Text generation system and text generation method

Abstract

To provide a document generation system for supporting the understanding of both an inventor and a patent engineer and supporting the preparation of a proper claim to an invention in a patent application.SOLUTION: This system is provided with a function for performing processing using a language model, a function for preparing the gist of the invention from invention data, a function for preparing a claim plan from the gist of the invention and a function for preparing a retrieval expression from the claim plan. The sentence generation system has a function of performing patent retrieval using the retrieval expression.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a language model, in particular to a sentence generation system and a sentence generation method that utilize a generative AI model.

[0002] Note that one embodiment of the present invention is not limited to the above technical field. Examples of the technical field of one embodiment of the present invention disclosed in this specification and the like include semiconductor devices, display devices, light-emitting devices, power storage devices, memory devices, electronic devices, lighting devices, input devices, input / output devices, driving methods thereof, and manufacturing methods thereof. A semiconductor device refers to any device that can function by utilizing semiconductor characteristics. [Background technology]

[0003] In recent years, there has been active development of language models using neural networks, with large-scale language models (LLMs) attracting particular attention. Large-scale language models are natural language processing models trained using large amounts of data. Large-scale language models can realize, for example, dialogue models that respond to user instructions. Non-Patent Document 1 discloses GPT-4 (Generative Pre-trained Transformer 4) (registered trademark) as a large-scale language model, and ChatGPT as a dialogue model. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Summary of ChatGPT / GPT-4 Research and Perspective Towards the Future of Large Language Models, Yiheng Liu et al. (Submitted on 4 Apr 2023, [online], Internet<URL:https: / / arxiv.org / abs / 2304.01852> Summary of the Invention [Problem to be solved by the invention]

[0005] When filing a patent application, an inventor can explain the technical content of his or her invention, but in order to clearly and completely describe that technical content in the claims, the assistance of patent engineers is required. However, patent engineers often do not have sufficient background knowledge of the technical content of the invention, making it difficult for them to fully understand the invention.

[0006] In view of the above problems, an object of the present invention is to support the understanding of both inventors and patent engineers and to support the creation of appropriate claims for inventions.

[0007] Note that the description of these problems does not preclude the existence of other problems. Note that one embodiment of the present invention does not necessarily solve all of these problems. Note that problems other than these can be extracted from the description of the specification, drawings, claims, etc. [Means for solving the problem]

[0008] In view of the above problems, one aspect of the present invention is a text generation system having a first information processing device, a second information processing device, and a third information processing device. The first information processing device has a function of receiving at least first to third instruction sentences, performing processing using a language model, and outputting at least first to third response sentences. The second information processing device has a function of receiving a search formula, performing a patent search, and outputting patent search results. The third information processing device is a text generation system having a function of accepting invention materials, a function of creating a first instruction sentence from the invention materials, a function of sending the first instruction sentence to the first information processing device and receiving a first response sentence including the gist of the invention, a function of outputting the gist of the invention, a function of creating a second instruction sentence from the invention materials and the gist of the invention, a function of sending the second instruction sentence to the first information processing device and receiving a second response sentence including a proposed claim, a function of creating a third instruction sentence from the proposed claim, a function of sending the third instruction sentence to the first information processing device and receiving a third response sentence including a search formula, and a function of sending the search formula to the second information processing device and receiving patent search results.

[0009] In the above-mentioned text generation system, it is preferable that the third information processing device has the functions of searching the invention materials for the contents of the invention and obtaining the invention material search results, creating a fourth instruction sentence from the invention material search results, sending the fourth instruction sentence to the first information processing device and receiving a fourth response sentence including element information of the invention from the first information processing device, storing the element information of the invention, and creating the gist of the invention from the element information of the invention.

[0010] In the above-mentioned text generation system, it is preferable that the third information processing device has a function of creating a fifth instruction sentence from the proposed claim, a function of sending the fifth instruction sentence to the first information processing device and receiving a fifth response sentence including search elements and components from the first information processing device, and a function of optimizing the proposed claim from the search elements and components using combinatorial optimization.

[0011] In the above-mentioned text generation system, it is preferable that the third information processing device has the following functions: a function of creating a fifth instruction sentence from the proposed claim; a function of sending the fifth instruction sentence to the first information processing device and receiving a fifth response sentence including a component and a search element from the first information processing device; a function of creating a sixth instruction sentence from pairs of components and search elements and the number of patent search results searched for each search element; and a function of sending the sixth instruction sentence to the first information processing device and receiving the sixth response sentence including the optimized proposed claim from the first information processing device.

