Document recognition support device, document recognition support method, and document recognition support program

Abstract

The objective is to provide document comprehension support technology to assist in understanding the contents of patent application documents. [Solution] One aspect of the present disclosure relates to a document recognition support device comprising: a generation unit that generates correlation information showing the correspondence between a first description in a first invention-related document for the invention to be filed and a second description in a second invention-related document for the invention to be filed that corresponds to the first description; and a display unit that displays the correlation information in one or both of the first invention-related document and the second invention-related document.

Description

Technical Field

[0001] The present disclosure relates to a document comprehension support apparatus, a document comprehension support method, and a document comprehension support program.

Background Art

[0002] Large Language Models (LLMs) such as ChatGPT are increasingly being utilized in society and economy. A typical large language model, when receiving an input question sentence, generates an answer sentence corresponding to the input question. Such a question sentence is called a "prompt", and the quality of the answer sentence generated by the large language model may vary depending on how the question sentence or the prompt is described. For this reason, research and development of technologies for enabling large language models to generate more useful answer sentences (referred to as "prompting technologies") are progressing.

[0003] In addition to prompting technologies, research and development of generative Artificial Intelligence (AI)-related technologies for improving the accuracy of answer sentences output by large language models are also underway. As one such generative AI-related technology, RAG (Retrieval Augmented Generation), which causes a large language model to generate an answer sentence by referring to data stored in a database, is known. Such a database stores data that is not used for training the large language model. Data related to a question sentence or a prompt is extracted from the database, and the extracted data is used to assist in preparing the question sentence or the prompt. This enables the large language model to generate an answer sentence using not only the data used for training the model itself, but also untrained data stored in the database.

[0004] In intellectual property practice, attempts are being made to utilize generative models such as large-scale language models. For example, attempts are being made to use large-scale language models to support the creation of patent application documents (specifications, claims, drawings, etc.), i.e., patent application documents. For example, Japanese Patent Publication No. 2023-7268 describes how to create a patent application specification for an invention that is subject to application, taking into account the similarity with prior art documents. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2023-7268 [Overview of the project] [Problems that the invention aims to solve]

[0006] According to Article 36, Paragraph 6, Item 1 of the Patent Act, the claims must be described in such a way that "the invention for which a patent is sought is described in the detailed description of the invention." Furthermore, according to Article 36, Paragraph 4, Item 1 of the Patent Act, the detailed description of the invention in the specification must be described in such a way that "a person with ordinary skill in the art to which the invention belongs can implement it, in accordance with the provisions of the Ministry of Economy, Trade and Industry Ordinance."

[0007] Therefore, in a patent application, the invention described in each claim must be described clearly and sufficiently in the detailed description of the invention in the specification to the extent that a person skilled in the art can implement it. Thus, the invention described in each claim is described in one of the paragraphs of the detailed description of the invention, and it is considered that there is a correspondence or relationship between the invention described in each claim and the description in the detailed description of the invention.

[0008] However, the claims and specifications of a patent application do not explicitly state any information indicating such a correspondence or relationship between the invention described in each claim and the detailed description of the invention in the specification. Therefore, when understanding the invention being applied for from a patent application, the reader must examine the claims, drawings, and / or specification to determine and identify which paragraphs and / or phrases in the specification correspond to, relate to, or describe the claimed invention.

[0009] One objective of this disclosure is to provide document comprehension support technology to assist in understanding the contents of patent application documents. [Means for solving the problem]

[0010] One aspect of this disclosure relates to a document recognition support device, comprising: a generation unit that generates correlation information showing the correspondence between a first description in a first invention-related document for the invention to be filed and a second description in a second invention-related document for the invention to be filed that corresponds to the first description; and a display unit that displays the correlation information in one or both of the first invention-related document and the second invention-related document. [Effects of the Invention]

