Information processing device

Abstract

The present invention provides an information processing device affording excellent convenience, usefulness, and reliability. The information processing device includes a reception unit, a documentation data extraction unit, a prompt data generation unit, an information extraction unit, and a display data generation unit. The reception unit has a function for receiving intellectual property data. The intellectual property data includes a specification, claims, a written notice of reasons for refusal, and cited documents. The documentation data extraction unit has a function for extracting at least one of the above pieces of documentation. The prompt data generation unit has a function for generating prompt data that includes instruction data and the extracted piece of documentation, and for outputting the prompt data to a language model. The information extraction unit has a function for extracting information from response data generated by the language model. The display data generation unit has a function for generating display data on the basis of the information and outputting the display data to an information terminal.

Description

Information processing device

[0001] 1. Field of the Invention The present invention relates to an information processing device, an information processing system, and an information processing method, and more particularly to an information processing system and an information processing method that utilize a language model, or an intellectual property business support device, an intellectual property business support system, and an intellectual property business support method.

[0002] Note that one embodiment of the present invention is not limited to the above technical field. The technical field of one embodiment of the invention disclosed in this specification and the like relates to an object, a method, or a manufacturing method. Alternatively, one embodiment of the present invention relates to a process, a machine, manufacture, or a composition of matter. Therefore, more specifically, examples of the technical field of one embodiment of the present invention disclosed in this specification include a semiconductor device, a display device, a light-emitting device, a power storage device, a memory device, a driving method thereof, or a manufacturing method thereof.

[0003] In order to obtain intellectual property rights such as patent rights, it is necessary to differentiate the invention from those disclosed in cited documents. Patent Document 1 discloses a method for determining whether patent documents similar to the patent to be evaluated are publicly known.

[0004] Furthermore, in recent years, there has been active development of language models using neural networks, with large-scale language models (LLMs) attracting particular attention. A large-scale language model is a natural language processing model trained using a large amount of data. A large-scale language model can realize, for example, a dialogue model that responds to user instructions. Non-Patent Document 1 discloses GPT-4 (Generative Pre-trained Transformer 4) (registered trademark) as a large-scale language model, and also discloses ChatGPT as a dialogue model.

[0005] International Publication No. 2020 / 26366

[0006] 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>

[0007] When an intellectual property application, such as a patent application, is rejected, it is necessary to respond based on the reasons for refusal. For example, when an intellectual property application is rejected due to a lack of novelty or an inventive step, the cited documents must be reviewed to determine whether the reasons for refusal are correct. Determining whether the reasons for refusal are correct requires a deep understanding of the invention and cited documents, which requires a great deal of effort and in-depth expertise.

[0008] In view of the above, an object of one embodiment of the present invention is to provide an information processing device that enables a user to perform intellectual property work with little effort. Another object of one embodiment of the present invention is to provide an information processing device that is highly convenient, useful, or reliable. Another object of one embodiment of the present invention is to provide a novel information processing device. Another object of one embodiment of the present invention is to provide an information processing system including the information processing device. Another object of one embodiment of the present invention is to provide an information processing method that can be applied to the information processing device.

[0009] 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 will become apparent from the description of the specification, drawings, claims, etc., and it is possible to extract other problems from the description of the specification, drawings, claims, etc.

[0010] One aspect of the present invention is an information processing device having a reception unit, a document data extraction unit, a prompt data generation unit, an information extraction unit, and a display data generation unit, wherein the reception unit has a function of receiving intellectual property data, the intellectual property data having first document data having a specification, second document data having claims, third document data having a notice of rejection, and fourth document data having cited documents, the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating prompt data having instruction statement data and the extracted document data and outputting it to a language model, the information extraction unit has a function of extracting information from response data generated by the language model based on the prompt data, and the display data generation unit has a function of generating display data based on the information and outputting it to an information terminal.

[0011] Alternatively, in the above aspect, the document data extraction unit may have a function of extracting first document data and fourth document data, and the instruction sentence data may have an instruction sentence for causing the language model to present features that are disclosed in the first document data but not disclosed in the fourth document data.

[0012] Alternatively, in the above aspect, the document data extraction unit may have a function of extracting third document data and fourth document data, and the instruction sentence data may have an instruction sentence for causing the language model to determine whether the reason for rejection included in the third document data is appropriately stated taking into account the content of the fourth document data.

[0013] Alternatively, in the above aspect, the intellectual property data may include a plurality of third document data, the document data extraction unit may have a function of extracting the plurality of third document data, and the instruction sentence data may include an instruction sentence for causing the language model to summarize and present the contents of the plurality of third document data.

[0014] Alternatively, one aspect of the present invention is an information processing device comprising: a reception unit, a document data extraction unit, a prompt data generation unit, and an information extraction unit, wherein the reception unit has a function of receiving intellectual property data, the intellectual property data including first document data having a description, second document data having claims, third document data having a notice of rejection, and fourth document data having cited documents, the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating first prompt data and second prompt data each having instruction statement data and at least one of the extracted document data, and outputting the first prompt data to a language model, the information extraction unit has a function of extracting first information from first response data generated by the language model based on the first prompt data, and the information extraction unit has a function of extracting second information from second response data generated by the language model based on the second prompt data, the second prompt data having the first information.

[0015] Alternatively, in the above aspect, the document data extraction unit may have a function of extracting first document data and fourth document data, and the instruction sentence data held by the first prompt data may have an instruction sentence for causing the language model to present features that are disclosed in the first document data but are not disclosed in the fourth document data, and the instruction sentence data held by the second prompt data may have an instruction sentence for causing the language model to determine whether the first information is disclosed in the first document data and whether the first information is not disclosed in the fourth document data.

[0016] Alternatively, in the above aspect, the display device may further include a display data generation unit, which has a function of generating display data based on the second information and outputting the display data to the information terminal.

[0017] Alternatively, one aspect of the present invention is an information processing device comprising: a reception unit, a document data extraction unit, a prompt data generation unit, and an information extraction unit, wherein the reception unit has a function of receiving intellectual property data, the intellectual property data including first document data having a description, second document data having claims, third document data having a notice of reasons for refusal, and fourth document data having cited documents, the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating prompt data having instruction statement data and the extracted document data, the prompt data generation unit has a function of outputting the prompt data to a language model k times (k is an integer greater than or equal to 2), the information extraction unit has a function of extracting first to k-th pieces of information from first to k-th response data generated by the language model based on the prompt data, each of the first to k-th pieces of information having one or more items, and the information extraction unit has a function of further extracting items of the same content included in two or more of the first to k-th pieces of information.

[0018] Alternatively, in the above aspect, the information extraction unit may have a function of calculating a similarity for each of the items included in the first to k-th information, and extracting items whose similarity is equal to or greater than a predetermined value.

[0019] Alternatively, the above aspect may further include a display data generation unit, which has a function of generating display data based on the extracted items and outputting the display data to an information terminal.

[0020] According to one aspect of the present invention, an information processing device that enables a user to perform intellectual property work with little effort can be provided. Alternatively, according to one aspect of the present invention, an information processing device that is highly convenient, useful, or reliable can be provided. Alternatively, according to one aspect of the present invention, a novel information processing device can be provided. Alternatively, according to one aspect of the present invention, an information processing system including the information processing device can be provided. Alternatively, according to one aspect of the present invention, an information processing method applicable to the information processing device can be provided.

[0021] The effects of one embodiment of the present invention are not limited to the effects listed above. The effects listed above do not preclude the existence of other effects. The other effects are described below and are not mentioned in this section. Effects not mentioned in this section can be derived by a person skilled in the art from the description in the specification or drawings, and can be extracted as appropriate from these descriptions. One embodiment of the present invention has at least one of the effects listed above and other effects. Therefore, one embodiment of the present invention may not have the effects listed above in some cases.

[0022] FIG. 1 is a schematic diagram showing an example of the configuration of an information processing system. FIG. 2 is a block diagram showing an example of the configuration of an information processing system. FIG. 3 is a flowchart showing an example of an information processing method. FIG. 4 is a flowchart showing an example of an information processing method. FIG. 5 is a schematic diagram showing an example of intellectual property data. FIGS. 6A and 6B are schematic diagrams showing examples of prompt data. FIG. 7 is a schematic diagram showing examples of prompt data. FIGS. 8A to 8C are schematic diagrams showing examples of response data. FIG. 9 is a flowchart showing an example of an information processing method. FIG. 10 is a flowchart showing an example of an information processing method. FIG. 11A is a schematic diagram showing an example of prompt data. FIG. 11B is a schematic diagram showing an example of response data. FIG. 11C is a schematic diagram showing an example of display data. FIG. 12 is a flowchart showing an example of an information processing method. FIG. 13 is a flowchart showing an example of an information processing method. FIG. 14A is a schematic diagram showing an example of response data. FIG. 14B is a schematic diagram showing an example of display data. FIG. 15 is a flowchart showing an example of an information processing method. FIG. 16 is a flowchart showing an example of an information processing method. Fig. 17A is a schematic diagram showing an example of prompt data. Fig. 17B is a schematic diagram showing an example of response data. Fig. 17C is a schematic diagram showing an example of display data. Fig. 18A is a schematic diagram showing an example of prompt data. Fig. 18B is a schematic diagram showing an example of response data. Fig. 18C is a schematic diagram showing an example of display data.

[0023] The embodiments will be described in detail with reference to the drawings. However, the present invention is not limited to the following description, and it will be readily understood by those skilled in the art that various changes can be made in the form and details without departing from the spirit and scope of the present invention. Therefore, the present invention should not be interpreted as being limited to the description of the embodiments shown below.

[0024] 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.

