segments into the language model, greatly improving the accuracy and efficiency of the model's generation of preliminary solutions. Through segmented processing, priority setting, and real-time feedback mechanisms, the system can dynamically adjust the task execution process and optimize the generated solutions to ensure that the tasks can be completed effectively, especially in complex multi-level task scenarios.

[0181] In one embodiment, the above S40 includes:

[0182] S401, identifying and extracting key elements of the preliminary solution;

[0183] S402, extracting a plurality of potential solution strategies from the preliminary solution based on the key elements, each potential solution strategy corresponding to one or more task objectives;

[0184] S403, classifying multiple potential solution strategies based on the task goal corresponding to each potential solution strategy, and generating multiple classified solution strategies.

[0185] In this embodiment, multiple potential solution strategies are extracted by parsing the preliminary solution. First, the system identifies key elements from the preliminary solution, which are the core components of the solution, such as task objectives, problem description, resource allocation or role allocation. Then, the system extracts potential solution strategies based on these key elements, each of which aims to achieve the corresponding task objectives. Finally, the system classifies these potential solution strategies for better analysis and processing in subsequent steps.

[0186] The system first analyzes the preliminary solution, identifies and extracts the core elements of the solution. These elements include task objectives, problem descriptions, resource allocation or role allocation, etc. They are the main components of the solution and the basis for the subsequent extraction of potential solution strategies. The system analyzes the ultimate goal of the task, such as improving efficiency, reducing risks, optimizing resources, etc., to clarify the purpose of the solution. The system identifies the specific problems that need to be solved in the task and ensures that the extracted strategies can directly solve these problems. The system analyzes the resource allocation required for the task, such as manpower, financial resources and materials, and ensures that the extracted strategies can effectively allocate these resources. The system identifies the different roles and responsibilities for performing the task, and ensures that the solution strategy can reasonably allocate the task. The system parses the text content of the preliminary solution through natural language processing (NLP) technology, and automatically identifies and extracts the key elements related to task execution.

[0187] Based on the identified key elements, the system extracts multiple potential solution strategies corresponding to each task objective. Each potential solution strategy is designed to achieve the corresponding task objective and solve a specific problem. The system analyzes possible execution paths or operation methods based on the key elements and extracts different potential solution strategies. Each strategy solves problems at different levels and may involve different resource and role allocations. The system ensures that each extracted solution strategy corresponds to a specific task objective to ensure the relevance and effectiveness of the strategy. The system automatically generates multiple potential solution strategies based on the key elements in the preliminary solution through a pattern matching algorithm or a strategy generation module. These strategies cover different task objectives and execution paths to ensure that the task can be solved from multiple angles.

[0188] The system classifies the extracted multiple potential solutions according to the task objectives of each potential solution. This classification helps to refine the analysis of different strategies and flexibly select the best solution according to the task requirements. The system classifies the extracted potential solutions according to the different goals of the solutions, task priorities, and resource usage. The classification of these strategies can be based on different dimensions, such as cost-effectiveness, risk control, execution time, etc. The system ensures that each classified solution is highly correlated with the task objectives to avoid invalid or redundant strategy generation. The classification algorithm groups the strategies based on the priority of the task, the urgency of the goal, or the resource requirements, and assigns labels to each group of strategies for subsequent analysis and execution.

[0189] Through the above steps, the system of this embodiment can efficiently extract multiple potential solution strategies from the preliminary solution and classify these strategies according to the task objectives. This process reduces the complexity of manual intervention through automated key element identification and strategy extraction, while improving the accuracy and efficiency of task execution. By classifying and managing potential solution strategies, the system can flexibly select the optimal strategy during the subsequent task execution process to ensure the efficient realization of task objectives.

[0190] In one embodiment, the above S50 includes:

[0191] S501, identify the core content of each solution strategy;

[0192] S502, determining a plurality of review dimensions for review;

[0193] S503, constructing review prompt information for each solution strategy based on the multiple review dimensions;

[0194] S504: Input the review prompt information into a preset language model to obtain a scoring result for each solution strategy.

