Generation method and electronic device

Abstract

The application provides a generation method and an electronic device, and relates to the technical field of terminals. The method comprises the following steps: in response to target content input, the target content comprising question content; generating a target answer by using a target model, the target answer comprising first content and second content, the first content comprising first partial answer content of the question content in the target answer, and the second content comprising second partial answer content of the question content in the target answer; the target model generates a first display area with a visualized border and a second display area with a visualized border, the first content being displayed in the first display area, and the second content being displayed in the second display area.

Description

Technical Field

[0001] This application relates to the field of terminal technology, and in particular to a generation method and an electronic device. Background Technology

[0002] With the rapid development of artificial intelligence technology, more and more users are beginning to use it to handle tasks. Summary of the Invention

[0003] This application provides a generation method and an electronic device, the technical solution of which is implemented as follows: This application provides a generation method, the method comprising: Responding to the target content input, which includes the question content; The target model is used to generate a target answer, which includes a first part and a second part. The first part includes the first part of the target answer's response to the question, and the second part includes the second part of the target answer's response to the question. A first display area with a visual border and a second display area with a visual border are generated using the target model. The first content is displayed in the first display area and the second content is displayed in the second display area.

[0004] This application provides an electronic device, including at least one processor and a target model running on the processor. The processor invokes the target model to perform the following operations: Responding to the target content input, which includes the question content; The target model is used to generate a target answer, which includes a first part and a second part. The first part includes the first part of the target answer's response to the question, and the second part includes the second part of the target answer's response to the question. The target model generates a first display area with a visual border and a second display area with a visual border. The first content is displayed in the first display area and the second content is displayed in the second display area.

[0005] This application provides a computer-readable storage medium storing a computer program or computer-executable instructions for implementing the generation method described in any of the foregoing embodiments when executed by a processor.

[0006] This application provides a computer program product, including a computer program or computer executable instructions, which, when executed by a processor, implements the generation method described in any of the foregoing embodiments.

[0007] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and are not intended to limit the technical solutions of this application. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of a dynamic interactive interface provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 1 ; Figure 3 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 1 ; Figure 4 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 2 ; Figure 5 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 2 ; Figure 6 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 3 ; Figure 7 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 4 ; Figure 8 This is a schematic diagram of a custom display area sub-template provided in an embodiment of this application; Figure 9 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 5 ; Figure 10 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 6 ; Figure 11 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 1 ; Figure 12 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 2 ; Figure 13 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 3 ; Figure 14 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 3 ; Figure 15 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 7 ; Figure 16This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 8 ; Figure 17 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 4 ; Figure 18 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 5 ; Figure 19 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 6 ; Figure 20 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 9 ; Figure 21 This is a schematic diagram illustrating the changes in a visual interface provided in an embodiment of this application; Figure 22 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 10 ; Figure 23 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 4 ; Figure 24 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 10 one; Figure 25 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 5 ; Figure 26 This is a logic block diagram of a generation device provided in an embodiment of this application. Detailed Implementation

[0009] In order to gain a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for reference and illustration only and are not intended to limit the embodiments of this application.

[0010] Unless otherwise defined, the technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit the application. In the following description, references to "some embodiments" describe a subset of all possible embodiments; however, it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0011] It should also be noted that the terms "first, second, and third" used in the embodiments of this application are only used to distinguish similar objects and do not represent a specific order of objects. It can be understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0012] Furthermore, the reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0013] With the rapid development of artificial intelligence technology, more and more users are beginning to use it to handle tasks.

[0014] For example, users can send tasks to an artificial intelligence (AI) model, which then processes the tasks and provides feedback to the user. However, the content generated by the AI ​​model is displayed in a continuous interface, requiring users to spend a considerable amount of time browsing, thus increasing the user's time cost.

[0015] To address the aforementioned technical problems, this application provides a generation method. The method includes responding to target content input, whereby the target content includes question content; generating a target answer using a target model, whereby the target answer includes first content and second content. The first content includes a first portion of the target answer's response to the question content, and the second content includes a second portion of the target answer's response to the question content. The target model generates a first display area with a visual border and a second display area with a visual border, whereby the first content is displayed in the first display area and the second content is displayed in the second display area. In this application embodiment, after generating the target answer, the target model also generates a first display area and a second display area with visual borders. The first portion of the target answer's response is displayed in the first display area, and the second portion of the target answer's response is displayed in the second display area. This allows for a structured display of the target answer, enabling users to intuitively and selectively view content in different display areas, resulting in higher browsing efficiency compared to a continuous interface display.

[0016] On the one hand, displaying different content of the target answer in different display areas enables modular presentation of AI content, which helps improve user browsing efficiency. On the other hand, users can edit the content displayed in different display areas, enabling human-computer co-creation and improving creation efficiency and controllability.

[0017] Please refer to Figure 1 , Figure 1 This is a schematic diagram of a dynamic interactive interface provided in an embodiment of this application. Figure 1 The input port of the AI ​​model is displayed. The input port can include input commands such as "Tell me what you want" or "How can I help you?". Users can input target content into the input port, such as: generate a song, generate a weight loss plan, generate a weekend travel plan, etc. In response to the target content, the AI ​​model can generate a target answer. The target answer includes a first content and a second content. The AI ​​model can also generate a first display area and a second display area with visual borders. The first content is displayed in the first display area, and the second content is displayed in the second display area.

[0018] The various embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0019] The execution subject of the generation method provided in this application embodiment can be an electronic device, such as a user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, artificial intelligence (AI) phone, AI terminal, personal digital assistant (PDA), handheld device, computing device, vehicle-mounted device, or wearable device. In some possible implementations, the electronic device can also be a server, server cluster, cloud server, etc.

[0020] In this embodiment, the electronic device may deploy the target model provided in this embodiment, which may also be a cloud-based model. The electronic device can execute the generation method provided in this embodiment by calling the target model.

[0021] The target model can be a large or small AI (Artificial Intelligence) model. The large model can be a general-purpose large model, such as a general-purpose large language model (LLM), a visual large model, a multimodal large model, etc., or it can be an expert large model obtained by fine-tuning a general-purpose large model based on requirements. This application does not limit this. The large model learns the features and patterns of the input information by training on a large amount of diverse data. It typically has hundreds of millions to hundreds of billions of model parameters (such as the weights corresponding to the model network layers, used to control model behavior), and can capture complex relationships and patterns in the input information for accurate and efficient reasoning / computation.

[0022] In addition, the target model can be a local model built into the agent, or it can be an external model such as a cloud model, and there are no restrictions on this.

[0023] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 1 The method may include: Step 201: In response to the target content input, the target answer can be generated using the target model.

[0024] The target content can include the problem content.

[0025] In this embodiment, the target model can be an artificial intelligence model. The target content input can refer to the input content received by the target model's input port. The target content can be a question input by the user, a question received by the target model from a third party, or a question determined by the target model based on autonomous detection, autonomous perception, or other methods. In this embodiment, the target model can extract the question content from the received target content. The question content can include the user's question information.

[0026] As an example, the question could be something like: generate a song, plan a weekend getaway, or create a weight loss plan.

[0027] In the embodiments of this application, during the process of the electronic device calling the target model to process the question content, the target model can call one or more applications to process the received question content and generate the target answer.

[0028] In this embodiment, after obtaining the question content, the target model can break down the question content into sub-questions, obtain the answers corresponding to each sub-question, and combine the answers to each sub-question to obtain the target answer. Each answer to a sub-question is a component of the target answer.

