Information processing device, information processing method, and program

Abstract

This invention provides an information processing device that can reduce the time required to complete a specification. [Solution] The information processing device includes a receiving unit that receives information about the invention and feedback regarding the first specification; an instruction unit that gives a language model a first instruction that instructs the creation of the first specification based on the information about the invention and a second instruction that instructs the creation of a second specification which is more detailed than the first specification based on the feedback; and an output unit that outputs the results generated by the language model in response to the first and second instructions.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] In the process of filing a patent application, a technique of creating an application specification using a Large Language Model (LLM) is known. By using the LLM, it is possible to automatically generate a specification based on information related to the invention input by a user. Patent Document 1 discloses a specification creation support apparatus that can significantly reduce the labor of a specification creator.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of Invention

Problem to be Solved by the Invention

[0004] In a specification creation system using an LLM, the generated specification may differ from the user's intention. In such a case, the user has to perform the work of checking the completed specification and pointing out corrections, which may take more time than necessary.

[0005] An object of the present disclosure is to provide an information processing apparatus capable of solving the above-described problems.

Means for Solving the Problem

[0006] This disclosure provides an information processing device comprising: information relating to an invention; receiving means for receiving feedback relating to a first specification; first instructions for instructing a language model to prepare the first specification based on the information relating to the invention; second instructions for preparing a second specification that is more detailed than the first specification, based on the feedback; and output means for outputting the results generated by the language model in response to the first and second instructions.

[0007] Furthermore, this disclosure provides an information processing method that receives information about an invention and feedback regarding a first specification, gives a first instruction to a language model instructing it to create the first specification based on the information about the invention, and gives a second instruction based on the feedback instructing it to create a second specification that is more detailed than the first specification, and outputs the results generated by the language model in response to the first and second instructions.

[0008] Furthermore, this disclosure provides a program that causes a computer to perform the following processes: a process of receiving information about an invention and feedback about a first specification; a first instruction that instructs a language model to create the first specification based on the information about the invention; a second instruction that instructs a language model to create a second specification that is more detailed than the first specification, based on the feedback; and a process of outputting the results generated by the language model for the first and second instructions. [Effects of the Invention]

[0009] This disclosure can reduce the time required to complete the specification. [Brief explanation of the drawing]

[0010] [Figure 1] This is a block diagram showing the hardware configuration of the information processing device relating to this disclosure. [Figure 2] Block diagram showing another example of the information processing device related to this disclosure. [Figure 3]This diagram illustrates an example of how the reception desk receives information about inventions through a chat-style UI. [Figure 4] This diagram illustrates an example of how the reception desk receives feedback through a chat-style UI. [Figure 5] This flowchart shows another example of the processing procedure performed by the information processing device related to this disclosure. [Figure 6] Block diagram showing another example of the information processing device related to this disclosure. [Figure 7] This flowchart shows another example of the processing procedure performed by the information processing device related to this disclosure. [Figure 8] Block diagram showing another example of the information processing device related to this disclosure. [Figure 9] This flowchart shows another example of the processing procedure performed by the information processing device related to this disclosure. [Figure 10] Block diagram showing another example of the information processing device related to this disclosure. [Figure 11] This figure shows an example of a query that outputs whether to apply general technology to describe areas where information is lacking. [Figure 12] This flowchart shows another example of the processing procedure performed by the information processing device related to this disclosure. [Figure 13] Block diagram showing another example of the information processing device related to this disclosure. [Figure 14] This flowchart shows another example of the processing procedure performed by the information processing device related to this disclosure. [Modes for carrying out the invention]

[0011] (Hardware configuration) The embodiments of this disclosure will be described below with reference to the drawings. First, the hardware configuration for realizing the information processing device according to this disclosure will be described. Figure 1 is a block diagram showing the hardware configuration of the information processing device. The information processing device is realized by a computer 10.

[0012] 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 performing data input and output, and a communication interface 16 for network connection. The respective components are connected via a bus 17.

[0013] The processor 11 operates an operating system to control the entire computer. Examples of the processor 11 include a CPU (Central Processing Unit), a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), and the like. The processor 11 loads, for example, a program stored in the ROM 12, the storage device 14, or the like. Then, the processor 11 executes each process coded in the program. The processor 11 may execute the processes or instructions in the illustrated flowchart based on the program.

[0014] The ROM 12 stores application programs, programs according to the respective embodiments, and the like. Further, the RAM 13 is used as a work area for the processor 11.

[0015] Examples of the storage device 14 include a semiconductor memory such as a flash memory, an HDD (Hard Disk Drive), and the like. The storage device 14 stores, for example, an OS (Operating System) program, an application program, a program according to each embodiment, and the like.

[0016] The input / output interface 15 is connected to a peripheral device (not shown). The connection method may be a wired network or a wireless network.

[0017] The communication interface 16 is connected to a communication network (not shown), such as a LAN (Local Area Network) or WAN (Wide Area Network), via a wireless or wired network. The communication network may consist of multiple communication networks. This allows the computer to connect to an external large-scale language model server (LLM) via the communication network. The computer 10 may have components other than those shown in Figure 1. For example, the computer 10 may have a drive device. For example, the processor 11 may be mounted on a drive device or the like and read programs and data stored on a non-temporary tangible recording medium into the RAM 13.

[0018] The computer 10 implementing the information processing device may be configured to include some of the components described in Figure 1. For example, the computer 10 implementing the information processing device may be configured to include only the processor 11, ROM 12, and RAM 13 from the components described above. In this case, the storage device 14, input / output interface 15, and communication interface 16 may each be provided as external devices. In addition, some of the arithmetic functions may be implemented by an external server or cloud. The language model 101 is provided outside the information processing device 100 and is connected to the information processing device 100 via a communication network. However, the information processing device 100 may have the language model 101 internally, and the specific form of connection is not limited.

[0019] When using LLM to create a patent application specification, the output specification may differ from what was expected. In this case, the user has to review the completed specification and point out any necessary corrections, which can be time-consuming.

[0020] The information processing device 100 relating to this disclosure, when preparing a specification using an LLM, gives instructions to the LLM to create a preliminary, simplified specification. The information processing device then receives a specification preparation request from the user based on the content of the preliminary, simplified specification. Based on the received request, the information processing device 100 gives detailed specification preparation instructions to the LLM to prepare the specification. This makes it possible to prepare a specification that conforms to the user's intentions and reduces the time required to complete the specification.

[0021] In the following, the first specification is a provisional specification created at the initial stage of specification preparation by the information processing device 100. The first specification only needs to allow the user to grasp the general outline and direction of the specification's description, and does not need to satisfy the description requirements under the Patent Law. In other words, the first specification does not need to contain all the items of the specification, and the explanation may be partial. The first specification may also be called a provisional specification.

[0022] The second specification is a specification prepared based on the first specification. The second specification is a specification to which additions and modifications have been made to the first specification in accordance with the user's intentions, based on instructions from the user who referred to the first specification. It is assumed that the second specification satisfies the description requirements and other matters to the extent that it can be used in a patent application as is, or with further necessary additions and modifications. The second specification may also be called this specification.

