Plan management method and apparatus, electronic device, and storage medium

Abstract

The application relates to the technical field of artificial intelligence, and provides a plan management method and device, electronic equipment and a storage medium, the method comprising the following steps: determining a demand text and user individual information; based on an expert rule model, applying the demand text and the user individual information to generate a current plan; based on a logic reasoning model and the user individual information, performing feasibility reasoning on the current plan; and based on a reasoning result of the current plan, adjusting the current plan until a stage plan is obtained; the expert rule model and the logic reasoning model are obtained based on large language model training; and based on an execution situation of the stage plan, the stage plan is managed.The plan management method and device, electronic equipment and storage medium provided by the application can discover and correct unreasonable places in a plan in a timely manner, ensure smooth execution of the plan, and thus enhance the feasibility and reliability of the plan.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a planning and management method, apparatus, electronic device, and storage medium. Background Technology

[0002] With the rapid development of natural language processing technology, especially the emergence of large language models, artificial intelligence has significantly improved its ability to assist humans in managing complex tasks. These technological breakthroughs enable machines to understand and generate natural language text, thereby helping users efficiently complete planning tasks.

[0003] Related technologies for plan formulation can be broadly categorized into two types: The first type relies on manual user operation, where users manually input the specific execution steps and details for each plan on the software interface. While this method directly reflects the user's intentions, it requires users to have extensive knowledge and lacks intelligent assistance, making it difficult to adapt to rapidly changing planning needs. The second type utilizes the understanding capabilities of large-scale language models, automatically generating plans based on the user's natural language descriptions and producing corresponding record files for user review. This approach improves the efficiency of plan formulation to some extent, but it cannot meet users' personalized customization needs, leading to frequent abandonment during execution due to unreasonable plans. Summary of the Invention

[0004] This invention provides a planning management method, apparatus, electronic device, and storage medium to address the shortcomings of existing technologies where planning is mismatched with users' personalized customization needs, leading to unreasonable plans that are easily abandoned.

[0005] This invention provides a planning management method, comprising:

[0006] Determine the requirements text and user personalization information;

[0007] Based on the expert rule model, the current plan is generated by applying the demand text and user personalized information. Based on the logical reasoning model and the user personalized information, the feasibility of the current plan is reasoned, and the current plan is adjusted based on the reasoning results until a phase plan is obtained. The expert rule model and the logical reasoning model are trained based on a large language model.

[0008] Manage the phase plan based on its execution status.

[0009] According to the planning management method provided by the present invention, adjusting the current plan based on the reasoning result of the current plan until a phase plan is obtained includes:

[0010] If the reasoning result indicates that the current plan does not match the user's personalized information, the current plan is adjusted until the adjusted current plan matches the user's personalized information, and the adjusted current plan is determined to be the stage plan.

[0011] According to the plan management method provided by the present invention, the step of adjusting the current plan if the inference result indicates that the current plan does not match the user's personalized information includes:

[0012] If the inference result indicates that the current plan does not match the user's personalized information, then the inference result is sent to the expert rule model so that the expert rule model adjusts the current plan based on the inference result; and / or,

[0013] Receive user feedback on the inference results, and adjust the current plan based on the user feedback.

[0014] According to the planning management method provided by the present invention, managing the phase plan based on the execution status of the phase plan includes:

[0015] The execution status and deviations from the aforementioned phase plan are analyzed to obtain the current execution status and the current deviation status.

[0016] Based on the deviation from the current situation, update the user's personalized information;

[0017] Based on the current execution status and the updated user personalization information, the phase plan is adjusted.

[0018] According to the planning management method provided by the present invention, the step of analyzing the execution status and deviation status of the phase plan to obtain the execution status and deviation status includes:

[0019] Based on at least one of user interaction behavior, behavior monitoring data, and knowledge question and answer results, the execution status of the phase plan is analyzed to obtain the execution status.

[0020] Based on the changes in user status monitoring indicators corresponding to the type of the phase plan, the deviations from the plan are analyzed to obtain the current deviation status.

[0021] According to the planning management method provided by the present invention, managing the phase plan based on the execution status of the phase plan includes:

[0022] If a virtual coach is required based on the execution status of the phase plan, or in response to user input, a virtual coach model is invoked for encouraging interaction, the virtual coach model being trained based on a large language model.

