Dynamic content generation method and device

Abstract

The invention provides a content dynamic generation method and device. The method comprises the steps of obtaining interaction data generated by multiple rounds of interaction between a user and an intelligent model; target cue words are determined according to the interaction data, and the target cue words comprise a target scene, a target theme, a target style and a target preset template; according to the target prompt word, a material file in a material library is called through an intelligent model to dynamically generate target content, and the target content meets the following conditions that the target content is applied to a target scene, used for expressing a target theme, matched with a target style and used for using a target preset template; wherein in the process of dynamically generating the target content by the intelligent model, the style of the target content and the material file used by the target content are adjusted in real time according to the feedback of the user. By adopting the technical means, the problems of low user satisfaction and low efficiency of the content generated by the model in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of content generation technology, and in particular, to a method and device for dynamically generating content. Background Art

[0002] Model-Generated Content (MGC) refers to the use of artificial intelligence and machine learning technologies, especially natural language processing (NLP) models, to automatically generate content such as text, images, audio, or video. This technology has been widely used in many fields, and with the continuous progress of technology, its application scope and effect are also constantly improving. For example, using models to generate news articles, promotional posters, review articles, etc. In current applications of using models to generate content, the user instructions are directly used as prompt words or the results of simple processing of the user instructions are used as prompt words, which results in low user satisfaction with the generated content. In addition, users cannot adjust the content in real time during the process of model-generated content. If the user is already dissatisfied during the process of model-generated content, they still need to wait until the model finishes generating the content, then modify the user instructions, and then the model regenerates the content, which results in low efficiency of generating content. Summary of the Invention

[0003] In view of this, embodiments of this application provide a method, device, electronic device, and computer-readable storage medium for dynamically generating content to solve the problems of low user satisfaction and low efficiency of the content generated by the model in the prior art.

[0004] In a first aspect of the embodiments of this application, a method for dynamically generating content is provided, including: obtaining interaction data generated by a user's multi-round interaction with an intelligent model; determining a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template; generating target content dynamically by using the intelligent model to call material files in a material library according to the target prompt word, where the target content meets the following conditions: applied to the target scenario, used to express the target theme, conform to the target style, and use the target preset template; where, during the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to the user's feedback.

[0005] In the second aspect of the embodiments of the present application, a content dynamic generation device is provided, including: an acquisition module configured to acquire interaction data generated by a user's multi-round interaction with an intelligent model; a determination module configured to determine a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template; a generation module configured to dynamically generate target content by using the intelligent model to call material files in a material library according to the target prompt word, where the target content meets the following conditions: being applied to the target scenario, being used to express the target theme, fitting the target style, and using the target preset template; the generation module is further configured to, during the process of the intelligent model dynamically generating the target content, adjust the style of the target content and the material files used by the target content in real time according to the user's feedback.

[0006] In the third aspect of the embodiments of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above method are implemented.

[0007] In the fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0008] The beneficial effects of the embodiments of the present application compared with the prior art are as follows: acquiring interaction data generated by a user's multi-round interaction with an intelligent model; determining a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template; dynamically generating target content by using the intelligent model to call material files in a material library according to the target prompt word, where the target content meets the following conditions: being applied to the target scenario, being used to express the target theme, fitting the target style, and using the target preset template; among them, during the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to the user's feedback. By adopting the above technical means, the problems of low user satisfaction and low efficiency in the content generated by the model in the prior art can be solved, thereby improving user satisfaction and generation efficiency. Description of the Drawings

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0010] Figure 1It is a schematic flowchart of a method for dynamically generating content provided by an embodiment of the present application;

[0011] Figure 2 It is a schematic flowchart of another method for dynamically generating content provided by an embodiment of the present application;

[0012] Figure 3 It is a schematic structural diagram of a device for dynamically generating content provided by an embodiment of the present application;

[0013] Figure 4 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0014] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0015] A method and device for dynamically generating content according to an embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0016] Figure 1 It is a schematic flowchart of a method for dynamically generating content provided by an embodiment of the present application. Figure 1 The method for dynamically generating content can be executed by a computer or a server, or software on a computer or a server. As Figure 1 shown, the method for dynamically generating content includes:

[0017] S101, obtaining interaction data generated by a user's multi-round interaction with an intelligent model;

[0018] S102, determining a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template;

[0019] S103, according to the target prompt word, using the intelligent model to dynamically generate target content by calling material files in a material library, where the target content meets the following conditions: applied to the target scenario, used to express the target theme, conforms to the target style, and uses the target preset template;

[0020] S104, during the process of the intelligent model dynamically generating the target content, adjusting the style of the target content and the material files used by the target content in real time according to the user's feedback.