[0011] According to this disclosure, it is possible to provide document comprehension support technology to assist in understanding the contents of patent application documents. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a schematic diagram illustrating correlation information according to one embodiment of the present disclosure. [Figure 2] Figure 2 is a schematic diagram showing a document recognition support process according to one embodiment of the present disclosure. [Figure 3] Figure 3 is a block diagram showing the hardware configuration of a document recognition support device according to one embodiment of the present disclosure. [Figure 4]FIG. 4 is a block diagram showing the functional configuration of a document comprehension support apparatus according to an embodiment of the present disclosure. [Figure 5] FIG. 5 is a diagram showing a correlation information generation prompt according to an embodiment of the present disclosure. [Figure 6] FIG. 6 is a diagram showing a display example of correlation information according to an embodiment of the present disclosure. [Figure 7] FIG. 7 is a diagram showing a display example of correlation information according to an embodiment of the present disclosure. [Figure 8] FIG. 8 is a flowchart showing document comprehension support processing according to an embodiment of the present disclosure. [Figure 9] FIG. 9 is a diagram showing document comprehension support processing according to a modified example of the present disclosure. [Figure 10] FIG. 10 is a diagram showing a correlation information generation prompt according to a modified example of the present disclosure. MODE FOR CARRYING OUT THE INVENTION

[0013] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0014] In the following embodiment, a document comprehension support apparatus for understanding patent application documents is disclosed.

[0015] SUMMARY OF THE PRESENT DISCLOSURE In the embodiment described later, the document comprehension support apparatus 100 generates correlation information indicating a correspondence relationship or association between an invention according to each claim recited in the claims for an invention to be applied for and a description content such as a paragraph corresponding to each claim in the specification, and displays the generated correlation information in the claims and / or the specification.

[0016] Specifically, as shown in FIG. 1, such correlation information may indicate the correspondence or interrelationship between an invention according to each claim recited in the claims or an invention-specifying matter constituting the invention, and a paragraph or wording in the specification that describes said invention or invention-specifying matter. In the illustrated example, correlation information indicating that the invention-specifying matter "an acquisition unit that acquires a determination target image showing a product" of claim 1 corresponds to paragraph

[0034] of the specification, that the invention-specifying matter "a processing unit that inputs the determination target image to a trained product determination model, and acquires a determination result indicating the type and quantity of the product from the trained product determination model" corresponds to paragraph

[0036] , and that the invention according to claim 2 corresponds to "specifically, the acquisition unit 110 acquires an image obtained by capturing the appearance of the product to be determined, and provides the acquired image to the processing unit 120" in paragraph

[0034] of the specification is shown by bidirectional arrows.

[0017] In such correlation information generation processing, for example, as shown in FIG. 2, when the document comprehension support apparatus 100 receives a correlation information generation instruction designating the claims and the specification of a patent application document to be processed from the user terminal 200, it creates a prompt for determining which paragraph or wording in the specification the invention according to each claim of the designated claims or each invention-specifying matter corresponds to, and transmits the created prompt to the generation model 30.

[0018] When acquiring the prompt from the document comprehension support apparatus 100, the generation model 30 generates correlation information indicating the paragraph or wording in the specification corresponding or related to the invention according to each claim or each invention-specifying matter as an answer to the prompt, and returns the generated correlation information to the document comprehension support apparatus 100. The generation model 30 herein may typically be any large language model (LLM) such as OpenAI's ChatGPT, Microsoft's Copilot, or the like.

[0019] The document recognition support device 100 provides the user terminal 20 with correlation information obtained from the generation model 30. The output format of the correlation information may be any appropriate format that shows the correspondence or association between the invention or specific feature of each claim and the corresponding paragraph or content of the specification. For example, the correlation information may show the corresponding paragraph or wording of the specification as an annotation for each claim or specific feature of the invention in the claims file, and / or show the paragraph or wording of the specification corresponding to each claim or specific feature of the invention as an annotation for the content of the specification.

[0020] Here, the document recognition support device 100 is implemented by a computing device such as a server, personal computer, smartphone, or tablet, and may have a hardware configuration such as that shown in Figure 3. Specifically, the document recognition support device 100 has a memory 101, storage 102, processor 103, communication device 104, user interface (UI) 105, and peripheral device interface 106 that are interconnected via bus B.

