Document-drafting assistance device and program
The document creation support device and program leverage a specialized language model to normalize patent literature data, generating tailored support information for creating high-quality documents submitted to the Japan Patent Office, addressing the limitations of existing technologies in providing relevant support.
Patent Information
- Application Number
- PCT/JP2024/017039
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-09
- Filing Date
- 2024-05-08
- Publication Date
- 2024-11-14
AI Technical Summary
Existing technologies for creating documents submitted to the Patent Office, such as application documents for patents and utility models, are limited in providing relevant support information, leading to challenges in producing high-quality documents.
A document creation support device and program utilizing a specialized language model for natural language processing, which generates support information by normalizing patent and utility model literature data into a unified format, providing suggestions for terminology, grammar corrections, and references to similar documents, tailored to the user's intentions.
Enables users to easily create high-quality documents by providing various support information that matches their intentions, improving the quality and efficiency of document preparation for submission to the Japan Patent Office.
Smart Images

Figure JP2024017039_14112024_PF_FP_ABST
Abstract
Description
Document creation support device and program
[0001] The present invention relates to a document creation support device and a program.
[0002] Regarding technology for assisting in the creation of documents to be submitted to the Patent Office, such as patent and utility model application documents, Patent Document 1 discloses a technology for updating the related information by associating a predetermined word from a first claim that has already been created and is later amended by the document creator with the amended word, storing superordinate and subordinate concepts of the pre-amended word in a storage unit as related information, and associating a predetermined word from a second claim that has already been created and is later amended by the document creator with the amended word, and then updating the related information and outputting related information for words entered when creating a third claim based on the updated related information.
[0003] According to the technology of Patent Document 1, when preparing application documents for a patent or utility model, related information such as superordinate or subordinate terms to the word before correction is output as support information, thereby assisting the document creator in preparing application documents using appropriate expressions by referring to the support information.
[0004] Japanese Patent Application Laid-Open No. 2021-086425
[0005] However, since the technology of Patent Document 1 uses information generated based on related information that associates words before and after correction as support information, the support information is limited to a range based on the relationship between the words before and after correction. As a result, application documents are created using limited support information as a reference, so there is room for further improvement in terms of easily creating high-quality application documents.
[0006] The present invention was made in consideration of the above circumstances, and its purpose is to enable users to easily create high-quality documents to be submitted to the Patent Office by providing various support information that meets their needs.
[0007] As a result of intensive research into solving the above-mentioned problems, the inventors have found that the above-mentioned object can be achieved by generating support information for assisting in the preparation of documents to be submitted to the Patent Office by utilizing a language model that has been trained and tuned specifically for the preparation of documents to be submitted to the Patent Office for patents and utility models, depending on the content of the prompt. The inventors have then completed the present invention. Specifically, the present invention provides the following:
[0008] The present invention is a document creation support device characterized by having: a literature database in which accumulated document information on patents and utility models is stored in a normalized state in a unified data format so that it can be adapted to the learning of a language model for natural language processing; and a support information generation unit that generates support information to support the creation of documents to be submitted to the Patent Office by combining one or more types of natural language processing based on the literature database and utilizing a language model that has been learned and adjusted to be specialized for the creation of documents to be submitted to the Patent Office for patents and utility models, in accordance with the content of a prompt.
[0009] According to the features of the present invention, a language model for natural language processing specialized for creating documents to be submitted to the Patent Office for patents and utility models is used to generate support information according to the content of the prompt, thereby providing various support information that matches the intentions of the user who entered the content of the prompt, making it easy to create high-quality documents to be submitted to the Patent Office.
[0010] According to the present invention, by providing various support information that meets the user's wishes, high-quality documents to be submitted to the Patent Office can be easily prepared.
[0011] FIG. 1 is an explanatory diagram showing the flow of information in the document creation support device of this embodiment. FIG. 2 is a block diagram showing an example of the hardware configuration and software configuration of the document creation support device of this embodiment. FIG. 3 is an explanatory diagram of a display screen showing a user interface. FIG. 4 is an explanatory diagram of a display screen showing the user interface. FIG. 5 is an explanatory diagram of a display screen showing the user interface. FIG. 6 is an explanatory diagram of a display screen showing the user interface.