[0021] Through multiple rounds of interaction, select the scenario databases required for the scenario from the underlying database, including scenario databases of types such as events, documents, and entities. And generate an association relationship table through an association relationship construction tool. During the process of the intelligent model dynamically generating the target content, automatically generate the code for the execution plan or call relevant tools, and feedback the sample results of the execution to the user in an interactive manner. Intermediate results are retained between the above processing steps. After the solution editing and generation are completed, the corresponding executable process is also arranged. After the solution editing and generation are completed, the application function corresponding to the solution can be published to solve the scenario problem.

[0022] The intelligent model can be any commonly used large language model for generating the target content required by the user. Interaction data is collected during the multiple rounds of interaction between the user and the intelligent model. Determine the target prompt words from the interaction data, where the target prompt words include: target scenario, target theme, target style, and target preset template. The material library stores a large number of materials such as texts, images, and videos. According to the target prompt words, use the intelligent model to call the material files in the material library to dynamically generate the target content. The generated target content meets the following conditions: applied to the target scenario, used to express the target theme, in line with the target style, and uses the target preset template. During the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to the user's feedback. By this method, if the user is already dissatisfied during the process of the model generating content, there is no need to wait until the model finishes generating the content and then the model regenerates the content.

[0023] According to the technical solution provided by the embodiment of the present application, obtain the interaction data generated by the multiple rounds of interaction between the user and the intelligent model; determine the target prompt words according to the interaction data, where the target prompt words include: target scenario, target theme, target style, and target preset template; according to the target prompt words, use the intelligent model to call the material files in the material library to dynamically generate the target content, where the target content meets the following conditions: applied to the target scenario, used to express the target theme, in line with the target style, and uses the target preset template; among them, during the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to the user's feedback. By adopting the above technical means, the problems of low user satisfaction and low efficiency in the content generated by the model in the prior art can be solved, thereby improving user satisfaction and generation efficiency.

[0024] Furthermore, during the process of the intelligent model dynamically generating the target content, the style of the target content is adjusted in real time according to the user's feedback, including: determining multiple candidate styles whose similarity to the target style is greater than a threshold from multiple styles of the content generated by the intelligent model; using the intelligent model to call the material files in the material library to generate the content of each candidate style, and displaying the content of each candidate style to the user through a pop-up window to receive feedback; determining the target content from the content of each candidate style according to the feedback.

[0025] The intelligent model can generate content in multiple styles. By calculating the cosine similarity between the descriptions of various styles and the description of the target style, multiple candidate styles whose similarity to the target style is greater than the threshold are determined from them. Then, the intelligent model is used to call the material files in the material library to generate the content of each candidate style, and at the same time, the content of each candidate style is displayed to the user through a pop-up window so that after the user sees the content of each candidate style, they can give or input feedback. The feedback is used to indicate which candidate style of content the user prefers more. Then, the target content can be determined from the content of each candidate style according to the feedback.

[0026] Furthermore, after determining multiple candidate styles whose similarity to the target style is greater than the threshold from multiple styles of the content generated by the intelligent model, the method further includes: using the intelligent model to call the material files in the material library to generate the content of the preset size of each candidate style, and displaying the content of the preset size of each candidate style to the user through a pop-up window to receive feedback; determining the target style from each candidate style according to the feedback; using the intelligent model to call the material files in the material library to generate the target content based on the content of the preset size of the target style.

[0027] To improve efficiency, it is not necessary to wait for the intelligent model to generate all the content of each candidate style and then determine the target content from it. The intelligent model can first be used to generate the content of the preset size (partial content) of each candidate style, and the content of the preset size of each candidate style is displayed to the user through a pop-up window. So that after the user sees the content of each candidate style, they can give or input feedback. The feedback is used to indicate which candidate style of content the user prefers more. The target style is determined from each candidate style according to the feedback, and then the intelligent model is used to call the material files in the material library to complete the generation of the subsequent content based on the content of the preset size of the target style, and thus the target content is obtained.