[0029] For example, the target problem can be broken down into a first problem and a second problem. Here, "first" and "second" are used to indicate different parts of the problem content and do not constitute a restriction on the quantity or order.

[0030] In one implementation, the target problem can be broken down into multiple sub-problems based on the semantic information of the target problem.

[0031] For example, the problem content includes multiple semantic information. One semantic information is to generate a travel plan, and another semantic information is to provide a detailed description of the scenic spot's features. Therefore, based on the semantic information, the problem content can be divided into two sub-problems: generating a travel plan and generating a description of the scenic spot's features within the travel plan.

[0032] In another implementation, the target problem can be broken down into multiple sub-problems based on the user intent corresponding to the target problem. This involves analyzing and understanding the problem content to determine the implicit user needs, which represent what the user wants to do. Then, the necessary conditions for achieving these needs are determined based on these user needs. Finally, the corresponding sub-problem content is generated based on these necessary conditions.

[0033] For example, the problem is: generate a song. The target model, based on its understanding of the problem, can determine which types of information are necessary to generate a song. For instance, the target model might believe that generating a song requires three aspects: lyrics, a cover image related to the lyrics, and music that matches the lyrics. Based on this, the target model can break down the problem into three sub-problems: generating lyrics, generating a cover image related to the lyrics, and generating audio that matches the lyrics.

[0034] In another possible implementation, multiple question types and multiple sub-questions corresponding to each question type can be pre-set. These pre-set types are then stored in a database.

[0035] For example, for questions about generating songs, multiple sub-questions could be set up, such as generating lyrics, generating illustrations, and generating music. Similarly, for questions about generating weight loss plans, multiple sub-questions could be set up, such as generating exercise plans, generating medication plans, and generating diet plans. And for questions about travel plans, multiple sub-questions could be set up, such as generating a travel plan for the first time period and generating a travel plan for the second time period, with the number and duration of the first and second time periods determined based on the specific question content.

[0036] Once the content of the current question is obtained, it can be filtered from the database to determine the sub-questions corresponding to the current question content.

[0037] For example, if the question is about generating songs, then by comparing data with the database, we can determine that the three sub-questions are: generating lyrics, generating illustrations, and generating music.

[0038] In some embodiments, different sub-problems (such as the first problem content and the second problem content) can be independent and complete problem content, and may not be related to each other. For example, the first problem content is to generate lyrics, and the second problem content is to generate music. Another example is that the first problem content is to generate a Saturday travel plan, and the second problem content is to generate a Sunday travel plan.

[0039] In other embodiments, different sub-problems (such as the first problem content and the second problem content) may be interrelated. For example, the second problem content may depend on the processing result of the first problem content. For instance, the first problem content might be generating lyrics, and the second problem content might be generating images related to the lyrics. Or, for example, the first problem content might be generating a Saturday travel plan, and the second problem content might be generating a Sunday travel plan based on popular attractions in the Saturday travel plan.

[0040] In this embodiment, for each sub-question, the target model can call the corresponding application or logic module to process it and obtain the corresponding answer. The processing of multiple sub-questions is independent. Different sub-questions may call the same application or logic module for processing, or they may call different application or logic modules for processing.

[0041] The target model generates the first content by processing the content of the first question, and the second content by processing the content of the second question. The first and second contents constitute the target answer. The first content includes the first part of the target answer's response to the question content, and the second content includes the second part of the target answer's response to the question content. Both the first and second parts of the response are answers to the question content, and the combined content of the first and second parts represents a complete answer to the question content.

[0042] It should be noted that, in this embodiment, the target answer can be obtained by combining the answers to multiple sub-questions. Each sub-question's answer is a component of the target answer. The answer to a single sub-question cannot be used as the target answer for the entire question content.

[0043] For example, the problem of generating a weight loss plan can be broken down into three sub-problems: generating an exercise plan, generating a medication plan, and generating a diet plan. The target model can then call the corresponding logic module to process the exercise plan sub-problem and obtain the exercise plan. Similarly, the target model can also obtain the medication plan and the diet plan.

[0044] The combination of exercise, medication, and diet plans constitutes the target answer to the question "Generate a weight loss plan." Individual "exercise," "medication," and "diet" plans constitute partial answers to the question "Generate a weight loss plan."

[0045] For example, the problem of generating a song can be broken down into two sub-problems: generating lyrics and generating music. The target model can then call the corresponding logic module to process the lyrics sub-problem and obtain the lyrics. Similarly, the target model can also obtain a piece of music.

[0046] The lyrics and music together constitute the target answer to the question "Generate a song". The lyrics and music alone are partial answers to the question "Generate a song".

[0047] Step 202: Generate a first display area with a visual border and a second display area with a visual border using the target model.

[0048] The first content is displayed in the first display area, and the second content is displayed in the second display area.

[0049] In the embodiments of this application, the first display area and the second display area may have the same shape and area. For example, the first display area and the second display area may be rectangular display areas with the same length and width. As another example, the first display area and the second display area may be circular display areas with the same radius. The first display area and the second display area may also be irregular shapes. This application does not limit the shape and size of the first display area and the second display area.

[0050] In the embodiments of this application, please refer to Figure 3 , Figure 3 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 1 . Figure 3 The example shows the input port of the target model and four display areas with visual borders. The input port of the target model can be located above the four display areas, which can be displayed side by side.

[0051] In this embodiment of the application, a first display area may display first content, and a second display area may display second content. The first content and the second content can together constitute the target answer.

[0052] In the generation method provided in this application embodiment, the target model generates a target answer and can also generate a first display area and a second display area with visual borders. The first part of the target answer is displayed in the first display area, and the second part of the target answer is displayed in the second display area. This allows for a structured display of the target answer, enabling users to intuitively and selectively view content in different display areas. Compared to a continuous interface display method, this results in higher browsing efficiency for the user.

[0053] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 4 , Figure 4 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 2 The method includes: Step 401: Obtain the problem details.

[0054] The target model can obtain the problem content from the input port.

[0055] Step 402: Divide the problem content into the first problem content and the second problem content.

[0056] In this application embodiment, one splitting method can be based on semantic information splitting, that is, obtaining semantic information in the question content, and then splitting one or more sub-question contents according to the semantic information.

[0057] If the question content includes multiple semantic information, such as generating a travel plan, where the travel plan includes two semantic information pieces (e.g., 5A scenic spots), then two sub-questions can be generated. If the question content includes three semantic information pieces, then three sub-questions can be generated, and so on.

[0058] The sub-problems that are separated out can be defined as the first problem content and the second problem content.

[0059] Another approach to decomposition is intent-based decomposition. This involves analyzing and understanding the question content to identify the implicit user needs, which represent what the user wants to do. Then, based on these user needs, the necessary conditions for achieving those needs are determined. Finally, based on these necessary conditions, corresponding sub-question contents are generated, resulting in the first and second question contents.

[0060] For example, the problem statement might be: Generate a song. The target model, based on its understanding of this problem statement, can determine which types of information generating a song must include, such as generating lyrics, generating a cover image related to the lyrics, generating audio that matches the lyrics, and so on. Based on this, the target model can generate multiple sub-problems, which can include: generating lyrics, generating a cover image related to the lyrics, generating audio that matches the lyrics, and so on.

