Information processing device, information processing method, and program

Abstract

An information processing device, an information processing method, and a program are provided that can easily identify the contents to be described when creating a specification. [Solution] The information processing device comprises a receiving means for receiving an invention idea and a viewpoint for writing a specification as input, an instruction means for issuing a first instruction to a language model to create a virtual personality, and a second instruction that includes the input and uses the virtual personality to identify the contents of the specification from the viewpoint, and an output means for receiving and outputting an output in response to the instruction from the language model.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] There is a known technology for creating specifications using large-scale language models (LLMs). When an LLM is used to write a specification, it may write the specification in an unintended direction. On the other hand, when trying to write the specification in the intended direction, a human must provide detailed instructions, which increases the amount of work required.

[0003] Patent Document 1 discloses a system for creating specifications using LLM. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2023-006493 Summary of the Invention [Problem to be solved by the invention]

[0005] When attempting to have an LLM write a specification, the LLM may not write the specification in the direction intended. On the other hand, when trying to write the specification in the intended direction, a human may have to provide detailed instructions, which may increase the amount of work required. In other words, it may be time-consuming to specify the content to be written when an LLM is creating a specification. [Means for solving the problem]

[0006] The present disclosure provides an information processing device comprising: a receiving means for receiving an idea for an invention and a viewpoint in writing a specification as input; an instruction means for issuing a first instruction to a language model to create a virtual personality; and a second instruction including the input and using the virtual personality to specify the contents of the specification from the viewpoint; and an output means for receiving and outputting an output in response to the instruction from the language model. The present disclosure also provides an information processing method that receives an idea for an invention and a viewpoint in writing a specification as input, issues a first instruction to a language model to create a virtual personality, and a second instruction that includes the input and uses the virtual personality to specify the content of the specification from the viewpoint, and receives and outputs an output in response to the instructions from the language model. The present disclosure also provides a program that causes a computer to execute the following processes: a process of accepting an idea for an invention and a viewpoint in writing a specification as input; a process of issuing a first instruction to a language model to create a virtual personality; and a process of issuing a second instruction that includes the input and uses the virtual personality to identify the contents of the specification from the viewpoint; and a process of receiving and outputting an output in response to the instruction from the language model. [Effects of the Invention]

[0007] According to the present disclosure, it is possible to easily specify the content to be described when preparing a specification. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram showing a hardware configuration of an information processing device according to the present disclosure. [Figure 2] FIG. 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 3] 10 is a flowchart showing the procedure of a process performed by an information processing device according to the present disclosure. [Figure 4] FIG. 10 illustrates an example of an output generated from an LLM. [Figure 5] FIG. 10 is a diagram showing an example of an instruction given by an instruction means to an LLM. [Figure 6] FIG. 1 is a block diagram illustrating a functional configuration of an information processing device according to the present disclosure. [Figure 7] 10 is a flowchart illustrating a processing procedure of an information processing device according to the present disclosure. [Figure 8] FIG. 10 is a diagram showing an example of a dialogue between virtual personalities generated by LLM. [Figure 9] FIG. 1 is a diagram showing an example of a specification item generated by LLM. [Figure 10] FIG. 10 is another block diagram showing the functional configuration of the information processing device according to the present disclosure. [Figure 11] 10 is a flowchart illustrating an operation of an information processing device according to the present disclosure. [Figure 12] FIG. 10 is a diagram illustrating an example of an output from an LLM. [Figure 13] FIG. 10 is a diagram showing an example of output of candidates for prior art documents output from an output unit. [Figure 14] FIG. 10 is yet another block diagram showing the functional configuration of an information processing device according to the present disclosure. [Figure 15] 10 is a diagram showing an example of setting a virtual personality by instruction means 420. FIG. [Figure 16] 10 is a diagram showing another example of setting a virtual personality by the instruction means 420. FIG. [Figure 17] 10 is a flowchart illustrating an example of a procedure of a process performed by an information processing device according to the present disclosure. [Figure 18] FIG. 10 is yet another block diagram showing the functional configuration of an information processing device according to the present disclosure. [Figure 19] 10 is a flowchart showing the operation of an information processing device according to the present disclosure. [Figure 20] FIG. 10 is a diagram showing an example of commentary by a virtual personality. [Figure 21] FIG. 10 is another block diagram showing the configuration of an information processing device according to the present disclosure. [Figure 22] 10 is another flowchart showing the operation of the information processing device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0009] (Hardware configuration) First, the hardware configuration of an information processing device according to the present disclosure will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the hardware configuration of the information processing device. The information processing device is realized by a computer 10.

[0010] The computer 10 includes a processor 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage device 14 such as a hard disk for storing programs, an input / output interface 15 for inputting and outputting data, and a communication interface 16 for network connection. Each component is connected via a bus 17.

[0011] The processor 11 runs an operating system to control the entire computer. Examples of the processor 11 include a CPU (Central Processing Unit), a DSP (Digital Signal Processor), and a GPU (Graphics Processing Unit). The processor 11 loads a program stored in, for example, the ROM 12 or the storage device 14. The processor 11 then executes each process coded in the program. The processor 11 may execute the processes or instructions in the illustrated flowchart based on the program.

[0012] The ROM 12 stores application programs, programs according to the embodiments, etc. The RAM 13 is used as a work area for the processor 11.

[0013] Examples of the storage device 14 include a semiconductor memory such as a flash memory, an HDD (Hard Disk Drive), etc. The storage device 14 stores, for example, an OS (Operating System) program, application programs, programs according to each embodiment, etc.

[0014] The input / output interface 15 is connected to peripheral devices (not shown) via a wired network or a wireless network.

[0015] The communication interface 16 is connected to a communication network (not shown), such as a LAN (Local Area Network) or a WAN (Wide Area Network), via a wireless or wired network. The communication network may be configured with multiple communication networks. As a result, the computer is connected to an external large-scale language model server that configures an LLM via the communication network. The computer 10 may have components other than those shown in FIG. 5. For example, the computer 10 may have a drive device or the like. For example, the processor 11 may be attached to a drive device or the like and read programs and data stored in a non-transitory tangible recording medium into the RAM 13.

[0016] The computer 10 that realizes the information processing device may be configured to include some of the components described in Fig. 1. For example, the computer 10 that realizes the information processing device may be configured to include only the processor 11, ROM 12, and RAM 13 among the components described above. In this case, the storage device 14, input / output interface 15, and communication interface 16 may each be provided as an external device. Furthermore, some of the computing functions may be realized by an external server, cloud, or the like.

[0017] The present disclosure relates to an information processing device for efficiently creating specifications using a large language model (LLM).

[0018] In related specification preparation systems, when an LLM is asked to write a specification, there is a problem that the LLM does not write the specification in the direction intended. On the other hand, if an LLM is asked to write the specification in the intended direction, a human would need to provide detailed instructions, which increases the amount of work required.