[0028] Furthermore, determining the target style from each candidate style according to the feedback includes: when the user inputs multiple feedbacks multiple times, dynamically switching the target style among each candidate style according to the multiple feedbacks.

[0029] For example, the user inputs the first feedback, the second feedback, and the third feedback in sequence. The first feedback, the second feedback, and the third feedback respectively indicate that the user hopes to generate content in the first candidate style, the second candidate style, and the third candidate style. Then, for the first time, the first candidate style is used as the target style, and the intelligent model generates content in the first candidate style. During this process (before the content generation is completed), the user gives the second feedback, indicating that they don't like the content in the first candidate style. Then the intelligent model generates content in the second candidate style. During this process (before the content generation is completed), the user gives the third feedback, indicating that they don't like the content in the second candidate style. Then the intelligent model generates content in the third candidate style.

[0030] Furthermore, during the process of the intelligent model dynamically generating the target content, the material files used for the target content are adjusted in real time according to the user's feedback, including: during the process of using the intelligent model to dynamically generate the target content, the correlation between the target content and the material files it uses is generated simultaneously; a visualization graph corresponding to the correlation between the target content and the material files it uses is generated, and the visualization graph is shown to the user through a pop-up window to receive feedback; the material files used for the target content are adjusted according to the feedback.

[0031] The correlation between the target content and the material files it uses includes the position of the material files in the target content, the proportion of the total content they occupy, the citation method, etc. The visualization graph is used to represent the correlation and is shown to the user through a pop-up window so that after seeing the visualization graph, the user can give or input feedback. The feedback is used to indicate adjusting the position of the material files in the target content, the proportion of the total content they occupy, the citation method, and replacing the material files, etc.

[0032] Furthermore, the target prompt word is determined according to the interaction data, including: determining the scene feature word, the theme feature word, the style feature word, and the template feature word from the interaction data; combining the scene feature word, the theme feature word, the style feature word, and the template feature word in a preset format to obtain the target prompt word.

[0033] The scene feature word is a feature word in the interaction data that can indicate the application scenario, such as news, promotional posters, etc. The theme feature word is a feature word in the interaction data that can indicate the theme of the target content, such as a certain accident, the promotion of a certain product, etc. The style feature word is a feature word in the interaction data that can indicate the style of the target content, such as simple and elegant, concise and clear, etc. The template feature word is a feature word in the interaction data that can indicate the template used for the target content. The preset format is the format or template of the target prompt word set in advance, such as combining the scene feature word, the theme feature word, the style feature word, and the template feature word in sequence to obtain the target prompt word.

[0034] Further, determining a target prompt word according to the interaction data includes: determining multiple scenario feature words, multiple theme feature words, multiple style feature words, and multiple template feature words from the interaction data; determining the view feature words corresponding to each scenario feature word, the view feature words corresponding to each theme feature word, the view feature words corresponding to each style feature word, and the view feature words corresponding to each template feature word from the interaction data; determining a target scenario feature word from each scenario feature word based on the view feature words corresponding to each scenario feature word; determining a target theme feature word from each theme feature word based on the view feature words corresponding to each theme feature word; determining a target style feature word from each style feature word based on the view feature words corresponding to each style feature word; determining a target template feature word from each template feature word based on the view feature words corresponding to each template feature word; combining the target scenario feature word, the target theme feature word, the target style feature word, and the target template feature word in a preset format to obtain the target prompt word.

[0035] Taking multiple scenario feature words as an example, there are multiple scenario feature words in the interaction data. Then which one is used as the target scenario feature word can be determined by using the view feature words corresponding to each scenario feature word. For example, the intelligent model sends "Do you want to generate a news release?" and the user sends "No", then "news" is excluded. "No" is the view feature word corresponding to "news". For example, the intelligent model sends "Do you want to generate a promotional poster?" and the user sends "Yes", then "promotional poster" is the target scenario feature word. "Yes" is the view feature word corresponding to "news".