[0195] In this embodiment, by constructing the review prompt information, the core content of the solution strategy and the review dimension are integrated, and the language model is input for evaluation to generate the scoring result. The review prompt information of each solution strategy needs to integrate the task objectives, execution content, required resources and potential risks of the strategy to ensure a comprehensive and accurate analysis of the feasibility, cost, expected effect and risk of each strategy during the review process. By inputting the prompt information into the language model, the system can quickly score each strategy, which is convenient for subsequent priority sorting and strategy selection.

[0196] The system first identifies the core content of each solution strategy. These core contents are the key factors in the implementation of the strategy and determine the implementation path of the strategy and the achievement of the task objectives. The core content usually includes the task objectives, execution steps, required resources and potential risks. The system identifies the ultimate goal or expected results of each strategy. Analyze the operation steps or implementation methods of the strategy to clarify the specific operation path in the implementation. Identify the various resources required to implement the strategy, including manpower, funds, technical equipment, etc. Identify the risks that may be faced during the implementation of the strategy. The system uses it as part of the evaluation dimension to measure the risk factor of the strategy. The system analyzes the strategy text through natural language processing technology to automatically identify and extract the core content of the strategy.

[0197] The system determines multiple review dimensions for each solution strategy to comprehensively evaluate the feasibility, execution cost, expected effect, and risk of each strategy. These dimensions provide a multi-angle analytical perspective for the strategy evaluation, ensuring that the system can objectively score the strategy. Evaluate the actual operability of the strategy and the possibility of successful implementation. Evaluate the resources and time investment required to execute the strategy, and generate a score based on its cost. Analyze whether the strategy can effectively achieve the expected task objectives and measure its implementation effect. Evaluate the potential risks that may be encountered during the execution of the strategy and consider its impact on the overall task execution. The system sets these review dimensions through the review rule engine and adjusts the weight of each dimension according to the priority of the task requirements to achieve flexible scoring rules.

[0198] The system generates specific review prompt information based on the core content and review dimensions. Review prompt information is a structured description of the strategy before the system inputs the language model to ensure that the model can correctly understand and evaluate the key information in each dimension. The mission objectives, execution content, required resources, potential risks and other information of each strategy are structured to meet the input requirements of the preset language model. The system ensures that each review dimension can match the relevant information in the core content so that the language model can effectively generate scores. Through the structured data processing module, the system organizes the key information of the strategy according to the review dimensions and generates standardized review prompt information.

[0199] The system inputs the constructed review prompt information into the preset language model, and the language model analyzes and scores based on the prompt information. The model automatically generates scoring results based on the various dimensions of the strategy and provides an overall evaluation of the strategy to help the system prioritize the strategy in subsequent steps. The language model generates scoring results for each dimension based on the strategy content in the prompt information. The scoring results may include a comprehensive assessment of feasibility, cost, effect, and risk. After the system obtains the scoring results from the language model, it automatically records these scores and uses them for subsequent strategy sorting and selection. Through the model input module, the system inputs the structured prompt information into the language model and receives the scoring data returned by the model in real time.

[0200] Through the above steps, this embodiment can automatically identify the core content of each solution strategy and comprehensively evaluate the strategy based on multi-dimensional review prompt information. The system generates a score for each solution strategy based on the feasibility, execution cost, expected effect and potential risk of the task, providing an objective basis for optimizing the priority of task execution.

[0201] In one embodiment, the above S60 includes:

[0202] S601, setting an initial scoring threshold based on the task requirements;

[0203] S602, acquiring historical project data related to the task requirement, analyzing the successfully executed solution strategies in the historical project data, adjusting the initial scoring threshold based on the analysis result, and generating a final scoring threshold;

[0204] S603: Filter out multiple target solution strategies with scores higher than the score threshold from all solution strategies for the task requirement.