[0023] Figure 2 shows an example of an information processing device 100. The information processing device 100 includes a receiving unit 110, an instruction unit 120, and an output unit 130. The information processing device 100 is also connected to a language model 101. In the example shown in Figure 1, the language model 101 is located outside the information processing device 100 and is connected to the information processing device 100 via a communication network. However, the information processing device 100 may also have the language model 101 built inside, and the specific configuration of the connection is not limited.

[0024] The reception unit 110 is a reception means for receiving information about the invention and feedback regarding the first specification from the user. The information about the invention is information that shows the content of the invention and is used to prepare the first specification. The information about the invention is not particularly limited in format or detail, as long as it is sufficient to understand the content of the invention. The information about the invention may be an invention description that explains the specific structure and operation of the invention, or it may be an idea summary that briefly shows the gist of the invention.

[0025] The reception unit 110 receives information about the invention through a user interface (UI). For example, the reception unit 110 displays the message "Please enter a summary of the invention" in a chat-style UI and accepts text input from the user. Figure 3 shows an example of the reception unit 110 receiving information about the invention through a chat-style UI.

[0026] The reception unit 110 may allow users to input information about the invention in a step-by-step manner. For example, the reception unit 110 may sequentially display specific questions in a chat-style UI, such as "What problem does this invention aim to solve?", "How does it solve that problem?", and "What are the effects of this invention?", and accept responses from the user. In this way, even if the user cannot adequately articulate the outline of the invention or is unfamiliar with using technical terms, appropriate invention information can be entered.

[0027] Feedback regarding the first specification is feedback that allows users to make corrections or additions to the first specification output by the output unit 130, as described later. The feedback is created by the user after they have viewed the outputted first specification. The reception unit 110 accepts feedback such as the content to be added or corrected, and the location where the addition or correction should be made. Figure 4 shows an example of the reception unit 110 receiving feedback through a chat-style UI.

[0028] Feedback may also be evaluation information. Evaluation information may include, for example, evaluation scores for each part of the specification or checklists of areas for improvement, and may be presented in a more structured format.

[0029] The reception unit 110 receives feedback on the first specification, for example, via a chat-style UI. The reception unit 110 may accept the feedback all at once, or it may accept it by having the user input it step by step.

[0030] Furthermore, the reception unit 110 may accept information and feedback regarding the invention from a UI that allows general text input, not just a chat-style UI. The reception unit 110 may also accept information and feedback regarding the invention from voice input or handwriting input, and is not particularly limited.

[0031] The instruction unit 120 is an instruction means that gives the language model 101 a first instruction that instructs the creation of a first specification based on information about the invention, and a second instruction that is based on feedback and instructs the creation of a second specification that is more detailed than the first specification. The first and second instructions are represented, for example, as prompts to be input to the language model 101.

[0032] The first instruction instructs the preparation of the first specification and includes, but is not limited to, referring to information based on the invention, instructing the preparation of descriptions for items to be included in the first specification, and specifying constraints on the content or format of the description. The first instruction is entered into language model 101 along with the information based on the invention. The information about the invention and the first instruction may be entered into language model 101 as a single prompt or as multiple prompts. A specific example of the first instruction is, "Based on the entered invention summary, prepare a specification including the following items: 'Background Art,' 'Problem to be Solved by the Invention,' 'Means for Solving the Problem,' and 'Effects of the Invention.' Each item should be described in 200 characters or more and include at least one specific example." However, it is not particularly limited.

[0033] Furthermore, the first instruction may consist of multiple instructions, each corresponding to a specific item in the specification (background art, problem, solution, effect). In this case, the instruction unit 120 may execute one of the individual instructions, wait for a response from the language model 101 to the transmitted instruction, and then execute the next instruction. If the first instruction consists of multiple instructions, the individual instructions are not limited to the examples described above and may be divided as appropriate.

[0034] The second instruction directs the preparation of the second specification. The second instruction includes, but is not limited to, referring to the first specification and feedback, instructing additions or modifications to the first specification in response to feedback, including matters not included in the first specification, and specifying constraints on the content or format of the description. The second specification and the second instruction may be entered into the language model 101 as a single prompt or as multiple prompts.

[0035] A concrete example of a second instruction would be: "Based on user feedback, revise the first specification and create a more detailed second specification. Pay particular attention to improving the areas pointed out in the feedback." The second instruction may be more specific.

[0036] Furthermore, similar to the first instruction, the second instruction may consist of multiple instructions separated by the item of the specification (background art, problem, solution, effect). In this case, the instruction unit 120 may issue one of the individual instructions, wait for the response of the language model 101 to the transmitted instruction, and then issue the next instruction. If the second instruction consists of multiple instructions, the individual instructions are not limited to the examples described above and may be separated as appropriate.

[0037] Furthermore, the first and second instructions are not limited to prompts. For example, tokenized information equivalent to the prompts described above may be input to the language model 101 as either the first or second instruction.

[0038] Furthermore, the language model 101 is, for example, various general language models, but is not particularly limited. The instruction unit 120 may give the first instruction and the second instruction to the same language model, or to different language models.

[0039] The output unit 130 is an output means that outputs the generation results of the language model 101 in response to the first and second instructions. When the instruction unit 120 gives the first instruction, the generation result from the language model 101 is the first specification. That is, in this case, the output unit 130 outputs the first specification. Similarly, when the instruction unit 120 gives the second instruction, the generation result from the language model 101 is the second specification. Therefore, in this case, the output unit 130 outputs the second specification. The output first and second specifications are displayed through a display, printer, or the like.

[0040] Furthermore, when outputting the second specification, the output unit 130 may output it in a way that allows for comparison with the first specification. This makes it easier to determine whether the user's feedback regarding the first specification has been reflected.

[0041] The operation of the information processing device 100 of the present invention will be explained with reference to Figure 5. Figure 5 is a flowchart showing the procedure (information processing method) of the processing performed by the information processing device 100.

[0042] First, the reception unit 110 receives information about the invention from the user through the user interface (UI) (step S101).

[0043] Next, the instruction unit 120 gives a first instruction to the language model 101 to prepare the first specification, based on the information about the invention received by the reception unit 110 (step S102).

[0044] When the language model 101 generates the first specification, the output unit 130 outputs the first specification generated by the language model 101 (step S103).

[0045] After the first specification is presented, the reception unit 110 receives feedback from the user regarding the first specification (step S104).

[0046] After receiving the feedback, the instruction unit 120 gives a second instruction to the language model 101, instructing it to prepare a second specification based on the feedback (step S105).

[0047] The output unit 130 outputs the second specification generated by the language model 101 (step S106). This allows the user to obtain a more complete specification through a two-step specification creation process starting from the initial input.

[0048] In this embodiment, as an example, the information processing device 100 accepts user input and starts the operation of step S101. For example, when the user clicks the "Start Specification Creation" button through the UI, the series of processes shown in the flowchart of Figure 2 are started. Furthermore, the information processing device 100 may automatically terminate the process when the output of the second specification in step S106 is completed. Moreover, the operations from step S101 to step S103, or from step S104 to step S106, may be repeated.