[0019] The present disclosure has been made to solve the above-mentioned problems, and aims to provide an information processing device that can easily identify the contents of a description when creating a specification.

[0020] 2 is a block diagram showing an example of the configuration of the information processing device 100. The information processing device 100 includes a receiving unit 110, an instruction unit 120, and an output unit .

[0021] The receiving means 110 receives input of an invention idea and a viewpoint for writing a specification via a user interface (UI). Specifically, the user can input the invention idea and viewpoint using an input device such as a touch panel display, keyboard, or mouse.

[0022] The instruction means 120 issues a first instruction to the LLM to create a virtual persona and a second instruction to specify the contents of the specification using the virtual persona, including the input viewpoint. The instruction means 120 is realized using hardware resources such as a CPU, memory, and a communication interface.

[0023] The output means 130 receives output from the LLM and presents it to the user via a display, printer, etc. The output means 130 is also realized using hardware resources such as a CPU, memory, a communication interface, and a display controller.

[0024] Next, the operation of the information processing device 100 will be described with reference to Fig. 3. Fig. 3 is a flowchart showing the procedure of the process (information processing method) performed by the information processing device 100.

[0025] First, the receiving means 110 receives an invention idea and a viewpoint for writing a specification from a user (step S101). For example, suppose the user inputs an invention idea of ​​"a face authentication system using a smartphone camera" and a viewpoint of "improving security."

[0026] Next, the instruction means 120 instructs the LLM to create a virtual personality (step S102). At this time, the instruction means 120 gives the LLM a specific instruction, such as "Please create a virtual personality of a security expert."

[0027] Next, the instruction means 120 instructs the LLM to identify the content of the description of the specification using the virtual persona, including the input viewpoint (step S103). Specifically, the instruction is given such as, "From the perspective of a security expert, please identify the content of the description of the specification that focuses on improving security for a face authentication system using a smartphone camera."

[0028] The LLM performs processing based on these instructions and outputs the results (step S104). For example, the LLM may generate an output such as that shown in FIG.

[0029] Finally, the output means 130 receives the output from the LLM and presents it to the user (step S105). Specifically, the output content is displayed on a display or printed out by a printer.

[0030] According to this embodiment, by having the LLM create a virtual personality and using that virtual personality to identify the content of the specification, the content of the specification can be obtained in the intended direction without the user having to provide detailed instructions. Furthermore, by using a virtual personality that mimics the perspective of an expert, the amount and quality of information in the description can be improved.

[0031] The LLM may display a message such as the following as an instruction to the user: "Enter your invention idea and the aspects you want to emphasize in writing the specification. For example, if you have an idea for a 'face recognition system using a smartphone camera,' and you enter the aspect of 'improving security,' we will suggest content for the specification from the perspective of a security expert."

[0032] In this way, by using the information processing device of the present disclosure, the user can obtain high-quality written content while significantly reducing the effort required to create a specification.

[0033] According to this embodiment, by having the LLM create a virtual personality and using that virtual personality to identify the contents of the specification, the contents of the specification can be obtained in the intended direction without the user having to provide detailed instructions. Furthermore, by using a virtual personality that mimics the perspective of an expert, the amount and quality of information in the description can be improved. This allows the contents to be identified without much effort when creating a specification. Furthermore, the amount of information in the description can be increased.

[0034] Furthermore, in this embodiment, the LLM automatically selects an appropriate virtual persona based on the invention idea and viewpoint entered by the user, and identifies the content of the specification from the virtual persona's perspective, allowing the user to focus on the essential parts of the invention and making the specification drafting process more efficient.

[0035] Furthermore, the information processing device 100 of this embodiment presents the output results of the LLM to the user, allowing the user to check the contents and provide corrections or additional instructions as necessary. This minimizes the gap between the user's intentions and the output results of the LLM, making it possible to create higher quality specifications.

[0036] In this embodiment, the receiving unit 110 receives an invention idea and a viewpoint for writing a specification as input via a user interface (UI). However, if the invention idea or viewpoint input by the user is unclear or insufficient, the LLM may not be able to select an appropriate virtual persona, and the expected high-quality description may not be obtained.

[0037] In this case, the receiving means 110 may provide a guidance function to assist in inputting ideas and viewpoints for the invention. For example, if it is determined that the information input by the user is insufficient, the receiving means 110 may display a prompt such as "Please explain your invention idea in more detail" or "Do you need to consider viewpoints other than security?" to elicit more detailed information from the user.

[0038] This allows users to more clearly express their invention ideas, enabling LLM to select more appropriate virtual personalities and generate higher-quality descriptions, further improving the efficiency and quality of the specification drafting process.

[0039] Furthermore, in this embodiment, the receiving means 110 is an interface that receives information to be sent to the LLM, and does not have to be on the same server as the LLM. For example, a distributed system configuration may be adopted in which the receiving means 110 is installed on a client terminal, and the instruction means 120 and output means 130 are installed on the server side. In this case, the user inputs the idea and perspective of the invention from the client terminal, and the information is sent to the server via the network. The LLM on the server performs processing and returns the results to the client terminal for display.

[0040] The receiving means 110 may also receive a designation of which personality to create. For example, the user may be able to select from virtual personalities such as "security expert," "patent attorney," and "software engineer." This allows the user to select the expert perspective that best suits their invention idea, resulting in more accurate description of the specification.

[0041] The instruction means 120 provides specific instructions to the LLM. For example, instructions such as those shown in Figure 5 can be provided for an invention idea such as "a face recognition system using a smartphone camera." These specific instructions enable the LLM to generate more focused and high-quality descriptions.

[0042] The output means 130 may not only display the content output by the LLM, but also present the content to the user and ask for their acceptance or rejection. For example, the user may be able to select from options such as "approve," "modify," or "delete" for each item of the content generated by the LLM. This allows the user to manage the generated content in more detail and create specifications that meet their own intentions.

[0043] In this embodiment, the perspectives for writing a specification include the key points of the invention, limitations on writing a specification, and requirements for writing a specification. For example, you can specify perspectives such as "The key point of the invention is a novel facial recognition algorithm," "Avoid the use of specific technical terms," ​​and "Include at least three working examples." This allows LLM to more accurately understand the user's intent and generate appropriate description content.

[0044] As described above, according to this embodiment, it is possible to provide an information processing device that can efficiently identify the contents of a specification using LLM and create a high-quality specification.

[0045] Second Embodiment The second embodiment will be described with reference to Fig. 6. Note that the information processing device 200 differs in some configurations and operations from the information processing device 100 described in the first embodiment, but other parts may be the same as the information processing device 100 of the first embodiment. Therefore, hereinafter, only the parts that differ from the first embodiment will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate.

[0046] 6 is a block diagram showing an example of the functional configuration of the information processing device 200. The information processing device 200 includes a receiving unit 210, an instruction unit 220, an output unit 230, and a creation unit 240.