[0061] For example, the problem might be: Plan a weekend getaway. Based on its understanding of the problem, the target model can determine that generating a weekend getaway plan must include a two-day itinerary, such as a plan for day one, a plan for day two, and so on. Building on this, the target model can generate multiple sub-problems, which could include: generating the itinerary for day one, generating the itinerary for day two, and so on.

[0062] It should be noted that this application does not specifically limit the application scenario of splitting the problem content into a first problem content and a second problem content; that is, those skilled in the art can set and adjust its type according to the actual situation. The above situation is only an illustrative description of the forms in which the splitting of the problem content into a first problem content and a second problem content can be applied in this application, but it is not limited to the situations described in the above embodiments.

[0063] Step 403: Process the content of the first question using the target model to obtain the first part of the answer content; process the content of the second question using the target model to obtain the second part of the answer content.

[0064] The first part of the answer is the response to the first question, and the second part of the answer is the response to the second question.

[0065] In this embodiment, the first question content and the second question content can be independent and complete questions. The target model can call one or more applications to process the first question content and obtain the first part of the answer content corresponding to the first question content. Similarly, the target model can call one or more applications to process the second question content and obtain the second part of the answer content corresponding to the second question content.

[0066] In some embodiments, there is no correlation between the first question content and the second question content. Therefore, the target model does not refer to each other when processing the first and second question content, and the first and second question content can be executed in parallel or without a necessary order. In this case, the first part of the answer content and the second part of the answer content may not be related.

[0067] In other embodiments, there is a correlation between the content of the first question and the content of the second question, and correspondingly, there is a correlation between the content of the first part of the answer and the content of the second part of the answer.

[0068] For example, the first question is to generate lyrics, and the second question is to generate a cover image containing elements from the lyrics. The target model first needs to determine the execution order based on the first and second question content. When processing the later question, it needs to refer to the answers to the earlier questions. In other words, when processing the second question, the target model can refer to the lyrics generated in the first question, extract at least one visual element from the lyrics, and then generate the cover image based on that visual element.

[0069] Please refer to Figure 5 , Figure 5 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 2 . Figure 5 The image exemplarily illustrates a first display area and a second display area. The first display area displays a first portion of the answer content, including lyrics containing elements such as flowers and plants. The second display area displays a second portion of the answer content, including a cover image containing visual elements such as flowers and plants from the lyrics.

[0070] Step 404: Generate a first display area based on the content of the first question, and generate a second display area based on the content of the second question.

[0071] Specifically, generating a first display area based on the content of the first question can mean generating a first display area based on the first part of the answer corresponding to the content of the first question. Generating a second display area based on the content of the second question can mean generating a second display area based on the second part of the answer corresponding to the content of the second question.

[0072] In this embodiment, the target model can display the first part of the answer content in a first display area and the second part of the answer content in a second display area. Thus, the target answer can be divided into different modules and displayed in different display areas.

[0073] This application embodiment breaks down the question content into different sub-questions, processes them separately, and obtains the answer for each sub-question. Each sub-question has its own dedicated display area, where the answer to each sub-question is displayed. The answers to these multiple sub-questions together constitute the complete target answer to the question content, and each individual sub-question's answer is a part of the complete target answer. This allows users to intuitively and selectively view content in different display areas, resulting in higher browsing efficiency compared to a continuous interface display.

[0074] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 6 , Figure 6 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 3 The method includes: Step 601: Obtain the problem content and determine the task type of the problem content.

[0075] In this application embodiment, the task type may include: generating songs, generating plans, generating work reports, generating dances, generating videos, etc.

[0076] After obtaining the problem content, the target model can analyze the problem content, determine the user needs, and then determine the task type based on the user needs. For example, if the user needs to generate a song, then the task type can be determined to be song generation.

[0077] In another implementation, after obtaining the question content, the target model can extract keywords from the question content and determine the task type based on the keywords. For example, if the question content contains the keyword "report," then the task type can be determined to be generating a work report.

[0078] In this embodiment, the task type may further include a custom type, which is a type specifically set by the user based on their own preferences or usage habits. When setting a custom type, the user can pre-set keywords or triggering conditions corresponding to the custom type. Triggering conditions may include, for example, specific semantics, specific user needs / user intentions, etc.

[0079] In some embodiments, the custom type may overlap with the system's built-in task types. For example, a user may customize the generated song type based on personal habits, while the built-in task types also include the generated song type.

[0080] Therefore, if the problem content is determined to be a custom type and also falls under one of the system's built-in task types, the user-defined type has higher priority. Thus, the problem content is ultimately determined to be a custom type. The system's built-in task types include, as mentioned above, song generation, plan generation, work report generation, dance generation, video generation, etc.

[0081] In some cases, if the issue content cannot be mapped to any of the system's built-in task types, it can be categorized as a custom type. For example, when the issue content cannot be categorized, a task type can be selected from the custom types as the issue content.

[0082] Step 602: Call the corresponding target task template based on the task type of the problem content.

[0083] In this embodiment, a corresponding task template can be pre-configured for each task type. Each task template includes multiple display area sub-templates. Different task templates have different presentation methods, such as different display area sub-templates within different task templates.

[0084] In some embodiments, a correspondence between multiple task types and multiple task templates can be established, whereby a task type can be bound to a task template, and a task template can be bound to one or more task types. In one possible implementation, a task type-task template mapping table can be established and stored in memory. The target model can then call the task type-task template mapping table to obtain the mapping relationship between task types and task templates.

[0085] In this embodiment of the application, the target model can determine the target task template corresponding to the task type of the problem content based on the task type of the problem content and the obtained mapping relationship between the task type and the task template.

[0086] For example, if the task type of the question is "generate a song", then the task template corresponding to the "generate a song" type is the target task template.

[0087] In the embodiments of this application, the presentation methods of the display area sub-templates in different task templates are different.

[0088] For example, the first type of question content corresponds to task template A. The presentation methods of the multiple display area sub-templates in task template A may include displaying text, displaying audio, displaying images, etc. The first type of question content is, for example, question content related to songs, music, etc.

[0089] For example, the second type of question content corresponds to task template B. The presentation methods of the multiple display area sub-templates in task template B may include displaying animations, displaying 3D images, etc. The second type of question content may be related to work reports, animations, etc.

[0090] For example, the third type of question content corresponds to task template C. The presentation of multiple display area sub-templates in task template C may include displaying the content of the first stage, displaying the content of the second stage, and so on. The third type of question content may be question content related to planning.

[0091] In some embodiments, the plurality of display area sub-templates may be of the same type. For example, task template D may include a plurality of display area sub-templates for displaying text.

[0092] In this embodiment, each task template is matched to the task type. The number of display area sub-templates in the task template is the same as the number of sub-question contents broken down into the question content, so that the result of each sub-question content can be mapped to the corresponding display area sub-template for display.

[0093] In some embodiments, after obtaining the target task template, the target model can split the problem content according to the multiple display area sub-templates included in the target task template, and split the problem content into sub-problem content that corresponds one-to-one with each display area sub-template, thereby obtaining the first problem content and the second problem content.

[0094] For example, if the task type is "generate a song," the corresponding task template could include sub-templates for displaying lyrics, audio, and cover art. Therefore, the problem can be broken down into sub-problems such as generating lyrics, generating audio, and generating cover art.

[0095] In this embodiment of the application, the target task template includes multiple display area sub-templates, and the number of these multiple display area sub-templates is the same as the number of the first problem content and the second problem content split from the problem content.