[0049] According to the information processing device 100 of this embodiment, the user can quickly obtain the first specification and provide feedback, thereby efficiently creating a more complete second specification. This reduces the time required to complete the specification. In other words, it reduces the time required for reviewing the completed specification and identifying areas for correction.

[0050] For example, using the information processing device 100 of this embodiment, the user inputs information about the invention that represents an outline of the invention, and the language model 101 generates a first specification in accordance with the first instruction of the instruction unit 120. By referring to this first specification, the user can easily grasp the outline and direction of the description.

[0051] Furthermore, users can provide feedback on the first specification to obtain a more detailed and accurate second specification. Users do not need to review the entire completed second specification from scratch and point out any necessary corrections; they only need to focus their feedback on the shortcomings and improvements of the first specification. This reduces the time and effort required to prepare the specification.

[0052] In addition, in response to a second instruction from the instruction unit 120, the language model 101 generates a second specification that reflects the user's feedback, thereby enabling the creation of a specification that aligns with the user's intentions. This reduces the time required for reviewing the completed specification and identifying areas for correction.

[0053] The second embodiment will be described below.

[0054] The information processing apparatus of the second embodiment will be described with reference to the drawings. Note that the information processing apparatus of the second embodiment differs from the information processing apparatus of the first embodiment described above in some configurations and operations, while other parts may be the same as those of the information processing apparatus of the first embodiment. Therefore, the parts that differ from the first embodiment will be described in detail below, and other overlapping parts will be omitted as appropriate.

[0055] Figure 6 is a block diagram showing an example of the information processing device 200. The information processing device 200 comprises a receiving unit 210, an instruction unit 220, and an output unit 230. The receiving unit 210, instruction unit 220, and output unit 230 include the functions of the receiving unit 110, instruction unit 120, and output unit 130 of the first embodiment. Also, similar to the example of the information processing device 100, the information processing device 200 is connected to a language model 201.

[0056] The reception unit 210 receives information about the invention and feedback regarding the first specification from the user through the user interface (UI).

[0057] The reception unit 210 may accept specifications regarding the level of detail of the first specification. These specifications identify the desired level of detail for the first specification. The level of detail indicates the degree of detail in the content of the first specification to be prepared. The level of detail can be expressed in three stages, for example, "detailed," "standard," and "summary." However, the expression of the level of detail is not limited to these; for example, it may be expressed in five stages, such as "very detailed," "detailed," "standard," "summary," and "abbreviated," and is not particularly restricted. The level of detail may also be expressed in text.

[0058] Furthermore, the description granularity may include a specification that allows the language model 201 to analyze the content of the invention and automatically adjust the granularity. In this case, the first instruction described later will include an instruction to the language model 201 to set the granularity to match the content of the invention. This makes it easier to create the first specification with granularity appropriate to the content of the invention, even if it is difficult for the user to specify an appropriate granularity.

[0059] For example, if the reception unit 210 receives a "standard" level of detail from the user, the basic function of each component will be explained in about 50 characters. For instance, the explanation for the reception unit 210 might look like this: "The reception unit 210 receives invention information and feedback via a user interface, and also has the function of enabling the selection of detail level and description items." Note that the number of characters and level of detail in the explanation for each level of detail are not limited to these examples and can be determined as appropriate.

[0060] The reception unit 210 displays, for example, the following in a chat-style UI to accept user specifications regarding the level of detail in the description.

[0061] Please select the granularity described in the first specification. 1. Details (Explain the operation of each component in 100 characters or more) 2. Standard (Explain the basic function of each component in about 50 characters) 3. Overview (Explain the overall system operation in approximately 200 characters)

[0062] The reception unit 210 may accept user instructions regarding the level of detail via a pull-down or button-type UI. The reception unit 210 may also accept instructions regarding the level of detail expressed as text.

[0063] The reception unit 210 may accept specifications regarding the items to be included in the first specification. Specifications regarding the items to be included are those that the user wishes to include in the first specification. These items are, for example, those specified in the specification format. However, the items to be included here are not limited to the above examples and may specify more specific content, such as a description of the configuration or operation in a mode for carrying out the invention.

[0064] For example, the reception unit 210 displays the following options in a chat-style UI and accepts the user's specification of the items to be included in the specification. The options may be something like, "Please select the items to be included in the first specification (multiple selections are possible): □ Background art, □ Problems to be solved by the invention, □ Means for solving the problems, □ Effects of the invention, □ Modes for carrying out the invention, □ Industrial applicability." The items to be included that the reception unit 210 accepts are predetermined, for example, in an item list, but are not limited to these, and users may add custom items or rearrange the order of items.

[0065] The reception unit 210 may also accept a request to generate a drawing outline.

[0066] During the preparation of the first specification, the general outline of the drawings to be created may be determined, but the drawings themselves may not have been created. In this case, it may be difficult to explain the contents of the drawings in detail, or there may be discrepancies between the detailed explanation of the drawings and the contents of the drawings that are actually created. The instruction unit 220 issues a first instruction to the language model 201, which includes an instruction to output an outline of the drawings to be created in the first specification, making it easier to include an appropriate explanation of the drawings in the second specification. In addition, by omitting the explanation of the drawings, it becomes easier to review the first specification.

[0067] For example, the reception unit 210 displays a checkbox in the UI to specify whether to generate an image summary and accepts the specification to generate a drawing summary. The reception unit 210 is not limited to this specification and may accept specifications in other formats. The reception unit 210 may also input the received information about the invention into the language model 201 to analyze the content of the invention and recommend appropriate items.

[0068] The instruction unit 220 issues a first instruction to the language model 201 to prepare the first specification, and a second instruction to the language model 201 to prepare the second specification.

[0069] The first instruction includes instructions for creating the first specification corresponding to the received level of detail and description items. For example, if the reception unit 210 receives both the level of detail and description items, the instruction unit 220 issues the following first instruction to the language model 201: "Based on the entered invention summary, please create the first specification including the following items: 'Background Art,' 'Problem to be Solved by the Invention,' 'Means for Solving the Problem,' and 'Effects of the Invention.' The level of detail should be 'Standard,' and the basic function of each component should be explained in approximately 50 characters."

[0070] The instruction unit 220 may issue a first instruction to the language model 201, which includes an instruction to generate a drawing outline, in response to instructions from the user received by the reception unit 210. The instruction unit 220 may issue a first instruction to the language model 201, which includes an instruction to omit a detailed description of the drawings when preparing the first specification, and instead output a drawing outline that is intended to be included in the second specification.

[0071] Instructions for generating a drawing summary might be something like, "Also, omit detailed descriptions of the drawings and instead output only the drawing summary." Alternatively, instructions for generating a drawing summary could specify more detail for each individual drawing to be created, such as, "In the first specification, omit detailed descriptions of the drawings and instead output only the following drawing summaries: Figure 1: Block diagram of the entire system, Figure 2: Flowchart for creating the first specification, Figure 3: Flowchart for creating the second specification. Briefly describe the contents of each figure in one or two sentences."