[0023] According to the planning management method provided by the present invention, the step of invoking the virtual coaching model for encouraging interaction includes:

[0024] Based on the execution status of the phase plan and / or the user input, determine the virtual coach prompts;

[0025] Based on the virtual coach prompts and the requirements corresponding to the type of the phase plan, the virtual coach model is invoked to generate interactive text, and encouraging interaction is performed based on the interactive text.

[0026] The present invention also provides a planning management device, comprising:

[0027] The determination unit is used to determine the requirement text and user personalization information;

[0028] The plan generation unit is used to generate a current plan based on an expert rule model, applying the requirement text and user personalized information; to perform feasibility reasoning on the current plan based on a logical reasoning model and the user personalized information; and to adjust the current plan based on the reasoning results until a phase plan is obtained. The expert rule model and the logical reasoning model are trained based on a large-scale language model.

[0029] The planning management unit is used to manage the phase plan based on its execution status.

[0030] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the planning management method as described above.

[0031] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the planning management method as described above.

[0032] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the planning management method as described above.

[0033] The planning management method, apparatus, electronic device, and storage medium provided by this invention generate the current plan through an expert rule model, ensuring that the plan takes into account industry best practices and potential risks during the planning process. The personalized customization combined with user-specific information not only improves plan satisfaction but also enhances user participation. On this basis, the logical reasoning model can evaluate the feasibility of the plan, promptly identify and correct any unreasonable aspects of the plan, and ensure the smooth execution of the plan, thereby enhancing the feasibility and reliability of the plan. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0035] Figure 1 This is one of the flowcharts of the planning management method provided by the present invention.

[0036] Figure 2 This is the second flowchart of the planning management method provided by the present invention.

[0037] Figure 3 This is one of the flowcharts for step 130 in the planning management method provided by the present invention.

[0038] Figure 4 This is the second flowchart of step 130 in the planning management method provided by the present invention.

[0039] Figure 5 This is a flowchart illustrating the virtual coaching model provided by the present invention.

[0040] Figure 6 This is a schematic diagram of the planning management device provided by the present invention.

[0041] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0043] Related technologies for planning can be broadly categorized into two types. The first type relies on manual user operation, where users must manually input the specific execution steps and details for each plan on the software interface. While this method directly reflects the user's intentions, it requires users to have extensive knowledge and lacks intelligent assistance, making it difficult to adapt to rapidly changing planning needs.

[0044] The second approach utilizes the understanding capabilities of large language models to automatically generate plans based on users' natural language descriptions and create corresponding record files for users to review. While this approach improves the efficiency of plan creation to some extent, it cannot meet users' personalized customization needs, leading to frequent abandonment during execution due to unreasonable plans.

[0045] To address the aforementioned issues, this invention proposes a plan management method. During the plan formulation stage, based on an expert rule model, the current plan is generated using demand text and user-personalized information. Based on a logical reasoning model and user-personalized information, the feasibility of the current plan is determined, and the current plan is adjusted based on the reasoning results until a phase plan is obtained.

[0046] Expert rule models leverage the experience and knowledge of domain experts to ensure that best practices and potential risks are considered during the planning process. Expert rules can quickly identify key elements that match user needs and personalized information, leading to greater plan satisfaction and enhanced user engagement.

[0047] Logical reasoning models can assess the feasibility of a plan and check for logical flaws or contradictions. Through logical reasoning, unreasonable aspects of the plan can be identified and corrected in a timely manner, ensuring the smooth execution of the plan and thus enhancing its feasibility and reliability.

[0048] Based on this, planning management is carried out according to the implementation of the established phased plans. Deviations and problems in the plan execution process can be identified in a timely manner, and the plan can be adjusted quickly. This flexibility and adaptability helps to ensure the smooth execution of the plan.

[0049] The embodiments of this invention can be applied to scenarios requiring planning and management, such as fitness, weight loss, learning, and travel. The executing entity of this method can be an electronic device such as a terminal device, computer, server, server cluster, or a specially designed planning management device, or a planning management device installed within such an electronic device. This planning management device can be implemented through software, hardware, or a combination of both.

[0050] Figure 1 This is one of the flowcharts illustrating the planning management method provided by the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0051] Step 110: Determine the requirements text and user personalization information.

[0052] Specifically, a requirements document is usually submitted by the user and is typically expressed in text form. It describes the specific needs of the plan, including but not limited to functional requirements, performance requirements, and time requirements. The requirements document forms the basis for plan development. For example, a requirements document might be, "I want to lose 10 pounds in a month; please help me create a weight loss plan." Another example is, "I want to pass the CET-4 (College English Test Band 4) this year; please help me create an English learning plan."