[0012] The following is a detailed description of an example of an embodiment of the present invention, with reference to the drawings. (Document Creation Support Device 1) As shown in Fig. 1, the document creation support device 1 is configured to generate support information for supporting the creation of documents to be submitted to the Patent Office by utilizing a language model for natural language processing specialized for the creation of documents to be submitted to the Patent Office for patents and utility models.
[0013] To give a specific example, the document creation support device 1 has a literature database in which accumulated document information for patents and utility models is stored in a normalized state in a unified data format so that it can be adapted to the learning of a language model for natural language processing, and a support information generation unit 115 that generates support information to support the creation of documents to be submitted to the Patent Office by combining one or more types of natural language processing based on the literature database and utilizing a language model that has been learned and adjusted to be specialized for the creation of documents to be submitted to the Patent Office for patents and utility models, depending on the content of the prompt.
[0014] Here, "patent and utility model document archives" include domestic documents such as published patent gazettes, published utility model gazettes, and registered gazettes (patent publications (patent gazettes), registered utility model gazettes), foreign documents such as U.S. patent specifications and international publications, and non-patent documents such as domestic technical journals and proceedings. Furthermore, "document archives" also include documents such as opinions, amendments, various trial decision documents (ex parte trials, inter partes trials), opposition decision documents, and judgment bulletins. This document archive is information collected from various search sites, such as J-PlatPat (registered trademark), a patent information search site; Espacenet (registered trademark), a patent information search site provided by the European Patent Office; WIPO PatentScope, an international patent information search site; the Intellectual Property High Court's case precedent search site; and data within each patent office.
[0015] "Natural language processing" enables a computer to understand text and audio data written in natural language and execute processing appropriate to the purpose. Specific examples include morphological analysis, which breaks natural language down into "morphemes," the smallest units that make up the language, and assigns information such as parts of speech; syntactic analysis, which analyzes the grammatical structure of natural language to clarify the structure and meaning of a sentence; semantic analysis, which analyzes the meaning of natural language to understand the meaning of words and sentences and make logical judgments and inferences; contextual analysis, which understands natural language while taking into account the context before and after a sentence; and intent analysis, which extracts the intention of a speaker or writer from a conversation or sentence using natural language. Thus, "natural language processing" processes natural language by combining processes such as morphological analysis, syntactic analysis, semantic analysis, contextual analysis, and intent analysis, and enables the generation of support information to assist in the preparation of documents to be submitted to the Patent Office, as well as machine translation, automatic summarization, question-answering systems, and speech recognition.
[0016] A "language model" is a type of probabilistic model used in natural language processing, and is a model that probabilistically predicts how likely a given word or sentence is to occur in natural language. Specifically, a language model calculates the occurrence probability of a given word sequence or sentence, or compares the occurrence probabilities of multiple word sequences or sentences, making it possible to automatically generate the most likely word or sentence based on the context when predicting the next word or sentence.
[0017] "Normalized into a unified data format" refers to the process of converting information stored in different formats for each document into a common format and unifying it when dealing with large amounts of data, such as patent documents. For example, patent documents contain various information, such as the names of applicants and inventors, application and publication numbers, technical fields, document summaries, and claims. However, this information is often stored in different formats for each country, document, or database. This can make it difficult to collect necessary information from multiple databases or handle information in different formats to create a language model for natural language processing. However, normalizing data into a unified data format facilitates data search and analysis. Furthermore, using normalized data as input for machine learning of document archive information improves processing efficiency and accuracy. "Normalization" is a processing method for organizing and structuring data in database tables according to certain rules, eliminating data redundancy and maintaining data consistency and integrity.
[0018] The "literature database" stores patent and utility model document information in a standardized data format suitable for training language models for natural language processing. Specifically, the specification database 121, which stores patent and utility model specification information, and the intermediate document database 122, which stores intermediate document information such as patent and utility model opinion and amendment information, correspond to the literature database. As a result, even if specifications written in different formats, such as those in Japan and the United States, are collected as literature information, the specification data is divided into normalized sections such as "problem to be solved" and "embodiment" and stored in each table of the specification database 121, thereby reducing data duplication and improving data consistency. Furthermore, the tables in the specification database 121 are linked to each other, enabling efficient retrieval of necessary information.