[0047] The receiving means 210 receives input of an invention idea and a viewpoint for writing a specification via a user interface (UI). Specifically, the receiving means 210 receives information such as an outline of the invention, a technical field, a problem to be solved, and an effect of the invention, which is input by the user using a keyboard or a touch screen.

[0048] The instruction means 220 issues a first instruction to the large-scale language model (LLM) to create a virtual personality, and a second instruction to specify the contents of the specification from a viewpoint using the virtual personality, including the input received by the reception means 210. For example, the instruction means 220 issues a specific instruction to the LLM such as "Please create a virtual personality as a patent attorney and create a background art section of the specification based on the input invention idea."

[0049] The output means 230 receives output from the LLM and presents it to the user via a display, printer, etc. For example, the contents of each item of a specification generated by the LLM or the contents of discussions between virtual personalities are displayed in text format on the display.

[0050] The creation means 240 issues a third instruction to the LLM to create a specification based on the content of the description identified by the LLM based on the second instruction issued by the instruction means 220. For example, the creation means 240 issues an instruction to the LLM such as, "Please create the overall structure of the specification using the identified background art, the problem that the invention aims to solve, and the content of the means for solving the problem."

[0051] Next, the operation of the information processing device 200 of the second embodiment will be described in more detail with reference to Fig. 7. Fig. 7 is a flowchart showing an example of a processing procedure of the information processing device 200.

[0052] First, the receiving means 210 receives an invention idea and a viewpoint for writing a specification from a user (step S201). At this time, the receiving means 210 displays a message to the user saying, "Please enter your invention idea. Please include the main components, how they work together, and the expected effects. Also, please enter the viewpoint for writing a specification (e.g., key points of the invention, limitations, necessary items)." to prompt the user to input. The user inputs the invention idea of ​​"a face authentication system using a smartphone camera" and the viewpoint of "improving security."

[0053] Next, the instruction means 220 instructs the LLM to create virtual personalities (first instruction) (step S202). Specifically, the instruction to the LLM is as follows: "Create virtual personalities for a patent attorney and an inventor, and reflect their respective specialized knowledge and experience. Set the patent attorney as an expert in patent law and specification writing, and the inventor as a person well versed in technical details."

[0054] Next, the instruction means 220 instructs the LLM (second instruction) to identify the contents of the specification using the virtual persona, including the input viewpoint (step S203). This instruction includes an instruction to conduct a simulated interview between the virtual personas, and includes specific content such as, "Using the virtual personas of the patent attorney and inventor you created, please conduct a simulated interview about the input invention idea, taking into account the specified viewpoint. During the interview, please discuss in detail the technical features of the invention, differences from the prior art, and the effects of the invention, and identify the contents of the specification." Or, the instruction may be, "Using the virtual personas of the patent attorney and inventor you created, please conduct a simulated interview about a facial recognition system using a smartphone camera from the perspective of improving security, and identify the contents of the specification."

[0055] The LLM performs processing based on these instructions and outputs the results (step S204). For example, the dialogue between the virtual personalities shown in FIG. 8 is generated. The output means 230 receives the output from the LLM and presents it to the user (step S205). For example, the content of the dialogue between the virtual personalities and the contents of the identified specification are displayed on the display. The user can check the discussion between the virtual personalities.

[0056] Next, the creation means 240 instructs the LLM to create a specification based on the identified description (third instruction) (step S206). This instruction includes detailed instructions such as, "Based on the identified description, please create a specification including the following items: 1. Technical field, 2. Background art, 3. Problem to be solved by the invention, 4. Means for solving the problem, 5. Effects of the invention, 6. Mode for carrying out the invention, 7. Industrial applicability. Please describe each item clearly and concisely in accordance with the Patent Act Enforcement Regulations." Or, the creation means 240 may instruct the LLM to, "Based on the content of the above dialogue, please create the following items: Background art, Problem to be solved by the invention, Means for solving the problem, and Effects of the invention."

[0057] The LLM creates a statement based on these instructions and outputs the result (step S207). The output means 230 presents the statement created by the LLM to the user (step S208). For example, statement items such as those shown in Figure 9 are generated and displayed on the display.

[0058] Through the above process, the information processing device 200 of the second embodiment can identify the content of the specification through a simulated interview between virtual personalities based on the invention idea and viewpoint input by the user, and then create the entire specification based on the identified content. This makes it possible to efficiently create a high-quality specification that meets legal requirements while reflecting the user's intentions.

[0059] According to the information processing device 200 of the second embodiment, by setting the personalities of a patent attorney and an inventor as virtual personalities and conducting a simulated interview between these personalities, it is possible to identify more specific and detailed content of the description of the specification. Through this dialogue between the virtual personalities, discussions are held from multiple perspectives on the technical features of the invention, differences from prior art, effects of the invention, etc., making it possible to extract all important information that should be included in the specification.

[0060] Furthermore, by creating a specification based on the identified content, users can efficiently create a specification that more appropriately expresses the essence of the invention. This allows users to significantly reduce the time and effort required to create a specification while still obtaining a high-quality specification. In particular, automatic generation of specifications that comply with the Patent Law Enforcement Regulations makes it easier to create specifications that meet legal requirements, contributing to improving the quality of patent applications.

[0061] In addition, in this embodiment, the content of the specification is determined taking into consideration the viewpoints input by the user (e.g., key points of the invention, limitations, necessary matters), making it easier to reflect the user's intentions and priorities in the specification. This makes it possible to create a specification that is highly satisfying to the user and accurately expresses the essence of the invention.

[0062] In addition, in the information processing device 200 of this embodiment, if the idea or perspective of the invention input by the user is unclear or insufficient, the LLM will not be able to set an appropriate virtual personality, and as a result, inappropriate or insufficient description content may be generated in the specification.

[0063] In this case, the instruction means 220 may instruct the LLM to analyze the input idea and viewpoint of the invention and identify any unclear or insufficient points, for example, "Analyze the input idea and viewpoint of the invention, and if any information necessary for writing the specification is lacking, please point out the points."

[0064] If the LLM performs an analysis based on this instruction and detects that there is insufficient information, the output means 230 may display a message prompting the user to enter additional information, such as "The technical field of the invention is unclear. Please enter the technical field to which the invention belongs."

[0065] The receiving means 210 may receive additional information input by the user in response to this message, and the instruction means 220 may again instruct the LLM, including this additional information. This makes it possible to set a more appropriate virtual personality and identify appropriate specification contents based on that.

[0066] By adding such processing, the information processing device 200 of the second embodiment can always identify appropriate description content for a statement and create a high-quality statement, regardless of the quality of the input information. This makes it possible to efficiently create consistently high-quality statements, regardless of the user's input skills or experience.

[0067] Furthermore, the information processing device 200 of this embodiment can dynamically adjust the instructions to the LLM to identify more appropriate content for the specification. For example, the instruction means 120 can change the instructions to the LLM depending on the technical field and complexity of the invention idea input by the user.