[0025] In this specification, the ordinal numbers "first" and "second" are used for convenience and do not limit the number of components or the order of the components (e.g., the order of steps). In addition, an ordinal number assigned to a component in one part of this specification may not match an ordinal number assigned to the same component in another part of this specification or in the claims.

[0026] Furthermore, in this specification, terms indicating positions such as "upper," "lower," "left," and "right" are used for convenience in describing the positional relationship between components with reference to the drawings. Furthermore, the positional relationship between components changes as appropriate depending on the direction in which each component is depicted. Therefore, the terms are not limited to those described in the specification, and can be rephrased appropriately depending on the situation.

[0027] Embodiment In this embodiment, an information processing system, an information processing device, and an information processing method according to one embodiment of the present invention will be described with reference to drawings.

[0028] One aspect of the present invention relates to an information processing device that supports intellectual property work. Another aspect of the present invention relates to an information processing system that includes the information processing device. Furthermore, another aspect of the present invention relates to an information processing method applicable to the information processing device. A user of the information processing device of one aspect of the present invention can, for example, respond to reasons for rejection of an application based on information generated by a language model. The application is an application related to intellectual property rights, such as a patent application. As described above, one aspect of the present invention can provide an information processing system, information processing device, and information processing method that are highly convenient, useful, or reliable.

[0029] A user of an information processing device according to one embodiment of the present invention may be, for example, a person in charge of intellectual property work. Furthermore, if there are multiple people in charge of intellectual property work, one of the people may be the user. Furthermore, the user may be different from the person in charge of intellectual property work.

[0030] In this specification and the like, an information processing device, an information processing system, and an information processing method that support intellectual property work can be referred to as an intellectual property work support device, an intellectual property work support system, and an intellectual property work support method, respectively.

[0031] <Configuration Example of Information Processing System> Fig. 1 is a schematic diagram showing a configuration example of an information processing system 10, which is an information processing system according to one embodiment of the present invention. The information processing system 10 includes an information terminal 20, an information processing device 100, and an information processing device 40. The information processing system 10 may also include a network 30. Two selected from the information processing device 100, the information terminal 20, and the information processing device 40 are connected via the network 30. Note that Fig. 1 shows information terminal 20a, information terminal 20b, information terminal 20c, and information terminal 20d as examples of the information terminal 20.

[0032] A user of the information processing system 10 can access the information processing device 100 from the information terminals 20a to 20d, etc. The user can then receive services using the information processing system of one aspect of the present invention. As described above, the user of the information processing system 10 can specifically be the user of the information terminal 20. Note that the user of the information processing system 10 may also be referred to as the user of the information processing device 100.

[0033] The information processing device 100 can generate prompt data using document data related to intellectual property rights. The document data related to intellectual property rights can be input from the information terminal 20 via the network 30. Examples of document data related to intellectual property rights include the application form, specification, claims, abstract, and drawings. Furthermore, examples of document data related to intellectual property rights include the notice of rejection for the application and the cited documents indicated in the notice of rejection. The prompt data includes instructional text data as well as at least one of the above-mentioned document data related to intellectual property rights.

[0034] In this specification, prompt data is data that indicates a prompt. A prompt corresponds to an input sentence that causes a language model to perform a desired operation. When a prompt is given to a language model, the language model generates a response sentence based on the prompt. The prompt has an instruction sentence. An instruction sentence can be rephrased as a question sentence, an instruction sentence, or the like.

[0035] The information processing device 100 can output prompt data to the information processing device 40 via the network 30. The information processing device 40 can generate response data based on the output prompt data. The information processing device 100 can extract information included in the response data and generate display data. The information processing device 100 can also output the display data to the information terminal 20. This allows the information terminal 20 to display the information included in the response data and show it to the user.

[0036] The information terminals 20a to 20d are each an information terminal device such as a computer used by a user, and can also be referred to as a client computer. FIG. 1 illustrates, as examples, the information terminal 20a, which is a desktop computer, the information terminal 20b, which is a notebook computer, the information terminal 20c, which is a smartphone, and the information terminal 20d, which is a tablet computer. The number of information terminals 20 connected to the information processing device 100 is not particularly limited. While FIG. 1 shows four information terminals 20, the number of information terminals 20 may be one, two, three, five or more. Examples of the information terminals 20 include desktop information terminals, notebook information terminals, tablet information terminals, and mobile information terminals such as smartphones.

[0037] The information processing device 100 is a device capable of executing an information processing method according to one embodiment of the present invention. The information processing device 100 is a large-scale computer such as a server computer or a supercomputer. The information processing device 100 is a computer with higher processing power than the information terminal 20. The information processing device 100 may be capable of performing processing using artificial intelligence (AI).

[0038] The information processing device 40 is a large computer such as a server computer or a supercomputer. The information processing device 40 is a computer with higher processing power than the information terminal 20 and the information processing device 100. The information processing device 40 can perform large-scale calculations required for AI learning and inference, for example.

[0039] The information processing device 40 can perform processing using a natural language processing model that uses AI. Examples of natural language processing models that use AI include BERT (Bidirectional Encoder Representations from Transformers) and T5 (Text-to-Text Transfer Transformer). The information processing device 40 can also perform processing using a large-scale language model. Specifically, the information processing device 40 can perform processing using a model that uses a large-scale language model (such as a sentence generation model or a dialogue model). Examples of large-scale language models include GPT-3, GPT-3.5, GPT-4 (registered trademark), LaMDA (Language Model for Dialogue Applications), PaLM (Pathways Language Model), and PaLM2, and it is preferable to use GPT-4 (registered trademark).

[0040] The network 30 may be, for example, a local network or a global network. Examples of local networks include an intranet and an extranet. Examples of global networks include the Internet, which is the foundation of the World Wide Web (WWW). The network 30 may be a computer network such as a personal area network (PAN), a local area network (LAN), a campus area network (CAN), a metropolitan area network (MAN), a wide area network (WAN), or a global area network (GAN).

[0041] Note that the network to which one of the information terminal 20, the information processing device 100, and the information processing device 40 is connected may be different from the network to which the other one is connected. For example, the information terminal 20 and the information processing device 100 may be connected via a local network, and the information processing device 40 and the information processing device 100 may be connected via a global network.

[0042] Furthermore, when wireless communication is performed, communication standards such as the fourth generation mobile communication system (4G), fifth generation mobile communication system (5G), and sixth generation mobile communication system (6G), or specifications standardized by the IEEE such as Wi-Fi (registered trademark) and Bluetooth (registered trademark), can be used as communication protocols or communication technologies.

[0043] The information processing system 10 shown in FIG. 1 is suitable for a case where a user who uses an information terminal 20 and a service provider belong to the same organization such as a company.

[0044] For example, it is preferable to use a network 30 established within an organization such as a company to transmit and receive data between the information terminal 20 and the information processing device 100, and between the information processing device 100 and the information processing device 40. This allows data to be transmitted and received more securely than when the information terminal 20 or the information processing device 100 is connected to the information processing device 40 via a global network such as the Internet. It also makes it possible to prevent confidential information within the organization from leaking to the outside.

[0045] Fig. 2 is a block diagram showing an example configuration of the information processing system 10. Fig. 2 shows a more specific example configuration of the information processing device 100 than Fig. 1. The information processing device 100 has a reception unit 101, an output unit 103, a storage unit 110, a document data extraction unit 121, a prompt data generation unit 123, an information extraction unit 125, and a display data generation unit 127. The document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 are collectively referred to as a processing unit.

[0046] The reception unit 101 and the output unit 103 are connected to the information terminal 20 via the network 30, and are also connected to the information processing device 40 via the network 30. Note that, although Fig. 2 shows an example in which the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 are all provided in the same information processing device 100, at least one of these may be provided in a different information processing device 100.

[0047] In the drawings accompanying this specification, the components are classified by function and shown as independent blocks in the block diagrams. However, in reality, it is difficult to completely separate the components by function, and one component may be involved in multiple functions.

[0048] The reception unit 101 has a function of receiving data from outside the information processing device 100. The reception unit 101 can receive data from the information terminal 20 via the network 30. The reception unit 101 can also receive data from the information processing device 40 via the network 30. The reception unit 101 can receive intellectual property data from the information terminal 20. The intellectual property data can include multiple pieces of document data related to the above-mentioned intellectual property rights. The reception unit 101 can also receive response data from the information processing device 40. The response data can be generated, for example, by a language model of the information processing device 40 based on prompt data.

[0049] The output unit 103 has a function of outputting data to the outside of the information processing device 100. The output unit 103 has a function of outputting data generated by processing performed by at least one of the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127, for example. The output unit 103 can output data to the information terminal 20 via the network 30. The output unit 103 can also output data to the information processing device 40 via the network 30. The output unit 103 can output display data to the information terminal 20. The output unit 103 can also output prompt data to the information processing device 40.

[0050] The storage unit 110 has a function of storing data used by at least one of the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127. The storage unit 110 has a function of storing, for example, instruction statement data. The instruction statement data can be included in the prompt data generated by the prompt data generation unit 123. The storage unit 110 can store two or more types of instruction statement data. In this case, for example, the user of the information terminal 20 can specify the instruction statement data to be included in the prompt data. This allows the information processing device 100 to execute one of multiple functions. Therefore, by storing multiple instruction statement data in the storage unit 110, the information processing device 100 can be a multi-functional information processing device.

[0051] The storage unit 110 also has a function of storing data (e.g., calculation results, analysis results, inference results, etc.) generated by at least one of the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127. The storage unit 110 can store, for example, data output by the output unit 103. The storage unit 110 can store, for example, prompt data generated by the prompt data generation unit 123. The storage unit 110 can store, for example, display data generated by the display data generation unit 127. The storage unit 110 may also have a function of storing data received by the reception unit 101. The storage unit 110 may have, for example, a function of storing intellectual property data or a function of storing response data.