[0012] In addition, one aspect of the present invention has first to eleventh steps, in which in the first step, invention materials are received, in the second step, a first instruction sentence is created from the invention materials, in the third step, the first instruction sentence is input into a language model and a first response sentence including the gist of the invention is obtained, in the fourth step, the gist of the invention is output to the user, in the fifth step, the gist of the invention revised by the user is received, in the sixth step, a second instruction sentence is created from the invention materials and the revised gist of the invention, and in the seventh step, In step 1, a second instruction sentence is input into a language model to obtain a second response sentence including a proposed claim; in step 8, a third instruction sentence is created from the proposed claim; in step 9, the third instruction sentence is input into the language model to obtain the third response sentence including a search formula; in step 10, the search formula is input into a search server to obtain patent search results; and in step 11, the patent search results and the proposed claim are output to the user, and steps 1 to 11 are executed by a processing unit in the above order.

[0013] In the above-mentioned sentence generation method, it is preferable that the method has a 12th step, which is executed after the 10th step, and the 12th step determines whether the number of documents in the patent search results is within a predetermined range, and if it is not within the predetermined range, returns to the 8th step, and if it is within the predetermined range, proceeds to the 11th step.

[0014] In the above-mentioned sentence generation method, it is preferable to have a thirteenth step, which is executed after the eleventh step, in which the user checks the patent search results and the proposed claims, and if the proposed claims need to be revised, return to the fifth step, and if the proposed claims do not need to be revised, end the execution of the steps.

[0015] In the above-mentioned sentence generation method, it is preferable to have steps 14 to 19 instead of steps 2 to 4, in which step 14 searches the invention materials with a question about the content of the invention and obtains the invention material search results, step 15 creates a fourth instruction sentence from the invention material search results, step 16 inputs the fourth instruction sentence into a language model to obtain a fourth answer sentence including element information of the invention, step 17 accumulates the element information of the invention, step 18 repeats steps 14 to 17 until the last question is asked, and step 19 outputs the accumulated element information of the invention as the main points of the invention.

[0016] In the above-mentioned sentence generation method, it is preferable that the eleventh step is followed by the twentieth to twenty-second steps, in which the twentieth step creates a fifth instruction sentence from the proposed claim, the twenty-first step inputs the fifth instruction sentence into a language model to obtain a fifth answer sentence including a search element and a component, and the twenty-second step optimizes the proposed claim from the search element and the component using combinatorial optimization.

[0017] In the above sentence generation method, it is preferable to have steps 23 and 24 instead of step 22, in which step 23 creates a sixth instruction sentence from a pair of a component and a search element and the number of patent search results searched for each search element, and step 24 inputs the sixth instruction sentence into a language model to obtain a sixth response sentence including an optimized claim proposal. [Effects of the Invention]

[0018] According to one aspect of the present invention, it is possible to support the understanding of both inventors and patent engineers, to support the creation of appropriate claims for inventions, and to perform appropriate patent searches for inventions.

[0019] Note that the description of these effects does not preclude the existence of other effects. Note that one embodiment of the present invention does not necessarily have all of these effects. Note that effects other than these can be extracted from the description in the specification, drawings, claims, etc. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of a sentence generation system. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of a sentence generation system. [Figure 3] FIG. 3 is a flowchart showing an example of a sentence generation method. [Figure 4] FIG. 4 is a flowchart showing an example of a sentence generation method. [Figure 5] FIG. 5 is a flowchart showing an example of a sentence generation method. [Figure 6] FIG. 6 is a flowchart showing an example of a sentence generation method. [Figure 7] FIG. 7 is a flowchart showing an example of a sentence generation method. [Figure 8] FIG. 8 is a flowchart showing an example of a sentence generation method. [Figure 9] FIG. 9 is a flowchart showing an example of a sentence generation method. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, embodiments will be described with reference to the drawings. However, it will be readily understood by those skilled in the art that the embodiments can be implemented in many different ways and that various changes in form and details can be made without departing from the spirit and scope of the present invention. Therefore, the present invention should not be interpreted as being limited to the following description of the embodiments.

[0022] In the configuration of the invention described below, the same parts or parts having similar functions are denoted by the same reference numerals in different drawings, and repeated explanations thereof will be omitted. In addition, when referring to similar functions, there are cases where no particular reference numerals are used.

[0023] In the drawings described in this specification, the size of each component, the thickness of a layer, or an area may be exaggerated for clarity, and therefore, the drawings are not necessarily limited to the scale.

[0024] Note that ordinal numbers such as "first" and "second" used in this specification are used to avoid confusion between components and do not limit the number. They do not indicate any order or ranking, such as the order of processes or stacking. Even if a term does not have an ordinal number in this specification, an ordinal number may be added in the claims to avoid confusion between components. Even if a term has an ordinal number in this specification, a different ordinal number may be added in the claims. Even if a term has an ordinal number in this specification, the ordinal number may be omitted in the claims.

