Content generation system

Abstract

Provided is a content generation system capable of efficiently generating content that is more easily understood by people using a large language model. Using communicating human language information input through the user interface, the large language model acquires communicating human language information to be included in the content. Meanwhile, the large language model acquires information for selecting visualization software capable of generating visual information contained in the content, using alternating human language information input through the user interface. And selecting visualization software based on the acquired communication human language information. And operating the selected visualization software based on the text information obtained from the large language model output according to the communication human language information input through the user interface. Visual information for being included in the content is acquired, and the content including at least a part of the acquired communication human language information and at least a part of the acquired visual information is generated and output.

Description

TECHNICAL FIELD

[0001] The present application relates to a content generation system. BACKGROUND

[0002] A text generation system using an LLM (Large Language Model) such as GPT-2 and GPT-3 has been proposed (Patent Literature 1). Recently, GPT-4 has been released, and various use cases are being actively discussed (Non-Patent Literature 1).

[0003] PRIOR ART DOCUMENTS PATENT LITERATURE Patent Literature 1: U.S. Patent Application Publication No. 2021 / 192140 NON-PATENT LITERATURE Non-Patent Literature 1: "GPT-4 finally released: explaining how to use GPT-4 and its performance", online, March 15, 2023, ChatGPT Research Institute, retrieved March 21, 2023, Internet <URL: https: / / chatgpt-lab.com / n / n7facbf0f8890> SUMMARY

[0004] Problem to be solved It is desirable to generate content that people can more easily understand in an efficient manner using a large language model GPT-3 or later version.

[0005] An object of the present application is to efficiently generate content that people can more easily understand using a large language model. More specifically, an object is to both promote user understanding of content and improve the efficiency of content generation, thereby reducing the load on hardware resources that constitute a content generation system.

[0006] Solution The present application can provide the following content generation system.

[0007] A content generation system comprising: at least one user interface, at least one memory, at least one processor configured to execute at least one program stored in the memory and connected to the memory, the at least one program being programmed to cause the at least one processor to perform: (A) using each of a large language model that operates using communication human language information input through the user interface and a visualization software that is programmed to output visual information using provided text information, (B) in the (A), acquiring, by the large language model, communication human language information for inclusion in content using at least a part of the communication human language information input through the user interface, (C) in the (A), acquiring, by the large language model, communication human language information for selecting a visualization software capable of generating visual information for inclusion in content using at least a part of the communication human language information input through the user interface, and selecting the visualization software based on the acquired communication human language information, (D) in the (A), acquiring, by operating the visualization software selected in the (C) using text information acquired based on the output of the large language model corresponding to the communication human language information input via the user interface, visual information for inclusion in content, (E) generating content including at least a part of the communication human language information acquired in the (B) or a modification thereof and at least a part of the visual information acquired in the (D) or a modification thereof, and (F) outputting, through the user interface, the content generated in the (E).

[0008] The content generation system efficiently generates content that people can more easily understand using a large language model. In this system, the user does not need to select a visualization software by himself or herself, and the content generation system selects an appropriate visualization software based on the user's needs. This both improves the user's understanding of the content and improves the efficiency of content generation. The content generation system does not assume the use of a specific visualization software. Since the content generation system selects a visualization software based on the user's needs, it is not necessary to configure the content generation system for each visualization software. As a result, the versatility of the system is improved, and the load of the hardware resources that constitute the content generation system can be reduced. In other words, if the same hardware resources are used, more advanced processing can be performed.

[0009] The content generation system includes at least one user interface, at least one memory, and at least one processor. The at least one user interface, the at least one memory, and the at least one processor are connected so that they can communicate with each other. The at least one user interface, the at least one memory, and the at least one processor can be installed on a single device to build a centralized processing system, or they can be distributed over multiple devices to build a distributed processing system. Here, the term "distributed processing system" can be broadly interpreted to include a local system, a cloud system, or a combination of both. If the content generation system is a distributed processing system, a user terminal device can be included in the multiple devices that make up the system. In addition, a program related to the content generation system can be installed on the user terminal device, which can function as the content generation system by executing the program on its at least one processor. If the content generation system includes a user terminal device, the user interface is an input device and an output device provided on the user terminal device. The input device can include a keyboard and a pointing device, and the output device can include a display, a projector, a printer, a speaker, and a headphone.