[0053] Personalized user information refers to specific information related to a user, such as their behavioral preferences, user profiles, and needs. Personalized user information helps to more accurately understand user needs, thereby enabling the development of plans that better meet user expectations.

[0054] User behavior preferences may include, for example, a user's daily routine, dietary preferences, and commonly used hardware devices; user profiles may include, for example, educational background, geographical location, and job type; demand characteristics may include the category of the user's current demand (fitness, weight loss, learning, travel, education, etc.), the average time to achieve the goal (1 day, 10 days, 20 days, 2 years, etc.), and external dependencies for achieving the demand (purchasing equipment, travel services, video services, etc.).

[0055] The requirement text and user-personalized information can be obtained by communicating with users, collecting and organizing the information, or it can be entered by the users themselves.

[0056] Step 120: Based on the expert rule model, apply the requirement text and user personalized information to generate the current plan. Based on the logical reasoning model and user personalized information, perform feasibility reasoning on the current plan, and adjust the current plan based on the reasoning results until the stage plan is obtained. The expert rule model and logical reasoning model are trained based on a large language model.

[0057] In this step, the expert rule model is a model based on domain expert knowledge and experience, used to generate plans that conform to industry best practices or specific requirements. An expert rule model typically contains a set of rules developed based on expert experience to guide the formulation and adjustment of the plan. A rule package is created by human experts, then organized and refined before training the model. This model is then merged with a general large-scale language model to obtain the expert rule model. The requirement text and user-personalized information are input into the trained expert rule model to generate the current plan.

[0058] Considering that related technologies do not take into account the degree of matching between plans and user personalization information, resulting in poor feasibility of the formulated plans or logical loopholes, this embodiment performs feasibility reasoning on the current plan based on a logical reasoning model and user personalization information.

[0059] A logical reasoning model is a model based on logical reasoning used to evaluate the feasibility of a plan. It checks the logical consistency and condition fulfillment within the plan, ensuring its rationality and executability. This model is also trained using a large language model. By inputting the current plan and user-personalized information into the trained logical reasoning model, the model outputs its reasoning results for the current plan.

[0060] If the reasoning results indicate that the current plan is unreasonable or not feasible, it is necessary to adjust the current plan until the adjusted plan is both in line with the user's personalized customization information and feasible, that is, to obtain a phased plan that meets the requirements.

[0061] Step 130: Manage the phase plan based on its execution status.

[0062] Specifically, the execution status of a phase plan refers to its actual performance during the execution process, including information on progress, deviations, etc. Execution status is the foundation for managing phase plans, used to evaluate the effectiveness of plan execution and make corresponding adjustments.

[0063] Here, managing the phase plan may include real-time monitoring of the phase plan's execution and timely adjustments and optimizations based on feedback from the execution.

[0064] The method provided in this invention generates the current plan through an expert rule model, ensuring that the plan takes into account industry best practices and potential risks during the planning process. The personalized customization combined with user-specific information not only improves plan satisfaction but also enhances user participation. On this basis, the logical reasoning model can evaluate the feasibility of the plan, promptly identify and correct any unreasonable aspects of the plan, and ensure the smooth execution of the plan, thereby enhancing the feasibility and reliability of the plan.

[0065] Based on any of the above embodiments, step 110, which adjusts the current plan based on the reasoning result of the current plan until a stage plan is obtained, includes:

[0066] Step 111: If the reasoning result indicates that the current plan does not match the user's personalized information, then adjust the current plan until the adjusted current plan matches the user's personalized information, and determine the adjusted current plan as the phase plan.

[0067] Specifically, the matching between the current plan and the user's personalized information can be identified and judged by interpreting and analyzing the reasoning results output by the logical reasoning model. Based on a series of rules and algorithms, the logical reasoning model evaluates the fit between the current plan and the user's personalized information and provides a score or matching index. When this index falls below a preset threshold, the current plan is considered to be mismatched with the user's personalized information.

[0068] For example, if a user is a first-grade elementary school student and the current study time is scheduled for 10 pm, it is clearly not in line with the daily routine of elementary school students. In this case, it can be considered that the current plan does not match the user's personalized information. As another example, if a user is an adult female and the current plan includes 100 push-ups every day, it can be preliminarily determined that the current plan does not match the user's personalized information.