[0096] In this embodiment, for user-defined types, the user simultaneously configures the corresponding task template when defining the task type. The process for customizing the task objective is described below.

[0097] Step 603: Map the first part of the answer content to the display area sub-template corresponding to the first question content to obtain the first display area; map the second part of the answer content to the display area sub-template corresponding to the second question content to obtain the second display area.

[0098] In this embodiment of the application, both the first part of the answer content and the second part of the answer content are answers to the question content, and the combined content of the first part of the answer content and the second part of the answer content represents a complete answer to the question content.

[0099] In this embodiment of the application, the target task template includes multiple display area sub-templates, which correspond one-to-one with the first problem content and the second problem content separated from the problem content.

[0100] For example, the target task template is a task template corresponding to the song generation type. This template can include sub-templates for displaying lyrics, audio, and cover images. The question content is broken down into sub-questions for generating lyrics, audio, and cover images. Specifically, the sub-question for generating lyrics corresponds to the sub-template for displaying lyrics, the sub-question for generating audio corresponds to the sub-template for displaying audio, and the sub-question for generating cover images corresponds to the sub-template for displaying cover images.

[0101] Based on the above correspondence, the first part of the answer to the first question is mapped into the display area sub-template corresponding to the first question, resulting in the first display area. Similarly, the second part of the answer to the second question is mapped into the display area sub-template corresponding to the second question, resulting in the second display area.

[0102] In this embodiment, the answer to a question is broken down into different sub-questions, and the answer to each sub-question is obtained. Each sub-question has its own dedicated display area, and the answer to each sub-question is displayed within its designated area. The answers to these multiple sub-questions together constitute the complete target answer to the question, with each individual sub-question's answer being a part of the complete target answer. This allows users to intuitively and selectively view content in different display areas, resulting in higher browsing efficiency compared to a continuous interface display.

[0103] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 7 , Figure 7 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 4 The method includes: Step 701: Receive user-defined instructions.

[0104] Among them, user-defined instructions include template information for custom task types and corresponding custom task templates.

[0105] In this embodiment of the application, in the custom mode, the user can input custom instructions and indicate information such as the name, keywords, and triggering conditions of the custom task type in the custom instructions. The user can also indicate the template information of the custom task template in the custom instructions.

[0106] The template information may include the number of display area sub-templates included in the custom task template, the type of each display area sub-template, the parameter information and layout information of each display area sub-template, etc. The parameter information of the display area sub-template is used to limit the display method of the display area sub-template, such as how to display video type display area sub-templates, how to display text type display area sub-templates, etc.

[0107] In one possible implementation, the target model can provide a configuration interface containing configuration items required for custom task types and custom task templates. Users can select or input these items in the configuration interface to generate template information for custom task types and custom task templates, and then pass it to the target model as a custom instruction.

[0108] In another possible implementation, custom instructions can be obtained through multi-turn dialogue. In custom mode, the target model can obtain the user-defined custom task type and template information of the custom task template through dialogue.

[0109] Step 702: In response to user-defined instructions, generate custom task types and custom task templates.

[0110] The custom task template includes multiple custom display area sub-templates.

[0111] In this embodiment, the target model can respond to user-defined instructions and determine the task type corresponding to the custom task template based on the instructions. The task type could be, for example, generating a paper, generating a song, generating a dance, etc. Multiple custom display area sub-templates can be created based on the template information.

[0112] The process of creating a custom display area sub-template may include, for example, generating a basic framework for the display area sub-template based on its type. This basic framework includes a visual border with a target shape and interactive controls. The internal structure of the basic framework is determined based on the parameter and layout information of the display area sub-template, thus obtaining the display area sub-template. This application does not limit the internal structure of the display area sub-template.

[0113] Please refer to Figure 8 , Figure 8 This is a schematic diagram of a custom display area sub-template provided in an embodiment of this application. The type of this custom display area sub-template is audio type, including a visual border, interactive controls, and an audio display area. Figure 8 The example shows three interactive controls located above a custom display area sub-template, and also shows three audio display areas, which are the areas for displaying subsequent audio content.

[0114] The custom task types and custom task goals provided in this application can enhance user autonomy and better meet users' personalized usage needs.

[0115] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 9 , Figure 9 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 5 The method includes: Step 901: Determine the task type of the problem content.

[0116] In this embodiment of the application, keywords can be extracted from the question content, and then the task type can be determined based on the keywords.

[0117] In one implementation, a mapping relationship between keywords and task types can be pre-established. This mapping relationship is then retrieved from the database to determine the task type for the question content.

[0118] Step 902: If the task type of the problem content is a custom task type, then the custom task template corresponding to the custom task type is determined as the target task template corresponding to the task type of the problem content.

[0119] In this embodiment of the application, each task type is bound to a task template. When the problem content belongs to a custom task type, the custom task template corresponding to that custom task type is the target task template.

[0120] Step 903: Break down the problem content into subtasks corresponding to multiple custom display area sub-templates in the target task template.

[0121] The target task template includes multiple custom display area sub-templates. These multiple custom display area sub-templates may be of the same type or different types.

[0122] In this embodiment of the application, the target model can split the problem content according to the multiple custom display area sub-templates, and split the problem content into sub-tasks that correspond one-to-one with each custom display area sub-template.

[0123] For example, the problem statement is: Generate a song. The target task template includes several custom display area sub-templates, including a display area sub-template for displaying lyrics and a display area sub-template for displaying an image. Therefore, the problem statement can be broken down into two sub-tasks: generating lyrics and generating a cover image.

[0124] Among them, the subtask of generating lyrics corresponds to the sub-template of the display area for displaying lyrics, and the subtask of generating cover image corresponds to the sub-template of the display area for displaying image.

[0125] Step 904: For each subtask, map the execution result of the subtask to the corresponding custom display area sub-template.

[0126] Continuing with the example above, the target model can perform the subtask of generating lyrics and map the generated lyrics to a display area sub-template to obtain the first display area. It can also perform the subtask of generating a cover image and map the generated cover image to a display area sub-template to obtain the second display area.

[0127] In this embodiment of the application, the variety of task types supported by the target model can be expanded by customizing task types and customizing task templates, which can also better meet the personalized needs of users.

[0128] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 10 , Figure 10 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 6 The method includes: Step 1001: Obtain the question content corresponding to the target content.

[0129] Step 1002: Determine the task type of the problem content, and determine the target task template based on the task type.

[0130] The target task template includes multiple display area sub-templates.

[0131] Step 1003: The problem content is split into the first problem content and the second problem content.

[0132] The content of the first question and the content of the second question correspond one-to-one with the multiple display area sub-templates in the target task template.

[0133] Step 1004: Use the target model to generate the first part of the answer content corresponding to the first question content, and generate the second part of the answer content corresponding to the second question content.

[0134] In this embodiment of the application, the first part of the answer content includes multiple first candidate contents, and the second part of the answer content includes multiple second candidate contents.

[0135] In this embodiment of the application, when processing the content of the first question, the target model can generate multiple answers based on different angles, different wording, different logic, etc., thereby obtaining multiple first candidate contents, which together constitute the first part of the answer content. That is to say, for a first question, the target model can output multiple answers, and these multiple answers exist in parallel and independently.

[0136] Similarly, when processing the second question content, the target model can also generate multiple second candidate contents.