[0072] Furthermore, the summary of the drawings may be described as follows: "Figure 1: Block diagram showing the overall configuration of the information processing device 200. It illustrates the relationship between the reception unit 210, the instruction unit 220, and the output unit 230. Figure 2: Flowchart of the first specification creation process. It illustrates the flow from the input of invention information to the output of the first specification. Figure 3: Flowchart of the second specification creation process. It illustrates the flow from the input of feedback to the output of the second specification." However, the summary of the drawings is not limited to these examples; any summary that corresponds to the instructions is acceptable.

[0073] The output unit 230 outputs the generation results from the language model 201 for the first and second instructions. That is, the output unit 230 outputs the first specification when the instruction unit 220 gives the first instruction, and outputs the second specification when the instruction unit 220 gives the second instruction. The first specification includes the specified items at the specified level of granularity according to the received specification, and may also include an outline of the drawings. The output first and second specifications are displayed via a display, printer, or the like.

[0074] Next, the operation of the information processing device 200 will be described. Figure 7 is a flowchart showing the operation of the information processing device 200.

[0075] First, the reception unit 210 receives information about the invention from the user (step S201). At this time, the reception unit 210 may receive instructions from the user regarding the level of detail of the first specification, or instructions from the user regarding the items to be included in the specification.

[0076] Next, the instruction unit 220 issues a first instruction to the language model 201 instructing it to prepare the first specification (step S202). The first instruction may include an instruction to output a summary of the drawings to the first specification.

[0077] Next, the output unit 230 outputs the first specification generated by the language model 201 in response to the first instruction (step S203). The output first specification has a granularity corresponding to the description granularity received in step S201, and contains items corresponding to the description items. In addition, the first specification includes a summary of the drawings in accordance with the first instruction.

[0078] Subsequently, the reception unit 210 receives feedback from the user regarding the first specification (step S204).

[0079] Next, the instruction unit 220 gives a second instruction to the language model 201, instructing it to prepare a second specification based on the feedback (step S205). In step S205, if the first instruction includes an instruction to generate a drawing outline, the second instruction may include an instruction to include a detailed description of the drawing.

[0080] The output unit 230 outputs the second specification generated by the language model 201 in response to the second instruction (step S206).

[0081] According to this embodiment, by using the granularity and items of description received by the reception unit 210, a first specification focusing on the necessary level of detail and items is created based on the first instruction from the instruction unit 220. This suppresses the creation of a first specification that is unnecessarily detailed and contains many items, thus reducing the time required to create the first specification. Furthermore, by referring to the first specification, which contains the necessary items at the desired level of detail, the user can easily consider feedback for creating the second specification. In other words, with the information processing device 200 of this embodiment, it becomes easier to obtain a final specification that more accurately reflects the user's intentions.

[0082] Furthermore, by generating detailed descriptions of the drawings in the second specification, the description in the first specification is simplified, making it easier for the user to understand the contents of the first specification. This makes it easier for the user to consider feedback for creating the second specification. In other words, the information processing device 200 of this embodiment can reduce the time required to complete the specification. Moreover, the information processing device 200 of this embodiment makes it easier to obtain a final specification that more accurately reflects the user's intentions.

[0083] Furthermore, a first specification, including an outline of the drawings, is generated in response to user instructions received by the reception unit 210. Including an outline of the drawings in the first specification makes it easier to include an appropriate description of the drawings in the second specification. Also, omitting the description of the drawings makes it easier to review the first specification.

[0084] The third embodiment will be described below.

[0085] The information processing apparatus of the third embodiment will be described with reference to the drawings. Note that the information processing apparatus of the third embodiment differs in some configurations and operations from the information processing apparatuses of the first and second embodiments described above, while other parts may be the same as those of the information processing apparatuses of the first and second embodiments. Therefore, the parts that differ from the first and second embodiments will be described in detail below, and other overlapping parts will be omitted as appropriate.

[0086] Figure 8 is a block diagram showing an example of the information processing device 300. The information processing device 300 comprises a receiving unit 310, an instruction unit 320, and an output unit 330. These components have the same basic functions as in the first and second embodiments, and the instruction unit 320 includes additional functions. Also, similar to the example of the information processing device 100, the information processing device 300 is connected to a language model 301.

[0087] The instruction unit 320 analyzes the content of the feedback regarding the first specification received by the reception unit 310. Then, based on the analysis results, the instruction unit 320 gives a second instruction to the language model 301 to create the second specification.

[0088] As explained earlier, feedback on the first specification is intended to allow users to make revisions or additions to the first specification. This feedback may include specific points such as "the description of the background art is insufficient," or direct revisions of the text.

[0089] The instruction unit 320 analyzes the content of the feedback regarding the first specification in order to identify the parts and specific contents that need to be amended or added to the first specification. For example, as part of the analysis of the content of the feedback, the instruction unit 320 analyzes, for example, the revision intention of the user who entered the feedback. The revision intention is the user's intention regarding the parts and specific contents that need to be amended or added to the first specification. For example, if the instruction unit 320 receives feedback that "the description of background art is insufficient," it analyzes this as including the revision intention that "the section on background art needs to be expanded to include more details and specific examples."

[0090] Furthermore, the instruction unit 320 may analyze whether the content of the feedback regarding the first specification is applicable to areas other than those pointed out. For example, if a correction instruction for a specific section of the description is applicable to other similar descriptions, the instruction unit 320 will analyze whether the correction instruction included in the feedback is applicable to other descriptions. In this case, the scope of applicability may be the entire specification, or it may be the same description item as the correction instruction included in the feedback. For example, if the point raised in the feedback concerns Embodiment 1, the instruction unit 320 may analyze whether the point raised is applicable to the description items of Embodiments, such as Embodiment 2 and Embodiment 3. By applying the content of the feedback to the entire specification, the description of the specification as a whole can be improved. Also, by applying the content of the feedback to the same item, the possibility of unexpected descriptions is reduced.

[0091] The instruction unit 320 analyzes the content of the feedback by, for example, giving the language model 301 a prompt instructing it to analyze the feedback regarding the first specification and the intention to modify the feedback. The instruction unit 320 may also analyze the content of the feedback using other machine learning methods, rule-based analysis methods, or text analysis methods based on natural language processing technology.

[0092] Furthermore, for example, feedback regarding the first specification may be vague or insufficient, such as "please write in more detail." Another example is when the content of the feedback regarding the first specification contradicts itself, such as when some part of a relevant description is given feedback to be "more concise," while other related descriptions are given feedback to be "more detailed," or when revisions are requested that are inconsistent with other parts of the specification.

[0093] In this case, the instruction unit 320 may further analyze the content of the feedback using additional feedback. For example, the instruction unit 320 first evaluates the clarity and specificity of the received feedback regarding the first specification using a language model 301 or the like. If the feedback regarding the first specification is evaluated as unclear or insufficient, the instruction unit 320 generates a message requesting additional information or specific instructions from the user.