[0036] Figure 2 It is a flowchart of another content dynamic generation method provided by an embodiment of the present application. As Figure 2 shown, the method includes:

[0037] S201, obtaining interaction data generated by a user's multi-round interaction with an intelligent model;

[0038] S202, determining a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template;

[0039] S203, according to the target prompt word, using the intelligent model to dynamically generate target content by calling material files in a material library, where the target content meets the following conditions: applied to the target scenario, used to express the target theme, conform to the target style, and use the target preset template;

[0040] S204, using the intelligent model to determine multiple candidate styles of the target style;

[0041] S205. Determine multiple candidate styles from multiple styles of content generated by the intelligent model, where the similarity between each candidate style and the target style is greater than a threshold value.

[0042] S206. Use the intelligent model to call the material files in the material library to generate content of a preset size for each candidate style, and display the content of the preset size for each candidate style to the user through a pop-up window to receive feedback.

[0043] S207. Determine the target style from each candidate style according to the feedback.

[0044] S208. Use the intelligent model to call the material files in the material library to generate the target content based on the content of the preset size in the target style.

[0045] S209. When the user inputs multiple feedbacks multiple times, dynamically switch the target style among each candidate style according to the multiple feedbacks.

[0046] S210. During the process of dynamically generating the target content by the intelligent model, simultaneously generate the association relationship between the target content and the material files used for it by the intelligent model.

[0047] S211. Generate a visualization graph corresponding to the association relationship between the target content and the material files used for it, and display the visualization graph to the user through a pop-up window to receive feedback.

[0048] S212. Adjust the material files used for the target content according to the feedback.

[0049] All of the above optional technical solutions can be combined arbitrarily to form optional embodiments of the present application, which will not be elaborated here one by one.

[0050] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the present application.

[0051] Figure 3 is a schematic diagram of a content dynamic generation device provided by an embodiment of the present application. As Figure 3 shown, the content dynamic generation device includes:

[0052] An acquisition module 301, configured to acquire interaction data generated by a user's multiple rounds of interaction with the intelligent model;

[0053] A determination module 302, configured to determine a target prompt word according to the interaction data, where the target prompt word includes: a target scene, a target theme, a target style, and a target preset template;

[0054] A generation module 303, configured to dynamically generate target content by using an intelligent model to call material files in a material library according to a target prompt word, where the target content meets the following conditions: being applied to a target scenario, being used to express a target theme, conforming to a target style, and using a target preset template;

[0055] The generation module 303 is further configured to, during the process of the intelligent model dynamically generating the target content, adjust the style of the target content and the material files used by the target content in real time according to the user's feedback.

[0056] According to the technical solution provided by the embodiment of the present application, interaction data generated by a user's multi-round interaction with an intelligent model is obtained; a target prompt word is determined according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template; according to the target prompt word, an intelligent model is used to call material files in a material library to dynamically generate target content, where the target content meets the following conditions: being applied to a target scenario, being used to express a target theme, conforming to a target style, and using a target preset template; among them, during the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to the user's feedback. By adopting the above technical means, the problems of low user satisfaction and low efficiency in the content generated by the existing model can be solved, and thus the user satisfaction and generation efficiency can be improved.

[0057] In some embodiments, the generation module 303 is further configured to determine multiple candidate styles from multiple styles of the content generated by the intelligent model, where the similarity of each candidate style to the target style is greater than a threshold; use the intelligent model to call material files in the material library to generate content of each candidate style, and display the content of each candidate style to the user through a pop-up window to receive feedback; determine the target content from the content of each candidate style according to the feedback.

[0058] In some embodiments, the generation module 303 is further configured to use the intelligent model to call material files in the material library to generate content of a preset size for each candidate style, and display the content of the preset size of each candidate style to the user through a pop-up window to receive feedback; determine the target style from each candidate style according to the feedback; use the intelligent model to call material files in the material library to generate the target content based on the content of the preset size of the target style.

[0059] In some embodiments, the generation module 303 is further configured to, when the user inputs multiple feedbacks multiple times, implement dynamic switching of the target style among each candidate style according to the multiple feedbacks.

[0060] For example, the user inputs the first feedback, the second feedback, and the third feedback in sequence. The first feedback, the second feedback, and the third feedback respectively indicate that the user hopes to generate content in the first candidate style, the second candidate style, and the third candidate style. Then, for the first time, the first candidate style is used as the target style, and the intelligent model generates content in the first candidate style. During this process (before the content generation is completed), the user gives the second feedback, indicating that they don't like the content in the first candidate style. Then the intelligent model generates content in the second candidate style. During this process (before the content generation is completed), the user gives the third feedback, indicating that they don't like the content in the second candidate style. Then the intelligent model generates content in the third candidate style.