[0205] In this embodiment, by setting an initial scoring threshold, analyzing historical project data and making adjustments, the target solution strategies with scores exceeding the preset threshold are finally screened out. The system first sets the initial scoring threshold according to the current task requirements, then adjusts the scoring threshold by analyzing successful solution strategies in historical projects and combining historical data to ensure that the scoring threshold is reasonable and effective. Finally, the system screens out multiple target solution strategies with scores exceeding the threshold from all solution strategies for further execution and priority sorting.

[0206] The system sets an initial scoring threshold based on the specific requirements of the current task as a baseline for subsequent screening of solution strategies. The setting of the initial scoring threshold takes into account factors such as the difficulty, resource constraints, and urgency of the task. The system presets a reasonable scoring threshold based on factors such as the complexity of the task, execution priority, and expected results. This threshold is used to preliminarily screen out those strategies with higher scores. The setting of the threshold depends on the current task requirements. For example, when resources are tight, the threshold can be higher to ensure that the most efficient strategy is selected. The system generates an initial scoring threshold through the task requirement analysis module and combines the task characteristics as a benchmark for subsequent scoring evaluation.

[0207] The system extracts successful solution strategies related to the current task requirements from historical project data, analyzes these historical strategies, and finds strategies with high scores and successful execution as references. The system analyzes successful strategies in historical projects and evaluates their performance in similar tasks to obtain reference values ​​for adjusting the scoring threshold of the current task. The system analyzes the similarity between the current task and historical projects to ensure that the selected historical data is highly relevant to the current task to improve the accuracy of the analysis results. Through data mining and machine learning algorithms, the system can automatically extract relevant data from the historical project database, conduct in-depth analysis, and determine the scoring criteria for historical successful strategies.

[0208] After obtaining historical project data and completing the analysis, the system adjusts the initial scoring threshold according to the scoring criteria in the historical data to generate a final scoring threshold that better meets the current task requirements. This threshold will be more in line with the actual task execution requirements. Based on the results of the historical data analysis, the system reasonably adjusts the initial scoring threshold to make it more in line with the needs of the current task. Based on the analysis of the success rate of historical projects, scoring trends, and task requirements, the system generates the final scoring threshold for the current task to ensure that the selected strategy has high feasibility. Through the historical data analysis module, the system automatically calculates and adjusts the scoring threshold in combination with the current task characteristics and the historical project success criteria to generate the final threshold.

[0209] Among all the solution strategies for the task requirements, the system selects the solution strategies with scores higher than the generated score threshold as the final target solution strategies. Only strategies with scores exceeding the threshold will be further considered as execution strategies. Based on the score threshold, the system automatically selects strategies with higher scores as target strategies. These target strategies meet the task requirements in terms of execution feasibility, resource consumption, and risk management. The system can select multiple target solution strategies so that the best strategy can be further determined based on priority sorting. Through the strategy screening module, the system sorts all scores, automatically excludes strategies below the threshold, and retains multiple solution strategies above the threshold.

[0210] Through the above steps, this embodiment can automatically set the initial scoring threshold according to the task requirements, and reasonably adjust the threshold in combination with historical project data to ensure that the selected target solution strategy is highly feasible in the current task. The system uses historical data analysis and dynamic adjustment of scoring criteria to ensure that the generated scoring threshold can flexibly respond to various task scenarios and improve the accuracy and efficiency of task decision-making.