[0069] Once a mismatch is identified, it indicates that the current plan is not feasible, and adjustments are necessary. These adjustments can be made based on expert rule models. Expert rule models typically contain a set of rules based on domain knowledge and experience, which can help identify problems and provide corresponding solutions. In addition to expert rule models, user feedback is also an important basis for plan adjustments. For example, online feedback forms and multi-round dialogues can be set up to facilitate user feedback at any time, inviting users to participate in the plan's development and adjustment process.

[0070] Feasibility reasoning is performed on the adjusted current plan. If the adjusted current plan still does not match the user's personalized information, the current plan is adjusted again, forming an iterative loop, until the adjusted current plan fully matches the user's personalized information. At this point, the adjusted current plan that fully matches the user's personalized information can be identified as the phase plan.

[0071] The method provided in this invention, when the current plan does not match the user's personalized information, performs multiple rounds of adjustments through iterative loops until a phase plan that matches the user's personalized information is obtained, which can further enhance the executability and reliability of the plan.

[0072] Based on any of the above embodiments, if the reasoning result in step 111 indicates that the current plan does not match the user's personalized information, then the current plan is adjusted, including:

[0073] Step 111-1: If the inference result indicates that the current plan does not match the user's personalized information, then the inference result is sent to the expert rule model so that the expert rule model adjusts the current plan based on the inference result; and / or,

[0074] Step 111-2: Receive user feedback based on the reasoning results, and adjust the current plan based on the user feedback.

[0075] Specifically, when the logical reasoning model concludes that the current plan does not match the user's personalized information, the plan needs to be adjusted. Plan adjustments can be based on expert rule models and / or on user feedback.

[0076] The reasoning result is then passed as input data to the expert rule model. Upon receiving the result, the expert rule model analyzes the current plan based on its built-in rule set and algorithms. The rules may cover multiple aspects, including the plan's logic, feasibility, and user acceptance. Based on the analysis, the expert rule model proposes specific adjustments, such as modifying certain functions, optimizing processes, and adjusting timelines, thereby creating a new version of the plan.

[0077] Furthermore, the plan can be adjusted based on user feedback. After the logical reasoning model concludes a mismatch, this result is communicated to the user in an appropriate manner (such as email, SMS, in-app notifications, etc.). Simultaneously, users are encouraged to provide feedback based on the reasoning results, including their acceptance of the current plan and suggestions for improvement, and this feedback is received. The current plan is then adjusted accordingly based on the user feedback.

[0078] For example, a logical reasoning model determines that it is unreasonable for an adult woman to do 100 push-ups a day. However, user feedback indicates that users can accept 50 push-ups a day, so the plan can be adjusted based on user feedback.

[0079] The method provided in this invention adjusts the current plan based on expert rule models and / or user feedback, which can further improve the matching degree between the plan and the user's personalized information, thereby enhancing the plan's executability.

[0080] Based on any of the above embodiments Figure 2 This is the second flowchart of the planning management method provided by the present invention, as shown below. Figure 2 As shown, the plan generation process involves an expert rule model, a large-scale inference model, and multiple dialogues with the user. The user only participates when personalized information is required. The expert rule model uses a set of rules developed by human experts. These rules are then organized and trained into a model, which is then merged with a general large-scale model to generate the plan, incorporating personalized information. The logical inference model performs feasibility reasoning on the plan generated by the rule model, taking into account personalized information. Feedback can be provided on plans that the user cannot achieve, and these are submitted to the expert rule model for correction and adjustment. For plans requiring user confirmation, the user is guided to make a selection. After multiple adjustments, a highly feasible plan that meets the user's current conditions is obtained—this is the phased plan.

[0081] Based on any of the above embodiments Figure 3This is one of the flowcharts illustrating step 130 in the planning management method provided by this invention, such as... Figure 3 As shown, based on the execution status of the phase plan, the management of the phase plan includes:

[0082] Step 131: Analyze the execution status and deviations from the phase plan to obtain the current execution status and deviation status.

[0083] Step 132: Update user personalized information based on deviations from the current situation;

[0084] Step 133: Adjust the phase plan based on the current execution status and updated user personalization information.

[0085] Specifically, considering the relevant technologies, it is impossible to fine-tune the plan according to the user's personalized characteristics and progress during the execution process. Many new problems arise that the user did not consider when making the plan, resulting in the plan not conforming to the user's current situation.

[0086] The execution status of a phase plan refers to the actual progress made during the execution of that phase plan. This includes completed tasks, ongoing tasks, incomplete tasks, and any factors that may affect the plan's progress. For example, push-ups were completed for the day, and a running task is currently underway.