[0025] In this specification, a language model is based on the Transformer architecture and has undergone additional learning to become an interactive (also called conversational) model. A typical language model is a large-scale language model (LLM). An LLM processes data based on given text and is specialized for text generation functions, while a generative AI not only generates text but also has image generation functions that process data based on image data. In other words, a large-scale language model is a type of generative AI.

[0026] (Embodiment 1) In this embodiment, a configuration example of a sentence generation system according to one embodiment of the present invention will be described with reference to FIG.

[0027] The text generation system and text generation method of one aspect of the present invention utilize a language model to output the main points of an invention from the invention documents held by the user. Furthermore, by creating a claim proposal from the main points of the invention and performing a patent search, it becomes possible to support the writing of claims for an invention and appropriate patent searches for the invention.

[0028] A more specific example will be described below with reference to the drawings.

[0029] <Example of text generation system configuration 1> 1, the sentence generation system of this embodiment preferably has a first information processing device 10, a second information processing device 20, a third information processing device 30, and an information terminal 40. The third information processing device 30 is connected to the first information processing device 10 and the second information processing device 20 via a network 50. The third information processing device 30 is also connected to the information terminal 40 via a network 60.

[0030] Furthermore, in the text processing device of this embodiment, the device configuration of Figure 1 is one example, and the third information processing device 30 may have the functions of at least one of the first information processing device and the second information processing device, or the functions of the third information processing device 30 may be distributed and implemented in multiple information processing devices.

[0031] <<Example of information terminal configuration>> In the configuration example 1 of the sentence generation system, the information terminal 40 is operated by a user and can also be called a client computer. While a desktop computer and a smartphone are shown as examples in Fig. 1, a notebook computer or a tablet computer can also be used as the information terminal 40. A tablet computer can be used by connecting it to a housing having an input unit (typically a keyboard).

[0032] <<Configuration Example of First Information Processing Device 10>> Next, an example of the configuration of the first information processing device 10 will be described.

[0033] The first information processing device 10 can perform processing using a language model. In particular, the first information processing device 10 can perform processing using a model that uses a large-scale language model (such as a sentence generation model or a dialogue model). For example, the processing can be performed using a large-scale language model such as GPT-4, Llama2, or Llama3. Note that in this specification and the like, a language model includes a large-scale language model.

[0034] The first information processing device 10 has a function of outputting a response sentence to a command sentence using a language model, and is capable of executing various natural language processing tasks such as translation and summarization.

[0035] A person who provides a service using the sentence generation system according to one embodiment of the present invention does not necessarily have to own the first information processing device 10. For example, a service provider can use part of a service provided by another business operator as the first information processing device 10.

[0036] <<Configuration Example of Second Information Processing Device 20>> The second information processing device 20 has a function of accepting a search formula and conducting a search for patents.

[0037] In this specification, "search" means to find highly relevant text data from among a plurality of text data.

[0038] When performing a search, an index may be created from multiple pieces of text data. An "index" is an index created from text data. Creating an index enables high-speed searches.

[0039] When performing a search, text data may be converted into vector data, and the degree of similarity between the vector data may be determined. "Vector data" refers to a multidimensional array of numbers from 0 to 9, as opposed to text data consisting of character strings (natural language) such as sentences. Vector data can also be said to be data in a format that can be processed numerically. Specific methods for converting text data into vector data include bag of words, distributed representation, and embedded representation. Cosine similarity is one index that indicates the degree of similarity between vector data.

[0040] In this specification, a "search formula" is a character string that includes at least a search word. A search formula may include multiple search words. A search formula may also include search conditions, classification codes, etc.

[0041] A person who provides a service using the sentence generation system according to one embodiment of the present invention does not necessarily need to own the second information processing device 20. For example, a service provider can use part of a service provided by another business operator as the second information processing device 20.

[0042] <<Configuration example of third information processing device 30>> Next, an example of the configuration of the third information processing device 30 will be described with reference to FIG.

[0043] As shown in Fig. 2, the third information processing device 30 has a reception unit 110, an output unit 120, a storage unit 130, a processing unit 140, and a transmission path 150. In Fig. 2, in addition to the first information processing device 10, a second information processing device 20 and an information terminal 40 are shown, and arrows indicate transmission and reception of data. The reception unit and output unit may be collectively referred to as a communication unit. The communication unit enables the third information processing device 30 to transmit and receive data to and from the outside.