[0211] In one embodiment, a task solution strategy generation device is provided, and the task solution strategy generation device corresponds one-to-one to the task solution strategy generation method in the above embodiment. Figure 3 , Figure 3 This is a functional module diagram of a preferred embodiment of the task solution strategy generation device of the present invention. Task requirement acquisition module 10, preliminary prompt information generation module 20, preliminary solution generation module 30, solution strategy extraction module 40, review prompt information generation module 50, solution strategy screening module 60, execution content generation module 70, execution content analysis module 80 and priority sorting module 90. The detailed description of each functional module is as follows:

[0212] A task requirement acquisition module 10 is used to acquire task requirements;

[0213] A preliminary prompt information generating module 20 is used to construct preliminary prompt information including roles, task requirements and solution generating instructions based on the task requirements;

[0214] A preliminary solution generation module 30, used to input the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0215] A solution strategy extraction module 40 is used to analyze the preliminary solution and extract multiple solution strategies from the preliminary solution;

[0216] The review prompt information generating module 50 is used to construct review prompt information for each solution strategy, and input the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0217] A solution strategy screening module 60 is used to determine a plurality of target solution strategies having scores exceeding a preset score threshold;

[0218] An execution content generation module 70 is used to construct execution prompt information for each target solution strategy, input the execution prompt information into the preset language model, and generate specific execution content corresponding to each target solution strategy;

[0219] An execution content analysis module 80 is used to analyze the specific execution content and determine an execution score for each specific execution content;

[0220] The priority sorting module 90 is used to prioritize the corresponding target solution strategies based on the execution score of each specific execution content and generate a priority list of the target solution strategies.

[0221] In one embodiment, the content analysis module 80 is executed to:

[0222] Analyze the execution cost of each specific execution content based on the resource consumption, time investment and complexity of each specific execution content, and generate an execution cost score;

[0223] Based on the possibility of each specific implementation content achieving the expected effect and the stability of the implementation process, the expected effect of each specific implementation content is scored to generate an expected effect reliability score;

[0224] Based on the coverage of user groups or business scenarios of each specific execution content, the scope of application of each specific execution content is scored to generate an applicability score;

[0225] Based on the execution cost score, expected effect reliability score and applicability score of each specific execution content, a comprehensive execution score for each specific execution content is generated.

[0226] In one embodiment, the preliminary prompt information generating module 20 is specifically used to:

[0227] Analyze the task requirements and generate task requirement analysis results;

[0228] Based on the task requirement analysis results, set the roles required for task execution and generate a description of the responsibilities of each role;

[0229] Generate task requirements for each role based on the job description of each role;

[0230] Based on the task requirement analysis results and the task requirements of each role, generate a solution generation guide for guiding a preset language model to generate a preliminary solution;

[0231] The task requirement analysis results, the task requirements of each role and the solution generation guidance are integrated into preliminary prompt information.

[0232] In one embodiment, the preliminary solution generation module 30 is specifically used to:

[0233] Determining a preset language model based on the task requirements;

[0234] Formatting the preliminary prompt information so that the preliminary prompt information complies with the input format requirement of the preset language model;

[0235] Prioritize specific content in the processed preliminary prompt information based on the task requirements;

[0236] Segmenting the preliminary prompt information based on the priority to generate a plurality of priority information segments;

[0237] According to the order of priority, sequentially inputting a plurality of priority information segments into the preset language model;

[0238] After each priority information segment is input, the generated intermediate solution is obtained;

[0239] After all priority information segments are entered, a complete preliminary solution is generated based on the corresponding intermediate solutions.

[0240] In one embodiment, the solution strategy extraction module 40 is specifically used to:

[0241] Identify and extract key elements of the preliminary solution;

[0242] Extracting a plurality of potential solution strategies from the preliminary solution based on the key elements, each potential solution strategy corresponding to one or more task objectives;

[0243] Based on the task objective corresponding to each potential solution strategy, multiple potential solution strategies are classified to generate multiple classified solution strategies.

[0244] In one embodiment, the review prompt information generating module 50 is specifically used to:

[0245] Identify the core elements of each solution strategy;

[0246] Identify multiple review dimensions for review;

[0247] Based on the multiple review dimensions, construct review prompt information for each solution strategy;

[0248] The review prompt information is input into a preset language model to obtain a scoring result for each solution strategy.