[0068] Specifically, if the invention idea involves complex technology, the instruction means 220 instructs the LLM, "Please discuss the technical details in more depth. If necessary, please include explanations of technical terms and specific numerical examples." On the other hand, if the invention idea involves relatively simple technology, the instruction means 220 instructs the LLM, "Please explain the features of the invention concisely and clearly, and discuss the differences from the prior art with a focus on the differences."

[0069] By dynamically adjusting the instructions in this way, it is possible to identify the appropriate depth and breadth of specification content depending on the nature of the invention, ensuring that specifications of appropriate quality are always produced, regardless of the technical field or complexity of the invention.

[0070] Furthermore, the information processing device 200 of this embodiment can also provide a function that allows the user to check the content of the specification generated by the LLM and make corrections or additions as necessary. For example, the output means 230 displays the content of the specification generated by the LLM on a display and provides an interface that allows the user to directly edit it.

[0071] When the user makes edits, the receiving means 210 receives the edits, and the instruction means 220 issues new instructions to the LLM that reflect the edits. For example, the instruction may be, "Please rewrite the relevant part of the specification, taking into consideration the technical explanation added by the user."

[0072] This functionality allows users to easily reflect their intentions and additional technical information in the specifications generated by LLM, further improving the quality of automatically generated specifications and increasing user satisfaction.

[0073] As described above, the information processing device 200 of the second embodiment identifies the content of a specification through simulated interviews between virtual personalities and automatically generates a specification based on that content. Furthermore, by combining various functions, such as analyzing input information and requesting additional information, dynamically adjusting instructions, and providing user editing functions, it becomes possible to efficiently create high-quality specifications. This significantly reduces the effort required for specification creation while enabling the creation of high-quality specifications that adequately express the essence of the invention, significantly contributing to the efficiency and quality improvement of the entire patent application process.

[0074] <Third embodiment> An information processing device 300 of the third embodiment will be described with reference to Fig. 10. The information processing device 300 of the third embodiment differs in part in configuration and operation from the information processing device 100 of the first embodiment described above, but other parts may be similar to the information processing device 300 of the first embodiment. Therefore, hereinafter, only the parts that differ from the first embodiment will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate.

[0075] 10 is a block diagram showing an example of the functional configuration of the information processing device 300. The information processing device 300 includes a receiving unit 310, an instruction unit 320, and an output unit 330.

[0076] The receiving means 310 receives as input via a user interface (UI) the idea for the invention and the viewpoint for writing the specification. Furthermore, the receiving means 310 has a function for receiving as input prior art documents. For example, if a user inputs the invention idea of ​​"a face authentication system using a smartphone camera" and specifies the viewpoint of "security," the receiving means 310 receives this information. It also receives patent documents relating to "a face authentication system using a monocular camera" as related prior art documents.

[0077] The instruction means 320 issues a first instruction to the large-scale language model (LLM) to create a virtual persona, and a second instruction to identify the contents of the specification using the virtual persona, including the input viewpoint. In the third embodiment, the instruction means 320 includes an instruction to have the virtual personas hold a discussion based on prior art documents. The instruction means 320 also includes an instruction to have the virtual personas, including a virtual persona skilled in identifying system vulnerabilities, identify the system vulnerabilities and propose improvements.

[0078] For example, the instruction means 320 issues the following instruction to the LLM: "Please identify the contents of the specification from a security perspective regarding a facial recognition system using a smartphone camera. Also, please hold a discussion between a security expert and a fictitious personality of the inventor based on the provided prior art document, 'Facial Recognition System Using a Monocular Camera.' Furthermore, please include a fictitious personality that points out the system's vulnerabilities and proposes improvements."

[0079] The output means 330 receives output from the LLM and presents it to the user via a display, printer, etc. For example, the contents of discussions between virtual personalities, identified vulnerabilities, proposed improvements, etc. are output.

[0080] Next, the operation of the information processing device 300 of the third embodiment will be described.

[0081] The operation of the information processing device 300 will be described with reference to the flowchart in Fig. 11. First, the receiving means 310 receives an invention idea and a viewpoint for writing a specification from a user (step S301). Next, the receiving means 310 receives prior art documents as input (step S302).

[0082] Next, the instruction means 320 instructs the LLM to create virtual personalities (first instruction) (step S303). At this time, virtual personalities such as a security expert, an inventor, and an expert who points out system vulnerabilities are created.

[0083] Next, the instruction means 320 instructs the LLM to identify the contents of the specification using the virtual personas, including the input viewpoint (second instruction) (step S304). This instruction includes an instruction to have the virtual personas discuss the content of the specification based on prior art documents, and an instruction to point out vulnerabilities in the system and propose improvements.

[0084] The LLM processes these instructions and outputs the results (step S305). For example, a security expert's virtual personality might point out that "facial recognition using a monocular camera only uses two-dimensional images, so there is a risk of impersonation using photographs, etc." The inventor's virtual personality might respond by saying, "By combining multiple sensors on a smartphone, three-dimensional facial features can be captured." Furthermore, a virtual expert's virtual personality might point out system vulnerabilities, pointing out that "depending on the method of storing facial recognition data, there is a risk of data leakage," and propose an improvement measure of "encrypting and distributing the storage of facial recognition data." This discussion then continues.

[0085] The output means 330 receives the output from the LLM and presents it to the user (step S306). For example, an output such as that shown in FIG.

[0086] As described above, the information processing device 300 of the third embodiment can identify more specific and practical content for a specification by considering prior art documents and including a virtual personality that points out system vulnerabilities. This method allows for the creation of a more comprehensive and reliable specification that takes into account not only the idea of ​​an invention but also related prior art and potential vulnerabilities.

[0087] The information processing device 300 of the third embodiment can identify more specific and practical content for the specification by considering prior art documents and including a virtual personality that points out system vulnerabilities. This method allows for the creation of a more comprehensive and reliable specification that takes into account not only the idea of ​​the invention but also related prior art and potential vulnerabilities.

[0088] Specifically, by having virtual characters engage in discussions based on prior art documents, the novelty and inventive step of an invention can be clearly demonstrated. For example, the advantages of using a combination of multiple sensors compared to a facial recognition system using a monocular camera become clear. Furthermore, by including virtual characters that point out system vulnerabilities, potential problems with the invention can be identified in advance and solutions can be incorporated into the specification. This allows for a more robust and practical description of the invention.

[0089] Furthermore, this method significantly improves the efficiency of specification drafting. Conventional methods require inventors or patent experts to research prior art and analyze potential vulnerabilities, but in this embodiment, these tasks are automatically performed by the LLM and virtual personality. This makes it possible to produce high-quality specifications while reducing the time and effort required for specification drafting.

[0090] However, in the information processing device 300 of this embodiment, if the user does not select the appropriate prior art document, the discussion between the virtual personalities may not be carried out appropriately, and as a result, the contents of the specification may become inappropriate.