[0044] [Reception Section 110] The receiving unit 110 has a function of receiving data from outside. For example, the receiving unit 110 can receive a reply sentence from the first information processing device 10. As the receiving unit 110, for example, a communication port or an input device such as a personal computer equipped with a communication function may be used.

[0045] The data received by the receiving unit 110 from the first information processing device 10 includes, for example, the gist of the invention output by the language model.

[0046] The receiving unit 110 can supply the received data to one or more selected from the storage unit 130 and the processing unit 140 via the transmission path 150 .

[0047] [Output section 120] The output unit 120 has a function of outputting calculation results and the like to the outside. For example, the output unit 120 can transmit an instruction statement to the first information processing device 10. As the output unit 120, for example, a communication port may be used, or a device such as a personal computer equipped with a communication function may be used.

[0048] [Storage section 130] The storage unit 130 has a storage function. The storage unit 130 is a memory area and can store programs and / or data. A representative example of the program is a program executed by the processing unit 140. The data includes data received by the reception unit 110 (e.g., invention documents, patent search results). The data also includes data generated by a language model (e.g., gist of an invention, proposed claims).

[0049] The storage unit 130 may have a database. Furthermore, the first information processing device 10 may have a database separate from the storage unit 130. The first information processing device 10 may have a function to retrieve data from a database that exists outside the storage unit 130, outside the first information processing device 10, or outside the information processing system. Furthermore, the first information processing device 10 may have a function to retrieve data from both its own database and an external database.

[0050] Either or both of a storage and a file server can be used as the storage unit 130. Also, a database that records paths of files stored in a file server can be used as the storage unit 130.

[0051] The storage unit 130 includes at least one of a volatile memory and a nonvolatile memory. Examples of the volatile memory include a dynamic random access memory (DRAM) and a static random access memory (SRAM). Examples of the nonvolatile memory include a resistive random access memory (ReRAM), a phase change random access memory (PRAM), a ferroelectric random access memory (FeRAM), a magnetoresistive random access memory (MRAM), and a flash memory. The storage unit 130 may also include at least one of NOSRAM (registered trademark) and DOSRAM (registered trademark). The storage unit 130 may also include a recording media drive. Examples of the recording media drive include a hard disk drive (HDD) and a solid state drive (SSD).

[0052] NOSRAM is an abbreviation for "Nonvolatile Oxide Semiconductor Random Access Memory (RAM)." NOSRAM is a type of memory in which memory cells are two-transistor (2T) or three-transistor (3T) gain cells, and the transistors are transistors (also called OS transistors) that use metal oxides in the channel formation region. OS transistors have extremely low leakage current, i.e., the current that flows between the source and drain in the off state. NOSRAM can be used as a nonvolatile memory by retaining a charge corresponding to data in the memory cell using its extremely low leakage current. NOSRAM is particularly suitable for arithmetic processing that requires repeated large amounts of data read operations because it can read stored data without destroying it (nondestructive read). NOSRAM can increase its data capacity by stacking layers, so it can be used as a large-scale cache memory, main memory, or storage memory, thereby improving the performance of semiconductor devices.

[0053] DOSRAM is an abbreviation for "Dynamic Oxide Semiconductor RAM" and refers to RAM with 1T (transistor) 1C (capacitance) type memory cells. DOSRAM is a DRAM formed using OS transistors, and is a memory that temporarily stores information sent from outside. DOSRAM is a memory that takes advantage of the small off-current of OS transistors.

[0054] In this specification and the like, the term "metal oxide" refers to an oxide of a metal in a broad sense. Metal oxides are classified into oxide insulators, oxide conductors (including transparent oxide conductors), oxide semiconductors (also referred to as oxide semiconductors or simply as OSs), and the like. For example, when a metal oxide is used in a semiconductor layer of a transistor, the metal oxide may be referred to as an oxide semiconductor.

[0055] The metal oxide contained in the channel formation region preferably contains indium (In). When the metal oxide contained in the channel formation region contains indium, the carrier mobility (electron mobility) of the OS transistor is increased. For example, indium oxide can be suitably used as the metal oxide contained in the channel formation region. Furthermore, the metal oxide contained in the channel formation region is preferably an oxide semiconductor containing element M. Element M is preferably at least one of aluminum (Al), gallium (Ga), and tin (Sn). Other elements applicable to element M include boron (B), silicon (Si), titanium (Ti), iron (Fe), nickel (Ni), germanium (Ge), yttrium (Y), zirconium (Zr), molybdenum (Mo), lanthanum (La), cerium (Ce), neodymium (Nd), hafnium (Hf), tantalum (Ta), and tungsten (W). However, the element M may be a combination of two or more of the above elements. The element M is, for example, an element having a high bond energy with oxygen. For example, it is an element having a higher bond energy with oxygen than indium. Furthermore, the metal oxide contained in the channel formation region is preferably a metal oxide containing zinc (Zn). Metal oxides containing zinc may be easily crystallized. For example, indium gallium zinc oxide (IGZO) or indium tin oxide (ITZO (registered trademark)) can be used as the metal oxide contained in the channel formation region.