[0052] Furthermore, the storage unit 110 has the function of storing the programs executed by the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127.

[0053] The storage unit 110 may function as a database. In this case, the user of the information terminal 20 can specify data stored in the storage unit 110, and the specified data can be accepted by the acceptance unit 101. For example, if multiple pieces of intellectual property data are stored in the storage unit 110, the user of the information terminal 20 can specify intellectual property data, and the acceptance unit 101 can accept the specified intellectual property data.

[0054] The database may be provided outside the information processing device 100 or outside the information processing system 10. Alternatively, the database may be provided in two or more locations among inside the information processing device 100, outside the information processing device 100 and inside the information processing system 10, and outside the information processing system 10.

[0055] The storage unit 110 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 110 may also include at least one of NOSRAM (registered trademark) and DOSRAM (registered trademark). The storage unit 110 may also include a recording media drive. Examples of recording media drives include a hard disk drive (HDD) and a solid state drive (SSD).

[0056] NOSRAM is an abbreviation for "Nonvolatile Oxide Semiconductor Random Access Memory (RAM)." NOSRAM refers to a memory in which memory cells are two-transistor (2T) or three-transistor (3T) gain cells and transistors (also referred to as OS transistors) that use metal oxide in their channel formation regions. OS transistors have an extremely small leakage current, i.e., a current that flows between the source and drain in an off state. NOSRAM can be used as a nonvolatile memory by retaining a charge corresponding to data in the memory cell using its extremely small leakage current characteristic. In particular, NOSRAM can read stored data without destroying it (nondestructive readout), and is therefore suitable for arithmetic processing in which only data read operations are repeated a large number of times. NOSRAM can increase its data capacity by stacking layers, and therefore can be used as a large-scale cache memory, main memory, or storage memory to improve the performance of semiconductor devices.

[0057] DOSRAM is an abbreviation for "Dynamic Oxide Semiconductor RAM" and refers to a RAM having 1T (transistor) 1C (capacitor) type memory cells. DOSRAM is a DRAM formed using OS transistors. DOSRAM is a memory that temporarily stores information sent from the outside. DOSRAM is a memory that takes advantage of the low off-state current of OS transistors.

[0058] In this specification and the like, a 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 for a semiconductor layer of a transistor, the metal oxide may be referred to as an oxide semiconductor.

[0059] The information processing device 40 may also include a memory using OS transistors, such as NOSRAM or DOSRAM. As described above, OS transistors have extremely low off-state current. Therefore, the memory can be suitably used as a nonvolatile memory for storing analog data. Storing weighting data for a neural network in the analog memory allows multiple product-sum operations to be performed in parallel using an analog arithmetic circuit. Therefore, using an analog memory including a transistor having a metal oxide in the information processing device 40 can significantly reduce power consumption when processing using a large-scale language model.

[0060] The metal oxide included in the channel formation region preferably contains indium (In). When the metal oxide included in the channel formation region contains indium, the carrier mobility (electron mobility) of the OS transistor is increased. Furthermore, the metal oxide included in the channel formation region is preferably an oxide semiconductor containing element M. The 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, a combination of two or more of the above elements may be used as element M. 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. The metal oxide contained in the channel formation region is preferably a metal oxide containing zinc (Zn), since zinc-containing metal oxides may be easily crystallized.

[0061] 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.

[0062] The document data extraction unit 121 has a function of extracting at least one piece of document data contained in the intellectual property data, and extracts the document data to be included in the prompt data.

[0063] The prompt data generation unit 123 has a function of generating prompt data. The prompt data includes instruction sentence data and document data extracted by the document data extraction unit 121. As described above, one of the multiple instruction sentence data stored in the storage unit 110 can be included in the prompt data. The prompt data generation unit 123 can generate prompt data using a prompt data generation script or program. The prompt data is data containing instructions written in natural language. The information processing device 40 can generate response data based on the prompt data. The response data is generated using a language model.

[0064] The information extraction unit 125 has a function of extracting information from response data, for example, information that serves as a response to instruction statement data.

[0065] The display data generating unit 127 has a function of generating display data based on the information extracted by the information extracting unit 125. The display data includes information to be displayed on the information terminal 20, for example.

[0066] The document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 can perform at least one of calculations using a central processing unit (CPU), a graphics processing unit (GPU), and a neural processing unit (NPU), for example. The document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 can be included in, for example, a calculation circuit. Note that the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 may be included in the same calculation circuit or in different calculation circuits.

[0067] The transmission unit 130 has a function of transmitting data. Data can be transmitted and received between the reception unit 101, the output unit 103, the storage unit 110, the document data extraction unit 121, the prompt data generation unit 123, the information extraction unit 125, and the display data generation unit 127 via the transmission unit 130.

[0068] <Example 1 of Information Processing Method> An example of an information processing method according to one embodiment of the present invention will be described below. Specifically, an example of an information processing method using the information processing system 10 shown in FIG.

[0069] 3 and 4 are flowcharts illustrating an example of an information processing method according to one embodiment of the present invention. In FIG. 4, it is indicated whether each process is performed by the information terminal 20, the information processing device 100, or the information processing device 40.

[0070] 3 is performed, step S21 shown in Fig. 4 is performed. In step S21, the user of the information terminal 20 inputs intellectual property data IPD to the information processing device 100. In other words, in step S21, the information terminal 20 accepts the intellectual property data IPD and outputs it to the information processing device 100.

[0071] 5 is a schematic diagram showing an example of intellectual property data IPD. In the example shown in FIG. 5, the intellectual property data IPD includes document data DOCD01, document data DOCD10, document data DOCD11, document data DOCD12, document data DOCD13, document data DOCD20, and document data DOCD21[1] to document data DOCD21[m] (m is an integer equal to or greater than 1. In the example shown in FIG. 5, m is an integer equal to or greater than 3). When m is 1, document data DOCD21[1] is referred to as document data DOCD21. When m is 2 or greater, document data DOCD21[1] to document data DOCD21[m] are collectively referred to as document data DOCD21.

[0072] The document data DOCD01, DOCD10, DOCD11, DOCD12, and DOCD13 may include, for example, documents related to intellectual property applications. Examples of intellectual property applications include patent applications, utility model registration applications, international applications under the Patent Cooperation Treaty (PCT), design registration applications, international applications for design registration under the Hague Agreement, trademark registration applications, and international applications for trademark registration under the Madrid Protocol. Figure 5 shows an example in which the intellectual property data IPD includes document data related to patent applications.

[0073] In this specification, etc., a patent application includes an international patent application. Also, a utility model registration application includes an international utility model registration application. An international patent application refers to an international application that has entered the national phase as a patent application. An international utility model registration application refers to an international application that has entered the national phase as a utility model registration application.

[0074] In the example shown in Figure 5, document data DOCD01 includes an application, document data DOCD10 includes a specification, document data DOCD11 includes claims, document data DOCD12 includes an abstract, and document data DOCD13 includes drawings. In this case, document data DOCD10, document data DOCD11, document data DOCD12, and document data DOCD13 can be said to be attached to document data DOCD01. Here, document data DOCD11 can be said to be the claims in the case of a patent application, or the claims in the case of a utility model registration application. Note that Figure 5 shows an example in which document data DOCD10 includes

[0001] and

[0002] . Also, Figure 5 shows an example in which document data DOCD11 includes [Claim 1] and [Claim 2]. Furthermore, Figure 5 shows an example in which document data DOCD13 includes "Figure 1" and "Figure 2."

[0075] The document data DOCD20 also includes a notice of reasons for refusal. Specifically, the document data DOCD20 may include a notice of reasons for refusal for an application including document data DOCD01, document data DOCD10, document data DOCD11, document data DOCD12, and document data DOCD13 as application documents. The document data DOCD20 includes a "reason" that is the reason for refusal and a "list of cited documents, etc." that serve as the basis for, for example, a violation of novelty and / or a violation of inventive step. Examples of cited documents include unexamined patent publications, patent gazettes, utility model registration publications, design publications, papers, books, and magazines. FIG. 5 shows an example in which the reasons for refusal include "1) ccc.", "2) ddd.", "3) eee.", and "4) fff." 5 shows an example in which cited documents include "1) Publication No. xxx," "2) Patent Publication No. yyy," and "m) Paper zzz." In the case of an international application based on the PCT, the document data DOCD20 may include a written opinion of the International Searching Authority (ISA).

[0076] Document data DOCD21 has cited documents included in the "list of cited documents, etc." of document data DOCD20. In the example shown in FIG. 5, document data DOCD21 includes document data DOCD21[1], document data DOCD21[2], and document data DOCD21[m]. Document data DOCD21[1] can have "1) Publication No. xxx." Document data DOCD21[2] can have "2) Patent Publication No. yyy." Document data DOCD21[m] can have "m) Paper zzz."

[0077] If the document data DOCD21 is a publication of a publicly disclosed patent, patent application, or utility model registration, the document data DOCD21 includes a description, claims, an abstract, and drawings. Note that drawings may not be included in the document data DOCD21. If the document data DOCD21 is a publication of a design, the document data DOCD21 includes an application and drawings.

[0078] 5 shows an example in which document data DOCD20 includes one notice of rejection. Here, if two or more rejections have been issued in an application relating to intellectual property data IPD, two or more document data DOCD20 can be included in intellectual property data IPD. In this case, for example, a cited document included in at least one of the two or more document data DOCD20 can be included in intellectual property data IPD as document data DOCD21. Furthermore, if at least one of document data DOCD01, document data DOCD10, document data DOCD11, document data DOCD12, and document data DOCD13 is amended after filing, the amended document data DOCD01, document data DOCD10, document data DOCD11, document data DOCD12, and document data DOCD13 can be included in intellectual property data IPD.