[0019] "Patent and utility model documents submitted to the Patent Office" refer to documents submitted to the Patent Office, such as patent and utility model specifications, amendments, arguments, requests for appeals, and patent oppositions. In the field of natural language processing, a "language model" refers to a mathematical model that learns language patterns and grammatical rules to generate and understand natural language. For example, a language model used in preparing documents submitted to the Patent Office is trained specifically for the preparation of documents submitted to the Patent Office, and is used to generate patent application specifications and arguments, which are types of documents submitted to the Patent Office, in response to prompts, as well as to generate support information for preparing specifications and arguments.
[0020] The "learning" in "language model training and tuning" includes supervised learning (for example, the task of predicting the next word) and unsupervised learning (for example, the task of learning vector representations of words). Furthermore, "tuning" refers to improving the performance of a language model by evaluating the trained language model and adjusting parameters such as the weights of the model as needed, or by appropriately adjusting the values of hyperparameters such as the number of hidden layers and units of the neural network, the learning rate, and the strength of regularization.
[0021] "Prompts" are information that displays to User 4 the content to be entered and operation methods, etc., and provides instructions for User 4 to perform operations. Examples of "support information to support the preparation of documents to be submitted to the Patent Office" include suggestions for appropriate terminology and expressions when entering the necessary information for documents, suggestions for correcting grammar and expressions, providing information useful for preparing documents by referring to similar patent documents, and information and suggestions regarding document formats. Such prompts and support information will be explained in detail in the section on user interfaces below.
[0022] According to the above configuration, a language model for natural language processing specialized for creating documents to be submitted to the Patent Office for patents and utility models is used to generate support information according to the content of the prompt, thereby providing various support information that matches the intentions of the user 4 who inputs the content of the prompt, making it possible to easily create high-quality documents to be submitted to the Patent Office.
[0023] 2, the document creation support device 1 has, in addition to the support information generation unit 115, an activation unit 111, a data collection unit 112, a model learning unit 113, an input processing unit 114, a terminal control unit 116, and a communication unit 13. Some or all of the support information generation unit 115, the activation unit 111, the data collection unit 112, the model learning unit 113, the input processing unit 114, and the terminal control unit 116 may be configured as either hardware or software, and are included in the control unit 11, which is a computer.
[0024] The communication unit 13 is connected to the user terminal 2 operated by the user 4 via the Internet 5 so as to enable data communication. Data communication between the communication unit 13 and the user terminal 2 is not limited to the Internet 5, but may be via an information communication network such as a dedicated line or a local area network (LAN), or Bluetooth (registered trademark), a wireless communication standard for short-range communication. Data communication between the communication unit 13 and the user terminal 2 may also be via a dedicated line to prevent the leakage of highly confidential documents submitted to the Patent Office. Furthermore, when communicating data via the Internet 5, the communication unit 13 and the user terminal 2 preferably have a virtual private network (VPN) function. If the communication unit 13 and the user terminal 2 have a VPN function, data communication from the user terminal 2 passes through a VPN connection, ensuring secure communication and protecting it from unauthorized external access. The user terminal 2 is an information processing device such as a general personal computer, laptop computer, smartphone, or tablet terminal.
[0025] The activation unit 111 has a function of enabling data communication only with authenticated user terminals 2. Specifically, the activation unit 111 has a function of verifying whether an activation signal transmitted from a specific user terminal 2 matches pre-set authentication information. Examples of authentication methods include authentication modes that combine one or more authentication elements, such as password authentication, biometric authentication, one-time password authentication, and smart card authentication using a card with an embedded IC chip. This allows the document creation support device 1 to communicate data only with authenticated user terminals 2, making it difficult to leak highly confidential documents submitted to the Patent Office.
[0026] The data collection unit 112 has the function of collecting document accumulation information on patents and utility models from the patent information search site 6 and the like, normalizing the collected information into a unified data format, and storing it in a document database. As a result, the data collection unit 112 collects and organizes data necessary for learning and adjusting the language model, and stores it in the document database, thereby enabling the support information generation unit 115 to efficiently and accurately support the creation of documents to be submitted to the Patent Office. Note that the collection of document accumulation information may also be performed by the administrator of the document creation support device 1 operating the administrator terminal 7 to access the patent information search site 6 and the like.