[0087] Analyzing the execution of phased plans can identify which tasks have been successfully completed and which have been delayed or incomplete. The resulting execution status helps in understanding the current situation, and may include information such as completed tasks, progress percentages, and resource usage.

[0088] Deviation from the plan refers to the difference between the actual execution of a phase plan and the original plan. For example, if the phase plan requires a weight loss of 5 pounds by the current time point, but the actual weight loss is only 3 pounds, this indicates a deviation from the plan. Another example is that the phase plan requires studying Chapter 3 by the current time point, but the actual study progress is to Chapter 4, indicating a deviation from the plan and premature learning.

[0089] By comparing the phased plan with the actual implementation, the analysis reveals deviations from the current situation, including the reasons, degree, and impact of the deviations.

[0090] Then, based on the deviation from the current situation, the user's personalized information is updated. This can adjust user needs, preferences, and expectations. For example, the user's personalized information might show that they have time in the morning, but by observing the deviation, it is found that the user actually has more free time in the evening, so the user's personalized information can be updated.

[0091] Based on this, the phase plan is adjusted according to the current execution status and the updated user personalization information. Specific adjustments can be made using expert rule models and logical reasoning models until the adjusted phase plan aligns with the updated user personalization information.

[0092] Figure 4 This is the second flowchart of step 130 in the planning management method provided by the present invention, as shown below. Figure 4 As shown, during the execution of the plan, the best personalized plan that best matches the user's individual information can be achieved by analyzing the execution and deviation multiple times and making multiple adjustments.

[0093] Based on any of the above embodiments, the execution status and deviation from the plan are analyzed to obtain the current execution status and the current deviation status. Specifically, step 131 includes:

[0094] Step 131-1: Analyze the execution status of the phase plan based on at least one of user interaction behavior, behavior monitoring data, and knowledge question and answer results to obtain the current execution status;

[0095] Step 131-2: Based on the changes in user status monitoring indicators corresponding to the type of phase plan, analyze the deviation from the plan respectively to obtain the current deviation status.

[0096] Specifically, the current execution status can be obtained by analyzing the execution of the phase plan based on at least one of user interaction behavior, behavior monitoring data, and knowledge question and answer results.

[0097] User interaction behavior refers to the actions a user takes when interacting with a system while using a product or service, such as clicking, swiping, typing, and browsing. These behaviors can reflect the user's feedback and attitude towards the execution of a planned task for that day. For example, if a user clicks on or views recommended content, it can be determined that the user has completed the tasks in the plan for that day.

[0098] Behavioral monitoring data is user interaction data collected through hardware. This data can quantify user activity, engagement, and usage habits, providing objective evidence for analyzing the execution of planned tasks. For example, by monitoring movement trajectories and step counts through hardware, it can be determined whether a user has completed their planned tasks for the day.

[0099] The knowledge-based Q&A results refer to the knowledge points required for the task mentioned in casual conversation between the virtual avatar and the user through knowledge-based questions. The results can reflect the user's understanding of the phase plan, the degree to which needs are met, and potential problems and confusions. If the user answers correctly, it is determined that the user has completed the planned tasks for that day.

[0100] To address deviations from the current situation, the deviations can be analyzed based on changes in user status monitoring indicators corresponding to the type of phase plan.

[0101] Phased plans can be categorized in various ways, such as fitness services, learning and improvement services, and health management services. Each type of phased plan has corresponding user status monitoring indicators. These indicators are quantitative measures of changes in a user's status during the execution of the plan. For example, for weight and fitness plans, indicators might include weight and body fat percentage, monitored via scales, body fat scales, or webcams. These changes indicate whether the plan is being completed ahead of schedule or behind schedule. For learning and improvement plans, indicators might include the user's mastery of each learning unit, which can be verified through quizzes. For health management plans, indicators might include blood pressure, blood sugar, complexion, and mood, analyzed via cameras or health devices.

[0102] Based on any of the above embodiments, managing the phase plan based on its execution status, i.e., step 130 specifically includes:

[0103] Step 134: If a virtual coach is required based on the execution status of the phase plan, or in response to user input, the virtual coach model is invoked for encouraging interaction. The virtual coach model is trained based on a large language model.

[0104] Specifically, considering that users may encounter various difficulties during the plan execution phase, easily developing a fear of difficulty and leading to unsuccessful plan execution, this embodiment can autonomously determine whether virtual coaching intervention is needed based on the execution status of each phase of the plan. For example, if a user has not executed the plan for many consecutive days, or if the user's recent execution performance is not optimistic, it can be determined that virtual coaching intervention is necessary.