[0079] One intellectual property data IPD can include document data related to one application. Note that one intellectual property data IPD may also include document data related to two or more applications. For example, document data related to applications in the same family may be included in one intellectual property data IPD.

[0080] If the storage unit 110 functions as a database, multiple pieces of intellectual property data IPD can be registered in advance in the storage unit 110. In this case, the user of the information terminal 20 can specify intellectual property data IPD stored in the storage unit 110. The information terminal 20 can then accept the specified intellectual property data IPD and output it to the information processing device 100. The user of the information terminal 20 can specify the intellectual property data IPD using, for example, an application number, a publication number, or a management number assigned to the intellectual property data IPD within a company. The user of the information terminal 20 can also specify the intellectual property data IPD using, for example, one or more of the inventor, applicant, filing date, priority date, publication date, status, patent classification, category, and keyword. Furthermore, the user of the information terminal 20 can specify the intellectual property data IPD based on, for example, document data contained in the intellectual property data IPD. Specifically, the user of the information terminal 20 can specify the intellectual property data IPD based on the contents of the document data.

[0081] 5 shows an example in which an application related to the intellectual property data IPD is pending examination or trial, but this is not a limitation of the present invention. The application related to the intellectual property data IPD may be, for example, a registered application. The intellectual property data IPD may be, for example, data related to a registered utility model. In this case, the document data DOCD20 may include, for example, a utility model technical evaluation report.

[0082] The intellectual property data IPD may also be data relating to a registered patent that is the subject of a patent opposition. In this case, the document data DOCD20 may include, for example, a notice of grounds for revocation or a patent opposition. The intellectual property data IPD may also be data relating to a registered patent or registered utility model that is the subject of an invalidation trial. In this case, the document data DOCD20 may include a notice of grounds for invalidation or a request for trial.

[0083] The intellectual property data IPD may include document data other than the document data shown in Fig. 5. The intellectual property data IPD may include, for example, at least one of document data having a written opinion, document data having a written amendment, document data having a written statement, and document data having a request for correction.

[0084] [Step S11] In step S11, the reception unit 101 receives intellectual property data IPD. Specifically, the reception unit 101 receives the intellectual property data IPD output from the information terminal 20. The intellectual property data IPD received by the reception unit 101 can be stored in, for example, the storage unit 110. Hereinafter, an example of an information processing method when the reception unit 101 receives the intellectual property data IPD shown in FIG. 5 in step S11 will be described.

[0085] After step S11, step S22 shown in FIG. 4 is performed. In step S22, the user of the information terminal 20 selects a process to be performed by the information processing device 100 in a subsequent step. Each of these processes may be performed using a language model. For example, one example of this process is a first process in which the language model presents features that are disclosed in the document data DOCD10 but not in the document data DOCD21 shown in FIG. 5 . Another example of this process is a second process in which the language model determines whether the reason for rejection included in the document data DOCD20 shown in FIG. 5 is appropriately described in light of the content of the document data DOCD21. Another example of this process is a third process in which, if the intellectual property data IPD shown in FIG. 5 includes multiple document data DOCD20, the language model summarizes and presents the content of the multiple document data DOCD20.

[0086] [Step S12] In step S12, the document data extraction unit 121 extracts document data required in a subsequent process from the intellectual property data IPD. The document data extraction unit 121 extracts document data to be used in the process selected in step S22 from document data DOCD01, document data DOCD10, document data DOCD11, document data DOCD12, document data DOCD13, document data DOCD20, document data DOCD21, etc. shown in FIG. 5. The document data extraction unit 121 extracts, for example, document data to be included in prompt data generated in a subsequent process. The document data extracted by the document data extraction unit 121 can be stored, for example, in the storage unit 110.

[0087] [Step S13] In step S13, the prompt data generation unit 123 generates prompt data PPTD. The prompt data generation unit 123 reads, for example, instruction statement data corresponding to the process selected in step S22 and the document data extracted by the document data extraction unit 121 from the storage unit 110. The prompt data generation unit 123 then generates prompt data PPTD including the read data.

[0088] 6A, 6B, and 7 are schematic diagrams showing examples of prompt data PPTD_1, prompt data PPTD_2, and prompt data PPTD_3, respectively. Here, the prompt data PPTD_1, prompt data PPTD_2, and prompt data PPTD_3 are collectively referred to as prompt data PPTD.

[0089] The prompt data PPTD_1 shown in Figure 6A includes instruction sentence data DSD_1, document data DOCD10, and document data DOCD21[1] through DOCD21[m]. That is, when the prompt data generation unit 123 generates the prompt data PPTD_1, the document data extraction unit 121 extracts document data DOCD10 and document data DOCD21[1] through DOCD21[m] from the intellectual property data IPD shown in Figure 5. Note that in Figure 6A, document data DOCD21[1] through document data DOCD21[m] are shown under the heading "Cited Documents." In subsequent drawings, the document data DOCD21 included in the prompt data is also shown under the heading "Cited Documents."

[0090] The instruction data DSD_1 includes an instruction for the information processing device 100 to perform the first process described above. Specifically, the instruction data DSD_1 includes an instruction for causing the language model to present features that are disclosed in the document data DOCD10 but are not disclosed in the document data DOCD21. Here, the features may specifically be technical features. For example, the instruction may be, "Please indicate the features that are disclosed in the specification but are not disclosed in the cited documents. Please list multiple features in bullet points."

[0091] Prompt data PPTD_2 shown in Figure 6B includes instruction sentence data DSD_2, document data DOCD20, and document data DOCD21[1] to DOCD21[m]. That is, when prompt data generation unit 123 generates prompt data PPTD_2, document data extraction unit 121 extracts document data DOCD20 and document data DOCD21[1] to DOCD21[m] from intellectual property data IPD shown in Figure 5. Note that if intellectual property data IPD shown in Figure 5 includes multiple document data DOCD20, for example, document data DOCD20 with the most recent delivery date can be included in prompt data PPTD_2.

[0092] The instruction data DSD_2 has an instruction for when the information processing device 100 performs the second process described above. Specifically, the instruction data DSD_2 has an instruction for causing the language model to determine whether the reasons for refusal included in the document data DOCD20 are appropriately stated taking into account the contents of the document data DOCD21[1] to DOCD21[m]. The instruction can be, for example, "For each reason for refusal included in the notice of reasons for refusal, please indicate whether the arguments are made based correctly on the contents of the cited documents. Please itemize the reasons for refusal that are made based correctly on the contents of the cited documents and the reasons for refusal that are not made based correctly on the contents of the cited documents."

[0093] The prompt data PPTD_3 shown in FIG. 7 includes instruction data DSD_3 and document data DOCD20[1] through DOCD20[n] (n is an integer equal to or greater than 2; in the example shown in FIG. 7, n is an integer equal to or greater than 3). All of the document data DOCD20[1] through DOCD20[n] are included as document data DOCD20 in the intellectual property data IPD shown in FIG. 5. Here, document data DOCD20[1] can be defined as the document data DOCD20 with the oldest delivery date, and document data DOCD20[n] can be defined as the document data DOCD20 with the most recent delivery date. In the example shown in FIG. 7, document data DOCD20[n] includes the following rejection reasons: "1) ccc.", "2) ddd.", "3) eee.", and "4) fff." 5 also shows an example in which cited documents include "1) Publication No. xxx," "2) Patent Publication No. yyy," and "m) Paper zzz." When the prompt data generating unit 123 generates prompt data PPTD_3, the document data extracting unit 121 extracts n pieces of document data DOCD20 from the intellectual property data IPD shown in FIG.

[0094] The instruction sentence data DSD_3 has an instruction sentence for when the information processing device 100 performs the above-mentioned third process. Specifically, the instruction sentence data DSD_3 has an instruction sentence for causing the language model to summarize and present the contents of the document data DOCD20[1] to document data DOCD20[n]. The instruction sentence can be, for example, "Please summarize the reasons for rejection and present the progress of the reasons for rejection."

[0095] The prompt data PPTD can be stored, for example, in the storage unit 110. The prompt data generation unit 123 can generate, for example, any one of prompt data PPTD_1, prompt data PPTD_2, and prompt data PPTD_3 based on the content of the process selected in step S22, and store the generated data in the storage unit 110.

[0096] [Step S14] In step S14, the prompt data generation unit 123 outputs the prompt data PPTD to the information processing device 40. The prompt data PPTD is output to the information processing device 40 via the output unit 103 (step S14a shown in FIG. 4). The information processing device 40 generates response data RD based on the prompt data PPTD. The response data RD is generated using a language model of the information processing device 40. The information processing device 40 outputs the response data RD to the information processing device 100 (step S14b shown in FIG. 4). The reception unit 101 of the information processing device 100 receives the response data RD. As a result, the information processing device 100 acquires the response data RD (step S14c shown in FIG. 4). The response data RD can be stored in, for example, the storage unit 110.

[0097] 8A, 8B, and 8C are schematic diagrams showing examples of response data RD_1, response data RD_2, and response data RD_3, respectively. Response data RD_1 can be generated based on prompt data PPTD_1. Similarly, response data RD_2 can be generated based on prompt data PPTD_2. Furthermore, response data RD_3 can be generated based on prompt data PPTD_3. Here, response data RD_1, response data RD_2, and response data RD_3 are collectively referred to as response data RD.