[0027] The model learning unit 113 has a function of learning and adjusting a language model for natural language processing based on a literature database of patents and utility models. As a result, the model learning unit 113 learns and adjusts the language model to be specialized for creating documents to be submitted to the Patent Office, thereby performing more advanced natural language processing and enabling the support information generation unit 115 to generate high-quality support information to support the creation of documents to be submitted to the Patent Office.
[0028] The input processing unit 114 has a function of converting prompt information indicating the content of the prompt sent from the user terminal 2 into information that is easy to handle for a language model of natural language processing. The terminal control unit 116 has a function of setting the display screen of the user terminal 2 to a screen suitable for creating documents to be submitted to the Patent Office, based on data in the UI database 124 (user interface database 124). Details will be described later.
[0029] Furthermore, the document creation support device 1 includes an input device 15 connected to the input receiving unit 118 and a display device 14 connected to the display control unit 117. Examples of the input device 15 include a keyboard, a mouse, a touch panel, and a voice input device. Examples of the display device 14 include a liquid crystal display device. This allows the document creation support device 1 to be configured using information processing devices such as general personal computers, laptop computers, smartphones, and tablet terminals. The document creation support device 1 may lack at least one of the input device 15 and the display device 14. In this case, the input device 15 and the display device 14 are provided on an external terminal (not shown), allowing the document creation support device 1 to be used as a document creation support server. Furthermore, the document creation support device 1 may be used as a user terminal 2, and a user 4 may receive assistance from support information while the document creation support device 1 prepares documents to be submitted to the Patent Office.
[0030] In this embodiment, the case where the functions of the document creation support device 1 are installed in an information processing device is described, but the present invention is not limited to this, and the functions of the document creation support device 1 may be installed in cloud computing. In this case, cloud computing allows computer resources to be added as needed, making it possible to process large amounts of document data, enabling faster and more efficient processing, and also making it possible to easily expand processing capacity to accommodate a significant increase in the number of users 4.
[0031] (Memory Unit 12, etc.) As shown in FIG. 1 , the specification database 121, the interim document database 122 storing documents associated with assessments, such as written opinions and amendments, and the UI database 124 are stored in the memory unit 12. The memory unit 12 is connected to the document creation support device 1 so that data can be communicated. The memory unit 12 may be configured with a hard disk or a combination of a hard disk and memory. In the case of a combination of a hard disk and memory, some of the data, indexes, etc. used by the database are cached in the memory as needed, thereby speeding up access to the database. The memory unit 12 also stores a user database 123 storing various information, including identification information for identifying the user 4 and authentication information for authenticating the user 4. The activation unit 111 is configured to perform processes such as authentication based on the data in the user database 123.
[0032] The storage unit 12 may be a data server connected to an information communication network such as the Internet 5, separate from the document creation support device 1. The storage unit 12 may also be configured with a plurality of data servers, each provided for a different database.
[0033] (User Interface) As shown in FIG. 3 , the terminal control unit 116 has a function to set the display screen of the user terminal 2 to a screen suitable for preparing documents to be submitted to the Patent Office based on the UI database 124. To give a specific example, the terminal control unit 116 has a function to control the display screen 21 of the user terminal 2 to display a prompt section 212 that accepts prompt content, a support information section 213 that displays support information generated by the support information generation unit 115, and a document creation section 211, which is an area for creating documents to be submitted to the Patent Office. This allows the user 4 using the user terminal 2 to simultaneously check the content of the prompt section 212 and the support information in the support information section 213 while creating documents to be submitted to the Patent Office, thereby facilitating the creation of documents to be submitted to the Patent Office. Furthermore, the user terminal 2 may display entry items in a format conforming to regulations in the document creation section 211 at the start of preparing documents to be submitted to the Patent Office, such as specifications.
[0034] Furthermore, as shown in Figure 4, the terminal control unit 116 has the function of allowing the user to specify a standard phrase indicating a predetermined instruction content such as a similar word, antonym, or role, and controlling the display of a prompt operation unit 214 on the display screen 21 of the user terminal 2, which combines the specified standard phrase with a word or phrase specified in the document creation unit 211 and uses the information input to the prompt unit 212.