[0056] The metal oxide contained in the channel formation region is not limited to a metal oxide containing indium, but may be, for example, a metal oxide containing zinc but not indium, such as zinc tin oxide or gallium tin oxide, a metal oxide containing gallium, or a metal oxide containing tin.

[0057] [Processing section 140] The processing unit 140 has a function of performing processing such as calculation and analysis using data supplied from one or both of the receiving unit 110 and the storage unit 130. The processing unit 140 can supply the processed data to one or both of the storage unit 130 and the output unit 120.

[0058] The processing unit 140 may include, for example, an arithmetic circuit. The processing unit 140 may include, for example, a central processing unit (CPU). The processing unit 140 may also include a graphics processing unit (GPU) in addition to or instead of the CPU. The processing unit 140 may also include an NPU (neural processing unit / neural network processing unit).

[0059] The processing unit 140 may have a register and a main memory in addition to a CPU. The register and main memory may also be said to be owned by the CPU. The main memory is capable of sending and receiving data to and from a secondary cache, etc. The main memory has at least one of a volatile memory such as RAM (Random Access Memory) and a non-volatile memory such as ROM (Read Only Memory). The main memory may also have at least one of NOSRAM and DOSRAM. The main memory can have one or both of OS transistors and Si transistors. The configuration of the register and main memory can be understood by replacing CPU in this paragraph with GPU.

[0060] Examples of RAM include DRAM and SRAM. The DRAM or SRAM can also be used as a working space for the processing unit 140 by virtually allocating memory space thereto. The operating system, application programs, program modules, program data, lookup tables, and the like stored in the storage unit 130 are loaded into the RAM immediately before execution. The operating system, application programs, program modules, program data, and lookup tables loaded into the RAM can be accessed by the processing unit 140.

[0061] ROM can store systems that do not require rewriting. Examples of systems that do not require rewriting include BIOS (Basic Input / Output System) and firmware. Examples of ROM include mask ROM, OTPROM (One-Time Programmable Read Only Memory), and EPROM (Erasable Programmable Read Only Memory). Examples of EPROM include UV-EPROM (Ultra-Violet Erasable Programmable Read Only Memory), which allows stored data to be erased by exposure to ultraviolet light, EEPROM (Electrically Erasable Programmable Read Only Memory), and flash memory.

[0062] The processing unit 140 may have a microprocessor such as a DSP (Digital Signal Processor) in addition to a CPU or GPU. Since a DSP is specialized for digital signal processing, it is preferable to install it to control peripheral circuits of the CPU or GPU. The microprocessor may be realized by a PLD (Programmable Logic Device) that operates on hardware such as an FPGA (Field Programmable Gate Array) or an FPAA (Field Programmable Analog Array).

[0063] The processing unit 140 allows the third information processing device 30 to have a function of creating directives from text data. Specifically, it has a function of creating a first directive from invention materials, a function of creating a second directive from the main points of the invention, a function of creating a third directive from a proposed claim, etc.

[0064] Furthermore, the processing unit 140 may have a function in which the third information processing device 30 searches invention materials for the contents of the invention and creates a fourth instruction sentence from the invention material search results.

[0065] [Transmission Line 150] The transmission path 150 has a function of transmitting data. Data can be transmitted and received between the reception unit 110, the output unit 120, the storage unit 130, and the processing unit 140 via the transmission path 150. The transmission path 150 may be, for example, an external bus, a LAN (Local Area Network), or the Internet, which is the foundation of the World Wide Web (WWW).

[0066] <Example of sentence generation method> 3 and 4 are examples of flowcharts showing steps in a sentence generation method according to one aspect of the present invention.

[0067] <Step S101> In step S101, the third information processing device 30 receives invention materials from a user.

[0068] Invention documents are a collection of materials and data related to an invention. For example, experimental reports on the invention are examples. It is also preferable that invention documents include at least one of the following: technical field, problem, effect, and specific example.

[0069] The user's input operation corresponding to step S101 is performed at the information terminal 40. The invention material is transmitted to the third information processing device 30 via the reception unit 110 and stored in the storage unit .

[0070] <Step S102> In step S102, the processing unit 140 of the third information processing device 30 creates a first instruction sentence from the invention material.