[0098] The response data RD_1 includes information 231_1. Similarly, the response data RD_2 includes information 231_2. Furthermore, the response data RD_3 includes information 231_3. Here, the information 231_1, the information 231_2, and the information 231_3 are collectively referred to as information 231.

[0099] FIG. 8A shows an example in which response data RD_1 includes a sentence saying, "The following features are disclosed in the specification but not in the cited documents." Information 231_1 includes features that are disclosed in document data DOCD10 shown in FIG. 6A but are not disclosed in document data DOCD21[1] to document data DOCD21[m]. FIG. 8A shows an example in which information 231_1 includes "1) aaa.", "2) bbb.", and "3) kkk." In this case, "aaa.", "bbb.", and "kkk" can be considered to be features that are disclosed in document data DOCD10 but are not disclosed in document data DOCD21[1] to document data DOCD21[m].

[0100] FIG. 8B shows an example in which response data RD_2 includes a sentence saying, "Among the reasons for refusal included in the notice of reasons for refusal, the following are appropriately stated taking into account the contents of the cited documents," and a sentence saying, "The following are not appropriately stated taking into account the contents of the cited documents." Information 231_2 includes information 231_2a indicating reasons for refusal that are appropriately stated taking into account the contents of document data DOCD21[1] to document data DOCD21[m], and information 231_2b indicating reasons for refusal that are inappropriately stated taking into account the contents of the cited documents. FIG. 8B shows an example in which information 231_2a includes "1) ccc." and "2) ddd.", and information 231_2b includes "1) eee." and "2) fff." Information 231_2b may include, for example, reasons for refusal issued by an examiner or administrative judge without properly understanding the contents of the cited documents.

[0101] FIG. 8C shows an example in which response data RD_3 includes the sentence, "This application has been rejected for the following reasons." Information 231_3 includes summaries of the reasons for rejection included in document data DOCD20[1] through document data DOCD20[n] shown in FIG. 7. FIG. 8C shows an example in which "ggg." and "hhh." are included as summaries of the reasons for rejection. Here, information 231_3 can be arranged in order from the reason for rejection included in document data DOCD20 with the oldest delivery date. That is, in the example shown in FIG. 8C, "ggg." can be the reason for rejection included in document data DOCD20 with a delivery date older than "hhh." For example, "ggg." can be the reason for rejection included in document data DOCD20[1], and "hhh." can be the reason for rejection included in document data DOCD20[2].

[0102] [Step S15] In step S15, the information extraction unit 125 extracts information 231 from the response data RD. The information extraction unit 125 can extract, for example, itemized sentences from sentences included in the response data RD as the information 231. The extracted information 231 can be stored in, for example, the storage unit 110.

[0103] [Step S16] In step S16, the display data generation unit 127 generates display data DSPD1 based on the information 231. Thereafter, the display data generation unit 127 outputs the display data DSPD1 to the information terminal 20. The display data DSPD1 is output to the information terminal 20 via the output unit 103. The display data DSPD1 indicates a screen layout. The screen can present the information 231 to the user of the information terminal 20. The display data DSPD1 can be stored in the storage unit 110, for example.

[0104] 4 is performed after step S16 is performed. In step S23, the information terminal 20 displays the display data DSPD1. This allows the user of the information terminal 20 to check the information 231.

[0105] The above is an information processing method according to one embodiment of the present invention. In the information processing method according to one embodiment of the present invention, for example, a response to a rejection reason for an application can be made based on information generated by a language model. For example, when the prompt data generation unit 123 generates prompt data PPTD_1 shown in FIG. 6A, a user of the information terminal 20 can amend the claims based on information 231_1 shown in FIG. 8A. For example, a technical feature included in information 231_1 can be added to the claims as an invention-specifying matter.

[0106] When the prompt data generation unit 123 generates the prompt data PPTD_2 shown in Fig. 6B, the user of the information terminal 20 can argue to the examiner or administrative judge that the reason for refusal included in the information 231_2b shown in Fig. 8B is inappropriate. Also, the user of the information terminal 20 can determine that the reason for refusal included in the information 231_2a is appropriate and amend the claims to avoid that reason for refusal.

[0107] When the prompt data generation unit 123 generates the prompt data PPTD_3 shown in Fig. 7, the user of the information terminal 20 can understand the examination progress based on the information 231_3 shown in Fig. 8C. This allows the user of the information terminal 20 to understand the examination progress in a shorter time than if, for example, the user were to read all of the notices of reasons for refusal.

[0108] As described above, by using the information processing method of one embodiment of the present invention, intellectual property work can be performed with less effort. Furthermore, the information processing method of one embodiment of the present invention can be an information processing method with excellent convenience, usefulness, and reliability. Furthermore, the information processing device and information processing system of one embodiment of the present invention can be an information processing device and information processing system that allow a user to perform intellectual property work with less effort. Furthermore, the information processing device and information processing system of one embodiment of the present invention can be an information processing device and information processing system with excellent convenience, usefulness, and reliability.

[0109] <Example 2 of Information Processing Method> An example of an information processing method according to one embodiment of the present invention will be described below. Specifically, an example of an information processing method using the information processing system 10 shown in FIG. 2 that is different from, for example, the examples shown in FIGS. 3 and 4 will be described.

[0110] 9 and 10 are flowcharts showing an example of an information processing method according to one embodiment of the present invention. Fig. 10 indicates whether each process is performed by the information terminal 20, the information processing device 100, or the information processing device 40. In the information processing method shown in Fig. 9 and 10, the information processing device 100 acquires first response data. Then, the information processing device 100 acquires second response data to verify whether the first response data includes desired information.

[0111] For example, steps S31, S32, S33, S34, and S35 shown in Figure 9 can perform the same processing as steps S11, S12, S13, S14, and S15 shown in Figure 3, respectively. Here, prompt data PPTD and response data RD are read as prompt data PPTD1 and response data RD1, respectively. Note that step S34 includes steps S34a, S34b, and S34c. Steps S34a, S34b, and S34c can perform the same processing as steps S14a, S14b, and S14c shown in Figure 4, respectively.

[0112] 10 shows an example in which the information terminal 20 performs step S21 shown in FIG. 4 before the information processing device 100 performs step S31. Also, FIG. 10 shows an example in which the information terminal 20 does not perform step S22. Note that the information terminal 20 may perform step S22. Specifically, the information processing device 100 may perform step S31, and then the information terminal 20 may perform step S22. Thereafter, the information processing device 100 can perform step S32 and subsequent steps.

[0113] [Step S36] In step S36, the prompt data generation unit 123 generates prompt data PPTD2. The prompt data generation unit 123 reads the instruction statement data, the information 231, and the document data extracted by the document data extraction unit 121 from the storage unit 110. The prompt data generation unit 123 then generates prompt data PPTD2 that includes the read data.

[0114] 11A is a schematic diagram showing an example of prompt data PPTD2. The prompt data PPTD2 shown in FIG. 11A includes instruction sentence data DSD2, information 231_1, and document data DOCD21[1] through DOCD21[m]. In the example shown in FIG. 11A, information 231_1 is shown under the heading "Features."

[0115] The instruction data DSD2 includes an instruction for performing processing related to the information 231_1. Specifically, the instruction data DSD2 includes a first instruction for causing the language model to determine whether the information 231_1 is disclosed in the document data DOCD10. The instruction data DSD2 also includes a second instruction for causing the language model to determine whether the information 231_1 is not disclosed in the document data DOCD21[1] through DOCD21[m]. The instruction data DSD2 may include both the first instruction and the second instruction. For example, the instruction may be, "Please determine whether the extracted features are disclosed in the specification. Also, please determine whether they are not disclosed in the cited documents. Then, please present the features disclosed in the specification and the features not disclosed in the cited documents. Please list the features in itemized form." The instruction data DSD2 does not have to include the first instruction or the second instruction.

[0116] The prompt data PPTD2 can be stored, for example, in the storage unit 110. While Fig. 11A shows an example in which the prompt data PPTD2 includes information 231_1, the prompt data PPTD2 may also include information 231_2 or information 231_3. In this case, by replacing information 231_1 with information 231_2 or information 231_3, the explanations of the information processing methods shown in Figs. 9, 10, etc. can be referred to.

[0117] [Step S37] In step S37, the prompt data generation unit 123 outputs the prompt data PPTD2 to the information processing device 40. The prompt data PPTD2 is output to the information processing device 40 via the output unit 103 (step S37a shown in FIG. 10). The information processing device 40 generates response data RD2 based on the prompt data PPTD2. The response data RD2 is generated using a language model of the information processing device 40. The information processing device 40 outputs the response data RD2 to the information processing device 100 (step S37b shown in FIG. 10). The reception unit 101 of the information processing device 100 receives the response data RD2. As a result, the information processing device 100 acquires the response data RD2 (step S37c shown in FIG. 10). The response data RD2 can be stored in, for example, the storage unit 110. Note that the response data RD1 and the response data RD2 may be generated using different language models.

[0118] 11B is a schematic diagram showing an example of response data RD2, which can be generated based on prompt data PPTD2.

[0119] The response data RD2 includes information 232. In Fig. 11B, the information 232 includes information 232a and information 232b.

[0120] 11B shows an example in which response data RD2 includes a sentence "The following is disclosed in the specification" and a sentence "The following is not disclosed in the cited document." Information 232a includes, among the features included in information 231_1, features that are disclosed in document data DOCD10. Information 232b includes, among the features included in information 231_1, features that are not disclosed in document data DOCD21[1] to document data DOCD21[m].