[0035] For example, when the "predetermined instruction content" includes multiple instruction elements such as a synonym, a superordinate concept, an antonym, and a role, standard phrases including these instruction elements are displayed as buttons, and a first instruction button 2141 for standard phrases whose instruction content is a synonym, a second instruction button 2142 for standard phrases whose instruction content is a superordinate concept, a third instruction button 2143 for standard phrases whose instruction content is an antonym, and a fourth instruction button 2144 for standard phrases whose instruction content is a role are displayed on the prompt operation unit 214. For example, when the phrase "A spring is provided between the top board and the base" is written in the document creation unit 211, if the user 4 drags the word "spring" to the first instruction button 2141 while designating it, or if the user 4 clicks the first instruction button 2141 after designating it, the word "spring" is stored in the memory of the user terminal 2 and is combined with the prompt data of the standard phrase associated with the first instruction button 2141, which has been stored in advance in the memory. The synthesized data is then displayed on the prompt section 212 of the user terminal 2 .
[0036] As a result, the prompt unit 212 displays information combining a phrase and a fixed phrase, such as "Please give examples of synonyms for rubber," as the content of the prompt. This prompt content is sent to the document creation support device 1, and the support information generation unit 115 generates support information such as "(examples of synonyms): spring, coil, damper." This support information is sent to the user terminal 2 and displayed in the support information unit 213, allowing the user 4 to proceed with writing the specification using examples of synonyms for springs while visually checking or copying and pasting the support information in the support information unit 213. As a result, the user 4 using the user terminal 2 can easily proceed with the preparation of documents to be submitted to the Patent Office by using information combining predetermined instructions and specified phrases as the content entered in the prompt unit 212 while preparing documents to be submitted to the Patent Office.
[0037] Furthermore, as shown in Figures 5 and 6, the terminal control unit 116 has a function of allowing the user to specify a standard phrase indicating a predetermined instruction content, and controlling the display screen 21 to display a prompt operation unit 214 that combines the specified standard phrase with a phrase including a word specified in the document creation unit 211 as the content input to the prompt unit 212.
[0038] To give a specific example, if the phrase "A spring is provided between the top plate and the base" is written in the document creation unit 211, and user 4 designates the word "spring" and drags it to the fourth instruction button 2144, or if user 4 designates the word "spring" and clicks the fourth instruction button 2144, the phrase including the word "spring" is stored in the memory of the user terminal 2 and is combined with the prompt data of the fixed phrase that has been stored in advance in the memory and that is associated with the fourth instruction button 2144. The combined data is then displayed in the prompt unit 212 of the user terminal 2.
[0039] As a result, the prompt unit 212 displays the phrase "A spring is provided between the tabletop and the base." and the phrase "Please give examples of synonyms for 'rubber' in the context." as the prompt content. This prompt content is sent to the document creation support device 1, and the support information generation unit 115 generates support information such as "In the context of 'A spring is provided between the tabletop and the base,' the role of the spring is to create tension between the tabletop and the base and support the tabletop. The spring can undergo deformations such as compression and extension, and its elastic force causes the tabletop to..." This support information is sent to the user terminal 2 and displayed in the support information unit 213, allowing the user 4 to write the specification while visually checking or copying and pasting the support information in the support information unit 213 and assuming the role of the spring. As a result, during the preparation of documents to be submitted to the Patent Office, information combining predetermined instructions and phrases containing specified words is input to the prompt unit 212, thereby obtaining appropriate answers based on the context of the phrases, making it easier to prepare documents to be submitted to the Patent Office.
[0040] Furthermore, the terminal control unit 116 may have a function to configure the user interface of the user terminal 2 as follows: (Configuration 1) While writing a sentence in the document creation unit 211, the timing at which a phrase, a phrase, or a paragraph is completed may be automatically detected. At these timings, text information for the phrase, phrase, or paragraph may be transferred to the prompt unit 212. The document creation support device 1 may infer the text following the text information and display it as support information in the support information unit 213. At this time, the portion of the document creation unit 211 corresponding to the text information transferred to the prompt unit 212 may be highlighted. In this case, the user 4 can easily determine which phrase, phrase, or paragraph the support information corresponds to. Note that instead of automatically detecting the timing at which a phrase, phrase, or paragraph is completed, a fifth instruction button may be provided, and support information inferring the text following the phrase, phrase, or paragraph may be displayed when the fifth instruction button is clicked.