[0094] For example, if the feedback on the first specification is "please write in more detail," a message such as "Which parts should be described in more detail? Please tell us the specific items and the types of information that should be added (e.g., numerical data, specific examples, comparative explanations, etc.)" could be generated. Alternatively, a message such as "You have requested a more detailed explanation in the 'Background Art' section, but a concise description is requested in the 'Problem to be Solved by the Invention' section. To maintain overall consistency, which approach should be prioritized?" could be generated.

[0095] The instruction unit 320 further analyzes the content of the feedback using the additional feedback received in response to the message. In this case, the message to the user is output by the output unit 330. The additional feedback from the user regarding the message is received by the reception unit 310.

[0096] Based on the analysis results of the feedback regarding the first specification, the instruction unit 320 gives a second instruction to the language model 301 for creating the second specification. For example, if the instruction unit 320 analyzes that the feedback regarding the first specification includes an intention to revise the section on background art, which needs to be expanded to include more detail and specific examples, it gives a second instruction that includes details and specific examples of the background art. As another example, the second instruction may be something like, "Expand the section on background art to 500 characters or more and specifically list at least three prior art examples. For each prior art example, clearly explain its technical content and its relevance to the present invention."

[0097] The instruction unit 320 may issue a second instruction to the language model 301, which includes instructions to display the text before and after application side by side and to clarify the changes. In this case, the second instruction may include, for example, "Please revise the following text to include more specific examples. Display the text before and after revision side by side and underline the changes. Before revision: 'The present invention relates to document creation using AI.' After revision example: 'The present invention relates to the automatic creation of patent specifications using artificial intelligence (AI) technology, in particular large-scale language models (LLMs). Specifically, it relates to a system that generates high-quality patent specifications from inventor input using LLMs such as OpenAI's GPT-3 and Google's BERT.'"

[0098] The instruction unit 320 may, if the reception unit 310 has received only specific revision suggestions from the user, compare the original and revised text and analyze the user's intention to make the revisions. For example, the specific revision suggestion here refers to the first specification that the user has revised.

[0099] When outputting the second specification generated from the language model 301 in response to the second instruction, the output unit 330 may output the feedback analysis results and application results using graphs or charts. Using graphs or charts allows the user to more intuitively understand the impact of the feedback and the degree of improvement.

[0100] The operation of the information processing device 300 in this embodiment will be described with reference to the drawings. In the flowchart shown in Figure 9, the operations of steps S301 to S303 are the same as the operations of steps S101 to S103 in Embodiment 1, and a detailed explanation will be omitted.

[0101] In step S304, the reception unit 310 first receives feedback from the user regarding the first specification.

[0102] Next, in step S305, the instruction unit 320 analyzes the content of the received feedback. Based on the analysis results, the instruction unit 320 issues a second instruction to the language model 301.

[0103] Then, in step S306, the output unit 330 outputs a second specification that reflects the analyzed feedback, which is the result of generation by the language model 301.

[0104] Thus, in the third embodiment, the information processing device 300 makes it easy for the instruction unit 320 to analyze the user's feedback regarding the first specification and appropriately reflect that feedback throughout the specification. As a result, a second specification that more accurately reflects the user's intentions can be efficiently created.

[0105] The fourth embodiment will be described below.

[0106] The information processing apparatus of the fourth embodiment will be described with reference to the drawings. Note that the information processing apparatus of the fourth embodiment differs from the information processing apparatus of the first embodiment described above in some configurations and operations, while other parts may be the same as those of the information processing apparatus of the first embodiment. Therefore, the parts that differ from the first embodiment will be described in detail below, and explanations of other overlapping parts will be omitted as appropriate.

[0107] The information processing device of the fourth embodiment, in addition to the basic configuration of the first embodiment, has a function to identify and process areas where information is missing during the preparation of the first specification. Specifically, functions for identifying areas with missing information, outputting blank fields, and querying general technical applications have been added.

[0108] Figure 10 is a block diagram showing the functional configuration of the information processing device 400. As shown in Figure 10, the information processing device 400 of this embodiment includes a receiving unit 410, an instruction unit 420, and an output unit 430. The receiving unit 410, instruction unit 420, and output unit 430 each include the same basic functions as the receiving unit 110, instruction unit 120, and output unit 130 of the first embodiment. Also, similar to the example of the information processing device 100, the information processing device 400 is connected to a language model 401.

[0109] The instruction unit 420 gives a first instruction that further includes identifying areas where information is missing in the preparation of the first specification, and an instruction to output the areas where information is missing.

[0110] Areas lacking information are those where the content of the invention may become unclear due to insufficient information. For example, in the section on "Problems the invention aims to solve," if any of the following three points are missing, the description of that section in the specification may be insufficient: 1) the problems of the prior art, 2) the reasons why those problems occur, and 3) the necessity of solving them. Therefore, the first instruction includes content that identifies such areas as areas lacking information. In the example above, the first instruction might include something like, "Please identify areas lacking information by comparing them with the minimum required information list for each section. For example, in the section on 'Problems the invention aims to solve,' please check whether the following three points are included: 1) the problems of the prior art, 2) the reasons why those problems occur, and 3) the necessity of solving them." Note that areas lacking information are not limited to these; they may also include areas lacking content that could assert inventive step, areas that hinder the interpretation of rights, etc., and should be defined as appropriate.

[0111] Furthermore, the first instruction includes an instruction to output information about areas where information is missing. This information about areas where information is missing is a message that explains the items or paragraphs in the first specification that are lacking information and requests additional information, separate from the first specification. For example, a message explaining the items or paragraphs where information is missing might be a sentence like, "In the section 'Problems the invention aims to solve,' there is a lack of information about why the problem occurs." The message requesting additional information may be a message asking for information to be added to the areas where information is missing, or it may be a question about the content of the areas where information is missing.

[0112] Furthermore, information regarding areas where information is missing may be indicated as blanks in the first specification. For example, if the section "Problems the invention aims to solve" describes the problems of the prior art but does not explain why those problems occur, the first specification may include a blank like the following: "The problem the invention aims to solve is that in conventional specification creation systems using LLM, the generated specification may differ from the user's intent. The reason this problem occurs is __________." The blanks may be highlighted, or comments may be added. For example, areas where information is missing may be highlighted in yellow in the first specification, with a comment such as "Additional information is needed" displayed next to them.

[0113] Furthermore, information regarding areas where information is missing may be indicated in the first specification as a blank space, for example, along with a message such as, "In the section 'Problems the invention aims to solve,' there is insufficient information regarding the reason why the problem occurs. Please add appropriate information to the blank space." The instruction unit 420 gives the first instruction, which includes these contents.

[0114] The first instruction may further include an instruction to ask whether to apply general technical knowledge to fill in the gaps in the information. For example, the first instruction may output the message, "Do you want to fill in this gap in information by applying general technical knowledge?"