[0121] FIG. 11B shows an example in which information 232a includes "1) aaa." and "2) bbb." In this case, "aaa." and "bbb." can be considered to be features disclosed in document data DOCD10. FIG. 11B also shows an example in which information 232b includes "1) aaa." and "2) kkk." In this case, "aaa." and "kkk." can be considered to be features not disclosed in document data DOCD21[1] to document data DOCD21[m]. Note that response data RD2 may indicate features that are disclosed in document data DOCD10 but not disclosed in document data DOCD21[1] to document data DOCD21[m]. In other words, response data RD2 may indicate features included in both information 232a and information 232b.

[0122] [Step S38] In step S38, the information extraction unit 125 extracts information 232 from the response data RD2. The information extraction unit 125 can extract, for example, itemized sentences from the sentences included in the response data RD2 as the information 232. For example, the information extraction unit 125 can extract, as information 232a, the itemized sentences included between "The following is disclosed in the specification" and "The following is not disclosed in the cited documents." The information extraction unit 125 can also extract, as information 232b, the itemized sentences included below "The following is not disclosed in the cited documents." The extracted information 232 can be stored, for example, in the storage unit 110. Note that steps S36 to S38 may be performed multiple times. For example, steps S36 to S38 may be performed using prompt data PPTD2 having one of the above-mentioned first instruction sentence and second instruction sentence, and then steps S36 to S38 may be performed using prompt data PPTD2 having the other of the first instruction sentence and second instruction sentence.

[0123] [Step S39] In step S39, the display data generation unit 127 generates display data DSPD2 based on the information 232. Thereafter, the display data generation unit 127 outputs the display data DSPD2 to the information terminal 20. The display data DSPD2 is output to the information terminal 20 via the output unit 103. The display data DSPD2 indicates a screen layout. The screen can present the information 232 to the user of the information terminal 20. The display data DSPD2 can be stored in the storage unit 110, for example.

[0124] FIG. 11C is a schematic diagram showing an example of display data DSPD2. FIG. 11C shows an example in which the display data DSPD2 has the description "Features that can be Differentiating Points" and information 233 below it. The display data generation unit 127 can generate the information 233 based on the information 232 extracted by the information extraction unit 125. The display data generation unit 127 can include, for example, features included in both information 232a and information 232b shown in FIG. 11B in the information 233. In the example shown in FIG. 11B, it is assumed that "aaa." is included in both information 232a and information 232b. It is also assumed that "bbb." is included in information 232a but not in information 232b. It is also assumed that "kkk." is not included in information 232a but is included in information 232b. In this case, the display data generating unit 127 can include "aaa." in the information 233 and exclude "bbb." and "kkk." from the information 233. Note that the information extracting unit 125 may extract features included in both the information 232a and the information 232b shown in FIG. 11B, for example.

[0125] The display data generation unit 127 does not need to generate the information 233. In this case, the display data generation unit 127 can include in the display data DSPD2 all of the information included in the information 232. For example, not only features included in both the information 232a and the information 232b, but also features included in only one of the information 232a and the information 232b can be included in the display data DSPD2.

[0126] After step S39 is performed, step S23 shown in Fig. 10 is performed. In step S23, the information terminal 20 displays the display data DSPD2. This allows the user of the information terminal 20 to check the information 233. The user of the information terminal 20 can check, for example, features that are disclosed in the specification but not in the cited documents.

[0127] The above is an information processing method according to one embodiment of the present invention. In the information processing method according to one embodiment of the present invention, it is possible to verify whether response data RD1 contains desired information using prompt data PPTD2. This makes it easier to present desired information to a user of information terminal 20. For example, it is possible to prevent features not disclosed in the specification and features disclosed in cited documents from being presented to a user of information terminal 20. This prevents a user of information terminal 20 from adding, for example, technical features not disclosed in the specification to claims. It is also possible to prevent a user of information terminal 20 from adding, for example, technical features disclosed in cited documents to claims. As described above, this embodiment of the present invention can improve the convenience, usefulness, and reliability of an information processing device, information processing system, and information processing method.

[0128] <Example 3 of Information Processing Method> An example of an information processing method according to one embodiment of the present invention will be described below. Specifically, the example will be an information processing method using the information processing system 10 shown in FIG. 2 that is different from the examples shown in FIGS. 3, 4, 9, and 10.

[0129] 12 and 13 are flowcharts illustrating an example of an information processing method according to one embodiment of the present invention. Fig. 13 indicates whether each process is performed by the information terminal 20, the information processing device 100, or the information processing device 40. In the information processing method illustrated in Fig. 12 and 13, the information processing device 100 acquires response data RD k times (k is an integer equal to or greater than 2).

[0130] In steps S41, S42, and S43 shown in Fig. 12, the same processes as steps S11, S12, and S13 shown in Fig. 3 can be performed, respectively. As shown in Fig. 13, before the information processing device 100 performs step S41, the information terminal 20 performs step S21. After the information processing device 100 performs step S41 and before performing step S42, the information terminal 20 performs step S22.

[0131] [Steps S44 and S45] In step S44, the prompt data generating unit 123 prepares a variable i and inputs, for example, "1" as an initial value.

[0132] In step S45, for example, processing similar to step S14 shown in FIG. 3 is performed. In step S45, the prompt data generation unit 123 acquires response data RD[i]. For example, when i is "1", the prompt data generation unit 123 acquires response data RD[1]. Note that step S45 includes steps S45a, S45b, and S45c. In steps S45a, S45b, and S45c, processing similar to steps S14a, S14b, and S14c shown in FIG. 4 can be performed, respectively.

[0133] [Steps S46 and S47] In step S46, the prompt data generation unit 123 determines whether i is greater than or equal to k. If i is less than k, the prompt data generation unit 123 increments i by 1 in step S47. The prompt data generation unit 123 then performs steps S45 and S46 again. Thus, the prompt data generation unit 123 performs steps S45 and S46 k times.

[0134] 14A is a schematic diagram showing an example of response data RD[i]. Specifically, FIG. 14A shows examples of response data RD[1], response data RD[2], and response data RD[k]. Here, response data RD[1] to response data RD[k] are collectively referred to as response data RD. In the example shown in FIG. 14A, k can be an integer equal to or greater than 3.

[0135] 14A illustrates an example in which the structure of response data RD[1] through RD[k] is the same as that of response data RD_1 illustrated in FIG. 8A. That is, in step S43, the prompt data generator 123 generates prompt data PPTD having a structure similar to that of prompt data PPTD_1 illustrated in FIG. 6A. Note that the prompt data generator 123 may also generate prompt data PPTD having a structure similar to that of prompt data PPTD_2 illustrated in FIG. 6B or prompt data PPTD_3 illustrated in FIG. 7. In this case, the structure of response data RD[1] through RD[k] may be the same as that of response data RD_2 illustrated in FIG. 8B or response data RD_3 illustrated in FIG. 8C.

[0136] The response data RD includes information 231. Fig. 14A shows examples of information 231[1] included in the response data RD[1], information 231[2] included in the response data RD[2], and information 231[k] included in the response data RD[k].

[0137] Information 231 includes one or more items 241. The items 241 included in information 231[1], information 231[2], and information 231[k] are referred to as items 241[1], 241[2], and 241[k], respectively. In FIG. 14A , item 241[1]_1, item 241[1]_2, and item 241[1]_3 are shown as item 241[1]. Item 241[2]_1, item 241[2]_2, and item 241[2]_3 are shown as item 241[k]. Item 241[k]_1, item 241[k]_2, and item 241[k]_3 are shown as item 241[1] to 231[k] collectively.

[0138] In the example shown in Fig. 14A, the item 241 can be a feature. Specifically, the item 241 can be a feature that is disclosed in the document data DOCD10 shown in Fig. 6A but is not disclosed in the document data DOCD21[1] to DOCD21[m]. Here, for example, one feature can be one item 241.

[0139] 14A shows an example in which items 241[1]_1, 241[1]_2, 241[1]_3, 241[2]_1, 241[2]_2, 241[2]_3, 241[k]_1, 241[k]_2, and 241[k]_3 are "1) a1a1a1.", "2) b1b1b1.", "3) c1c1c1.", "1) a1a1a1.", "2) b1b1b1.", "3) c2c2c2.", "1) a1a1a1.", "2) c2c2c2.", and "3) d1d2d3." Here, "1)," "2)," and "3)" are referred to as item numbers.

[0140] The response data RD[1] through RD[k] can all be generated based on the same prompt data PPTD. However, even when the same prompt data is input to the language model, different response data may be generated. FIG. 14A illustrates an example in which the response data RD[1], response data RD[2], and response data RD[k] are different from one another. Specifically, the example illustrates an example in which at least some of the items 241[1] included in information 231[1], the items 241[2] included in information 231[2], and the items 241[k] included in information 231[k] are different from one another.

[0141] The information processing device 40 may generate the response data RD[1] through RD[k] based on prompt data PPTD that have mutually different portions. For example, the information processing device 40 may generate the response data RD[1] through RD[k] based on prompt data PPTD that have mutually different portions of instruction sentence data. In this case, the instruction sentence data may have, for example, instruction sentences that have the same content but at least a portion of the wording is different.

[0142] [Step S48] If i is greater than or equal to k, step S48 is performed. In step S48, the information extraction unit 125 extracts information 231[1] to information 231[k] from response data RD[1] to response data RD[k], respectively. That is, the information extraction unit 125 extracts information 231[i] from response data RD[i], for example. This allows the information extraction unit 125 to extract items 241[1] to 241[k] included in information 231[1] to information 231[k]. The extracted information 231[1] to information 231[k] can be stored in the storage unit 110, for example.