[0091] To address this issue, the instruction means 320 may have a function of automatically searching and suggesting related prior art documents based on the input invention idea. For example, the instruction means 320 may issue the following instructions to the LLM: "Please search for prior art documents related to the invention idea you entered, 'Facial Recognition System Using a Smartphone Camera,' and suggest the three most relevant documents. For each document, please provide the title, publication year, and a brief description of its relevance to the present invention."

[0092] The LLM processes the instructions and outputs candidates for relevant prior art documents. The output means 330 presents these candidates to the user, allowing the user to select the document that seems most appropriate.

[0093] For example, the output means 330 presents the output shown in FIG. 13 to the user.

[0094] The user selects the most appropriate document from these candidates and inputs the selection result into the reception means 310. The instruction means 320 instructs the LLM to have a discussion between virtual personalities based on the selected document.

[0095] In this way, by adding the automatic search and suggestion function for relevant prior art documents, discussions can be based on more appropriate prior art documents, and as a result, higher quality specification content can be identified.

[0096] An input interface may be used instead of the accepting unit 310. The input interface may be any of a touch panel, a keyboard, a mouse, a voice recognition device, an image recognition device, or the like.

[0097] Furthermore, a processing unit may be used instead of the instruction means 320. The processing unit can be realized by an arithmetic processing unit such as a CPU or a GPU, or an integrated circuit such as an FPGA or an ASIC.

[0098] Furthermore, a display device may be used instead of the output means 330. The display device may be any of a liquid crystal display, an organic EL display, a projector, a head-mounted display, or the like.

[0099] Instead of running on a cloud server, the LLM may run on a local computer, and a rule-based natural language processing system may be used instead of the LLM.

[0100] In setting virtual personalities, the personalities of related experts such as network engineers and data scientists may be set instead of security experts, and the personalities of white hat hackers and security consultants may be set instead of experts who point out system vulnerabilities.

[0101] When inputting prior art documents, academic papers and technical reports may be used instead of patent documents. Also, multiple prior art documents may be input simultaneously, and discussions may be conducted taking into account the relationships between those documents.

[0102] In specifying the contents of the specification, other viewpoints such as usability and processing speed may be used instead of the viewpoint of security. Also, multiple viewpoints may be considered simultaneously, and the trade-off relationship between them may be discussed.

[0103] The content of the specification may be output in a format that includes charts and formulas instead of text, or may be output as multiple options for the content to be selected by the user.

[0104] <Fourth embodiment> An information processing device 400 of the fourth embodiment will be described with reference to Fig. 14. Note that the information processing device 400 of the fourth embodiment differs in part in configuration and operation from the information processing device 100 of the first embodiment described above, but other parts may be similar to the information processing device 100 of the first embodiment. Therefore, hereinafter, only the parts that differ from the first embodiment will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate.

[0105] 14 is a block diagram showing an example of the functional configuration of the information processing device 400. As shown in FIG.

[0106] The receiving means 410 receives an invention idea and a viewpoint for writing a specification as input via a user interface (UI). Specifically, the receiving means 410 receives information input by a user using a keyboard, a touch panel, or the like. For example, if a user inputs an invention idea such as "a face authentication system using a smartphone camera," the receiving means 410 receives this information.

[0107] The instruction means 420 issues a first instruction to the large-scale language model (LLM) to create a virtual personality, and a second instruction to specify the contents of the specification using the virtual personality, including the input viewpoint. In the fourth embodiment, the instruction means 420 particularly sets a virtual personality according to the items of the specification and according to the technical field to which the idea of ​​the invention belongs.

[0108] For example, the instruction means 420 issues an instruction as shown in FIG. 15 to the LLM.

[0109] The instruction means 420 also sets a virtual personality according to the technical field. For example, the instruction means 420 issues instructions to the LLM as shown in FIG.

[0110] The output means 430 receives output from the LLM and presents it to the user via a display, printer, etc. For example, the explanation of the background art generated by the LLM and the explanation of the components and functions of the system are displayed in text format on the display.

[0111] In this way, the information processing device 400 of the fourth embodiment can set an appropriate virtual personality according to the item or technical field of the specification and identify the contents of the specification from that perspective. This makes it possible to create a specification that reflects specialized knowledge appropriate for each item, and to efficiently create a higher quality specification.

[0112] Next, the operation of the information processing device 400 of the fourth embodiment will be described with reference to Fig. 17. Fig. 17 is a flowchart showing an example of the procedure of processing performed by the information processing device 400.

[0113] First, the receiving unit 410 receives an invention idea and a viewpoint for writing a specification from a user (step S401). For example, the user inputs an invention idea such as "a face authentication system using a smartphone camera" and a viewpoint such as "improvements in terms of security."

[0114] Next, the instruction means 420 instructs the LLM to create virtual personalities (step S402). Specifically, the instruction means 420 instructs the LLM to set virtual personalities according to the items in the specification and the technical field to which the invention idea belongs. For example, the instruction means 420 instructs the LLM to set a virtual personality of "patent examiner" for the "background art" item, and virtual personalities of "image processing engineer" and "security expert" according to the technical field.

[0115] Next, the instruction means 420 instructs the LLM to identify the contents of the specification, including the input viewpoints, using the set virtual personas (step S403). For example, the instruction means 420 instructs the LLM to explain the background art from the viewpoint of a "patent examiner," and to explain the components and functions of the system from the viewpoints of an "image processing engineer" and a "security expert."

[0116] The LLM processes the instructions and outputs the results (step S404). For example, for background technology, it outputs information such as "Conventional face recognition systems have issues with reduced recognition accuracy due to changes in lighting conditions and facial angle," and for system components, it outputs elements and their functions such as "camera module, image processing unit, authentication algorithm execution unit, secure enclave."

[0117] Finally, the output means 430 receives the output from the LLM and presents it to the user (step S405). Specifically, the content of the generated specification is displayed on a display or printed out by a printer.

[0118] In this way, the information processing device 400 of the fourth embodiment can efficiently create higher quality specifications by setting an appropriate virtual personality according to the items and technical fields of the specification and identifying the contents of the specification from each specialized perspective.

[0119] The information processing device 400 of the fourth embodiment can set an appropriate virtual personality for each item and technical field of the specification and identify the content of the specification from that perspective. This enables the creation of specifications that reflect specialized knowledge appropriate for each item, resulting in the efficient creation of higher-quality specifications. For example, by utilizing the perspective of a patent examiner in the background art section and the perspectives of an image processing engineer and a security expert in the description of system components, it is possible to obtain content that takes advantage of each expertise. Descriptions from such specialized and multifaceted perspectives can more clearly express the technical features and advantages of an invention, thereby increasing the likelihood of obtaining a patent.