[0021] The programs or instructions that implement the various functions and processes described later in the document recognition support device 100 are stored on a removable storage medium such as a CD-ROM (Compact Disk-Read Only Memory) or flash memory, and are stored in the memory 101 and / or storage 102 via the peripheral device interface 106. However, the programs or instructions do not necessarily need to be installed from the storage medium; they may also be downloaded from any external device via a network or the like.

[0022] Memory 101 is implemented using random access memory, static memory, etc., and when a program or instruction is activated, it reads the program or instruction, data, etc. from storage 102 and stores it.

[0023] The storage 102 is implemented by a hard disk drive or the like and stores files, data, etc., used to execute the installed program or instructions, along with the program or instructions themselves. The memory 101, storage 102, and removable storage medium may be collectively referred to as non-transitory storage medium.

[0024] The processor 103 may be implemented by one or more CPUs (Central Processing Units), GPUs (Graphics Processing Units), processing circuits, etc., which may consist of one or more processor cores, and executes various functions and processes of the document recognition support device 100, as described later, according to data such as programs, instructions, and parameters necessary to execute said programs or instructions stored in the memory 101.

[0025] The communication device 104 is implemented by various communication circuits that perform communication processing with external devices, the Internet, LAN (Local Area Network), WLAN (Wireless LAN), WiFi, and other communication networks.

[0026] The user interface (UI) 105 may consist of input devices such as a keyboard, mouse, camera, and microphone, output devices such as a display, speaker, headset, and printer, and input / output devices such as a touch panel, and realizes an interface between the user and the document comprehension support device 100. For example, the user operates the document comprehension support device 100 by operating a GUI (Graphical User Interface) displayed on the display or touch panel with a keyboard, mouse, etc.

[0027] The peripheral device interface 106 provides various interfaces for connecting to peripheral devices such as USB (Universal Serial Bus) ports and HDMI (High-Definition Multimedia Interface) ports.

[0028] However, the hardware configuration described above is merely an example, and the document recognition support device 100 according to this disclosure may be implemented by any other suitable hardware configuration.

[0029] [Document comprehension support device] Next, a document recognition support device 100 according to one embodiment of the present disclosure will be described. Figure 4 is a block diagram showing the functional configuration of the document recognition support device 100 according to one embodiment of the present disclosure. As shown in Figure 4, the document recognition support device 100 has a generation unit 110 and a display unit 120. For example, one or more functional units of the generation unit 110 and the display unit 120 may be realized by one or more processors 103 executing one or more programs or instructions stored in the memory 101.

[0030] The generation unit 110 generates correlation information showing the correspondence between the content of the claims for the invention to be filed and the corresponding content in the specification for the invention to be filed. Specifically, the generation unit 110 uses a generation model 30, which is implemented as a large-scale language model (LLM) such as ChatGPT or Copilot, to generate correlation information showing the correspondence between the content of the claims and the corresponding content in the specification.

[0031] For example, a user specifies the claims file and the specification file of the patent application documents for the invention to be filed on the user terminal 20, and sends a correlation information generation instruction to the document recognition support device 100 to cause the generation model 30 to generate correlation information for these files. Upon receiving the correlation information generation instruction from the user terminal 20, the generation unit 110 creates a prompt to cause the generation model 30 to generate correlation information indicating which paragraph or wording in the specification corresponds to, is related to, is described in, or is supported by each claim or inventive feature described in the specified claims file, and sends the created prompt to the generation model 30.

[0032] Specifically, the generation unit 110 creates a prompt as shown in Figure 5, causing the generation model 30 to generate correlation information between the claims file (Claims.docx) and the specification file (Spec.docx). The illustrated prompt first states, "You are a patent attorney at a patent office. You are preparing patent application documents for an invention requested by a client," instructing the generation model 30 on its role.

[0033] Next, the following is written: "Based on the attached draft claims file and draft specification file, create correlation information indicating the paragraphs or content of the draft specification related to each claim described in the draft claims. If the claims are divided into multiple inventive features, create correlation information indicating the paragraphs or content of the draft specification related to each divided inventive feature." This instructs the generation model 30 to generate correlation information for the attached files "Claims.docx" and "Spec.docx" indicating the paragraphs or content of the specification that correspond to, relate to, or support the invention and each inventive feature related to each claim.