[0061] In some embodiments, the generation module 303 is further configured to, during the process of dynamically generating the target content using the intelligent model, simultaneously generate the association relationship between the target content and the material files it uses; generate a visualization graph corresponding to the association relationship between the target content and the material files it uses, and display the visualization graph to the user through a pop-up window to receive feedback; adjust the material files used for the target content according to the feedback.

[0062] In some embodiments, the determination module 302 is further configured to determine scene feature words, theme feature words, style feature words, and template feature words from the interaction data; combine the scene feature words, theme feature words, style feature words, and template feature words in a preset format to obtain the target prompt word.

[0063] In some embodiments, the determination module 302 is further configured to determine multiple scene feature words, multiple theme feature words, multiple style feature words, and multiple template feature words from the interaction data; determine the view feature words corresponding to each scene feature word, the view feature words corresponding to each theme feature word, the view feature words corresponding to each style feature word, and the view feature words corresponding to each template feature word from the interaction data; based on the view feature words corresponding to each scene feature word, determine the target scene feature word from each scene feature word; based on the view feature words corresponding to each theme feature word, determine the target theme feature word from each theme feature word; based on the view feature words corresponding to each style feature word, determine the target style feature word from each style feature word; based on the view feature words corresponding to each template feature word, determine the target template feature word from each template feature word; combine the target scene feature word, the target theme feature word, the target style feature word, and the target template feature word in a preset format to obtain the target prompt word.

[0064] In some embodiments, the generation module 303 is further configured to obtain the interaction data generated by the user's multi-round interaction with the intelligent model; determine a target prompt word according to the interaction data, where the target prompt word includes: a target scenario, a target theme, a target style, and a target preset template; according to the target prompt word, use the intelligent model to dynamically generate target content by calling the material files in the material library, where the target content meets the following conditions: applied to the target scenario, used to express the target theme, conforms to the target style, and uses the target preset template; use the intelligent model to determine multiple candidate styles of the target style; determine multiple candidate styles with a similarity to the target style greater than a threshold from the multiple styles of the content generated by the intelligent model; use the intelligent model to call the material files in the material library to generate content of a preset size for each candidate style, and display the content of the preset size for each candidate style to the user through a pop-up window to receive feedback; determine the target style from each candidate style according to the feedback; use the intelligent model to call the material files in the material library to generate the target content based on the content of the preset size in the target style; when the user inputs multiple feedbacks multiple times, dynamically switch the target style among each candidate style according to the multiple feedbacks; during the process of dynamically generating the target content by using the intelligent model, simultaneously generate the association relationship between the target content and the material files used by it; generate a visualization graph corresponding to the association relationship between the target content and the material files used by it, and display the visualization graph to the user through a pop-up window to receive feedback; adjust the material files used by the target content according to the feedback.

[0065] It should be understood that the sequence numbers of the steps in the above embodiments do not represent the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0066] Figure 4 is a schematic diagram of the electronic device 4 provided by the embodiments of the present application. As Figure 4 shown, the electronic device 4 in this embodiment includes: a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 403, it implements the steps in the above various method embodiments. Alternatively, when the processor 401 executes the computer program 403, it implements the functions of each module / unit in the above various device embodiments.

[0067] The electronic device 4 may be a desktop computer, a notebook, a palm computer, a cloud server, and other electronic devices. The electronic device 4 may include, but is not limited to, the processor 401 and the memory 402. Those skilled in the art can understand that Figure 4 is only an example of the electronic device 4, and does not constitute a limitation to the electronic device 4. It may include more or fewer components than shown in the figure, or different components.

[0068] The processor 401 may be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0069] The memory 402 may be an internal storage unit of the electronic device 4. For example, the hard disk or memory of the electronic device 4. The memory 402 may also be an external storage device of the electronic device 4. For example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the electronic device 4. The memory 402 may also include both an internal storage unit and an external storage device of the electronic device 4. The memory 402 is used to store computer programs and other programs and data required by the electronic device.