[0071] The instruction sentence is an input sentence that causes the language model to perform a desired operation, and the language model generates a response sentence based on the instruction sentence. The first instruction sentence is a string of characters intended to cause the language model to generate the gist of the invention. The first instruction sentence is, for example, a string of characters such as "Please extract the gist of the invention from the following invention materials" and a string of characters including the invention materials.

[0072] <Step S103> In step S103, the third information processing device 30 transmits the first instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the first instruction sentence into a language model and generates a first answer sentence including the gist of the invention. The third information processing device 30 acquires the first answer sentence via the reception unit 110.

[0073] <Step S104> In step S104, the third information processing device 30 transmits the gist of the invention to the information terminal 40 via the output unit 120, and the information terminal 40 displays the gist of the invention to the user.

[0074] <Step S105> In step S105, the third information processing device 30 acquires the gist of the invention from the information terminal 40 via the receiving unit 110. At this time, the user may modify the gist of the invention if necessary.

[0075] <Step S106> In step S106, the processing unit 140 of the third information processing device 30 creates a second directive from the invention materials and the gist of the invention.

[0076] The second instruction is a string intended to cause the language model to generate a proposed claim. For example, the second instruction is a string such as "Please generate a proposed claim from the following invention documents and gist of the invention" and a string including the invention documents and gist of the invention.

[0077] <Step S107> In step S107, the third information processing device 30 transmits the second instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the second instruction sentence into the language model and generates a second response sentence including the proposed claim. The third information processing device 30 acquires the second response sentence via the reception unit 110.

[0078] In step S107, the second response preferably includes a correspondence table between each element of the proposed claim and the corresponding part of the invention document, which makes it easier for both the inventor and the patent engineer to understand the content of the proposed claim.

[0079] <Step S108> In step S108, the processing unit 140 of the third information processing device 30 creates a third directive from the proposed claim.

[0080] The third directive is a string intended to cause the language model to generate a search query, such as "Please generate a search query for patents from the following claims" and a string containing proposed claims.

[0081] <Step S109> In step S109, the third information processing device 30 transmits the third instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the third instruction sentence to the language model and generates a third answer sentence including the search formula. The third information processing device 30 acquires the third answer sentence via the reception unit 110.

[0082] <Step S110> In step S110, the third information processing device 30 transmits the search formula to the second information processing device 20 via the output unit 120. The second information processing device 20 inputs the search formula to the search server and obtains the patent search results. The third information processing device 30 obtains the patent search results via the reception unit 110. The patent search results are a list of documents, for example, a list of patent numbers.

[0083] The patent search results obtained in step S110 are preferably stored in the storage unit 130.

[0084] <Step S111> In step S111, the third information processing device 30 transmits the proposed claim and the patent search results to the information terminal 40 via the output unit 120, and the information terminal 40 displays the proposed claim and the patent search results to the user.

[0085] The sentence generation method according to one aspect of the present invention can end after step S111.

[0086] By using this text generation method, it is possible to create a claim draft that is easy to understand for both inventors and patent engineers, and to obtain patent search results that are useful for inventions.

[0087] <Modification 1 of the sentence generation method> Fig. 5 shows a flow of a modified example of the sentence generation method, in which step S112 is added between step S110 and step S111.

[0088] <Step S112> In step S112, the third information processing device 30 determines whether the number of documents in the patent search results obtained in step S110 is within a predetermined range, and if it is within the range, proceeds to step S111, and if it is outside the range, returns to step S108.

[0089] When returning to step S108, the third information processing device 30 again creates a third instruction sentence from the proposed claim in the processing unit 140. At this time, it is preferable to add the search formula obtained in step S109 or a third response sentence including the search formula to the third instruction sentence. For example, if the number of documents in the patent search results is greater than a predetermined range, it is preferable to add an instruction to generate a search formula with a more limited search range than the search formula. Also, for example, if the number of documents is less than a predetermined range, it is preferable to add an instruction to generate a search formula with an expanded search range than the search formula.

[0090] In step S112, the range that serves as the criterion is preferably set by the user.

[0091] <Variation 2 of the sentence generation method> Fig. 6 shows a flow of a modified example of the sentence generation method, in which step S113 is added after step S111.

[0092] <Step S113> In step S113, the third information processing device 30 receives the user's confirmation results for the patent search results and the proposed claims output in step S111 via the reception unit 110. If the user determines that the gist of the invention needs to be revised, the process returns to step S105; if the user determines that no revision is necessary, the process ends after step S113.

[0093] <Variation 3 of the sentence generation method> Fig. 7 is a flow chart of a modified example of the sentence generation method, in which steps S114 to S119 are executed instead of steps S102 to S104.