[0034] Such claims and / or specifications may typically be prepared by the user themselves, the patent attorney entrusted with the patent application, the inventor, intellectual property staff, or other persons, and may be a pre-filing draft or may be referenced when responding to rejection reasons after filing. Alternatively, the claims and / or specifications may be prepared by a generative model 30 such as a large-scale language model. Note that the claims file and the specification file do not need to be provided separately; for example, they may be provided together in a single document file.

[0035] The display unit 120 displays correlation information in either the claims or the specification, or both. For example, as shown in Figure 6, the display unit 120 may display correlation information in the claims in annotation form, indicating which paragraph and / or wording in the specification corresponds to, is related to, or is described for each claim's invention and / or each inventive feature constituting the invention.

[0036] In the example shown in Figure 6, the invention-specific feature of claim 1, "an acquisition unit for acquiring a determination target image indicating a product," is indicated in footnote form as being related to, or described in paragraph

[0034] of the specification.

[0037] Similarly, the inventive feature "a processing unit that inputs the image to be judged into a trained product judgment model and obtains a judgment result from the trained product judgment model indicating the type and quantity of the product" is indicated in footnotes as corresponding to, related to, or described in paragraph

[0036] of the specification.

[0038] Similarly, the inventive feature "The trained product determination model is trained based on the error between the determination result of the product determination model being trained for a training image showing a product and the correct label indicating the type and quantity of the product in the training image" is indicated in footnotes as being related to or described in paragraph

[0038] of the specification.

[0039] Furthermore, it is indicated in footnote form that the invention of claim 2, "the image to be determined is obtained by imaging the appearance of the product," corresponds to or is related to the phrase in paragraph

[0034] of the specification, "Specifically, the acquisition unit 110 acquires an image of the appearance of the product to be determined and provides the acquired image to the processing unit 120."

[0040] Similarly, the display unit 120 may, for example, display correlation information in annotation form within the claims indicating which paragraph and / or wording of the specification corresponds to, is related to, or describes the invention and / or each inventive feature constituting the invention.

[0041] On the other hand, as shown in Figure 7, the display unit 120 may display correlation information in annotation form in the specification indicating whether a paragraph and / or wording of the specification corresponds to, is related to, or is described in relation to, the invention and / or inventive feature of, which claim described in the claims. In the example shown in Figure 7, it is indicated in annotation form that paragraph

[0034] of the specification corresponds to, is related to, or is described in relation to, the inventive feature of the invention of claim 1, "acquisition unit for acquiring a determination target image showing a product."

[0042] Similarly, the wording in paragraph

[0034] of the specification, "Specifically, the acquisition unit 110 acquires an image of the appearance of the product to be judged and provides the acquired image to the processing unit 120," is indicated in footnote form as corresponding to, related to, or described in the invention of claim 2.

[0043] Similarly, paragraph

[0036] of the specification is indicated in footnote form as being related to or describing the inventive feature of the invention of claim 1, "a processing unit that inputs the image to be determined into a trained product determination model and obtains a determination result from the trained product determination model indicating the type and quantity of the product."

[0044] Furthermore, paragraphs

[0038] and

[0039] of the specification are indicated in footnote form as being related to or describing the inventive feature of the invention according to claim 1: "The trained product determination model is trained based on the error between the determination result of the product determination model to be trained for a training image showing a product and the correct label indicating the type and quantity of the product in the training image."

[0045] Furthermore, the annotated claims shown in Figure 6 and the annotated specification shown in Figure 7 do not need to be mutually exclusive; they may be combined. In other words, the correlation information may be displayed in both the claims and the specification, not just one of them.

[0046] Furthermore, in the illustrated embodiment, the display unit 120 displays correlation information near the invention and / or inventive features of each claim in the claims, and / or near paragraphs and / or words in the specification, or both. However, the correlation information provided in this disclosure does not necessarily have to be displayed in such annotation format. The correlation information may be displayed in the claims and / or specification in any other appropriate format, or it may be embedded in the claims and / or specification in any appropriate format instead of being visibly displayed to the user, or it may be stored in another file (e.g., a correlation information file).