[0010] The content generation system can accept input and can output content based on the exchange of human language information through the user interface. The content generation system can be based on a large language model. In this case, the content generation system can be multi-modal. In one embodiment, the content generation system has the function of a chatbot that can input and output the exchange of human language information through the user interface. In this embodiment, the chatbot is based on a large language model based on a large language model. The content generation system can communicate the exchange of human language information iteratively through the user interface or with the large language model. The content generation system can maintain context during this iterative exchange of information. In addition, in one embodiment, the content generation system is not integrated into any particular visualization software and is not associated with only a particular visualization software. The visualization software to be selected is not determined before the content generation system begins to be used.

[0011] "Content" includes communicating human language information and visual information. Content is generated based on "communicating human language information input through a user interface." "Communicating human language information input through a user interface" is, for example, "please generate explanatory materials for technical matter X" in the following example, which includes a subject matter of content and is a query in nature. Here, a query includes a command, a request, a question, or a combination thereof to a content generation system (e.g., "please generate... explanatory materials"). A query can also include a command, a request, a question, or a combination thereof regarding a content specification (e.g., quantity, language, layout). A subject matter is, for example, a matter that a user wants to explain or communicate to someone (e.g., "technical matter X"). A subject matter can include, for example, a topic or a subject. "Communicating human language information input through a user interface" does not have to be input all at once; it can be input through multiple rounds of input and output between a user and a content generation system via a user interface. In addition, "communicating human language information input through a user interface" does not include information indicating a visualization software name. The selection of a visualization software is not made by a user, but by a content generation system. However, information indicating a visualization software name can be input into a content generation system.

[0012] Content is generated, for example, according to a query to describe or explain a subject matter. Communicating human language information and visual information in content are related to each other and to the subject matter. Content is generated to facilitate understanding of the subject matter and to describe or explain the subject matter in more detail. As one example, communicating human language information in content has more text than the communicating human language information input through a user interface. Visual information is added, for example, to supplement the explanation or description given by the communicating human language information.

[0013] Examples of "content" can include files, presentation materials, and videos. Regarding files, examples are technical specification files, intellectual property related files submitted to administrative agencies or courts, intellectual property assessment files, and intellectual property search files. A content generation system can generate these files or drafts thereof. Examples of intellectual property related files include patent specification, examination opinion reply files, and files for trials or litigation. Intellectual property assessment files can relate to infringement or non-infringement, or to the validity of rights. Intellectual property search files can relate to, for example, infringement or non-infringement, prior art search, or trend investigation. Note that things in which visual information is the main output and communicating human language information is included as an additional output do not belong to "content" as referred to here. For example, movies, television programs, video game guides, and sports videos do not qualify as "content" in this context.

[0014] Communicating human language information refers to language information that people can understand, recognize, and memorize. Communicating human language information is based on the language system used by people in their daily conversations. Communicating human language information can be conveyed as textual information or audio information. Programming languages do not belong to communicating human language information. The term "programming language" here includes not only low-level languages (machine languages and assembly languages) but also high-level languages (interpreted languages and compiled languages). The following types of information can utilize communicating human language information, and can include programming languages. In this context, programming languages are not used for execution by a processor, but for presenting information to people. In one embodiment, the following types of information do not include programming languages that are conveyed for execution by a processor: - information input into the content generation system through a user interface; - information input into a large language model to operate it; or - information included in the content together with visual information.

[0015] Visual information can be an image or a video. An image is static visual information, while a video is dynamic. An image can be a visual chart or a non-visual chart. A visual chart is an image derived from data or information. Data can be the raw material of information. Examples of visual charts include graphs, figures, charts, diagrams, drawings, histograms, tables, and matrices. For graphs, examples can be given such as contour maps, topographic maps, vector maps, equipotential maps, mechanical drawings, design drawings, and patent drawings. A non-visual chart can be, for example, a photograph actually taken. A visual chart can also be generated based on one or more non-visual charts. A video can be an animation or a simulation, or it can be a live capture. At least a part of the visual information in the content can be generated by visualization software, or it can be generated by visualization software and then modified by the content generation system. Visual information generated by visualization software in the content generation system does not belong to visual information obtained by searching through a network. An animation or a simulation can also be generated based on one or more live captures. An animation or a simulation can contain one or more non-visual contents and / or visual contents.

[0016] A user interface is a device and / or a device included in the content generation system that facilitates the mutual exchange of information between the system and the user. If the content generation system is configured not to include a user terminal device but to be able to communicate, the user interface can be, for example, a communication module that can communicate with a user terminal device. If the content generation system includes a user terminal device, the user interface can include the input and output devices of the user terminal device.