[0070] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0071] When the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-described embodiment methods of this application, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. The computer program can include computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0072] The above embodiments are only used to illustrate the technical solutions of this application, rather than to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A content dynamic generation method, characterized in that: include: Obtain interaction data generated by multiple rounds of interaction between users and intelligent models; Determining a target prompt word according to the interaction data, wherein the target prompt word includes: a target scene, a target theme, a target style, and a target preset template; According to the target prompt word, the intelligent model is used to call the material files in the material library to dynamically generate the target content, wherein the target content meets the following conditions: being applied to the target scene, being used to express the target theme, being in line with the target style, and using the target preset template; In the process of the intelligent model dynamically generating the target content, the style of the target content and the material files used by the target content are adjusted in real time according to user feedback.

2. The method according to claim 1, characterized in that In the process of the intelligent model dynamically generating the target content, adjusting the style of the target content in real time according to user feedback includes: Determine, from the multiple styles of the content generated by the intelligent model, multiple candidate styles whose similarity to the target style is greater than a threshold; Using the intelligent model to call the material files in the material library to generate content of each candidate style, and displaying the content of each candidate style to the user through a pop-up window to receive the feedback; The target content is determined from the contents of each candidate style according to the feedback.

3. The method according to claim 2, characterized in that After determining a plurality of candidate styles having a similarity with the target style greater than a threshold from the plurality of styles of the content generated by the intelligent model, the method further includes: Using the intelligent model to call the material files in the material library to generate content of preset sizes for each candidate style, and displaying the content of preset sizes for each candidate style to the user through a pop-up window to receive the feedback; Determining the target style from each candidate style according to the feedback; The intelligent model is used to call the material files in the material library to generate the target content based on the content of the preset size of the target style.

4. The method according to claim 3, characterized in that Determining the target style from each candidate style according to the feedback includes: When the user inputs a variety of the feedbacks for multiple times, the target style is dynamically switched among the candidate styles according to the multiple feedbacks.

5. The method according to claim 1, characterized in that In the process of dynamically generating the target content by the intelligent model, adjusting the material files used by the target content in real time according to user feedback includes: In the process of dynamically generating the target content using the intelligent model, the intelligent model is also used to generate an association relationship between the target content and the source files used by the target content; Generate a visualization diagram corresponding to the association relationship between the target content and the material files used by the target content, and display the visualization diagram to the user through a pop-up window to receive the feedback; The material files used by the target content are adjusted according to the feedback.

6. The method according to claim 1, characterized in that Determining a target prompt word according to the interaction data includes: Determining scene feature words, theme feature words, style feature words and template feature words from the interaction data; The scene feature words, the theme feature words, the style feature words and the template feature words are combined according to a preset format to obtain the target prompt word.

7. The method according to claim 1, characterized in that Determining a target prompt word according to the interaction data includes: Determining a plurality of scene feature words, a plurality of theme feature words, a plurality of style feature words, and a plurality of template feature words from the interaction data; Determine from the interaction data the viewpoint feature word corresponding to each scene feature word, the viewpoint feature word corresponding to each theme feature word, the viewpoint feature word corresponding to each style feature word, and the viewpoint feature word corresponding to each template feature word; Based on the viewpoint feature words corresponding to each scene feature word, a target scene feature word is determined from each scene feature word; Based on the viewpoint feature words corresponding to each topic feature word, a target topic feature word is determined from each topic feature word; Based on the viewpoint feature words corresponding to each style feature word, a target style feature word is determined from each style feature word; Based on the viewpoint feature words corresponding to each template feature word, a target template feature word is determined from each template feature word; The target scene feature word, the target theme feature word, the target style feature word and the target template feature word are combined according to a preset format to obtain the target prompt word.

8. A content dynamic generation device, characterized in that: include: An acquisition module, configured to acquire interaction data generated by multiple rounds of interaction between the user and the intelligent model; A determination module is configured to determine a target prompt word according to the interaction data, wherein the target prompt word includes: a target scene, a target theme, a target style and a target preset template; A generation module is configured to dynamically generate target content by calling the material files in the material library based on the target prompt word using the intelligent model, wherein the target content meets the following conditions: being applied to the target scene, being used to express the target theme, being in line with the target style, and using the target preset template; The generation module is also configured to adjust the style of the target content and the material files used by the target content in real time according to user feedback during the process of the intelligent model dynamically generating the target content.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images