[0047] As illustrated, the correlation information indicates the claim number or inventive feature wording of the claims, or one or more paragraphs or one or more phrases of the specification, but the correlation information provided in this disclosure is not necessarily limited thereto. For example, the correlation information may indicate the drawing number and / or reference numerals or elements described in the drawings corresponding to each claim or inventive feature of each claim, or one or more paragraphs or one or more phrases of the specification. That is, in the illustrated annotation, the drawing number and / or reference numerals or elements described in the drawings may be indicated in place of or in addition to the relevant claim number or inventive feature and / or relevant paragraph or phrase.

[0048] In another embodiment, the display unit 120 may display correlation information in response to a predetermined user operation. For example, the correlation information may be embedded in the claims and / or specification in a state invisible to the user, and when the user presses a predetermined button or the like while viewing the claims and / or specification on the user terminal 20, the display unit 120 may display the correlation information.

[0049] In one embodiment, the display unit 120 may also display the invention and / or inventive features related to the claims described in the claims for which no correlation information is generated. That is, if the display unit 120 cannot detect paragraphs, words, and drawings corresponding to the invention and / or inventive features related to the claims described in the claims, it may notify the user that the claim number and / or inventive features may be grounds for rejection.

[0050] According to the document recognition support device 100 of this embodiment, correlation information is generated showing the correspondence between the inventions and / or inventive features constituting these inventions described in the claims of a patent application document and the paragraphs and / or words described in the specification of the patent application document, and the generated correlation information can be displayed to the user in the claims and / or specification. This allows the user to easily confirm which paragraphs and / or words in the specification correspond to, are related to, or describe the inventions and / or inventive features described in each claim.

[0051] [Document comprehension support processing] Next, with reference to Figure 8, a document recognition support process according to one embodiment of the present disclosure will be described. Figure 8 is a flowchart of the document recognition support process according to one embodiment of the present disclosure. The document recognition support process is performed by the document recognition support device 100 described above, and more specifically, it may be realized by one or more processors 103 of the document recognition support device 100 executing one or more programs or instructions stored in one or more memories 101.

[0052] As shown in Figure 8, in step S101, the document recognition support device 100 receives the claims and specification for the invention to be filed from the user terminal 20. The claims and specification here may be a draft before the patent application is filed, or they may be application documents after the patent application has been filed.

[0053] For example, before filing a patent application, the user can use the generated correlation information to verify whether the draft claims and specification are appropriate, that is, whether the inventions related to each claim in the claims are properly described or supported in the specification. If an invention related to a claim that is not described or supported in the specification is detected, the user can add it to the specification to satisfy the support requirement.

[0054] Furthermore, after filing a patent application, for example, when considering how to respond to a notice of rejection, the user can identify the paragraphs and / or wording of the specification corresponding to the invention related to each claim based on the generated correlation information, and consider a strategy for amending the claims.

[0055] In step S102, the document recognition support device 100 creates a prompt to cause the generation model 30 to generate correlation information indicating the correspondence and association between the content of the claims and the content of the specification corresponding to said claims. Specifically, the document recognition support device 100 creates a prompt to cause the generation model 30 to generate correlation information indicating which paragraph and / or wording of the specification corresponds to, is related to, or describes the invention and / or specific feature of each claim described in the claims.

[0056] In step S103, the document recognition support device 100 inputs the created prompt to the generation model 30. Typically, the generation model 30 may be a general-purpose large-scale language model installed on a cloud server, or it may be a large-scale language model dedicated to the document recognition support process installed on a local server. For example, the document recognition support device 100 sends a prompt to such a cloud server and / or local server, causing the generation model 30 on the cloud server and / or local server to execute the generation process for that prompt.

[0057] In step S104, the document recognition support device 100 obtains correlation information from the generation model 30. Specifically, the document recognition support device 100 obtains correlation information from the generation model 30 that shows the invention and / or specific feature of each claim described in the claims and the corresponding paragraph and / or wording of the specification.

[0058] In step S105, the document recognition support device 100 displays the acquired correlation information on the user terminal 20. Specifically, the document recognition support device 100 may display the correlation information near each claim and / or each specific feature of the invention described in the claims and / or near the corresponding paragraph and / or wording of the specification.