[0017] Processors include central processing units, microprocessors, general-purpose processors, digital signal processors, graphics processors, controllers, microcontrollers, programmable logic devices, field-programmable gate arrays, and application-specific integrated circuits. Multiple processors can be configured in any combination of these. The execution of a program by at least one processor is not necessarily limited to execution by a single processor; it can also be parallel processing in a multi-processor configuration. In a multi-processor, symmetric or asymmetric multiprocessing can be employed. The multi-processor can be tightly or loosely coupled. The processor can also be a multi-core processor.

[0018] Memory includes random access memory, read-only memory, non-volatile random access memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, flash memory, magnetic data storage, optical data storage, registers, or any combination of these. Memory can be single or multiple.

[0019] Large language models refer to language models having a large number of parameters. Large language models are capable of performing natural language processing tasks. Large language models can also be natural language generation models based on large language models, capable of generating sentences from inputted human language information. Generating new sentences from inputted human language information is an example of a natural language processing task.

[0020] The number of parameters in a large language model can be, for example, more than 1 billion, more than 10 billion, or even more than 100 billion. A language model is something that models human language using the probability of language occurrence. In one embodiment, a large language model can perform inference without fine-tuning, using methods such as zero-shot learning, one-shot learning, or few-shot learning. In one embodiment, a large language model is configured to perform a task and produce an output based on an input prompt containing human language information. Examples of large language models include, but are not limited to, GPT-3, GPT-4, GShard, SwitchTransformer, Gopher, and HyperCLOVA. A large language model can or can not be included in a content generation system and can communicate with a content generation system.

[0021] The visualization software is not particularly limited and can be, for example, data visualization software or can include a visual base model. If multiple visual elements are included in a single content, each visual element can be generated by a different visualization software. In this case, the processor can select a visualization software for each visual element.

[0022] Process (A) includes processes (B) to (D). In the following embodiments, processes (B), (C), and (D) are executed in this order. Furthermore, in each of processes (B) to (D), the communicative human language information is provided from the content generation system to the large language model. However, the order of processes (B) to (D) is not particularly limited. Processes (B) to (D) do not need to be strictly distinguished from each other in time or content. Process (A) (i.e., processes (B) to (D)) and process (E) are completed before the content is output in process (F). If each of processes (B) to (D) includes a plurality of sub-processes, the sub-processes related to (B) to (D) can be executed interchangeably. The provision of the communicative human language information from the content generation system to the large language model can be common in all or any two of processes (B) to (D). In any case, the result of a previously executed process can be used in a subsequently executed process.

[0023] In process (B), a part or all of the communicative human language information input through the user interface can be used. The communicative human language information input through the user interface can be the same as or different from the communicative human language information provided to the large language model. In one embodiment, the information provided to the large language model in process (B) mainly includes the communicative human language information input through the user interface.

[0024] Based on the subject matter contained in the communicative human language information input through the user interface, tasks can be set, and prompts based on these tasks can be provided to the large language model. In the case where the tasks are predefined to elaborate or describe the input matter in more detail, the prompts can be provided to the large language model based on the subject matter contained in the input and the predefined tasks. In this context, the prompts can be interpreted as the input matter, and the communicative human language information for the content can be, for example, a detailed explanation of the input matter. Process (B) can be iteratively executed a plurality of times. In this case, first one or more matters (e.g., words, phrases, expressions, or sentences) are extracted from the explanation of the subject matter acquired from the large language model, and then the extracted matters can be provided to the large language model. In addition, if an explanation about these matters is acquired from the large language model, more specific sub-concepts related to the matters can be extracted from the explanation and provided to the large language model. This makes it possible to generate a more detailed explanation about the technical matter. The content generated by this process can contain a detailed explanation or description about the subject matter, and a deeper and more detailed explanation or description about the matters contained in the explanation or description.