[0115] Furthermore, the first instruction may also be an instruction to inquire whether to apply general technology to the description, depending on the relationship between the part where information is lacking and the content of the invention. That is, the first instruction may be an instruction to inquire whether to apply general technology to the part of the part where information is lacking that is outside the point of the invention, and to provide a blank space or a question to the part that is related to the point of the invention. In this case, the point of the invention may be the point that the reception unit 410 has received input from the user, or it may be the point that the language model 401 has determined. That is, the first instruction may be an instruction that includes the point of the invention input by the user, or it may include an instruction that causes the language model 401 to determine whether or not it is related to the point of the invention.

[0116] The first instruction may include instructions that cause language model 401 to determine the applicability of general art in areas where information is lacking. In this case, the first instruction may include instructions to generate content to be described in the first specification and to indicate the application areas if general art is applicable.

[0117] Thus, because the first instruction is a query asking whether to apply general technical principles to the description, it becomes possible to reduce the time users spend dealing with areas where information is lacking.

[0118] The output unit 430 outputs the first specification and information regarding the areas where information is missing. The first specification and the information regarding the areas where information is missing are generated by the language model 401 in response to the first instruction given by the instruction unit 420 as described above. In this case, the output first specification may indicate the areas where information is missing as blank spaces.

[0119] Furthermore, the output unit 430 may output a query asking whether to apply general technology to the areas where information is lacking. If the first instruction is an instruction to ask whether to apply general technology to the areas where information is lacking depending on the relationship between the areas where information is lacking and the content of the invention, the output unit 430 may output a message suggesting different responses depending on the importance. For example, for areas where information is lacking that relate to the key points of the invention, the output unit 430 may output a message such as, "This area relates to an essential part of the invention, so please input as much specific information as possible." The output unit 430 may also output a suggestion such as, "We recommend supplementing this with general technical knowledge" for areas that are outside the key points of the invention. This makes it easier for users with limited technical expertise to make appropriate judgments. This message may be generated by the output unit 430 in response to the output from the language model 401, or the first instruction may include an instruction to generate the message. Figure 11 shows an example of outputting a query asking whether to apply general technology to the areas where information is lacking through a chat-style UI.

[0120] Furthermore, the reception unit 410 accepts additional information as feedback regarding the first specification, specifically addressing any areas where information is lacking. This additional information may include, but is not limited to, answers to questions that have been output, sentences to be filled in the blanks shown in the first specification, or answers to inquiries about whether to apply general art.

[0121] Furthermore, when the receiving unit 410 receives additional information for areas where information is missing, the instruction unit 420 issues a second instruction to generate the second specification by including the received additional information and supplementing the areas where information is missing.

[0122] Furthermore, before issuing the second instruction, the instruction unit 420 may analyze the received input and evaluate the specificity and sufficiency of the information. In this case, the instruction unit 420 may issue the second instruction if it determines that the specificity and sufficiency of the information are appropriate. On the other hand, if the input information consists only of abstract expressions, the instruction unit 420 may output instructions to the user via the output unit 430 and the reception unit 410 to input more detailed information, and then issue the second instruction after receiving the information. In this case, the output unit 430 may output a suggestion such as, "Including more specific examples and numerical values ​​will improve the quality of the specification. Would you like to add specific examples?" This improves the quality of the user's input and enables the creation of a more comprehensive specification.

[0123] The output unit 430 outputs the second specification generated by the language model 401 in response to the second instruction. In this case, the second specification is a specification generated by supplementing the parts where information is missing.

[0124] Next, the operation of the information processing device of this embodiment will be described. Figure 12 is a flowchart showing the operation of the information processing device.

[0125] The reception unit 410 receives information regarding the invention (step S401). The instruction unit 420 gives a first instruction to the language model 401 instructing it to prepare a first specification based on information relating to the invention (step S402). In this case, the first instruction includes identifying areas where information is missing when preparing the first specification, and an instruction to output the areas where information is missing.

[0126] The output unit 430 outputs the first specification generated by the language model 401 (step S403). In step S403, the output unit 430 outputs information about the parts where information is missing from the language model 401. The first specification may also indicate the parts where information is missing as blank spaces. The output unit 430 may also output a query asking whether to apply general technology to the parts where information is missing. The query asking whether to apply general technology may be in the form of, for example, "Should we apply general technical knowledge to supplement the parts where information is missing? [Yes / No]".

[0127] The reception unit 410 receives feedback regarding the first specification (step S404). In step S404, the feedback includes additional information for areas where information is missing.

[0128] The instruction unit 420 gives a second instruction to the language model 401 instructing it to prepare the second specification (step S405). In step S405, the second instruction includes the information received in step S404 for the areas where the feedback is insufficient.

[0129] For example, suppose in step S403 the output unit 403 outputs a query asking whether to apply general art, and in step S404 the reception unit 410 receives a response that it does apply. In this case, the instruction unit 420 gives a second instruction, which includes an instruction to apply general art to fill in the missing information and generate the second specification. For example, the second instruction includes the instruction, "Regarding the missing 'Reason for the problem' in the 'Problem to be Solved by the Invention' section, please fill in the information considering the general challenges of specification creation systems using LLM." The language model 401, having received the second instruction, fills in the information based on this instruction. For example, "The reason for this problem is that it is difficult for LLM to fully understand human intent, and it is possible that LLM has not sufficiently learned the format and requirements specific to patent specifications."

[0130] On the other hand, if the reception unit 410 receives a response in step S404 indicating that it is not applicable, or if the missing information is related to the key point of the invention, the instruction unit 420 gives a second instruction to generate the second specification by supplementing the missing information based on other feedback received.

[0131] The output unit 430 outputs the second specification generated by the language model 401 in response to the second instruction (step S406).

[0132] Thus, in the fourth embodiment, the information processing device 400 has an instruction unit 420 that gives instructions to identify areas where information is missing when preparing the first specification. This makes it easier for the user to address areas where information is missing, and enables the efficient creation of a more complete specification.

[0133] Furthermore, the first instruction may include instructions to output blank spaces where information is missing during the creation of the first specification and generate a message requesting the user to provide additional information. This allows the user to easily identify missing information in the specification and efficiently add the necessary information.

[0134] Furthermore, the first instruction may include an instruction to inquire whether to apply general technology to describe areas where information is lacking. This makes it easier for the user to address areas where information is missing, thereby reducing their workload. In addition, the user may be asked to provide additional information regarding the key points of the invention, while other parts may be supplemented by applying general technology. In this way, it becomes possible to create a high-quality specification that meets the user's intentions while reducing the user's burden.

[0135] The fifth embodiment will be described below.

[0136] The information processing apparatus of the fifth embodiment will be described with reference to the drawings. Note that the information processing apparatus of the fifth embodiment differs in some configurations and operations from the information processing apparatus of the first embodiment described above, while other parts may be the same as those of the information processing apparatus of the first embodiment. Therefore, the parts that differ from the first embodiment will be explained in detail below, and other overlapping parts will be omitted as appropriate.