[0059] According to the document recognition support processing of this embodiment, correlation information is generated showing the correspondence between the inventions and / or inventive features constituting these inventions described in the claims of a patent application document and the paragraphs and / or words described in the specification of the patent application document. The generated correlation information can be displayed to the user in the claims and / or specification. This allows the user to easily confirm which paragraphs and / or words in the specification correspond to, are related to, or describe the inventions and / or inventive features described in each claim.

[0060] [Example 1] In the embodiments described above, the document recognition support device 100 generated correlation information for the claims and specification of the invention to be filed, provided by the user; however, this disclosure is not limited thereto. For example, the document recognition support device 100 may generate correlation information for the claims and / or specification generated by the generation model 30, instead of for the claims and / or specification provided by the user.

[0061] For example, when generating a specification based on the claims and drawings of an invention to be filed, obtained from the user terminal 20, the user transmits the claims and drawings to the document recognition support device 100, as shown in Figure 9. The generation unit 110 creates a prompt to generate a specification from the claims and drawings received from the user terminal 20, and transmits the created prompt to the generation model 30.

[0062] The generation model 30 prepares a specification for the invention based on the claims and drawings provided by the user in accordance with prompts created by the generation unit 110, and generates correlation information showing the correspondence between the contents of the claims and the contents of the specification. Specifically, the generation model 30 may display the correlation information in annotation form near the corresponding claims and inventions and / or near the corresponding paragraphs and phrases in the specification.

[0063] To achieve this, for example, the generation unit 110 may create a prompt as shown in Figure 10 and cause the generation model 30 to generate the specification and correlation information according to the prompt. The prompt shown in Figure 10 first states, "You are a patent attorney at a patent office. You have been asked by a client to prepare patent application documents for the invention to be filed," instructing the generation model 30 on its role.

[0064] Furthermore, the instructions state, "Prepare a draft specification based on the attached claims file and drawings file," instructing the generating model 30 to create the specification based on the attached files "Claims.docx" and "Drawings.pdf."

[0065] Furthermore, the instructions state, "Based on the attached draft claims file and the prepared draft specification, create correlation information indicating the paragraphs or content of the draft specification related to each claim or inventive feature described in the draft claims," ​​instructing the generation model 30 to create correlation information between the attached claims and the prepared specification.

[0066] Furthermore, the instruction states, "Display the correlation information in the format of annotations for the inventions and specific features of each claim, and in the format of annotations for the corresponding paragraphs and phrases in the specification," instructing the generation model 30 to display the generated correlation information in annotation format.

[0067] When the generated specification and correlation information are obtained from the generation model 30 in accordance with such prompts, the display unit 120 provides and displays the obtained specification and correlation information to the user terminal 20.

[0068] In the above-described modifications, the specification and correlation information were generated based on the claims and drawings, but this disclosure is not necessarily limited thereto. For example, instead of using a prescribed format for patent application documents such as claims, drawings and / or specification, if the claims, drawings and / or specification are created by the generation model 30 based on an invention proposal that freely describes the invention subject to application, correlation information may be generated for the claims and specification generated by the generation model 30.

[0069] As described above, correlation information can be generated showing the correspondence between the content of the claims and the content of the specification, not only for the claims and specification provided by the user, but also for the claims and / or specification generated by the generation model 30. Typically, it may not be easy to grasp the correspondence between the content of the claims and the content of the specification generated by the generation model 30, but by displaying the correlation information between the content of the claims and the content of the specification generated by the generation model 30 to the user, such a correspondence can be easily grasped.

[0070] [Differentiation 2] In the embodiments described above, the document recognition support device 100 generated correlation information between the claims and specification of the invention to be filed, provided by the user or generated by the generation model 30, but this disclosure is not limited thereto. For example, the document recognition support device 100 may generate correlation information between two or more invention-related documents, such as an invention proposal, invention description, invention manuscript (hereinafter collectively referred to as "invention proposal"), claims, drawings, and / or specification, in which the inventor or the like freely describes the invention to be filed. Invention-related documents here refer to any document that contains information related to the invention to be filed.