[0094] <Step S114> In step S114, the third information processing device 30 searches the invention materials with the question about the contents of the invention in the processing unit 140. As a result, the invention material search results are obtained.

[0095] Examples of questions about the content of the invention include "What is the technical field?" and "What is the problem of the invention?" It is preferable that the specific content of the questions be set by the user. It is also possible to set multiple questions about the content of the invention, and steps S114 to S117 are repeated for each question.

[0096] <Step S115> In step S115, the processing unit 140 of the third information processing device 30 creates a fourth instruction sentence from the invention material search results.

[0097] The fourth directive is a string intended to cause the language model to generate element information of the invention, such as a string containing a question about the content of the invention and a string containing the search results for invention documents.

[0098] <Step S116> In step S116, the third information processing device 30 transmits the fourth instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the fourth instruction sentence into the language model and acquires a fourth answer sentence including element information of the invention. The third information processing device 30 acquires the element information of the invention via the reception unit 110.

[0099] <Step S117> In step S117, the third information processing device 30 stores the element information of the invention in the storage unit .

[0100] <Step S118> In step S118, in the third information processing device 30, if the question about the contents of the invention in step S114 is not the last, the process returns to step S114, and if it is the last, the process proceeds to step S119.

[0101] <Step S119> In step S119, the third information processing device 30 creates a gist of the invention from the accumulated element information of the invention via the output unit 120. Here, the gist of the invention may be a list or table of the element information of the invention. The third information processing device 30 transmits the gist of the invention to the information terminal 40, and the information terminal 40 outputs the gist of the invention.

[0102] <Modification 4 of the sentence generation method> Fig. 8 is a flow chart of a modified example of the sentence generation method, in which steps S120 to S122 are executed between steps S110 and S111.

[0103] <Step S120> In step S120, the processing unit 140 of the third information processing device 30 creates a fifth directive from the proposed claim.

[0104] The fifth directive is a string intended to generate invention components and search elements from the language model, such as "Extract invention components from the following claims and create search elements corresponding to each component" and a string containing a proposed claim.

[0105] <Step S121> In step S121, the third information processing device 30 transmits a fifth instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the fifth instruction sentence into the language model and acquires a fifth answer sentence including the components of the invention and the search elements. The third information processing device 30 acquires the components of the invention and the search elements via the reception unit 110.

[0106] The elements of the invention are the technical features of the claims, and the search elements are character strings used to search for those technical features.

[0107] <Step S122> In step S122, the third information processing device 30 performs combinatorial optimization of the components using the search elements in the processing unit 140. For combinatorial optimization, it is preferable to use, for example, a genetic algorithm.

[0108] For example, in a sentence generation method, components of the invention are randomly selected and combined, and combinatorial optimization is performed to search for appropriate combinations of elements using a genetic algorithm to find the optimal combination.

[0109] For example, in the sentence generation method, it is preferable to use the number of search results for the patent search obtained in step S110 by the combination of search elements as the fitness of the genetic algorithm for claim draft optimization. In this case, it is preferable to search for a combination with the fewest elements that has the fewest search results, and to use the optimal combination of elements as an independent claim of the claim, and the unused combination as a dependent claim of the claim.

[0110] <Modification 5 of the sentence generation method> 9 is a flow chart of a modified example of the sentence generation method, in which steps S123 and S124 are executed instead of step S122.

[0111] <Step S123> In step S123, the processing unit 140 of the third information processing device 30 creates a sixth directive from the pair of the constituent element and the search element and the number of patent search results searched for each search element.

[0112] The sixth directive is a string intended to allow the language model to optimize the proposed claim, such as "Please optimize the proposed claim using pairs of components and search elements and the number of patent search results for each search element," and a string including the proposed claim, pairs of components and search elements, and the number of patent search results.

[0113] <Step S124> In step S121, the third information processing device 30 transmits the sixth instruction sentence to the first information processing device 10 via the output unit 120. The first information processing device 10 inputs the sixth instruction sentence into the language model and obtains a sixth response sentence including an optimized proposed claim. The third information processing device 30 obtains the optimized proposed claim via the reception unit 110.