[0137] Figure 13 is a block diagram showing an example of an information processing device 500. The information processing device 500 comprises a receiving unit 510, an instruction unit 520, and an output unit 530. The receiving unit 510, instruction unit 520, and output unit 530 each include the same basic functions as the receiving unit 110, instruction unit 120, and output unit 130 of the first embodiment. Also, similar to the example of the information processing device 100, the information processing device 500 is connected to a language model 501.

[0138] The instruction unit 520 further provides a first instruction that includes identifying parts that are expressed abstractly when preparing the first specification and outputting a question to the user regarding those parts that are expressed abstractly.

[0139] Abstract expressions are those related to expressions whose scope or degree is ambiguous, potentially making the content of the invention unclear. Abstract expressions may include, for example, expressions exemplified in laws and standards, or expressions generally considered abstract in patent practice, but they should be determined as appropriate.

[0140] The first instruction, which includes instructions to identify abstract expressions when drafting the first specification, is as follows: "When drafting the first specification, refer to the following keyword list to identify abstract expressions, such as 'efficient,' 'rapid,' 'high quality,' 'easy to use,' and 'innovative.' Where you find these words used, generate questions that ask for their specific meaning and numerical support." Note that the specific wording of the instructions is not limited to these. Also, the examples of abstract expressions listed in this example are not limited to these. Furthermore, the specific abstract expressions listed in the first instruction are, for example, maintained in advance as a keyword list. This keyword list is created by analyzing patent law examination guidelines and past rejection notices, and automatically extracting new abstract expressions and expressions that are likely to cause problems. Machine learning techniques are used as appropriate in the analysis. The keyword list may be updated periodically.

[0141] Furthermore, the first instruction includes instructions to output questions regarding sections that contain abstract expressions. These questions are intended to obtain information necessary to clarify the scope and degree of the expressions in the sections that are deemed to contain abstract expressions.

[0142] For example, if the description states, "The present invention allows for the efficient preparation of specifications," and "efficiently" is used as an abstract expression, a possible question might be, "What specific indicators measure 'efficient preparation of specifications,' and what degree of improvement can be expected? For example, please provide specific numerical examples, such as a percentage reduction in preparation time compared to conventional methods." The first instruction includes instructions to output such a question.

[0143] The output unit 530 outputs the first specification, a list of abstract expressions, and questions regarding the abstract expressions. The first specification and the questions regarding the abstract expressions are generated by the language model 501 in accordance with the first instruction given by the instruction unit 520, as described above.

[0144] Furthermore, the reception unit 510 accepts further feedback regarding the first specification, including answers to questions about abstract expressions. The instruction unit 520 uses the answers to questions about abstract expressions to issue a second instruction to the language model 501, instructing it to prepare the second specification.

[0145] The operation of the information processing device 500 of the present invention will be explained with reference to Figure 14. Figure 14 is a flowchart showing the procedure of processing performed by the information processing device 500.

[0146] First, the reception unit 510 receives information about the invention from the user (step S501).

[0147] The instruction unit 520 gives a first instruction to the language model 501 that instructs it to prepare a first specification based on information relating to the invention (step S502). In this case, the first instruction includes instructions to identify abstract expressions and to generate questions.

[0148] The output unit 530 outputs the first specification generated by the language model 501 (step S503). In step S503, the output unit 530 outputs a list of abstract expressions and a list of questions generated by the language model 501.

[0149] The reception unit 510 receives feedback regarding the first specification (step S504). Step S504 includes the user's answers to the questions output in step S503.

[0150] The instruction unit 520 instructs the language model 501 to prepare the second specification based on the responses and feedback (step S505).

[0151] The output unit 530 outputs the second specification generated by the language model 501 (step S506).

[0152] According to this embodiment, the information processing device 500 can identify abstract expressions when preparing the first specification and encourage the user to elaborate on them. This improves the quality of the final second specification and reduces the grounds for rejection during patent examination. Furthermore, the user has the opportunity to express their invention more concretely and communicate the essence of the invention more clearly.

[0153] For example, the information processing device 500 assists the user in clearly explaining the details of the invention by having the instruction unit 520 automatically detect abstract expressions and instruct the language model 501 to generate specific questions regarding them. This makes it possible to enrich the "modes for carrying out the invention" section of the patent specification and increase the likelihood of obtaining a patent.

[0154] The information processing device 500 identifies abstract expressions and generates questions after the first specification has been prepared, but is not limited to this. For example, the instruction unit 520 may use the language model 501 to directly identify abstract expressions and generate questions for the invention information input by the user, and request answers from the user via the output unit 530. The instruction unit 520 may then be configured to give a first instruction to prepare the first specification based on the answers received.

[0155] Furthermore, in the information processing device 500, there may be cases where the user cannot provide specific explanations or numerical support for abstract expressions. In such cases, it may be difficult to concretize the abstract expressions in the second specification as well.

[0156] In this case, the instruction unit 520 may have a function to suggest alternative approaches if the user is unable to provide specific explanations or numerical values. For example, the instruction unit 520 may include the following instructions in the first or second instruction to the language model 501: "If the user is unable to provide specific explanations or numerical values, please present the following alternatives: 1. Show examples of typical improvement rates or effects in similar technical fields. 2. Use relative expressions (e.g., 'approximately twice as efficient as conventional technology'). 3. Explain qualitative descriptions in more detail (e.g., change 'efficient' to 'simultaneously achieves reduced processing time and improved accuracy')."

[0157] This makes it easier for the information processing device 500 to create specifications with improved abstract representations.

[0158] Furthermore, the instruction unit 520 may also instruct the system to provide alternatives along with experimental and measurement methods for obtaining specific data in the future. This allows the user to prepare for amendments or additions to the examples after the patent application has been filed.

[0159] In this way, the information processing device 500 can improve the quality of the specification even in situations where it is difficult to concretize abstract expressions. This makes it possible to address inventions in various technical fields and development stages.

[0160] In addition, the timing of identifying abstract expressions and generating questions in the information processing device 500 may be performed sequentially at the stage when the user inputs the invention information, not just after the creation of the first specification. For example, the instruction unit 520 may instruct the language model 501 to identify abstract expressions and generate questions before issuing the first instruction. The instruction unit 520 may then receive the user's response to this via the reception unit 510 and use the result to issue the first instruction. This makes it possible to concretize abstract expressions at an earlier stage and further improve the quality of the first specification.

[0161] Furthermore, in the information processing device 500, the instruction unit 520 may be configured to extend the question generation function and instruct the language model 501 not only to request the concretization of abstract expressions, but also to generate questions regarding the novelty and inventiveness of the invention. For example, the instruction unit 520 can make the language model 501 generate a question such as, "What are the specific differences between this invention and the closest prior art?", thereby making the user aware of the features of the invention and supporting the acquisition of patent rights.

[0162] This disclosure may also be implemented by recording a program for realizing the above-described functions on a computer-readable recording medium, having a computer system read the program recorded on the recording medium, and executing it.