[0137] Step 1005: Map the first part of the answer content to the first display area corresponding to the first question content, and map the second part of the answer content to the second display area corresponding to the second question content.

[0138] The first and second display areas each include interactive controls. Users can independently operate the first and second display areas based on these interactive controls.

[0139] In one implementation, the user can determine the display method of the first part of the answer content in the first display area based on interactive controls. Specifically, multiple first candidate entries from the first part of the answer content can be mapped to the first display area, meaning that all multiple first candidate entries can be seen simultaneously in the first display area. Similarly, multiple second candidate entries can be mapped to the second display area, meaning that all second candidate entries can be seen simultaneously in the second display area.

[0140] Please refer to Figure 11 , Figure 11 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 1 The first display area shows multiple first-choice content items. Figure 11 The example shows five first-choice candidates, displayed side-by-side. Users can visually see all the first-choice candidates.

[0141] Please refer to Figure 12 , Figure 12This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 2 The first display area shows multiple candidate items, which partially overlap each other, with the top candidate item being visible. Users can use interactive controls within the first display area to select one of the candidate items to make visible.

[0142] In this embodiment of the application, the second display area can be displayed in the same way as the first display area, which will not be described in detail here.

[0143] In another implementation, the target candidate content can be displayed in a first display area and then in a second display area.

[0144] Please refer to Figure 13 , Figure 13 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 3 The first display area shows the target candidate content, which is the selected one among a plurality of first candidate content. In some embodiments, the target candidate content is selected by the target model from a plurality of first candidate content based on a score. In other embodiments, the target candidate content may be selected by the user.

[0145] The second display area can be displayed in the same way as the first display area, which will not be elaborated here.

[0146] In this embodiment, when processing the first problem content, the target model can generate multiple first candidate contents and provide these multiple first candidate contents to the user for selection. After receiving the user's selection instruction, the first candidate content selected by the user is used as the target candidate content and mapped to the first display area for display. Similarly, the user can select one from multiple second candidate contents as the target candidate content and map it to the second display area for display.

[0147] In another implementation, the first display area and the second display area can display content in different ways.

[0148] Users can select how the first part of the answer content is displayed in the first display area using interactive controls, and they can also select how the second part of the answer content is displayed in the second display area using interactive controls.

[0149] Please refer to Figure 14 , Figure 14 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 3 In this visualization interface, one target candidate is displayed in the first display area, and multiple second candidate items are displayed in the second display area.

[0150] In this embodiment, the user and the target model participate in the task processing together, realizing human-computer co-creation, improving interactivity, and facilitating the acquisition of results that better meet the user's needs.

[0151] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 15 , Figure 15 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 7 The method includes: Step 1501: Obtain the question content corresponding to the target content.

[0152] Step 1502: Generate the target answer using the target model.

[0153] The target answer includes a first part and a second part. The first part includes the first part of the target answer's response to the question, and the second part includes the second part of the target answer's response to the question.

[0154] The first part of the answer includes multiple first-choice options. The second part of the answer includes multiple second-choice options.

[0155] Step 1503: Using the target model, generate a first display area with a visual border based on the target answer, and generate a second display area with a visual border. The first display area includes interactive controls, and the second display area includes interactive controls.

[0156] Step 1504: Edit the first display area and the second display area using the target model.

[0157] In this embodiment, the user can perform independent editing operations on the first display area and the second display area based on interactive controls.

[0158] In this embodiment, editing the first display area / second display area can include two methods: one is to edit the entire first display area / second display area, and the other is to edit the displayed content of the first display area / second display area. These will be described separately below.

[0159] In one implementation, when multiple first candidate contents are displayed in the first display area, the user can select one from the multiple first candidate contents. In response to the user's selection operation, the target model can update the first display area, and the first candidate contents selected by the user can be displayed in the updated first display area, while other first candidate contents not selected by the user are no longer displayed.

[0160] In one implementation, editing the entire first display area / second display area may include: content regeneration processing, content re-layout processing, etc.

[0161] In some cases, the multiple candidate contents displayed in the first display area may not meet the user's needs. In such cases, the user can issue a content regeneration command by operating the interactive controls in the first display area.

[0162] Upon receiving a content regeneration instruction from the first display area, the target model determines the first question content corresponding to that area. It then calls an application or logic module capable of processing the first question content to regenerate the answer, resulting in a regenerated answer. This regenerated answer is the new first part of the answer. The target model can then map this new first part of the answer back to the first display area, overwriting the original first part of the answer in that area.

[0163] It's important to note that the target model can invoke different applications or logic modules when processing different sub-problems. Therefore, when regenerating the content of the first problem, the target model can only invoke the application or logic module corresponding to the content of the first problem and process only that content. This process does not involve the content of the second problem; therefore, regenerating the content of the first problem does not affect the existing answers to the second problem, nor does it require regenerating the answers to the second problem. The content of the first problem and the content of the second problem are decoupled.

[0164] In the embodiments of this application, when the target model undergoes content regeneration, it can typically be reprocessed as a sub-problem to obtain the regenerated result, without needing to regenerate the entire problem content.

[0165] The first part of the answer before regeneration differs from the first part of the answer after regeneration. In embodiments of this application, the new first part of the answer may include at least one first candidate content. In some embodiments, the first candidate content in the new first part of the answer may be partially the same as the first candidate content in the first part of the answer before regeneration.

[0166] Similarly, the target model can be edited in the same way for the second display area, which will not be elaborated here.

[0167] It should be noted that, in the embodiments of this application, the target model performs independent editing operations on the first display area and the second display area without interfering with each other. In some embodiments, the content of the second display area remains unchanged during the regeneration of the first display area. In other embodiments, the second display area may also be undergoing a regeneration operation during the regeneration of the first display area, with both operations running in parallel without interference. In still other embodiments, the second display area may be undergoing other editing operations during the regeneration of the first display area.

[0168] In this embodiment of the application, editing the first display area / second display area may further include: editing the layout of the first display area and the second display area in the visualization interface.

[0169] In this embodiment, the user can drag the first display area as a whole based on the operation of the interactive controls on the first display area. The target model can respond to the user's operation and drag the first display area to a certain position in the visualization interface. Similarly, the target model can respond to the user's drag operation on the second display area and drag the second display area to a certain position in the visualization interface.

[0170] In another implementation, editing the content displayed in the first display area / second display area may include adding, deleting, or modifying the displayed content.

[0171] Among them, the addition, deletion, and modification processing refers to the processing of adding, deleting, and modifying the displayed content.

[0172] In this embodiment, taking the first display area as an example, the user can trigger modifications to the content displayed within the first display area based on the interactive controls of the first display area. After receiving the user's modification operation, the target model controls the first display area to enter modification mode.

[0173] In one implementation, the target model can respond to user modification operations by deleting one or more of the multiple first candidate contents displayed in the first display area.

[0174] In another implementation, the target model can respond to the user's modification operation by selecting one of the multiple first candidate contents displayed in the first display area, using the selected first candidate content as the target candidate content, retaining the target candidate content, and deleting the remaining first candidate contents.

[0175] In another implementation, the target model can respond to user modifications by editing the text of the first candidate content. Text editing includes inputting text, deleting text, modifying the position of text paragraphs, and so on.

[0176] Similarly, the target model can respond to the user's operation on the interactive controls of the second display area and perform the above-mentioned editing processing on multiple second candidate contents in the second display area, which will not be elaborated here.