[0071] For example, the document recognition support device 100 may input a prompt to generate correlation information between the descriptions of the invention proposal, claims, drawings, and specification of the invention to be filed, along with the files of the invention proposal, claims, drawings, and specification, into the generation model 30, and obtain the correlation information between the descriptions of the invention proposal, claims, drawings, and specification from the generation model 30. In this case, if the user specifies any of the claims, the document recognition support device 100 may display the description of the invention proposal, the drawing number, and / or the paragraph of the specification related to the invention of the specified claim in annotation format within the claims.

[0072] In this case, the document recognition support device 100 may detect descriptions in the invention proposal, claims, drawings, and specification from which no correlation information is generated. If a description from which no correlation information is generated is detected, the document recognition support device 100 may display the detected description in the invention proposal, claims, drawings, or specification to the user and allow the user to confirm the description. For example, if such a description from which no correlation information is generated is detected in the invention proposal, the document recognition support device 100 can allow the user to confirm whether the matter corresponding to that description is not described in the claims, drawings, and specification and is therefore an omission. Alternatively, if a description from which no correlation information is generated is detected in any claim of the claims, the document recognition support device 100 can warn the user that the description corresponding to that claim is not described in the specification and drawings and may violate the support requirement. Alternatively, if a description from which no correlation information is generated is detected in any drawing, the document recognition support device 100 can allow the user to confirm whether the description corresponding to that drawing is not described in the specification and is therefore an omission.

[0073] In the various embodiments described above, we focused on the correlation information between the content of the claims and the content of the specification. However, it will be clear that this disclosure can be easily replaced with the correlation information between the content of the claims and the content of the specification of a utility model registration application. In other words, the claims described above can be replaced with the claims for utility model registration, and the claims and the claims for utility model registration may be collectively referred to simply as "claims."

[0074] Although embodiments of this disclosure have been described in detail above, this disclosure is not limited to the specific embodiments described above, and various modifications and changes are possible within the scope of the gist of this disclosure as described in the claims. [Explanation of Symbols]

[0075] 20 User Terminals 30 Generative Models 100 Document recognition support device 110 Production Department 120 indicates the department

Claims

1. A generation unit that generates correlation information showing the correspondence between the first description in the invention proposal for the invention subject to application and the second description corresponding to the first description in the claims, drawings, or specification for the invention subject to application, A display unit that displays the aforementioned correlation information in the invention proposal, the claims, the drawings, or the specification, or in one or both of the above, It has, The display unit is a document comprehension support device that displays the contents of the invention proposal in which it has been detected that no correlation information has been generated, and allows the user to confirm them.

2. The contents of the first item described above are the contents of the invention proposal document, The document recognition support device according to claim 1, wherein the second description is the invention or specific feature of the invention related to the claim described in the claim, the drawing number of the drawing, or the paragraph or wording of the specification.

3. The document recognition support device according to claim 1, wherein the correlation information indicates the description in the invention proposal, the invention or specific features of the invention as described in the claims, the drawing numbers of the drawings, or the paragraphs or words of the specification.

4. The document comprehension support device according to claim 1, wherein the display unit displays the correlation information near either or both of the first description and the second description.

5. The document comprehension support device according to claim 1, wherein the generation unit generates the correlation information using a generation model.

6. To generate correlation information showing the correspondence between the first description in the invention proposal for the invention subject to application and the second description corresponding to the first description in the claims, drawings, or specification for the invention subject to application, The aforementioned correlation information shall be displayed in the invention proposal and in the claims, drawings, or specification, or in either or both. The contents of the invention proposal document in which the aforementioned correlation information was detected not to have been generated are displayed and the user is asked to confirm them. A document comprehension support method performed by a computer.

7. To generate correlation information showing the correspondence between the first description in the invention proposal for the invention subject to application and the second description corresponding to the first description in the claims, drawings, or specification for the invention subject to application, The aforementioned correlation information shall be displayed in the invention proposal and in the claims, drawings, or specification, or in either or both. The contents of the invention proposal document in which the aforementioned correlation information was detected not to have been generated are displayed and the user is asked to confirm them. A document comprehension support program that has a computer perform the necessary actions.

Images