[0114] The sentence generation method according to one aspect of the present invention can support the process from invention to claim writing, and also allows for an appropriate patent search for an invention. [Explanation of symbols]

[0115] 10. Information processing equipment 20 Information processing equipment 30 Information processing equipment 40 Information terminal 50 Network 60 Network 110 Reception 120 Output section 130 Storage section 140 Processing section 150 Transmission Line

Claims

1. a first information processing device; a second information processing device; a third information processing device; the first information processing device has a function of receiving at least first to third instruction sentences, performing processing using a language model, and outputting at least first to third response sentences; the second information processing device has a function of receiving a search formula, conducting a patent search, and outputting a patent search result; The third information processing device A function to accept invention documents; a function of creating the first instruction sentence from the invention materials; a function of transmitting the first instruction sentence to the first information processing device and receiving the first reply sentence including the gist of the invention; a function of outputting the gist of the invention; a function of creating the second instruction sentence from the invention materials and the gist of the invention; a function of transmitting the second instruction sentence to the first information processing device and receiving the second reply sentence including the proposed claim; generating the third directive from the proposed claim; a function of transmitting the third instruction sentence to the first information processing device and receiving the third answer sentence including the search formula; A sentence generation system having a function of transmitting the search formula to the second information processing device and receiving the patent search results.

2. In claim 1, The third information processing device A function of searching the invention documents for the contents of the invention and obtaining the invention document search results; a function of creating a fourth instruction sentence from the invention material search results; a function of transmitting the fourth instruction sentence to the first information processing device and receiving a fourth response sentence including element information of the invention from the first information processing device; A function of accumulating element information of the invention; and a function of creating the gist of the invention from the element information of the invention.

3. In claim 1, The third information processing device generating a fifth directive from the proposed claim; a function of transmitting the fifth instruction sentence to the first information processing device and receiving a fifth answer sentence including a search element and a component from the first information processing device; and a function of optimizing the proposed claim from the search elements and the components using combinatorial optimization.

4. In claim 1, The third information processing device generating a fifth directive from the proposed claim; a function of transmitting the fifth instruction sentence to the first information processing device and receiving a fifth answer sentence including a search element and a component from the first information processing device; a function of creating a sixth instruction sentence from the pair of the search element and the component and the number of patent search results searched for each of the search elements; a function of transmitting the sixth instruction sentence to the first information processing device and receiving a sixth response sentence including the optimized proposed claim from the first information processing device.

5. The method includes steps 1 to 11, In the first step, invention materials are received, In the second step, a first instruction is created from the invention materials; In the third step, the first instruction sentence is input into a language model to obtain a first response sentence including a gist of the invention; In the fourth step, the gist of the invention is output to a user; In the fifth step, the gist of the invention as modified by the user is received, In the sixth step, a second instruction is created from the invention documents and the modified gist of the invention; In the seventh step, the second instruction sentence is input to the language model to obtain a second response sentence including a proposed claim; In the eighth step, a third directive is generated from the proposed claim; In the ninth step, the third instruction sentence is input to the language model to obtain a third answer sentence including the search formula; In the tenth step, the search formula is input into a search server to obtain patent search results; In the eleventh step, the patent search results and the proposed claims are output to the user; A sentence generation method, wherein the first to eleventh steps are executed by a processing unit in the above order.

6. In claim 5, the method further comprises a twelfth step, The twelfth step is performed after the tenth step; The twelfth step determines whether the number of documents in the patent search results is within a predetermined range, and if it is not within the predetermined range, returns to the eighth step; If the result is within the predetermined range, the method proceeds to the eleventh step.

7. In claim 5, the method further comprises a thirteenth step, The thirteenth step is performed after the eleventh step, In the thirteenth step, the user confirms the patent search results and the proposed claim, and if the proposed claim needs to be revised, the process returns to the fifth step, and if the proposed claim does not need to be revised, the process ends.

8. In claim 5, The method has fourteenth to nineteenth steps instead of the second to fourth steps, The fourteenth step is to search the invention documents with questions about the contents of the invention, and obtain the invention document search results; The fifteenth step creates a fourth instruction sentence from the invention material search result, the sixteenth step inputs the fourth instruction sentence into the language model to obtain a fourth answer sentence including element information of the invention; The seventeenth step is to store the element information of the invention; The eighteenth step repeats the fourteenth to seventeenth steps until the last question is asked; The nineteenth step is a sentence generation method for outputting the accumulated element information of the invention as the gist of the invention.

9. In claim 5, The method includes steps 20 to 22 after the eleventh step, The twentieth step is to create a fifth directive from the proposed claim; The twenty-first step inputs the fifth instruction sentence into the language model to obtain a fifth answer sentence including a search element and a component; The 22nd step is a sentence generation method for optimizing the proposed claim using combinatorial optimization of the search elements and the components.

10. In claim 9, Instead of the 22nd step, a 23rd step and a 24th step are included, The 23rd step creates a sixth instruction sentence from the pair of the component and the search element and the number of patent search results searched for each of the search elements; The 24th step of the sentence generation method includes inputting the sixth instruction sentence into the language model and obtaining a sixth response sentence including the optimized proposed claim.

Images