[0177] In this embodiment, users can edit the first display area / second display area through interactive controls to achieve human-computer co-creation.

[0178] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 16 , Figure 16 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 8 The method includes: Step 1601: Obtain the question content corresponding to the target content, and split the question content into the first question content and the second question content.

[0179] In this embodiment, the target model can analyze the question content to determine what the user-instructed action should be. Then, it determines what information the user-instructed action needs to include, and based on this, breaks it down into multiple first question contents and second question contents. These multiple first question contents and second question contents are used to obtain the required information.

[0180] For example, if the problem is to generate a song, the target model determines that generating a song requires lyrics, music, and images. Therefore, the problem can be broken down into: generating lyrics, generating audio, and generating images.

[0181] Step 1602: Use the target model to generate the first part of the answer content corresponding to the first question content, and generate the second part of the answer content corresponding to the second question content.

[0182] Step 1603: Based on the first part of the answer content, obtain the first target information from the first part of the answer content. Based on the second part of the answer content, obtain the second target information from the second part of the answer content.

[0183] The first target information includes at least one of the following: content object, content address, service information, order parameters, and payment port; the second target information includes at least one of the following: content object, content address, service information, order parameters, and payment port.

[0184] Here, "content object" refers to the target involved in the first target information, such as a cinema, scenic spot, park, shopping mall, etc. "Content address" refers to the address information corresponding to the content object. "Service information" refers to the services that the content object can provide, such as the types and types of services. Service information can also be the service content and types associated with the content object. "Order parameters" can include order amount, order instructions, and other order-related data. "Payment port" refers to the port through which payment is processed when a user places an order.

[0185] In this embodiment, after generating the first part of the answer content, the target model can analyze it. It can search for content that the user might be interested in from the first part of the answer content, such as cinemas, scenic spots, parks, shopping malls, restaurants, performances, etc. Then, it can extract target information from this potentially interesting content.

[0186] In some embodiments, if the first part of the answer includes a first candidate, the target model can analyze that first candidate. If the first part of the answer includes multiple first candidates, the target model can analyze each of the multiple first candidates separately.

[0187] Taking a first candidate item in the first part of the answer as an example, the target model can analyze this first candidate item. If it includes content that the user might be interested in, then the first target information can be extracted from this potentially interesting content. This potentially interesting content could include, for example, scenic spots, performances, restaurants, hotels, etc. It can also be user-defined content, such as amusement parks, cinemas, supermarkets, etc.

[0188] In some embodiments, the first candidate content may include one or more pieces of content that a user might be interested in. Specifically, if the first candidate content includes content that a user might be interested in, then the first target information corresponding to that content is obtained.

[0189] If the first candidate content contains multiple pieces of content that users may be interested in, then one first target information can be obtained based on each piece of content that users may be interested in, thus obtaining multiple independent first target information.

[0190] In this embodiment, the processing method for the first part of the answer can be used to process the second part of the answer to obtain the second target information. Further details are omitted here.

[0191] Step 1604: Generate first access port data based on the first target information and display the first access port data in the first display area. Generate second access port data based on the second target information and display the second access port data in the second display area.

[0192] Specifically, the first access port data is used to interact with the target device, enabling the target device to obtain first target information; the second access port data is used to interact with the target device, enabling the target device to obtain second target information. The target device is a third-party device.

[0193] In this embodiment, after obtaining the first target information, the first target information can be packaged into a secure, proprietary data string. The target model can use this proprietary data to generate first access port data. The first access port data can be, for example, a hyperlink, a QR code, a barcode, etc.

[0194] If the first candidate content includes multiple first target information entries, each first target information entry can be independently packaged and its corresponding first access port data can be generated independently. In other words, multiple first access port data entries can be generated based on the first candidate content.

[0195] In this embodiment, after the target model generates the first access port data, the first access port data can be displayed within a first display area. In some embodiments, the first access port data can be displayed at a target location within the first display area.

[0196] Please refer to Figure 17 , Figure 17 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 4 . Figure 17 The example displays first access port data at the bottom of the first display area. This first access port data is, for example, a QR code.

[0197] If there are multiple first access port data entries, all of them can be displayed at the target location. Please refer to [reference needed]. Figure 18 , Figure 18 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 5 . Figure 18 The example displays multiple first access port data at the bottom of the first display area. These multiple first access port data are, for example, multiple QR codes.

[0198] In other embodiments, since each piece of first access port data corresponds to a piece of first target information, and each piece of first target information corresponds to a piece of content that a user may be interested in, the first access port data can be associated with the content that a user may be interested in. This allows the first access port data to be displayed at the location of its corresponding content that the user may be interested in.

[0199] Please refer to Figure 19 , Figure 19 This is a schematic diagram of a first display area provided in an embodiment of this application. Figure 6 . Figure 19 The example shows two first access port data (e.g., QR codes), each of which is displayed at the location of its corresponding content that the user may be interested in.

[0200] In this embodiment of the application, when a user scans or triggers the first access port data using the target device, the first target information corresponding to the first access port data can be displayed on the target device in response to the triggering operation.

[0201] In this embodiment, the generation and display process of the second access port data can be the same as the generation and display process of the first access port data, and will not be described in detail here.

[0202] In this embodiment, the target device can directly obtain the first target information through the first access port data, realizing cross-device information acquisition. The first access port data is generated based on the first target information in the first part of the answer content generated by the target model, and therefore is secure, unique, and tamper-proof.

[0203] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 20 , Figure 20 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 9 The method includes: Step 2001: Obtain the question content and use the target model to process the question content and generate the target answer.

[0204] The target answer includes a first part and a second part. The first part includes the first part of the target answer's response to the question, and the second part includes the second part of the target answer's response to the question.

[0205] Step 2002: Generate a first display area with a visual border and a second display area with a visual border using the target model.

[0206] The first content is displayed in the first display area, and the second content is displayed in the second display area.

[0207] Step 2003: In response to the fusion command, generate a fusion display area template, which includes a first module and a second module.

[0208] The first module corresponds to the first display area, and the second module corresponds to the second display area. The positions and area proportions of the first and second modules within the merged display area template are determined based on the logical relationship and / or weight information between the first and second display areas.

[0209] In this embodiment, the user can trigger a fusion command through a visual interface. In one implementation, the visual interface can include interactive controls, which can be located in a first display area or within that area, or they can be displayed in both the first display area and areas outside the first display area. The user can trigger the fusion command through these interactive controls.

[0210] In this embodiment, the fusion instruction may include ID information of the first display area and the second display area, so that the target model can identify the display areas to be fused. In some embodiments, all first and second display areas in the visualization interface can be fused. In other embodiments, the first and second display areas selected by the user can be fused.

[0211] In this embodiment, if the first display area displays multiple first candidate contents and the second display area displays multiple second candidate contents, then the user can specify the selected first display area and the second display area in the fusion instruction, and can also specify the first candidate contents in the selected first display area to participate in the fusion, and specify the second candidate contents in the selected second display area to participate in the fusion.

[0212] During merging, the first candidate content selected in the first selected display area is mapped to the first module in the merged display area template, and the second candidate content selected in the second selected display area is mapped to the second module in the merged display area template.

[0213] The target model can obtain the logical relationship and weight information between the first display area and the second display area, and then determine the first module and the second module in the merged display area template based on the logical relationship and weight information. The first module and the second module can have the same shape and area, and the first module and the second module are displayed side by side.