[0025] In process (C), a portion or all of the conversational human language information input through the user interface can be utilized. With respect to process (C), in one embodiment, the information provided to the large language model in process (C) includes primarily the conversational human language information input through the user interface. Selecting the visualization software based on the conversational human language information obtained from the large language model can include the following processes. If the conversational human language information includes the name of the visualization software, the processor can directly select the software by the name. Alternatively, the processor can conduct a web search using the conversational human language information, obtain conversational human language and / or visual information, and then select the visualization software based on the findings. The web search can involve searching for visual information (e.g., image search) or conversational human language information (e.g., text search). The visualization software options available to the content generation system need not be identified prior to system use. For example, one or more visualization software options can be selected from software available through a network such as the Internet. In addition, multiple visualization software options can be identified prior to system use. For example, the visualization software can be selected based on a topic matter contained in the conversational human language information input through the user interface. As the content generation system performs the selection, the user need not select the visualization software based on the desired content.

[0026] With respect to process (D), the text information obtained based on the large language model output can be the same as the information output by the large language model, or can be a modification of the information output by the large language model. The text information corresponds to or is related to the conversational human language information input through the user interface, although the text information can be different from it. The processor performs the input of the text information in a format compatible with the visualization software without input from the user. In other words, "the text information obtained based on the large language model output" in process (D) means that between the input through the user interface and the output of the large language model, both of which can be sources of the text information in the large language model, the output of the large language model serves as the basis for the text information. However, in addition to the output of the large language model, the input through the user interface can also serve as the basis for the text information.

[0027] In process (E), the generated content includes at least a part of the exchanged human language information acquired in process (B) or the modification thereof, and at least a part of the visual information acquired in process (D) or the modification thereof. The phrase "at least a part" means that the content does not necessarily include all of the exchanged human language information acquired in process (B) or all of the visual information acquired in process (D). The term "modification" refers to changes made by the content generation system to the information acquired in process (B) or (D). In other words, the content can include modifications in addition to or instead of the information acquired in process (B) or (D). These modifications should not substantially change the information conveyed. The layout of the exchanged human language information and the visual information in the content is not particularly limited, for example, can be determined automatically by the processor or can be determined according to a request input through the user interface.

[0028] In process (F), the manner in which the content is output is not particularly limited. It can involve providing a file of the content, or it can involve displaying the content itself.

[0029] Inventive effect According to the present application, by utilizing a large language model, the content generation system can efficiently generate content that is easier for people to understand.

[0030] Brief description of drawings Figure 1 (a) is a system overview diagram explaining a content generation system related to the embodiments of the present disclosure; Figure 1 (b) is a flowchart explaining the processes performed by the content generation system.

[0031] Embodiments of the invention Figure 1 (a) is a system overview diagram explaining a content generation system 1 related to the embodiments of the present disclosure.

[0032] The content generation system 1 includes a processor 2, and a memory 3 and a communication module 4 communicatively connected to the processor 2. The memory 3 stores programs for executing processes (A) to (F). The processor 2 executes these programs. In this embodiment, the communication module 4 corresponds to a user interface. The number of processors 2 and memory units 3 is not particularly limited and can be one or more. The hardware configuration of the content generation system 1 is not limited. The content generation system 1 can be configured by a single server device or a plurality of server devices capable of communicating with each other. In this case, the plurality of server devices can be configured to provide a cloud computing service. The communication module 4 enables communication between the processor 2 and the user terminal device 11, the large language model 6, and the plurality of visualization software 7.

[0033] A plurality of user terminal devices 11 can communicate with the content generation system 1 through a network 12. The number of user terminal devices 11 is not limited. Examples of the user terminal devices 11 shown in the figure include personal computer devices, tablet devices, and mobile phones. However, the user terminal devices 11 are not limited to these examples. Various types of terminal devices that can be used by users can be used as the user terminal devices 11.

[0034] The network 12 enables communication between the plurality of user terminal devices 11 and the content generation system 1. The type of the network 12 is not limited, and can be constructed by various types of wired networks or wireless networks or a combination thereof. The communication method is not limited, and the communication protocol is not limited.

[0035] The large language model 6 is stored in one server device or a plurality of server devices capable of communicating with each other. This one or a plurality of server devices can communicate with the content generation system 1. The large language model 6 outputs to the content generation system 1 in response to an input from the content generation system 1. Such an input is, for example, communication of human language information. Such an output is, for example, communication of human language information in response to the above-mentioned input, and is also text information for input to the visualization software 7.

[0036] A plurality of visualization softwares 7 are stored in one server device or a plurality of server devices capable of communicating with each other, respectively. These server devices can communicate with the content generation system 1 through the network such as the Internet. In other words, each of the visualization softwares 7 is available to the content generation system 1 through the network 12 such as the Internet. From among the plurality of visualization softwares 7, the content generation system 1 can select one or a plurality of them. The selected visualization software 7 outputs to the content generation system 1 in response to an input from the content generation system 1. Such an input is, for example, text information acquired from the large language model 6. Such an output is, for example, visual information generated using the text information.