[0120] However, in the information processing device 400 of the fourth embodiment, if the idea or viewpoint of the invention input by the user is vague or insufficient, it may be difficult to set an appropriate virtual personality or identify the contents of the specification. For example, if the user inputs a vague idea such as "a new authentication method," it may be difficult to identify the technical field, and it may not be possible to set an appropriate virtual personality.

[0121] In this case, the instruction means 420 may generate a prompt requesting additional information from the user and present it to the user via the output means 430. For example, a prompt such as "Please tell us the specific technology of the authentication method (e.g., biometric authentication, two-factor authentication, etc.)" may be displayed to collect more detailed information from the user.

[0122] The receiving unit 410 may also receive input of additional information from the user. For example, if the user inputs additional information such as "a two-factor authentication system that combines face authentication and voice authentication," the instruction unit 420 can set a more appropriate virtual personality, such as "biometric authentication expert" or "voice processing engineer," based on this information.

[0123] Furthermore, the instruction means 420 may dynamically adjust the instructions to the LLM based on the collected additional information. For example, the instruction means 420 may generate more specific and appropriate instructions such as, "Please explain the synergistic effect of combining face recognition and voice recognition from the perspective of a biometrics expert."

[0124] In this way, even if the user's input is insufficient, by collecting additional information and dynamically adjusting the instructions, it is possible to set a more appropriate virtual personality and identify the details of the specification. As a result, the information processing device 400 of the fourth embodiment can support the creation of high-quality specifications under various circumstances.

[0125] The instruction means 420 may present the set virtual personalities to the user through a user interface and accept confirmation or modification by the user. For example, the instruction means 420 may display a message on the display saying, "Currently, virtual personalities of patent examiner, image processing engineer, and security expert are being set. Are these personality settings OK?", and obtain the user's approval.

[0126] Furthermore, if the user wishes to change or add a virtual personality, the receiving means 410 can receive the request. For example, if the user inputs "I would like the perspective of a patent attorney to be added," the instruction means 420 can add a virtual personality of a patent attorney and update the instructions to the LLM.

[0127] Furthermore, the instruction means 420 can automatically select the optimal combination of virtual personalities for each item in the specification. For example, for the item "Effects of the Invention," the virtual personalities of a patent examiner and an inventor can be combined to describe the effects of the invention from multiple perspectives.

[0128] In this way, the information processing device 400 of the fourth embodiment can flexibly set a virtual personality according to the user's intentions and the characteristics of the invention, and specify the contents of the specification, which enables the creation of a more appropriate and comprehensive specification and maximizes the value of the invention.

[0129] Although the fourth embodiment of the present disclosure has been described above, the present disclosure is not limited to the above embodiment and various modifications are possible without departing from the spirit of the present disclosure. For example, the method of setting the virtual personality, the instructions to the LLM, the method of presenting the output results, etc. can be appropriately modified as needed.

[0130] Fifth Embodiment The information processing device of the fifth embodiment will be described with reference to Fig. 18. The information processing device of the fifth embodiment differs in part in configuration and operation from the information processing device of the first embodiment described above, but other parts may be the same as the information processing device of the first embodiment. Therefore, the following will describe in detail the parts that differ from the first embodiment, and will omit explanations of other overlapping parts as appropriate.

[0131] 18 is a block diagram showing an example of the functional configuration of the information processing device 500. In FIG. 18, the information processing device 500 includes a receiving unit 510, an instruction unit 520, and an output unit 530.

[0132] The receiving means 510 receives input of an idea for an invention and a viewpoint for writing a specification via a user interface (UI). The receiving means 510 also has a function of receiving a selection of a virtual personality by the user, which will be described later.

[0133] The instruction means 520 issues a first instruction to the large-scale language model (LLM) to create a virtual character, and a second instruction to specify the content of the specification using the virtual character, including the input viewpoint. Furthermore, the instruction means 520 also has a function to issue an instruction to have the virtual character explain the idea of ​​the invention, and a function to set a virtual character selected by the user.

[0134] The output means 530 receives output from the LLM and presents it to the user via a display, printer, etc. The output means 530 also has a function of presenting commentary by the virtual personality to the user.

[0135] A characteristic operation of this embodiment is that instruction means 520 causes the virtual personality to explain the idea of ​​the invention, and output means 530 presents the explanation to the user. Thereafter, reception means 510 receives the user's selection of a virtual personality, and instruction means 520 sets the virtual personality selected by the user.

[0136] For example, if a user inputs an invention idea such as "a face authentication system using a smartphone camera," the instruction means 520 causes multiple virtual personalities (e.g., "AI engineer," "security expert," and "UX designer") to explain the idea from their respective perspectives. The output means 530 presents these explanations to the user, and the user can select the virtual personality they feel is most appropriate.

[0137] In this way, in this embodiment, the user can select an appropriate virtual personality after checking the explanation of the virtual personality, making it possible to more efficiently identify the contents of the specification.

[0138] Next, the operation of the information processing device 500 of this embodiment will be described in more detail with reference to the flowchart of FIG.

[0139] First, the receiving means 510 receives an idea for an invention and a viewpoint for writing a specification (step S501). At this time, the receiving means 510 displays a message to the user saying, "Please enter your idea for the invention. When doing so, please include the main components, how they work together, and the expected effects. Also, if there is similar existing technology, please specify the differences from that technology.", prompting input from the user.

[0140] Next, the instruction means 520 instructs the LLM to create virtual personalities (step S502). Specifically, the instruction means 520 instructs the LLM to create multiple virtual personalities such as an "AI engineer," an "security expert," and an "UX designer."

[0141] Next, the instruction means 520 instructs the virtual personalities to explain the invention idea (step S503). This instruction includes the following content: "Each virtual personality should analyze the invention idea from the perspective of their own field of expertise and specifically explain its features, advantages, potential problems, areas for improvement, etc."

[0142] The output means 530 presents an explanation by the virtual personality to the user (step S504). For example, for an invention idea of ​​"a face authentication system using a smartphone camera," an explanation such as that shown in FIG.

[0143] Thereafter, the reception means 510 receives the selection of the virtual personality by the user (step S505). The user checks the presented explanation and selects the virtual personality that he or she feels is most appropriate.

[0144] The instruction means 520 sets the virtual personality selected by the user (step S506). For example, if the user selects "security expert," the instruction means 520 instructs the LLM to identify the contents of the specification with emphasis on the perspective of a security expert.

[0145] Next, the instruction means 520 instructs the LLM to identify the contents of the specification, including the input viewpoint and using the set virtual persona (step S507). This instruction includes the content, "From the perspective of a security expert, please describe in detail the anti-spoofing technology, the method for protecting biometric data, and other security-related functions."

[0146] The LLM performs processing based on the instructions and outputs the results (step S508). The LLM generates the content of the specification that reflects the viewpoint of the security expert.

[0147] Finally, the output means 530 receives the output from the LLM and presents it to the user (step S509). The output means 530 displays the content of the generated specification on a display so that the user can confirm it.