[0214] In this embodiment, the number of selected first and second display areas is the same as the number of first and second modules in the merged display area template. The first and second modules represent different modules and do not differ in order.

[0215] Here, the logical relationship refers to the dependency between the first question content corresponding to the first display area and the second question content corresponding to the second display area. The weight information refers to the relative weights of the first question content corresponding to the first display area and the second question content corresponding to the second display area. The target model can determine the dependency and weight information between the first and second question content when splitting the question content.

[0216] For example, the first question is to generate lyrics, and the second question is to generate a cover image containing elements from the lyrics. In this case, processing the second question depends on the answer to the first question; therefore, the second question is strongly dependent on the first. In this embodiment, the weight information between the two questions can be determined based on their dependency relationship.

[0217] For example, since the content of the second question is strongly dependent on the content of the first question, it can be determined that the weight of the content of the first question is greater than the weight of the content of the second question. Accordingly, the weight of the first display area is greater than the weight of the second display area.

[0218] For example, the first problem is to generate lyrics, and the second problem is to generate an image. In this case, there is no explicit dependency between the first and second problem contents. Therefore, the weights of the first and second problem contents can be the same, and correspondingly, the weights of the first and second display areas can be the same.

[0219] Step 2004: Map the first content displayed in the first display area to the first module, map the second content displayed in the second display area to the second module, and display the resulting merged display area in the visualization interface.

[0220] Please refer to Figure 21 , Figure 21 This is a schematic diagram illustrating the changes in a visualization interface provided in an embodiment of this application. Before fusion, the visualization interface exemplarily displays a first display area and a second display area. The user can issue a fusion command, and upon receiving the fusion command, the target model updates the visualization interface in response. The updated visualization interface includes a first display area, a second display area, and a fused display area.

[0221] The fused display area includes a first module and a second module; the first content in the first display area is mapped to the first module in the fused display area for display, and the second content in the second display area is mapped to the second module in the fused display area for display.

[0222] The embodiments of this application can output the fused content as multimodal files such as audio, lyrics, music videos, and visual reports, and support export and sharing.

[0223] In this embodiment of the application, the target answer can be merged into one display area by fusing the first display area and the second display area, which facilitates the user to output the final version.

[0224] In this embodiment, if the user modifies the content in the first display area or the second display area, a new fusion command can be issued to merge the modified first and second display areas. The new fusion result will then overwrite the original fusion result.

[0225] In this embodiment, the merged display area may include an interactive control, which the user can trigger to edit the merged display area. Editing includes deleting, dragging, or rearranging at least one of the first and second modules within the merged display area.

[0226] For example, you can zoom in or out on the first / second module, delete the first or second module, adjust the relative positions of the first and second modules, and so on.

[0227] It should be noted that, in this embodiment of the application, the editing of the fused display area can be applied to the first module and the second module, but cannot modify the content in the first module and the second module.

[0228] In this embodiment of the application, if the user needs to modify the content in the first module and the second module, the content in the first display area or the second display area can be modified, and then the modified first display area or the second display area can be re-fused, and the new fusion result can be used to cover the original fusion result, so that the content in the first module and the second module in the fused display area changes.

[0229] In this embodiment, after obtaining the first display area and the second display area, a fusion command can be used to merge the first display area and the second display area into a single merged display area. Furthermore, the first module / second module within the merged display area can be edited, enabling human-computer co-creation.

[0230] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 22 , Figure 22 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 10 The method includes: Step 2201: Obtain the problem content. The task type of the problem content is "generate song".

[0231] The question could be something like: Generate a song.

[0232] Step 2202: Obtain the song task template corresponding to the generated song type. The song task template includes at least: lyrics template, image template and audio template.

[0233] In this embodiment, the lyrics template, image template, and audio template belong to different types of display area sub-templates.

[0234] Step 2203: Break down the problem content into lyrics task, image task, and audio task.

[0235] Step 2204: Call the first submodule to perform the lyrics task and obtain at least one lyric; call the second submodule to perform the image task and obtain at least one image; call the third submodule to perform the audio task and obtain at least one audio.

[0236] Among them, the first submodule, the second submodule, and the third submodule are different types of logical modules.

[0237] In the embodiments of this application, the target model can call different applications or different logic modules to independently handle each sub-problem.

[0238] Step 2205: Map at least one lyric to a lyric template to obtain a lyric display area; map at least one image to an image template to obtain an image display area; map at least one audio to an audio template to obtain an audio display area.

[0239] In one implementation, all obtained lyrics can be mapped to a lyrics template to obtain a lyrics display area. Similarly, all obtained image content can be mapped to an image template to obtain an image display area, and all obtained audio content can be mapped to an audio template to obtain an audio display area.

[0240] In another implementation, target lyrics can be selected from the at least one set of lyrics, and the target lyrics can be mapped onto a lyrics template to obtain a lyrics display area. Similarly, target image content can be selected from the at least one set of image content, and the target image content can be mapped onto an image template to obtain an image display area. Finally, target audio content can be selected from the at least one set of audio content, and the target audio content can be mapped onto an audio template to obtain an audio display area.

[0241] Please refer to Figure 23 , Figure 23 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 4 The visualization interface includes a lyrics display area, an image display area, and an audio display area. The lyrics display area shows the lyrics generated by the target model, the image display area shows the cover image, and the audio display area shows the background music for the lyrics.

[0242] In this embodiment, different parts of the target answer are displayed in different display areas, allowing users to intuitively and selectively view the content of different display areas, thereby improving user browsing efficiency.

[0243] Based on the above embodiments, this application also provides a generation method, please refer to... Figure 24 , Figure 24 This is a flowchart illustrating a generation method provided in an embodiment of this application. Figure 10 1. The method includes: Step 2401: Obtain the problem content. The task type of the problem content is the generation plan type.

[0244] The question could be something like: Generate a weekend travel plan.

[0245] Step 2402: Obtain the planned task template corresponding to the generated plan type. The planned task template includes at least: a first time period template and a second time period template.

[0246] In this embodiment, the first time period template and the second time period template belong to the same type of display area sub-template.

[0247] It should be noted that the first time period template and the second time period template represent templates for different time periods and do not constitute a limitation on the number of time periods.

[0248] Step 2403: Divide the problem content into planned tasks for the first time period and planned tasks for the second time period.

[0249] The first phase of the plan includes tasks such as departing on Saturday and planning the trip, while the second phase includes tasks such as traveling on Sunday and returning.

[0250] Step 2404: Call the planning submodule to execute the first time period planning task and obtain at least one first time period planning result; call the planning submodule to execute the second time period planning task and obtain at least one second time period planning result.

[0251] In this embodiment, the target model can call the planning submodule to handle the planned tasks for the first time period and the planned tasks for the second time period, respectively. That is, the target model can call a single logic module to handle multiple sub-problems.

[0252] Step 2405: Map at least one first time period plan result to the first time period template to obtain the first time period plan display area; map at least one second time period plan result to the second time period template to obtain the second time period plan display area.

[0253] In one implementation, all the obtained first-period plan results can be mapped onto the first-period template to obtain the first-period plan display area. Similarly, all the obtained second-period plan results can be mapped onto the first-period template to obtain the second-period plan display area.