[0041] (Configuration 2) The document creation unit 211 may be provided with a sixth instruction button for requesting advice on drafting a patent or utility model specification based on the input invention information, such as an outline of the invention and invention documents. For example, when meeting documents are input and the sixth instruction button is clicked, the document creation support device 1 extracts and summarizes key points from the documents and displays them as support information in the support information unit 213. Furthermore, when the sixth instruction button is clicked, for example, the prompt unit 212 displays, for example, "Please read the following document 1 and tell us what you think is not known to those skilled in the art. (Two line breaks) (Document 1 body)." Support information in which important points that may be novel are listed in bullet points, spaced two lines apart, is displayed in the support information unit 213. Next, when the sixth instruction button is clicked, the prompt unit 212 displays, for example, "Please read the following document 1 and tell us what you think would be known to those skilled in the art. (Two line breaks) (Document 1 body)." Support information in which important points that may be novel are listed in bullet points, spaced two lines apart, is displayed in the support information unit 213.
[0042] Furthermore, when user 4 considers and inputs the key points of the invention with reference to the support information (important points) and then clicks the sixth instruction button, the document creation support device 1 extracts prior art documents from the specification database 121, identifies elements that may be novel and / or involve an inventive step, and elements that are suspected to be well-known art, and displays each element as support information. Note that in configuration 2, in order to improve operability for user 4, a user interface is provided in which successive clicks of the sixth instruction button sequentially perform the novelty extraction process, the inventive step extraction process, and the element evaluation process. However, this is not limited to this, and an instruction button may be provided for each evaluation process. Furthermore, multiple types of prompts indicating the novelty extraction process, the inventive step extraction process, and the element evaluation process may be selectably displayed in the prompt section 212 each time the sixth instruction button is clicked.
[0043] (Configuration 3) The document creation unit 211 may be provided with a seventh instruction button for creating a proposed claim. For example, when invention points are input and the seventh instruction button is subsequently clicked, the document creation support device 1 generates a proposed claim based on the specification database 121 and displays it as support information. Furthermore, when a proposed claim is input and the seventh instruction button is clicked, a proposed overall structure of the specification may be displayed as support information. Specifically, when a proposed claim is input and the seventh instruction button is clicked, the prompt unit 212 displays "Please indicate the proposed overall structure of the specification corresponding to the following claims. (2 line breaks) (Claims)" and support information indicating the proposed overall structure of the specification is displayed in the support information unit 213. Furthermore, when the seventh instruction button is clicked, the prompt unit 212 displays "Please indicate an example of the description of the following components in an embodiment. If there are any references to materials, please also indicate them. (2 line breaks) (Components indicated in the proposed overall structure)" and the support information unit 213 displays the example of the embodiment and the references to materials.
[0044] Furthermore, after the user terminal 2 creates a draft claim, clicking the seventh instruction button may display a message in the prompt section 212 saying, "Please check whether the following claims include symbols. (2 line breaks) (claims)," and the check results may be displayed in the support information section 213. Alternatively, the prompt section 212 may display a message saying, "For each term in the following claims, please provide a candidate for a generic term. (2 line breaks) (claims)," and the candidate for a generic term may be displayed in the support information section 213. Note that instead of the seventh instruction button, separate instruction buttons may be provided according to the content of each prompt. Furthermore, clicking the seventh button may display multiple types of prompt content that can be selected.
[0045] (Configuration 4) The document creation unit 211 may be provided with an eighth instruction button for checking the format, grammar, etc. of the specification and for checking for inconsistencies in terminology, symbols, etc. For example, when the eighth instruction button is clicked during or after the specification is completed, the document creation support device 1 checks the format, grammar, etc. of the specification and for inconsistencies in terminology, symbols, etc., and displays the results as support information. Furthermore, suggested revisions for various check targets may be displayed as support information.
[0046] (Configuration 5) The document preparation section 211 may be provided with a ninth instruction button for checking whether the proposed amendments and draft opinion can resolve the reasons for refusal (whether opinions are written for all of the reasons for refusal, etc.). Specifically, when the ninth instruction button is clicked, "Please summarize and memorize the following notice of reasons for refusal. (Line breaks x 2) (Notice of Reasons for Refusal)" and "Please check whether the following draft opinion can resolve the reasons for refusal notified in the notice of reasons for refusal that you memorized earlier. Please also indicate suggested amendments. (Line breaks x 2) (Draft opinion)" are displayed in the prompt section 212, and the check results and suggested amendments are displayed in the support information section 213.