[0037] Figure 1 (b) of FIG. 1 is a flowchart that explains the process executed by the content generation system 1. Hereinafter, an example case in which a user inputs "Please generate explanatory material of technical matter X" as communication of human language information to the user terminal device 11 will be described. However, this is only an example, and the input of communication of human language information is not limited to this example.

[0038] First, the user inputs the communication of human language information request of "Generate explanatory material of technical matter X" to the user terminal device 11 (step S111).

[0039] The user terminal device 11 can run a software or an application related to the service provided by the content generation system 1, or display a relevant website on a web browser. In this state, communication human language information is input by the user. The input communication human language information has a subject matter ("generate explanatory material of technical matter X"). The input communication human language information is sent by the user terminal device 11 to the content generation system 1 through the network 12 (step S112). In the content generation system 1, the processor 2 receives the communication human language information through the communication module 4 (step S11).

[0040] Process (A) After step S111, the processor 2 utilizes both the large language model 6 and the visualization software 7 (process (A)). In the process (A), the selection and management of the visualization software 7 is performed by the content generation system 1 with the output of the large language model 6. This management is directed to the selected visualization software 7, and involves the use of the visualization software 7 to explain or describe the subject matter contained in the communication human language information. The process (A) includes the following processes (B) to (D). In the process (A), an iterative process can be performed between the content generation system 1 and the large language model 6 and / or the visualization software 7.

[0041] Process (B) After step S111, the processor 2 acquires communication human language information for inclusion in the content based on the input received through the communication module 4 (process (B)). In this process (B), first, the processor 2 provides the communication human language information to the large language model 6 (step S12). After step S12, the communication human language information for the content is provided from the large language model 6 to the content generation system 1 (step S62). As a result, the processor 2 acquires the communication human language information for inclusion in the content through the large language model 6 (step S13). Although the process (B) is iteratively performed in this embodiment, this is not an exclusive example. In this embodiment, the processor 2 acquires a detailed explanation about the technical matter X as the communication human language information for inclusion in the content through multiple iterations of the process (B). This explanation includes a general formula for explaining the technical matter X, and variables contained in the general formula.

[0042] Process (C) Next, the processor 2 acquires communication human language information for selecting the visualization software 7 capable of generating visual information for inclusion in the content from the large language model 6 using the communication human language information input through the communication module 4. Based on the acquired communication human language information, the processor 2 selects the visualization software 7 (process (C)).

[0043] In this process (C), first, the processor 2 provides the conversational human language information to the large language model 6 (step S14). In this embodiment, since the process (B) is executed before the process (C), the conversational human language information acquired in the process (B) is used in the process (C). As described above, in the process (B), the processor 2 acquires the general formula for explaining the technical matter X, and the variable included in the formula. In step S14, the processor 2 provides the large language model 6 with a query about the visualization software 7 capable of generating visual information using the general formula and the variable as the conversational human language information.

[0044] After step S14, the conversational human language information for selecting the visualization software 7 is provided from the large language model 6 to the content generation system 1 (step S64). As a result, the processor 2 acquires the conversational human language information for selecting the visualization software 7 from the large language model 6 (step S15). In this embodiment, the conversational human language information acquired in step S15 includes the name of the visualization software 7 capable of generating visual information using the general formula and the variable. Based on the acquired conversational human language information, the processor 2 selects the visualization software 7 (step S16).

[0045] Process (D) Next, the processor 2 acquires visual information to be included in the content based on the text information acquired from the output of the large language model 6 in response to the conversational human language information input through the communication module 4 by operating the visualization software 7 selected in step S16 (process (D)).

[0046] In the process (D), first, the processor 2 provides the conversational human language information to the large language model 6 (step S17). In this embodiment, the processes (B) and (C) are executed before the process (D), and the conversational human language information acquired in the processes (B) and (C) is used in the process (D). Specifically, the processor 2 submits a query to the large language model 6 about the data type and value that should be input into the visualization software 7 to acquire visual information representing the technical matter X. As a result, the text information is provided from the large language model 6 to the content generation system 1 (step S67). Therefore, the processor 2 acquires the text information from the large language model 6 (step S18), which includes the data type and value input in a format compatible with the visualization software 7.