[0163] In this context, "computer system" includes hardware such as the operating system and peripheral devices. Furthermore, "computer-readable recording media" refers to portable media such as flexible disks, magneto-optical disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computer systems.

[0164] Furthermore, "computer-readable recording media" may include those that dynamically hold programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or communication lines such as telephone lines, as well as those that hold programs for a fixed period of time, such as volatile memory inside computer systems that act as servers or clients in such cases.

[0165] Furthermore, the above program may be for the purpose of realizing some of the functions described above, or it may be for the purpose of realizing the above functions in combination with a program already recorded in the computer system, or it may be for the purpose of realizing using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0166] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure are possible, as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0167] Some or all of the above embodiments may also be described as follows, but are not limited to the following: (Note 1) Information regarding the invention and a means for receiving feedback regarding the first specification, An instruction means for giving instructions to a language model, which includes a first instruction for instructing the creation of the first specification based on information relating to the invention, and a second instruction for instructing the creation of a second specification that is more detailed than the first specification, based on the feedback, Output means for outputting the results generated by the language model for the first instruction and the second instruction, An information processing device characterized by comprising: (Note 2) The first instruction includes an instruction to omit the description of the drawing and an instruction to output an outline of the drawing intended to be included in the second specification. The information processing device described in Appendix 1. (Note 3) The receiving means receives instructions regarding the granularity described in the first specification, The first instruction includes instructions for preparing the first specification in accordance with the granularity described in the first specification, The information processing device described in Appendix 1 or 2. (Note 4) The aforementioned receiving means receives instructions regarding the items to be described in the specification, The first instruction includes instructions for preparing the first specification that correspond to the items described above, An information processing device as described in any one of the items 1 to 3 of the appendix. (Note 5) The second instruction includes instructions for preparing the second specification based on the analysis results of the feedback, An information processing device as described in any one of the items 1 to 4 of the appendix. (Note 6) The second instruction includes instructions to apply the content of the feedback to areas other than the feedback portion to prepare the second specification, The information processing device described in Appendix 5. (Note 7) The second instruction includes instructions to prepare the second specification by applying the content of the feedback to descriptions other than the location of the feedback within the same description item, The information processing device described in Appendix 6. (Note 8) The first instruction includes identifying areas where information is missing in preparing the first specification, and an instruction to output the areas where information is missing. An information processing device as described in any one of the items 1 to 6 of the appendix. (Note 9) The first instruction includes an instruction to output a blank space where the information is missing, The information processing device described in Appendix 8. (Note 10) The first instruction includes an instruction to inquire whether to apply general technology to describe the parts where the information is lacking, The information processing device described in Appendix 8. (Note 11) The first instruction includes an instruction to inquire whether to apply general art to the portion where the information is missing if the portion where the information is missing is related to the invention, The information processing device described in Appendix 10. (Note 12) The first instruction includes identifying sections that will be abstract when the first specification is being prepared, and providing instructions to output questions to the user regarding the sections that will be abstract. The information processing device described in Appendix 1. (Note 13) We accept information regarding the invention and feedback regarding the first specification. A first instruction is given to the language model to prepare the first specification based on information relating to the invention, and a second instruction is given based on the feedback to prepare a second specification that is more detailed than the first specification. Output the results generated by the language model for the first instruction and the second instruction. Information processing methods. (Note 14) On the computer, A process for receiving information about the invention and feedback regarding the first specification, A process that gives to a language model a first instruction that instructs the creation of the first specification based on information relating to the invention, and a second instruction that instructs the creation of a second specification which is more detailed than the first specification, based on the feedback, A process to output the generation results by the language model for the first instruction and the second instruction, A program that executes the command.

[0168] Furthermore, some or all of the configurations described in Appendices 2 to 12, which are subordinate to Appendice 1 above, may also be subordinate to Appendices 13 and 14 in the same way as those described in Appendices 2 to 12. Moreover, not limited to Appendices 1, 13, and 14, some or all of the configurations described as appendices may also be subordinate to various hardware, software, various recording means for recording software, or systems, without departing from the embodiments described above. [Explanation of Symbols]

[0169] 100, 200, 300, 400, 500 Information Processing Devices 101, 201, 301, 401, 501 language models 110, 210, 310, 410, 510 Reception Desk 120, 220, 320, 420, 520 Indicator Output section 130, 230, 330, 430, 530

Claims

1. A means for receiving information about the invention and feedback regarding the first specification, An instruction means for giving instructions to a language model, which includes a first instruction for instructing the creation of the first specification based on information relating to the invention, and a second instruction for instructing the creation of a second specification that is more detailed than the first specification, based on the feedback, Output means for outputting the generation results by the language model for the first instruction and the second instruction, An information processing device characterized by comprising:

2. The first instruction includes an instruction to omit the description of the drawing and an instruction to output an outline of the drawing intended to be included in the second specification. The information processing apparatus according to claim 1.

3. The receiving means receives instructions regarding the particle size described in the first specification, The first instruction includes instructions for preparing the first specification in accordance with the granularity described in the first specification, The information processing apparatus according to claim 1.

4. The aforementioned receiving means receives instructions regarding the items to be described in the specification, The first instruction includes the instruction to prepare the first specification corresponding to the item described above, The information processing apparatus according to claim 1.

5. The second instruction includes instructions for preparing the second specification based on the results of the analysis of the feedback, The information processing apparatus according to claim 1.

6. The second instruction includes an instruction to apply the content of the feedback to areas other than the feedback area to prepare the second specification, The information processing apparatus according to claim 5.

7. The second instruction includes instructions to prepare the second specification by applying the content of the feedback to descriptions other than the location of the feedback within the same description item, The information processing apparatus according to claim 6.

8. The first instruction includes identifying a location where information is missing in preparing the first specification, and an instruction to output the location where information is missing. The information processing apparatus according to claim 1.

9. The first instruction includes an instruction to output a blank space where the information is missing, The information processing apparatus according to claim 8.

10. The first instruction includes an instruction to inquire whether to apply general technology to describe the parts where the information is lacking, The information processing apparatus according to claim 8.

11. The first instruction includes an instruction to inquire whether to apply general art to the portion where the information is missing if that portion is related to the invention, The information processing apparatus according to claim 10.

12. The first instruction includes identifying a section that will be expressed abstractly when the first specification is being prepared, and an instruction to output a question to the user regarding the section that will be expressed abstractly. The information processing apparatus according to claim 1.

13. We accept information regarding the invention and feedback regarding the first specification. A first instruction is given to the language model to prepare the first specification based on information relating to the invention, and a second instruction is given to the language model based on the feedback, to prepare a second specification that is more detailed than the first specification. Output the generation results by the language model for the first instruction and the second instruction. Information processing methods.

14. On the computer, A process for receiving information about the invention and feedback regarding the first specification, A process that gives to a language model a first instruction that instructs the creation of the first specification based on information relating to the invention, and a second instruction that instructs the creation of a second specification which is more detailed than the first specification, based on the feedback. A process to output the generation results by the language model for the first instruction and the second instruction, A program that executes the command.

Images