[0047] (Configuration 6) The document creation section 211 may be provided with a tenth instruction button for generating a draft drawing such as a flowchart or block diagram or a draft data table corresponding to the content of the input claim or part of the content of the specification, or may be provided with a tenth instruction button for generating a draft claim or a draft specification corresponding to the content of the drawing such as a flowchart or block diagram or the content of the data table.
[0048] (Document Creation Support Program) The activation unit 111, data collection unit 112, model learning unit 113, input processing unit 114, support information generation unit 115, terminal control unit 116, display control unit 117, and input acceptance unit 118 may be configured as either hardware or software. Each of these units 111 to 118 constitutes at least a part of the control unit 11. When each of the units 111 to 118 is configured as software, the control unit 11 is caused to execute a document creation support program that generates support information to support the creation of documents to be submitted to the Patent Office by utilizing a language model for natural language processing specialized for the creation of documents to be submitted to the Patent Office for patents and utility models.
[0049] As a specific example, the document creation support program is a program for causing a computer of a user terminal 2, which is a terminal device accessible to a document creation support device 1, to execute a display processing step (S1) that displays on the user terminal 2: a prompt section 212 that accepts the contents of the prompts, a support information section 213 that displays the support information generated by the support information generation section 115, a document creation section 211 that is an area for creating documents to be submitted to the Patent Office, and a prompt operation section 214 that allows a user 4 to specify fixed phrases indicating predetermined instructions. According to the program, the contents of the prompt section and various support information in the support section can be simultaneously checked while creating documents to be submitted to the Patent Office.
[0050] The document creation support program may also be configured to cause the computer of the user terminal 2 to execute a prompt operation step (S2) in which, when a phrase is specified in the document creation unit 211 and a fixed phrase is specified in the prompt operation unit 214, information combining the specified fixed phrase and the phrase specified in the document creation unit 211 is set as the content input to the prompt unit 212. According to the above program, during the creation of documents to be submitted to the Patent Office, information combining a predetermined instruction content and the specified phrase is set as the content input to the prompt unit 212, thereby facilitating the creation of high-quality documents to be submitted to the Patent Office.
[0051] Furthermore, the document creation support program may be configured to cause the computer of the user terminal 2 to execute processing steps that realize the various functions of the terminal control unit 116.
[0052] The document creation support program configured as described above can support the user 4 in preparing documents to be submitted to the Patent Office in various forms with high accuracy, thereby enabling documents to be submitted to the Patent Office to be prepared quickly and with high quality. Furthermore, simply by installing the program on an information processing device such as a personal computer or tablet terminal, the information processing device can function as the document creation support device 1. The program may be distributed in a state recorded on a recording medium such as a CD-ROM or USB memory, or may be distributed via a two-way communication network or communication line such as the Internet or a one-way communication network such as television broadcasting.
[0053] (Document Creation Support Method) Some or all of the steps in the document creation support program may be implemented by hardware. That is, a user terminal 2 accessible to the document creation support device 1 may be configured to execute the document creation support method. Specifically, the document creation support method is executed on a user terminal 2 accessible to the document creation support device 1, which includes a document database in which patent and utility model document information is stored in a standardized data format so as to be adapted for training a language model for natural language processing, and a support information generation unit 115 that generates support information for supporting the preparation of documents to be submitted to the Patent Office by combining one or more types of natural language processing based on the document database and utilizing a language model that has been trained and adjusted to be specialized for the preparation of documents to be submitted to the Patent Office for patents and utility models, in accordance with the content of a prompt. The document creation support method includes a display processing step (S1). Furthermore, the document creation support method may include a prompt operation step (S2) and processing steps for implementing various functions of the terminal control unit 116.
[0054] In the above description, the document creation support device 1 generates support information by utilizing a language model for natural language processing specialized for creating documents to be submitted to the Japan Patent Office for patents and utility models, but the present invention is not limited to this. That is, the terminal control unit 116 that sets the user interface may be provided in a document creation support device that generates support information using a language model for general-purpose natural language processing such as Chat GPT, and a document creation support program or document creation support method may be applied to a user terminal 2 that can access such a document creation support device.