[0249] In one embodiment, the priority sorting module 90 is specifically configured to:

[0250] Setting an initial scoring threshold based on the task requirements;

[0251] Acquire historical project data related to the task requirements, analyze successfully executed solution strategies in the historical project data, adjust the initial scoring threshold based on the analysis results, and generate a final scoring threshold;

[0252] A plurality of target solution strategies having scores higher than the score threshold are screened out from all solution strategies required by the task.

[0253] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external user terminal via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a task solution strategy generation method service side.

[0254] In one embodiment, a computer device is provided. The computer device may be a user terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via 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 non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the user side of a task solution strategy generation method

[0255] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are implemented:

[0256] Obtain task requirements;

[0257] Based on the task requirements, construct preliminary prompt information including roles, task requirements and solution generation guidance;

[0258] Inputting the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0259] parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution;

[0260] Constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0261] Identifying multiple target resolution strategies with scores exceeding a preset score threshold;

[0262] Constructing execution prompt information for each target solution strategy, inputting the execution prompt information into the preset language model, and generating specific execution content corresponding to each target solution strategy;

[0263] Analyze the specific execution contents and determine the execution score of each specific execution content;

[0264] Based on the execution score of each specific execution content, the corresponding target solution strategies are prioritized to generate a priority list of target solution strategies.

[0265] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0266] Obtain task requirements;

[0267] Based on the task requirements, construct preliminary prompt information including roles, task requirements and solution generation guidance;

[0268] Inputting the preliminary prompt information into a preset language model to generate a corresponding preliminary solution;

[0269] parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution;

[0270] Constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy;

[0271] Identifying multiple target resolution strategies with scores exceeding a preset score threshold;

[0272] Constructing execution prompt information for each target solution strategy, inputting the execution prompt information into the preset language model, and generating specific execution content corresponding to each target solution strategy;

[0273] Analyze the specific execution contents and determine the execution score of each specific execution content;

[0274] Based on the execution score of each specific execution content, the corresponding target solution strategies are prioritized to generate a priority list of target solution strategies.

[0275] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the user side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0276] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an 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 (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0277] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by 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.

[0278] It should be noted that if software tools or components other than those of the Company appear in the embodiments of the present application, they are only used for illustration and do not represent actual use. The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the above-mentioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the above-mentioned embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents; and these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for generating a task solution strategy, characterized in that: The following steps are involved: Obtain task requirements; Based on the task requirements, construct preliminary prompt information including roles, task requirements and solution generation guidance; Inputting the preliminary prompt information into a preset language model to generate a corresponding preliminary solution; parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution; Constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, so that the preset language model generates a score for each solution strategy; Identifying multiple target resolution strategies with scores exceeding a preset score threshold; Constructing execution prompt information for each target solution strategy, inputting the execution prompt information into the preset language model, and generating specific execution content corresponding to each target solution strategy; Analyze the specific execution contents and determine the execution score of each specific execution content; Based on the execution score of each specific execution content, the corresponding target solution strategies are prioritized to generate a priority list of target solution strategies.

2. The task solution strategy generation method according to claim 1, characterized in that: Analyze the specific execution contents and determine the execution score of each specific execution content, including: Based on the resource consumption, time investment and complexity of each specific execution content, the execution cost of each specific execution content is analyzed to generate an execution cost score; Based on the possibility of each specific implementation content achieving the expected effect and the stability of the implementation process, the expected effect of each specific implementation content is scored to generate an expected effect reliability score; Based on the coverage of user groups or business scenarios of each specific execution content, the applicability scope of each specific execution content is scored to generate an applicability score; Based on the execution cost score, expected effect reliability score and applicability score of each specific execution content, a comprehensive execution score for each specific execution content is generated.