[0105] A virtual coach is a simulated coach that utilizes artificial intelligence technology. It can mimic the behavior and thought processes of a real coach and interact with users through software or applications, providing guidance, advice, and encouragement. For example, in fitness program management, a virtual coach can provide voice guidance or feedback during training based on the user's progress, offering reassurance and helping to boost their confidence.

[0106] In another embodiment, the user can proactively initiate the virtual coaching service during the execution of the plan, that is, the user actively requests to interact with the virtual coach.

[0107] At this point, a pre-trained virtual coach model can be invoked to provide personalized encouragement and guidance based on the user's actual situation (such as progress of the plan, physical condition, and mental state). For example, if the user feels tired or frustrated during exercise, the virtual coach can use positive and encouraging words to motivate the user; if the user needs to improve a certain movement or technique, the virtual coach can provide detailed guidance and suggestions.

[0108] The virtual coach model is trained using a large-scale language model. A large amount of domain-specific text data is used to train the language model. During training, the model learns how to understand and generate natural language text relevant to the planning domain and how to provide appropriate guidance and suggestions based on the user's situation. After training, the model can be used as a virtual coach in the application to interact with the user in real time. Preferably, to further enhance the user experience, a virtual avatar can be generated to interact with the user.

[0109] The method provided in this invention utilizes a virtual coach trained with a large language model to assist users in completing phased plans and provides encouragement and guidance when necessary, thereby improving user participation and satisfaction.

[0110] Based on any of the above embodiments, invoking the virtual coach model for encouraging interaction includes:

[0111] Based on the execution status of the phase plan and / or user input, determine the virtual coach prompts;

[0112] Based on the virtual coach's prompts and the corresponding needs characteristics of the phase plan, the virtual coach model is invoked to generate interactive text, and encouraging interaction is carried out based on the interactive text.

[0113] Specifically, virtual coach prompts are system-generated information used to guide the interaction between the virtual coach and the user. Virtual coach prompts may include the execution status of the phase plan, i.e., the user's current progress, as well as questions and concerns entered by the user, or both execution status and user input.

[0114] Demand characteristics refer to specific needs or features related to the type of phase plan (such as fitness, learning, gaming, etc.). Demand characteristics determine what aspects a virtual coach should focus on when interacting with users, and how to provide personalized guidance and advice. Demand characteristics corresponding to different types of phase plans are shown in Table 1. For example, in an English speaking phase plan, the demand characteristic is typically that users can't stick to it from day 10 onwards; at this point, the virtual coach can directly guide them to engage in spoken conversations in daily life. For a learning phase plan, the demand characteristic is usually the existence of a learning setback curve; the virtual coach can encourage users to rest for 2-3 days and then continue.

[0115] Table 1

[0116]

[0117] Input text can be generated based on virtual coach prompts and the corresponding needs characteristics of the phase plan type. This input text is then fed into the virtual coach model to generate interactive text. The interactive text may include explanations of why the user encountered the problem and specific suggestions. The virtual coach can then engage in encouraging interaction with the user based on this interactive text to ensure the successful completion of the plan.

[0118] Figure 5 This is a flowchart illustrating the virtual coaching model provided by the present invention, as shown below. Figure 5 As shown, when a user inputs, "Is there no way to improve my writing skills? Am I stupid?", the user proactively initiates a virtual coaching encouragement service. The system automatically determines that the user has not written for four consecutive days. Combining the user's input and the current status of the plan's execution, it automatically supplements relevant information and obtains virtual coaching prompts. Then, considering the characteristic that writing exercises tend to regress in the early stages of the plan, the model outputs interactive text. The interactive text can cite examples of how historical writers honed their writing skills, using concrete examples to boost the user's writing confidence.

[0119] Based on any of the above embodiments, a plan management method is provided, including:

[0120] S1, determine the requirements text and user personalization information.

[0121] S2, based on the expert rule model, applies the requirement text and user personalized information to generate the current plan, and performs feasibility reasoning on the current plan based on the logical reasoning model and user personalized information.

[0122] If the inference result indicates that the current plan does not match the user's personalized information, the inference result is sent to the expert rule model so that the expert rule model can adjust the current plan based on the inference result; and / or, user feedback on the inference result is received, the current plan is adjusted based on the user feedback, and the current plan is further adjusted based on the inference result of the current plan until the adjusted current plan matches the user's personalized information, and the adjusted current plan is determined as the stage plan. The expert rule model and the logical inference model are trained based on a large language model.