[0047] In the process (D), the processor 2 provides the text information to the visualization software 7 (step S19). The visualization software 7 is stored in a manner that allows it to operate on one or more server devices. The visualization software 7 operates using the text information to generate visual information. As described previously, in this embodiment, the text information includes the type and value of data that should be input. In addition, the text information is acquired in a format that is compatible with the visualization software 7. Thus, the processor 2 can provide this text information to the visualization software 7, and the processor 2 can operate the visualization software 7. As a result, visual information is generated by the visualization software 7. In this embodiment, the generated visual information includes simulation results and graphs that provide specific examples of the technical matter X. The generated visual information is provided from the visualization software 7 to the content generation system 1 (step S79). Thus, the processor 2 acquires the visual information from the visualization software 7 to be included in the content (step S20).

[0048] Process (E) After the processes (B) to (D), the processor 2 generates content that includes at least a portion of the exchanged human language information acquired in the process (B) and at least a portion of the visual information acquired in the process (D) (process (E)).

[0049] Process (F) The processor 2 outputs the content generated in the process (E) from the communication module 4 to the user terminal device 11 via the network 12 (process (F)). The user terminal device 11 receives the content (step S121) and outputs the content (step S122). The manner in which the content is output is not particularly limited. The output content can be displayed on a display equipped on the user terminal device 11, or can be provided as a data file.

[0050] The present application is not limited to the above-described embodiments. The application can be implemented in other embodiments, and various modifications can be added.

[0051] Explanation of Symbols 1: Content generation system 2: Processor 3: Memory 4: Communication module 6: Large language model 7: Visualization software 11: User terminal device 12: Network

Claims

1. A content generation system comprising: at least one user interface; at least one memory; at least one processor configured to execute at least one program stored in the memory and connected to the memory, the at least one program being programmed to cause the at least one processor to perform: (A) utilizing each of a large language model that operates using communicated human language information input through the user interface and a visualization software that is programmed to output visual information using provided textual information, (B) in the (A), acquiring, by the large language model, communicated human language information for inclusion in content using at least a portion of the communicated human language information input through the user interface, (C) in the (A), acquiring, by the large language model, communicated human language information for selecting a visualization software capable of generating visual information for inclusion in content using at least a portion of the communicated human language information input through the user interface and selecting the visualization software based on the acquired communicated human language information, (D) in the (A), acquiring, by operating the visualization software selected in the (C) using textual information acquired based on an output of the large language model corresponding to the communicated human language information input via the user interface, visual information for inclusion in content, (E) generating content including at least a portion of the communicated human language information acquired from the (B) or a modification thereof and at least a portion of the visual information acquired from the (D) or a modification thereof, and (F) outputting, through the user interface, the content generated in the (E).

2. The content generation system according to claim 1, wherein the (A) includes: (A1) in the (A), acquiring, by the large language model, communicated human language information for inclusion in content using at least a portion of the communicated human language information input through the user interface, (A2) in the (A), acquiring, by the large language model, communicated human language information for selecting a visualization software capable of generating visual information for inclusion in content using at least a portion of the communicated human language information input through the user interface and selecting the visualization software based on the acquired communicated human language information, (A3) in the (A), acquiring, by the large language model, textual information for inclusion in content using at least a portion of the communicated human language information input through the user interface, and (A4) in the (A), acquiring, by the large language model, textual information for selecting a visualization software capable of generating visual information for inclusion in content using at least a portion of the communicated human language information input through the user interface and selecting the visualization software based on the acquired textual information.

3. The content generation system according to claim 1 or 2, wherein the (B) includes: (B1) in the (B), acquiring, by the large language model, communicated human language information for inclusion in content using at least a portion of the communicated human language information input through the user interface, (B2) in the (B), acquiring, by the large language model, communicated human language information for selecting a visualization software capable of generating visual information for inclusion in content using at least a portion of the communicated human language information input through the user interface and selecting the visualization software based on the acquired communicated human language information, (B3) in the (B), acquiring, by the large language model, textual information for inclusion in content using at least a portion of the communicated human language information input through the user interface, and (B4) in the (B), acquiring, by the large language model, textual information for selecting a visualization software capable of generating visual information for inclusion in content using at least a portion of the communicated human language information input through the user interface and selecting the visualization software based on the acquired textual information.

4. The content generation system according to any one of claims 1

Images