[0254] In another implementation, one of the at least one first time period plan results can be selected as the target first time period plan result, and the target first time period plan result can be mapped to the first time period template to obtain the first time period plan display area. Similarly, one of the at least one second time period plan results can be selected as the target second time period plan result, and the target second time period plan result can be mapped to the second time period template to obtain the second time period plan display area.

[0255] Please refer to Figure 25 , Figure 25 This is a schematic diagram of a visual interface provided in an embodiment of this application. Figure 5 The visualization interface includes a first-time period display area and a second-time period display area. The first-time period display area shows the Saturday travel plan generated by the target model, and the second-time period display area shows the Sunday travel plan.

[0256] In this embodiment, different parts of the target answer are displayed in different display areas, allowing users to intuitively and selectively view the content of different display areas, thereby improving user browsing efficiency.

[0257] It should be noted that although the steps of the method in this application are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps; or steps from different embodiments may be combined into a new technical solution.

[0258] In another embodiment of this application, a generating apparatus is provided; please refer to [reference needed]. Figure 26 , Figure 26 This is a logic block diagram of a generation device provided in an embodiment of this application. The generation device 2600 may include: a receiving module 2601 and a generation module 2602. Wherein: Receiver module 2601 receives the target content; The generation module 2602 is used to generate a target answer using a target model in response to the target content input. The target content includes the question content, and the target answer includes first content and second content. The first content includes the first part of the answer to the question content in the target answer, and the second content includes the second part of the answer to the question content in the target answer. A first display area with a visual border and a second display area with a visual border are generated using the target model. The first content is displayed in the first display area and the second content is displayed in the second display area.

[0259] Each module in the aforementioned generating device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the electronic device, or stored in the memory of the electronic device in software form, so that the processor can call and execute the operations corresponding to each module.

[0260] In another embodiment of this application, an electronic device is also provided, which includes at least one processor and a target model running on the processor, wherein the processor invokes the target model to execute the steps of the generation method provided in the embodiments of this application.

[0261] In another embodiment of this application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the generation method described in the foregoing embodiments.

[0262] In another embodiment of this application, a computer program product is also provided, including a computer program or instructions that, when executed by a processor, implement the steps of the generation method as described in the foregoing embodiments.

[0263] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, devices, or computer program products. Therefore, this application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage) containing computer-usable program code.

[0264] It should be noted that, in this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0265] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0266] The features disclosed in the several product embodiments provided in this application can be arbitrarily combined to obtain new product embodiments without conflict. Similarly, the features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined to obtain new method or device embodiments without conflict. The above descriptions are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A generation method, comprising: In response to the input of target content, which includes question content; A target answer is generated using a target model. The target answer includes first content and second content. The first content includes a first part of the target answer's response to the question content, and the second content includes a second part of the target answer's response to the question content. The target model is used to generate a first display area with a visual border and a second display area with a visual border. The first content is displayed in the first display area and the second content is displayed in the second display area.

2. The method according to claim 1, wherein both the first part of the answer content and the second part of the answer content are answers to the question content, and the combined content of the first part of the answer content and the second part of the answer content represents a complete answer to the question content.

3. The method according to claim 1, comprising: The question content is divided into a first question content and a second question content. The first part of the answer content is the answer to the first question content, and the second part of the answer content is the answer to the second question content. The first display area is generated based on the content of the first question, and the second display area is generated based on the content of the second question.

4. The method according to claim 3, comprising: Determine the task type of the problem content; The target task template corresponding to the task type is invoked. Different task types correspond to different task templates, and different task templates are presented in different ways. The target task template includes multiple display area sub-templates. The number of the multiple display area sub-templates is the same as the number of the first question content and the second question content. The first part of the answer content is mapped to the display area sub-template corresponding to the first question content to obtain the first display area, and the second part of the answer content is mapped to the display area sub-template corresponding to the second question content to obtain the second display area.

5. The method according to claim 4, comprising: Receive user-defined instructions, wherein the user-defined instructions include template information of a custom task type and a custom task template corresponding to the custom task type; In response to the user-defined instruction, the custom task type and the custom task template are generated, and the custom task template includes multiple custom display area sub-templates.

6. The method according to claim 5, comprising: If the task type of the problem content is the custom task type, the custom task template corresponding to the custom task type will be determined as the target task template corresponding to the task type of the problem content; The target task template includes multiple custom display area sub-templates; The problem content is broken down into sub-tasks corresponding to multiple custom display area sub-templates in the target task template; For each subtask, the execution result of the subtask is mapped to the custom display area sub-template corresponding to the subtask.

7. The method of claim 4, further comprising one of the following: The task type of the problem content is song generation, and the target task template is a song task template, which includes at least the following: Lyrics templates, image templates, and audio templates; The problem content is broken down into lyrics task, image task, and audio task; The first submodule is invoked to execute the lyrics task, resulting in at least one set of lyrics. The second submodule is invoked to execute the image task, resulting in at least one image content. The third submodule is invoked to execute the audio task, resulting in at least one audio content; the first submodule, the second submodule, and the third submodule are logic modules of different types; the at least one lyric content is mapped to the lyric template to obtain a lyric display area; the at least one image content is mapped to the image template to obtain an image display area; Map the at least one audio content to the audio template to obtain an audio display area; The task type of the problem content is a plan generation type, and the target task template is a plan task template. The plan task template includes at least a first time period template and a second time period template. The problem content is split into a first time period plan task and a second time period plan task. The plan submodule is called to execute the first time period plan task to obtain at least one first time period plan result. The plan submodule is called to execute the second time period plan task to obtain at least one second time period plan result. The at least one first time period plan result is mapped to the first time period template to obtain a first time period plan display area. The at least one second time period plan result is mapped to the second time period template to obtain a second time period plan display area.

8. The method according to claim 3, wherein the first part of the answer content includes a plurality of first candidate contents, the first display area includes interactive controls, and in response to an operation on the interactive controls of the first display area, at least one of the following operations is performed: The plurality of first candidate contents are displayed in the first display area; The target candidate content is displayed in the first display area, and the target candidate content is one of the selected first candidate contents; The first question content corresponding to the first display area is regenerated, and the first part of the answer content displayed in the first display area is updated based on the regeneration result. The first part of the answer content displayed in the first display area differs before and after the update, while the content in the second display area remains unchanged; The content of the multiple first candidate contents is added, deleted, or modified.

9. The method according to any one of claims 1-8, comprising: Based on the first part of the answer content, first target information is obtained from the first part of the answer content. The first target information includes at least one of the following: content object, content address, service information, order parameters, and payment port. First access port data is generated based on the first target information and displayed in the first display area; wherein, the first access port data is used to interact with the target device, enabling the target device to obtain the first target information; Based on the second part of the answer content, second target information is obtained from the second part of the answer content. The second target information includes at least one of the following: content object, content address, service information, order parameters, and payment port. The second access port data is generated based on the second target information and displayed in the second display area; wherein, the second access port data is used to interact with the target device, so that the target device can obtain the second target information.

10. The method according to any one of claims 1-8, comprising: In response to the fusion command, a fusion display area template is generated. The fusion display area template includes a first module and a second module. The first module corresponds to the first display area, and the second module corresponds to the second display area. The positions and area ratios of the first module and the second module within the fusion display area template are determined based on the logical relationship and weight information between the first display area and the second display area. The first content displayed in the first display area is mapped to the first module, and the second content displayed in the second display area is mapped to the second module, and the resulting fused display area is displayed in the visualization interface.

Images