[0123] S3 manages the phase plan based on its execution status. Specifically, it includes:

[0124] S31. Analyze the execution status of the phase plan based on at least one of user interaction behavior, behavior monitoring data, and knowledge question-and-answer results to obtain the execution status; analyze the deviation from the plan based on the changes in user status monitoring indicators corresponding to the type of phase plan to obtain the deviation status; update user personalized information based on the deviation status; and adjust the phase plan based on the execution status and the updated user personalized information.

[0125] S32, if the execution status of the phase plan determines that a virtual coach is needed, or in response to user input, the virtual coach model is invoked for encouraging interaction. The virtual coach model is trained based on a large language model. Invoking the virtual coach model for encouraging interaction includes:

[0126] Based on the execution status of the phase plan and / or user input, determine the virtual coach prompts; based on the virtual coach prompts and the demand characteristics corresponding to the type of phase plan, call the virtual coach model to generate interactive text, and conduct encouraging interactions based on the interactive text.

[0127] The planning management device provided by the present invention is described below. The planning management device described below and the planning management method described above can be referred to in correspondence.

[0128] Based on any of the above embodiments Figure 6 This is a schematic diagram of the planning management device provided by the present invention, as shown below. Figure 6 As shown, the device includes:

[0129] Unit 610 is used to determine the requirement text and user personalization information;

[0130] The plan generation unit 620 is used to generate a current plan based on an expert rule model, applying the requirement text and user personalized information; perform feasibility reasoning on the current plan based on a logical reasoning model and the user personalized information; and adjust the current plan based on the reasoning results until a phase plan is obtained; the expert rule model and the logical reasoning model are trained based on a large language model.

[0131] The planning management unit 630 is used to manage the phase plan based on the execution status of the phase plan.

[0132] The apparatus provided in this invention generates the current plan through an expert rule model, ensuring that the plan takes into account best practices and potential risks in the industry during the planning process. The personalized customization combined with user-specific information not only improves the satisfaction of the plan but also enhances user participation. On this basis, the logical reasoning model can evaluate the feasibility of the plan, promptly identify and correct any unreasonable aspects in the plan, and ensure the smooth execution of the plan, thereby enhancing the feasibility and reliability of the plan.

[0133] Based on any of the above embodiments, the planning generation unit is specifically used for:

[0134] If the reasoning result indicates that the current plan does not match the user's personalized information, the current plan is adjusted until the adjusted current plan matches the user's personalized information, and the adjusted current plan is determined to be the stage plan.

[0135] Based on any of the above embodiments, the planning generation unit is specifically used for:

[0136] If the inference result indicates that the current plan does not match the user's personalized information, then the inference result is sent to the expert rule model so that the expert rule model adjusts the current plan based on the inference result; and / or,

[0137] Receive user feedback on the inference results, and adjust the current plan based on the user feedback.

[0138] Based on any of the above embodiments, the planning management unit is specifically used for:

[0139] The execution status and deviations from the aforementioned phase plan are analyzed to obtain the current execution status and the current deviation status.

[0140] Based on the deviation from the current situation, update the user's personalized information;

[0141] Based on the current execution status and the updated user personalization information, the phase plan is adjusted.

[0142] Based on any of the above embodiments, the planning management unit is specifically used for:

[0143] Based on at least one of user interaction behavior, behavior monitoring data, and knowledge question and answer results, the execution status of the phase plan is analyzed to obtain the execution status.

[0144] Based on the changes in user status monitoring indicators corresponding to the type of the phase plan, the deviations from the plan are analyzed to obtain the current deviation status.

[0145] Based on any of the above embodiments, the planning management unit is specifically used for:

[0146] If a virtual coach is required based on the execution status of the phase plan, or in response to user input, a virtual coach model is invoked for encouraging interaction, the virtual coach model being trained based on a large language model.

[0147] Based on any of the above embodiments, the planning management unit is specifically used for:

[0148] Based on the execution status of the phase plan and / or the user input, determine the virtual coach prompts;

[0149] Based on the virtual coach prompts and the requirements corresponding to the type of the phase plan, the virtual coach model is invoked to generate interactive text, and encouraging interaction is performed based on the interactive text.