[0143] [Step S49] In step S49, the information extraction unit 125 extracts items 241 with the same content that are included in two or more of the information 231[1] through 231[k]. For example, if an item 241 with the same wording except for the item number is included in two or more of the information 231[1] through 231[k], the information extraction unit 125 extracts the item 241. In the example shown in FIG. 14A , items 241[1]_1, 241[2]_1, and 241[k]_1 are all the same, "a1a1a1." Furthermore, items 241[1]_2 and 241[2]_2 are all the same, "b1b1b1." Furthermore, items 241[2]_3 and 241[k]_2 are all the same, "c2c2c2." From the above, the information extraction unit 125 extracts “a1a1a1.”, “b1b1b1.”, and “c2c2c2.” from the item 241 .

[0144] The information extraction unit 125 may extract items 241 with the same content that are included in three or more pieces of information 231[1] to 231[k], or may extract items 241 with the same content that are included in four or more pieces of information 231, or may extract items 241 with the same content that are included in more than that number of pieces of information 231. For example, let r be the smallest integer greater than k / 2. In this case, the information extraction unit 125 may extract items 241 with the same content that are included in r or more pieces of information 231 from information 231[1] to 231[k], or may extract items 241 with the same content that are included in more than that number of pieces of information 231. For example, when k is 4 or 5, r can be 3. When k is 6 or 7, r can be 4.

[0145] Here, the information extraction unit 125 may calculate the similarity of each item 241. In this case, the information extraction unit 125 calculates, for example, the similarity between item 241[1] and items 241[2] to 241[k]. The information extraction unit 125 also calculates the similarity between item 241[2] and items 241[3] (not shown) to 241[k]. In the example shown in FIG. 14A , the information extraction unit 125 can calculate the similarity between item 241[1]_1 and items 241[2]_1 to 241[2]_3, and items 241[k]_1 to 241[k]_3. Similarly, the information extraction unit 125 can calculate the similarity of items 241[1]_2 and 241[1]_3 to items 241[2]_1 to 241[2]_3 and items 241[k]_1 to 241[k]_3, respectively. In addition, the information extraction unit 125 can calculate the similarity of items 241[2]_1 to 241[2]_3 to items 241[k]_1 to 241[k]_3, respectively.

[0146] The information extraction unit 125 can then extract items 241 whose similarity is, for example, a predetermined value or greater as items 241 with the same content. Here, items 241 with the same content included in three of the information 231[1] to information 231[k] can also be extracted based on the similarity between the items 241. For example, assume that information 231[1] has a first item (one of items 241[1]), information 231[2] has a second item (one of items 241[2]), and information 231[k] has a third item (one of items 241[k]). Assume that the similarity between the first item and the second item, the similarity between the first item and the third item, and the similarity between the second item and the third item are all greater than a predetermined value. In this case, the information extraction unit 125 can extract the first item, the second item, and the third item as items 241 with the same content. Items 241 with the same content that are included in four or more of the information 231[1] to 231[k] can also be extracted in a similar manner.

[0147] The similarity between two items 241 can be calculated based on, for example, the degree of similarity of the characters (which can also be said to be the degree of similarity of character strings). Alternatively, the items 241 may be vectorized (quantified) and the similarity between the two vectors may be calculated. Note that the similarity between two items 241 may also be calculated by calculating the distance between the two vectors.

[0148] There are various methods for vectorizing the item 241. For example, the item 241 may be divided into phrases or words by performing morphological analysis and / or compound word analysis. Then, the item 241 may be vectorized from the divided phrases or words. Examples of such vectorization methods include Bag-of-Words and TF-IDF (Term Frequency-Inverse Document Frequency).

[0149] Furthermore, for example, a model that converts a sentence into a distributed representation can be learned by machine learning, and the distributed representation of item 241 can be obtained by the model. For example, a neural network can be used as the model. Examples of neural network models that can be used to obtain the distributed representation include Doc2Vec, fastText, BERT (Bidirectional Encoder Representations from Transformer), and SentenceBERT.

[0150] Methods for calculating the similarity between two vectors include cosine similarity, covariance, unbiased covariance, Pearson's product moment correlation coefficient, etc. Among these, it is particularly preferable to use cosine similarity.

[0151] Methods for finding the distance between two vectors include Euclidean distance, standard (normalized, average) Euclidean distance, Mahalanobis distance, Manhattan distance, Chebyshev distance, Minkowski distance, and the like.

[0152] In this way, the information extraction unit 125 can extract items 241 that have the same content but at least some of the wording is different. Here, a language model may be used to find items 241 with the same content.

[0153] [Step S50] In step S50, the display data generation unit 127 generates display data DSPD3 based on the extracted items 241. Thereafter, the display data generation unit 127 outputs the display data DSPD3 to the information terminal 20. The display data DSPD3 is output to the information terminal 20 via the output unit 103. The display data DSPD3 indicates a screen layout. The extracted items 241 can be presented to the user of the information terminal 20 using this screen. The display data DSPD3 can be stored in the storage unit 110, for example.

[0154] 14B is a schematic diagram showing an example of display data DSPD3. FIG. 14B shows an example in which display data DSPD3 has a description "Potentially Differentiating Features" and information 234 below it. The information 234 includes extracted items 241. In the example shown in FIG. 14B, the information 234 includes "1) a1a1a1.", "2) b1b1b1.", and "3) c2c2c2." as items 241.

[0155] Here, in step S49, if multiple items 241 that indicate the same content but have at least some different wording are extracted, the display data generation unit 127 can include only one of these items 241 in the information 234. For example, the item 241 with the largest number of characters or words, or the item 241 with the smallest number of characters or words, can be included in the information 234.

[0156] 13 is performed after step S50 is performed. In step S23, the information terminal 20 displays the display data DSPD3. This allows the user of the information terminal 20 to check the information 234.

[0157] The above is an information processing method according to one embodiment of the present invention. As described above, even when the same prompt data is input to a language model, different response data may be generated. Therefore, by acquiring response data RD multiple times, it becomes easier to present desired information to the user of the information terminal 20. For example, an item with a small number of occurrences of response data RD is more likely to be inconsistent with the instruction sentence than an item with a large number of occurrences of response data RD. Therefore, by not including items with a small number of occurrences of response data RD in the display data DSPD3, it is possible to prevent information different from the desired information from being presented to the user of the information terminal 20. Similarly, even if response data RD is acquired based on prompt data with at least a different instruction sentence, it is possible to prevent information different from the desired information from being presented to the user of the information terminal 20.

[0158] In the example shown in FIG. 14A , items 241 with fewer occurrences of response data RD are more likely to be features not disclosed in the specification or features disclosed in cited documents than items 241 with more occurrences of response data RD. The information processing methods shown in FIGS. 12 and 13 can prevent such items 241 from being included in the display data DSPD3. This can prevent, for example, features not disclosed in the specification and features disclosed in cited documents from being presented to the user of the information terminal 20. This can prevent the user of the information terminal 20 from adding, for example, technical features not disclosed in the specification to the scope of claims. Furthermore, it can prevent the user of the information terminal 20 from adding, for example, technical features disclosed in cited documents to the scope of claims. As described above, one aspect of the present invention can improve the convenience, usefulness, and reliability of an information processing device, an information processing system, and an information processing method.

[0159] <Example 4 of Information Processing Method> An example of an information processing method according to one embodiment of the present invention will be described below. Specifically, the information processing method using the information processing system 10 shown in Fig. 2 is different from, for example, the examples shown in Figs. 3, 4, 9, 10, 12, and 13.

[0160] 15 and 16 are flowcharts illustrating an example of an information processing method according to one embodiment of the present invention. In FIG. 16, it is indicated whether each process is performed by the information terminal 20, the information processing device 100, or the information processing device 40.

[0161] Before step S61 shown in Fig. 15 is performed, step S81 shown in Fig. 16 is performed. In step S81, similar to step S21 shown in Fig. 4, for example, the user of the information terminal 20 inputs intellectual property data IPD to the information processing device 100.

[0162] [Steps S61 and S62] In step S61, the receiving unit 101 receives the intellectual property data IPD, similar to step S11 shown in Fig. 3, for example. Subsequently, in step S62, the document data extraction unit 121 extracts document data DOCD10 from the intellectual property data IPD. The document data DOCD10 may represent, for example, a specification, as shown in Fig. 5. The document data DOCD10 may be stored in the storage unit 110, for example.

[0163] [Step S63] In step S63, the prompt data generation unit 123 generates prompt data PPTD3. The prompt data generation unit 123 reads, for example, instruction statement data and the document data DOCD10 extracted by the document data extraction unit 121 from the storage unit 110. The prompt data generation unit 123 then generates prompt data PPTD3 including the read data. The prompt data PPTD3 can be stored, for example, in the storage unit 110.

[0164] 17A is a schematic diagram showing an example of prompt data PPTD3. The prompt data PPTD3 includes instruction sentence data DSD3 and document data DOCD10. The instruction sentence data DSD3 includes an instruction sentence for presenting features disclosed in the document data DOCD10. The instruction sentence can be, for example, "Please indicate the features disclosed in the specification. Please list multiple features in bullet points."

[0165] [Step S64] In step S64, the prompt data generation unit 123 outputs the prompt data PPTD3 to the information processing device 40. The prompt data PPTD3 is output to the information processing device 40 via the output unit 103 (step S64a shown in FIG. 16). The information processing device 40 generates response data RD3 based on the prompt data PPTD3. The response data RD3 is generated using a language model of the information processing device 40. The information processing device 40 outputs the response data RD3 to the information processing device 100 (step S64b shown in FIG. 16). The reception unit 101 of the information processing device 100 receives the response data RD3. As a result, the information processing device 100 acquires the response data RD3 (step S64c shown in FIG. 16). The response data RD3 can be stored in, for example, the storage unit 110.