[0055] Specifically, a document creation support device that is capable of data communication with information sources such as websites and databases via the Internet 5 and generates support information by utilizing a language model for general-purpose natural language processing in accordance with the content of a prompt may be provided with a terminal control unit 116 that sets the user interface of the user terminal 2. Furthermore, the document creation support program may be a program that causes a computer of the user terminal 2 that is capable of data communication with the document creation support device that generates support information using a language model for general-purpose natural language processing to execute various processing steps of this embodiment. The document creation support method may be a method executed on the user terminal 2 that is capable of data communication with the document creation support device that generates support information using a language model for general-purpose natural language processing.
[0056] It should be noted that within the scope of the concept of the present invention, those skilled in the art may conceive of various modifications and alterations. Therefore, it is understood that such modifications and alterations fall within the scope of the present invention. For example, even if a person skilled in the art appropriately adds, deletes, or modifies components of the above-described embodiment, or adds, omits, or modifies the conditions of a process, such modifications are also included within the scope of the present invention as long as they maintain the gist of the present invention.
[0057] 1 Document creation support device 11 Control unit 111 Activation unit 112 Data collection unit 113 Model learning unit 114 Input processing unit 115 Support information generation unit 116 Terminal control unit 117 Display control unit 118 Input reception unit 12 Storage unit 121 Specification database 122 Intermediate document database 123 User database 124 UI database 13 Communication unit 14 Display device 15 Input device 2 User terminal 21 Display screen 211 Document creation unit 212 Prompt unit 213 Support information unit 214 Prompt operation unit 4 User 5 Internet 6 Patent information search site 7 Administrator terminal
Claims
1. A document creation support device comprising: a literature database in which accumulated information on patent and utility model documents is stored in a standardized data format that is adapted to the learning of a language model for natural language processing; and a support information generation unit that generates support information to support the creation of documents to be submitted to the Patent Office by combining one or more types of natural language processing based on the literature database and utilizing a language model that has been trained and adjusted to be specialized for the creation of documents to be submitted to the Patent Office for patents and utility models, in accordance with the content of a prompt.
2. The document creation support device as described in claim 1, characterized in that it has a terminal control unit that controls the display on a user terminal of a prompt unit that accepts the contents of the prompt, a support information unit that displays the support information generated in the support information generation unit, and a document creation unit that is an area for creating documents to be submitted to the Patent Office.
3. The document creation support device described in claim 2, characterized in that the terminal control unit allows the user to specify a standard phrase indicating a predetermined instruction content, and controls the user terminal to display a prompt operation unit in which information combining the specified standard phrase with a word specified in the document creation unit is the content input to the prompt unit.
4. The document creation support device described in claim 2, characterized in that the terminal control unit allows the user to specify a standard phrase indicating a predetermined instruction content, and controls the user terminal to display a prompt operation unit in which information combining the specified standard phrase with a phrase including a word specified in the document creation unit is the content input to the prompt unit.
5. A program for executing, on a computer of a terminal device accessible to a document creation support device having a literature database in which accumulated document information on patents and utility models is stored in a state normalized into a unified data format so as to be adapted to the training of a language model for natural language processing, and a support information generation unit that generates support information to support the preparation of documents to be submitted to the Patent Office by combining one or more types of natural language processing based on the literature database and utilizing a language model that has been trained and adjusted to be specialized for the preparation of documents to be submitted to the Patent Office for patents and utility models according to the content of a prompt, a display processing step for displaying on the terminal device a prompt unit that receives the content of the prompt, a support information unit that displays the support information generated in the support information generation unit, a document creation unit that is an area for creating documents to be submitted to the Patent Office, and a prompt operation unit that allows a user to specify a standard phrase indicating a predetermined instruction content, and a prompt operation step for setting information combining the specified standard phrase and the word or phrase specified in the document creation unit as the content input to the prompt unit when the standard phrase is specified in the prompt operation unit with a phrase specified in the document creation unit.
Citation Information
Patent Citations
Sentence structure drawing device
JP2021120790A
Patent text generation device, patent text generation method, and patent text generation program
JP7086424B1