[0150] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include: a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, the communications interface 720, and the memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a plan management method, which includes: determining demand text and user personalization information; generating a current plan based on an expert rule model, applying the demand text and user personalization information; performing feasibility reasoning on the current plan based on a logical reasoning model and the user personalization information; adjusting the current plan based on the reasoning results until a phase plan is obtained; the expert rule model and the logical reasoning model are trained based on a large-scale language model; and managing the phase plan based on its execution status.

[0151] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the plan management method provided by the above methods. The method includes: determining demand text and user personalized information; generating a current plan based on an expert rule model, applying the demand text and user personalized information; performing feasibility reasoning on the current plan based on a logical reasoning model and the user personalized information; and adjusting the current plan based on the reasoning result until a phase plan is obtained; the expert rule model and the logical reasoning model are trained based on a large language model; and managing the phase plan based on the execution status of the phase plan.

[0153] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the plan management method provided by the above methods. The method includes: determining demand text and user personalization information; generating a current plan based on an expert rule model, applying the demand text and user personalization information; performing feasibility reasoning on the current plan based on a logical reasoning model and the user personalization information; and adjusting the current plan based on the reasoning results until a phase plan is obtained; wherein the expert rule model and the logical reasoning model are trained based on a large-scale language model; and managing the phase plan based on the execution status of the phase plan.

[0154] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0155] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A planning management method, characterized in that, include: Determine the requirements text and user personalization information; Based on the expert rule model, the current plan is generated by applying the demand text and user personalized information. Based on the logical reasoning model and the user personalized information, the feasibility of the current plan is reasoned, and the current plan is adjusted based on the reasoning results until a phase plan is obtained. The expert rule model and the logical reasoning model are trained based on a large-scale language model; Manage the phase plan based on its execution status; Managing the phase plan based on its execution status includes: The execution status and deviations from the aforementioned phase plan are analyzed to obtain the current execution status and the current deviation status. Based on the deviation from the current situation, update the user's personalized information; Based on the current execution status and the updated user personalization information, the phase plan is adjusted.

2. The planning management method according to claim 1, characterized in that, The process of adjusting the current plan based on the reasoning results until a phase plan is obtained includes: If the reasoning result indicates that the current plan does not match the user's personalized information, the current plan is adjusted until the adjusted current plan matches the user's personalized information, and the adjusted current plan is determined to be the stage plan.

3. The planning management method according to claim 2, characterized in that, If the inference result indicates that the current plan does not match the user's personalized information, then adjusting the current plan includes: If the inference result indicates that the current plan does not match the user's personalized information, then the inference result is sent to the expert rule model so that the expert rule model adjusts the current plan based on the inference result; and / or, Receive user feedback on the inference results, and adjust the current plan based on the user feedback.

4. The planning management method according to claim 1, characterized in that, The analysis of the execution status and deviation from the phase plan is performed to obtain the current execution status and the current deviation status, including: Based on at least one of user interaction behavior, behavior monitoring data, and knowledge question and answer results, the execution status of the phase plan is analyzed to obtain the execution status. Based on the changes in user status monitoring indicators corresponding to the type of the phase plan, the deviations from the plan are analyzed to obtain the current deviation status.

5. The planning management method according to any one of claims 1 to 3, characterized in that, Managing the phase plan based on its execution status includes: If a virtual coach is required based on the execution status of the phase plan, or in response to user input, a virtual coach model is invoked for encouraging interaction, the virtual coach model being trained based on a large language model.

6. The planning management method according to claim 5, characterized in that, The invocation of the virtual coaching model for encouraging interaction includes: Based on the execution status of the phase plan and / or the user input, determine the virtual coach prompts; Based on the virtual coach prompts and the requirements corresponding to the type of the phase plan, the virtual coach model is invoked to generate interactive text, and encouraging interaction is performed based on the interactive text.

7. A planning management device, characterized in that, include: The determination unit is used to determine the requirement text and user personalization information; The plan generation unit is used to generate a current plan based on an expert rule model, applying the requirement text and user personalized information, perform feasibility reasoning on the current plan based on a logical reasoning model and the user personalized information, and adjust the current plan based on the reasoning results until a phase plan is obtained. The expert rule model and the logical reasoning model are trained based on a large-scale language model; The planning management unit is used to manage the phase plan based on its execution status. Managing the phase plan based on its execution status includes: The execution status and deviations from the aforementioned phase plan are analyzed to obtain the current execution status and the current deviation status. Based on the deviation from the current situation, update the user's personalized information; Based on the current execution status and the updated user personalization information, the phase plan is adjusted.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the planning management method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the planning management method as described in any one of claims 1 to 6.

Images