[0148] As described above, in this embodiment, the user can select an appropriate virtual personality after checking the explanation of the virtual personality, and then use the specialized knowledge of that virtual personality to identify the contents of the specification, thereby creating a more effective and professional specification.

[0149] According to this embodiment, a user can review explanations of an invention idea from multiple virtual personalities and select the most appropriate virtual personality, thereby enabling the content of the specification to be specified from an appropriate perspective depending on the characteristics of the invention and the technical field. This improves the quality of the specification and enables more accurate patent applications. For example, for a security-related invention, selecting the perspective of a security expert allows for a detailed description of security aspects, leading to the acquisition of stronger patent rights. Furthermore, even when a user is writing a specification for an invention in a field outside their own expertise, selecting the perspective of an appropriate expert allows for a more specialized and detailed description. In this way, this embodiment can simultaneously improve the efficiency and quality of specification writing.

[0150] In this embodiment, the instruction means 520 causes multiple virtual personalities to explain the idea of ​​the invention, and the output means 530 presents the explanation to the user. However, there are cases where the user cannot properly understand the presented explanation, or where the expertise of the virtual personalities is not suited to the idea of ​​the invention.

[0151] In this case, the instruction means 520 may suggest to the user that an additional virtual personality be created, and have the new virtual personality provide commentary. For example, a message may be displayed to the user saying, "If you cannot fully understand the presented commentary, or if you feel that you need the perspective of a more appropriate expert, you can request the creation of an additional virtual personality. What kind of expertise do you need?", and a new virtual personality may be created based on the user's input, and commentary may be provided.

[0152] The instruction means 520 may also suggest a combination of virtual personalities to the user, and allow the user to provide an explanation that integrates the viewpoints of multiple virtual personalities. For example, an explanation that combines the viewpoints of a "security expert" and a "UX designer" may be generated to analyze an invention idea from both the viewpoints of security and usability.

[0153] Furthermore, the output means 530 may present the user with questions to assess the user's level of understanding and interest in the explanations of each virtual personality, and recommend the most suitable virtual personality based on the user's answers. For example, a prompt such as "Please rate the level of understanding and interest in each explanation on a scale of 1 to 5" may be displayed, and the most suitable virtual personality may be suggested based on the user's answers.

[0154] These measures enable users to select the most appropriate virtual personality and identify the contents of a high-quality specification, even when they are unable to fully understand the virtual personality's explanation or when more appropriate expertise is required.

[0155] In this embodiment, a user input device may be used instead of the accepting unit, and the user input device may include an input device such as a keyboard, a mouse, a touch panel, or a voice recognition device.

[0156] A control unit may be used instead of the instruction means. The control unit is configured as a computer system including a CPU, memory, and storage, and can realize the functions of the instruction means by executing a software program.

[0157] A display device may be used instead of the output means, and the display device may include a visual information presentation device such as a liquid crystal display, an organic EL display, or a projector.

[0158] An alternative to large-scale language models (LLMs) is a natural language processing engine, which is a software component capable of understanding and generating natural language using machine learning algorithms.

[0159] Instead of a virtual personality, an expert knowledge model may be used, which is a knowledge base system that stores expert knowledge in a specific field in a database and has the function of generating appropriate answers to questions.

[0160] A technical analysis report may be used in place of an invention idea description. A technical analysis report is a document that analyzes and documents the technical features, marketability, feasibility, etc. of an invention idea.

[0161] The specification may be substituted for the formulation of claims.

[0162] As described above, each component of the present embodiment can be replaced with another component having an equivalent function, and the effects of the present disclosure can be similarly obtained by such replacement.

[0163] Sixth Embodiment An information processing device 600 of the sixth embodiment will be described with reference to Fig. 21. The information processing device 600 of the sixth embodiment differs in part in configuration and operation from the information processing device 100 of the first embodiment described above, but other parts may be similar to the information processing device 100 of the first embodiment. Therefore, hereinafter, only the parts that differ from the first embodiment will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate.

[0164] 21 is a block diagram showing an example of the configuration of the information processing device 600. The information processing device 600 includes a receiving unit 610, an instruction unit 620, an output unit 630, and a restriction setting unit 650.

[0165] The receiving means 610 receives an invention idea and a viewpoint for writing a specification as input via a user interface (UI). For example, the receiving means 610 receives an invention idea such as "a face authentication system using a smartphone camera" and a viewpoint such as "improving security performance" via a touch screen or keyboard.

[0166] The instruction means 620 issues a first instruction to the LLM to create a virtual personality and a second instruction including the input viewpoint to instruct the LLM to identify the contents of the description in the specification using the virtual personality. For example, the instruction may be "Create a virtual personality of a security expert, and from that viewpoint, identify the contents of the description related to improving the security performance of a face authentication system using a smartphone camera."

[0167] The output means 630 receives the output from the LLM and presents it to the user via a display, printer, etc. For example, the description "Anti-spoofing function using 3D face maps in face authentication systems" generated by the LLM is displayed to the user.

[0168] The restriction setting means 650 sets the description output by the output means 630 and not accepted by the user as a restriction for that user. For example, if the user does not want to use the term "3D face map," this term is set as a restriction.

[0169] FIG. 22 is a flowchart showing an example of the operation of the information processing device 600.

[0170] The receiving means 610 receives an idea for an invention and a viewpoint for writing a specification from a user (step S601).

[0171] The instruction means 620 instructs the LLM to create a virtual personality (first instruction) (step S602).

[0172] The instruction means 620 instructs the LLM to identify the contents of the specification using the virtual persona, including the input viewpoint (second instruction) (step S603).

[0173] The LLM performs the processing based on the instruction and outputs the result (step S604).

[0174] The output means 630 receives the output from the LLM and presents it to the user (step S605).

[0175] The user determines whether or not to accept the presented description (step S606). If the user accepts the description (step S606: Yes), the process proceeds to step S610, and if the user does not accept the description (step S606: No), the process proceeds to step S607.

[0176] The restriction setting means 650 sets the description that the user did not accept as a restriction (step S607). For example, a restriction such as "Do not use the term '3D face map'" may be set.

[0177] The instruction means 620 issues a second instruction in consideration of the set restrictions (step S608). For example, the instruction may be "identify the description content regarding the improvement of the security performance of the face authentication system without using the term '3D face map'."

[0178] The LLM performs processing based on the new instructions and outputs the results (step S609). For example, it generates a new description such as "Anti-spoofing function that analyzes three-dimensional facial information using facial images taken from multiple angles."

[0179] Subsequently, the output means 630 receives the new output from the LLM and presents it to the user (step S610).

[0180] In this way, the information processing device 600 of this embodiment can efficiently create a specification that meets the user's intentions by setting the content of the specification that the user does not accept as a restriction and specifying the content of the specification taking into account the restriction.