3. The task solving strategy generation method according to claim 1, characterized in that: Based on the task requirements, construct preliminary prompt information including roles, task requirements and solution generation guidance, including: Analyze the task requirements and generate task requirement analysis results; Based on the task requirement analysis results, set the roles required for task execution and generate a description of the responsibilities of each role; Generate task requirements for each role based on the job description of each role; Based on the task requirement analysis results and the task requirements of each role, generate a solution generation guide for guiding a preset language model to generate a preliminary solution; The task requirement analysis results, the task requirements of each role and the solution generation guidance are integrated into preliminary prompt information.

4. The method for generating a task solution strategy as claimed in claim 1, characterized in that: The preliminary prompt information is input into a preset language model to generate a corresponding preliminary solution, including: Determining a preset language model based on the task requirements; Formatting the preliminary prompt information so that the preliminary prompt information complies with the input format requirement of the preset language model; Prioritize specific content in the processed preliminary prompt information based on the task requirements; Segmenting the preliminary prompt information based on the priority to generate a plurality of priority information segments; According to the order of priority, sequentially inputting a plurality of priority information segments into the preset language model; After each priority information segment is input, the generated intermediate solution is obtained; After all priority information segments are entered, a complete preliminary solution is generated based on the corresponding intermediate solutions.

5. The task solution strategy generation method according to claim 1, characterized in that: Analyze the preliminary solution and extract multiple solution strategies from the preliminary solution, including: Identify and extract key elements of the preliminary solution; Extracting a plurality of potential solution strategies from the preliminary solution based on the key elements, each potential solution strategy corresponding to one or more task objectives; Based on the task objective corresponding to each potential solution strategy, multiple potential solution strategies are classified to generate multiple classified solution strategies.

6. The method for generating a task solution strategy according to claim 1, wherein: Constructing review prompt information for each solution strategy, and inputting the review prompt information into the preset language model, and generating a score for each solution strategy by the preset language model, including: Identify the core elements of each solution strategy; Identify multiple review dimensions for review; Based on the multiple review dimensions, construct review prompt information for each solution strategy; The review prompt information is input into a preset language model to obtain a scoring result for each solution strategy.

7. The task solution strategy generation method according to claim 1, characterized in that: Identify multiple target resolution strategies that score above a preset score threshold, including: Setting an initial scoring threshold based on the task requirements; Acquire historical project data related to the task requirements, analyze successfully executed solution strategies in the historical project data, adjust the initial scoring threshold based on the analysis results, and generate a final scoring threshold; A plurality of target solution strategies having scores higher than the score threshold are screened out from all solution strategies required by the task.

8. A task solution strategy generation device, characterized in that: The task solution strategy generating device comprises: Task requirement acquisition module, used to obtain task requirements; A preliminary prompt information generation module, used to construct preliminary prompt information including roles, task requirements and solution generation guidance based on the task requirements; A preliminary solution generation module, used to input the preliminary prompt information into a preset language model to generate a corresponding preliminary solution; A solution strategy extraction module, used for parsing the preliminary solution and extracting a plurality of solution strategies from the preliminary solution; An evaluation prompt information generation module, used to construct evaluation prompt information for each solution strategy, and input the evaluation prompt information into the preset language model, so that the preset language model generates a score for each solution strategy; A solution strategy screening module, for determining multiple target solution strategies with scores exceeding a preset score threshold; An execution content generation module is used to construct execution prompt information for each target solution strategy, input the execution prompt information into the preset language model, and generate specific execution content corresponding to each target solution strategy; An execution content analysis module is used to analyze the specific execution content and determine the execution score of each specific execution content; The priority sorting module is used to prioritize the corresponding target solution strategies based on the execution score of each specific execution content and generate a priority list of the target solution strategies.

9. A computer device, characterized in that: The computer device includes a memory, a processor, and a task solution strategy generation program stored in the memory and executable on the processor. When the task solution strategy generation program is executed by the processor, the steps of the task solution strategy generation method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The storage medium stores a task solution strategy generation program, and when the task solution strategy generation program is executed by the processor, the steps of the task solution strategy generation method according to any one of claims 1 to 7 are implemented.

Images