[0181] In this embodiment, an example has been shown in which the restriction setting means 650 automatically sets restrictions, but a configuration in which the user explicitly selects restrictions is also possible. For example, the output means 630 may display a list of descriptions that the user has rejected along with a message asking "Do you want to set the following descriptions as restrictions?", and the user may be able to select from the list the items that they want to set as restrictions.

[0182] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the present disclosure.

[0183] The present disclosure may be realized by recording a program for realizing the above-described functions on a computer-readable recording medium, and reading and executing the program recorded on the recording medium into a computer system.

[0184] The term "computer system" here includes hardware such as the OS and peripheral devices.

[0185] In addition, "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, CD-ROMs, and storage devices such as hard disks built into computer systems.

[0186] Furthermore, "computer-readable recording medium" may also include something that dynamically stores a program for a short period of time, such as a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line, or something that stores a program for a certain period of time, such as volatile memory inside a computer system that serves as the server or client in that case.

[0187] Furthermore, the above program may be one that realizes part of the above-mentioned functions, or may be one that can realize the above-mentioned functions in combination with a program already recorded in a computer system, or may be one that is realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0188] Although an embodiment of the present invention has been described above in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention.

[0189] Some or all of the above embodiments can be described as, but are not limited to, the following supplementary notes. (Appendix 1) a receiving means for receiving an idea for an invention and a viewpoint for writing a specification as an input; an instruction means for issuing a first instruction to a language model to create a virtual character, and a second instruction including the input, to specify the contents of the description of the specification from the viewpoint using the virtual character; output means for receiving and outputting an output in response to the instruction from the language model; An information processing device comprising: (Appendix 2) The viewpoint is at least one of the following: the key points of the invention, the limitations on the specification, and the requirements for the specification. 2. The information processing device according to claim 1. (Appendix 3) The invention further includes a third instruction generating means for instructing the creation of a specification based on the description of the specification from the viewpoint. 10. The information processing device according to claim 1 or 2. (Appendix 4) The virtual personalities are those of a patent attorney and an inventor, The second instruction includes an instruction to conduct a simulated interview between the virtual personalities. 10. The information processing device according to claim 1 or 2. (Appendix 5) The second instructions include instructions to accept prior art documents as input and to have a discussion between the virtual personalities based on the prior art documents. 10. The information processing device according to claim 1 or 2. (Appendix 6) The virtual personality includes a personality that is skilled in identifying vulnerabilities in the system constituting the invention, The second instruction includes a virtual personality pointing out a vulnerability of the system and an instruction for a remedy. 10. The information processing device according to claim 1 or 2. (Appendix 7) The first instruction includes an instruction to set a virtual personality according to the item of the specification. 10. The information processing device according to claim 1 or 2. (Appendix 8) The first instruction includes an instruction to set a virtual personality according to the technical field to which the invention idea belongs. 10. The information processing device according to claim 1 or 2. (Appendix 9) The first instruction is an instruction to cause the virtual persona to explain the idea of ​​the invention; and instructions for setting the virtual personality based on a selection of the virtual personality input by a user who has been presented with the commentary. 10. The information processing device according to claim 1 or 2. (Appendix 10) The information processing device further comprises a restriction setting means for setting the content of the description output by the output means and not accepted by the user as a restriction for the user, the instruction means issues the second instruction using the restriction. 10. The information processing device according to claim 1 or 2. (Appendix 11) the restriction setting means sets the description content that has been output by the output means and not accepted by the user and that has been selected by the user as a restriction for the user; 11. The information processing device according to claim 10. (Appendix 12) It accepts inputs from the inventor's ideas and perspectives on specification writing, issuing a first instruction to a language model to create a virtual persona, and a second instruction including the input, to specify the content of the description of the specification from the viewpoint using the virtual persona; receiving and outputting an output for the instruction from the language model; Information processing methods. (Appendix 13) On the computer, A process for receiving an invention idea and a viewpoint for writing a specification as input; a process of executing a first instruction to instruct a language model to create a virtual persona, and a second instruction including the input, to instruct the virtual persona to identify the contents of the description of the specification from the viewpoint; receiving and outputting an output from the language model in response to the instruction; A program that executes the following.

[0190] Note that part or all of the configurations described in Supplementary Notes 2 to 11, which are dependent on Supplementary Note 1, may also be dependent on Supplementary Notes 12 and 13 in the same dependent relationship as Supplementary Notes 2 to 11. Furthermore, not limited to Supplementary Notes 1, 12, and 13, part or all of the configurations described as Supplements may be made dependent on various hardware, software, various recording means for recording software, or systems, within the scope of each of the above-mentioned embodiments. [Explanation of symbols]

[0191] 100, 200, 300, 400, 500, 600 Information processing equipment 110, 210, 310, 410, 510, 610 Reception methods 120, 220, 320, 420, 520, 620 Indication means 130, 230, 330, 430, 530, 630 Output means 240 Creation Method 650 Restriction setting means

Claims

1. a receiving means for receiving as inputs an idea for an invention, a viewpoint for writing a specification, and a virtual personality designated by a user; instruction means for issuing to a language model a first instruction to create the virtual personality and a second instruction to specify the content of the specification based on the idea and the viewpoint of the invention using the virtual personality; and an output unit that receives an output corresponding to the instruction from the language model and outputs the output.

2. The information processing device according to claim 1 , wherein the viewpoint is at least one of the key points of the invention, limitations on writing a specification, and requirements on writing a specification.

3. 3. The information processing apparatus according to claim 1, further comprising a creating unit that issues a third instruction to create a specification based on the description of the specification from said viewpoint.

4. The virtual personalities are those of a patent attorney and an inventor, The information processing device according to claim 1 , wherein the second instruction includes an instruction to conduct a simulated interview between the virtual personalities.

5. The information processing device according to claim 1 , wherein the second instruction includes an instruction to accept prior art documents as input and to have the virtual personalities conduct a discussion based on the prior art documents.

6. a restriction setting means for setting the content of the specification output by the output means and not accepted by the user as a restriction for the user; The information processing apparatus according to claim 1 , wherein the instruction means issues the second instruction using the restriction.

7. 7. The information processing device according to claim 6, wherein the restriction setting means sets the content of the statement selected by the user that was output by the output means and not accepted by the user as a restriction for the user.

8. A computer comprising: The invention idea, the viewpoint of the specification drafting, and the virtual personality designated by the user are accepted as inputs; a first instruction to create the virtual personality and a second instruction to identify the content of the specification based on the idea and the viewpoint of the invention using the virtual personality, to a language model; receiving and outputting an output for the instruction from the language model.

9. On the computer, A process for receiving as input an invention idea, a specification drafting viewpoint, and a virtual persona designated by a user; a process of issuing to a language model a first instruction to create the virtual personality and a second instruction to identify the content of the specification based on the idea and the viewpoint of the invention using the virtual personality; receiving an output from the language model in response to the instruction and outputting the output.

Images