[0166] FIG. 17B is a schematic diagram showing an example of response data RD3. Response data RD3 includes information 235. Information 235 includes h items 243 (h is an integer equal to or greater than 1). In FIG. 17B, items 243[1], 243[2], and 243[h] are shown as items 243. In the example shown in FIG. 17B, h can be an integer equal to or greater than 3. Items 243 can be features disclosed in document data DOCD10, such as technical features. Here, for example, one item 243 can be provided for each feature.

[0167] 17B shows an example in which response data RD3 includes the sentence, "This document discloses the following features." Also, FIG. 17B shows an example in which items 243[1], 243[2], and 243[h] are "1) aaa.", "2) bbb.", and "h) ddd.", respectively. Here, "1)," "2)," "h)," etc. are referred to as item numbers.

[0168] [Step S65] In step S65, the information extraction unit 125 extracts information 235 from the response data RD3. The information extraction unit 125 may extract, for example, itemized sentences from the sentences included in the response data RD3 as the information 235. Also in step S65, the document data extraction unit 121 extracts document data DOCD21 from the intellectual property data IPD. As described above, the document data DOCD21 may include, for example, cited documents. The extracted information 235 and the document data DOCD21 may be stored in, for example, the storage unit 110. The document data extraction unit 121 may also extract the document data DOCD21 in step S62. Also, the information extraction unit 125 may extract the document data DOCD21 in step S65.

[0169] [Step S66] In step S66, the display data generation unit 127 generates display data DSPD4 based on the information 235 and the document data DOCD21. Thereafter, the display data generation unit 127 outputs the display data DSPD4 to the information terminal 20. The display data DSPD4 is output to the information terminal 20 via the output unit 103. The display data DSPD4 indicates a screen layout. The information 235 and the document data DOCD21 can be presented to the user of the information terminal 20 using this screen. The display data DSPD4 can be stored in the storage unit 110, for example.

[0170] 17C is a schematic diagram showing an example of display data DSPD4. Fig. 17C shows an example in which display data DSPD4 includes document data DOCD10, information 235, and document data DOCD21[1] through DOCD21[m]. In the example shown in Fig. 17C, items 243[1] through 243[h] are shown as information 235. Note that Fig. 17C shows an example in which information 235 is shown under the heading "Features," and document data DOCD21[1] through DOCD21[m] are shown under the heading "Literature List."

[0171] After step S66 is performed, step S82 shown in Fig. 16 is performed. In step S82, the information terminal 20 displays display data DSPD4. This allows the user of the information terminal 20 to check the document data DOCD10, information 235, and document data DOCD21[1] to document data DOCD21[m].

[0172] In step S82, the user of information terminal 20 selects, for example, one item 243. The user of information terminal 20 also selects, for example, one piece of document data DOCD21. The item 243 and document data DOCD21 selected by the user of information terminal 20 are referred to as item 244 and document data DOCD22, respectively. Note that the user of information terminal 20 may select two or more items 243, or may select two or more pieces of document data DOCD21.

[0173] Selection of item 243 and document data DOCD21 can be performed using, for example, a mouse or a keyboard. Furthermore, if information terminal 20 is provided with a touch sensor, the user of information terminal 20 can select item 243 and document data DOCD21 by touching the display unit. Note that when document data DOCD21 is selected, the contents of that document data DOCD21 may be displayed. For example, when document data DOCD21[1] is selected, the specification, claims, abstract, and drawings included in document data DOCD21[1] may be displayed.

[0174] [Step S67] In step S67, the reception unit 101 receives the item 244 and the document data DOCD 22. That is, the reception unit 101 receives the item 243 selected by the user of the information terminal 20 in step S82 and the document data DOCD 21. The item 244 and the document data DOCD 22 received by the reception unit 101 can be stored in the storage unit 110, for example.

[0175] [Step S68] In step S68, the prompt data generation unit 123 generates prompt data PPTD4. The prompt data generation unit 123 reads, for example, instruction statement data, item 244, and document data DOCD22 from the storage unit 110. The prompt data generation unit 123 then generates prompt data PPTD4 including the read data. The prompt data PPTD4 can be stored, for example, in the storage unit 110.

[0176] FIG. 18A is a schematic diagram showing an example of prompt data PPTD4. The prompt data PPTD4 includes instruction sentence data DSD4, item 244, and document data DOCD22. In the example shown in FIG. 18A, item 244 is shown under the heading "Features." FIG. 18A also shows an example in which item 243[2] and document data DOCD21[2] were selected in step S82. Therefore, in FIG. 18A, item 244 is "bbb." and document data DOCD22 is "Patent Publication No. yyy."

[0177] The instruction data DSD4 has an instruction for detecting and presenting the location where the content of the item 244 is described from the document data DOCD22. The instruction can be, for example, "Please provide the paragraph from the cited document in which the following characteristics are described. Please indicate the paragraph in bullet points."

[0178] [Step S69] In step S69, the prompt data generation unit 123 outputs the prompt data PPTD4 to the information processing device 40. The prompt data PPTD4 is output to the information processing device 40 via the output unit 103 (step S69a shown in FIG. 16). The information processing device 40 generates response data RD4 based on the prompt data PPTD4. The response data RD4 is generated using a language model of the information processing device 40. The information processing device 40 outputs the response data RD4 to the information processing device 100 (step S69b shown in FIG. 16). The reception unit 101 of the information processing device 100 receives the response data RD4. As a result, the information processing device 100 acquires the response data RD4 (step S69c shown in FIG. 16). The response data RD4 can be stored in, for example, the storage unit 110.

[0179] 18B is a schematic diagram showing an example of response data RD4. In FIG. 18B, the response data RD4 includes a sentence such as "Patent Publication No. yyy describes the features as follows."

[0180] Response data RD4 includes information 237. Information 237 includes the location where the content of item 244 is described. In the example shown in FIG. 18B , information 237 includes the paragraph number in "Patent Gazette No. yyy" where the content of item 244 is described. Specifically, FIG. 18B shows an example in which information 237 includes the statement "1) In paragraph [00xy], it says ppp" and the statement "2) In paragraph [0xyz], it says that it is qqq."

[0181] [Step S70] In step S70, the information extraction unit 125 extracts information 237 from the response data RD4. The information extraction unit 125 can extract, as the information 237, for example, itemized sentences from the sentences included in the response data RD4.

[0182] [Step S71] In step S71, the display data generation unit 127 generates display data DSPD5 based on the information 237. Thereafter, the display data generation unit 127 outputs the display data DSPD5 to the information terminal 20. The display data DSPD5 is output to the information terminal 20 via the output unit 103. The display data DSPD5 indicates a screen layout. The screen can present the information 237 to the user of the information terminal 20. The display data DSPD5 can be stored in the storage unit 110, for example.

[0183] FIG. 18C is a schematic diagram showing an example of display data DSPD5. FIG. 18C shows an example in which display data DSPD5 includes document data DOCD10, information 235, and information 237. In the example shown in FIG. 18C, items 243[1] to 243[h] are shown as information 235. In FIG. 18C, item 243, which has been set as item 244, i.e., item 243 selected in step S82, is hatched. This makes it easier for the user of information terminal 20 to confirm the selected item 243. In the example shown in FIG. 18C, item 243[2] is hatched. Note that FIG. 18C shows an example in which information 235 is displayed below the heading "Features." In addition, FIG. 18C shows an example in which "Patent Publication No. yyy," which is the name of the document contained in document data DOCD22, is displayed below the heading "Literature List," and information 237 is displayed below that.

[0184] 16 is performed after step S71 is performed. In step S83, the information terminal 20 displays the display data DSPD5. This allows the user of the information terminal 20 to check the document data DOCD10, the information 235, and the information 237.

[0185] The above is an information processing method according to one embodiment of the present invention. With the information processing method according to one embodiment of the present invention, for example, a response to a rejection reason for an application can be made based on information generated by a language model. A user of the information terminal 20 can, for example, include content not described in the location indicated by information 237 in the claims as an invention-specifying matter. As described above, the information processing method according to one embodiment of the present invention can be an information processing method that is highly convenient, useful, and reliable. Furthermore, the information processing device and information processing system according to one embodiment of the present invention can be an information processing device and information processing system that are highly convenient, useful, and reliable.

[0186] A plurality of configuration examples shown in this embodiment mode can be combined as appropriate.

[0187] DOCD10: document data, DOCD11: document data, DOCD12: document data, DOCD13: document data, DOCD20: document data, DOCD21: document data, DOCD22: document data, DSD_1: instruction data, DSD_2: instruction data, DSD_3: instruction data, IPD: intellectual property data, PPTD: prompt data, PPTD_1: prompt data, PPTD_2: prompt data, PPTD_3: prompt data, RD[1]: response data, RD[2]: response data, RD[i]: response data, RD[k]: response data, RD: response data, RD_1: response data, RD_2: response data, RD_3: response data, 10: information processing system, 20: information terminal, 20a: information terminal, 20b: information terminal, 20c: information terminal, 20d: information terminal, 30: Network, 40: Information Processing Device, 100: Information Processing Device, 101: Reception Unit, 103: Output Unit, 110: Storage Unit, 121: Document Data Extraction Unit, 123: Prompt Data Generation Unit, 125: Information Extraction Unit, 127: Display Data Generation Unit, 130: Transmission Unit, 231[1]: Information, 231[2]: Information, 231[i]: Information, 231[k]: Information, 231: Information, 231_1: Information, 231_ 2: Information, 231_2a: Information, 231_2b: Information, 231_3: Information, 232: Information, 232a: Information, 232b: Information, 233: Information, 234: Information, 235: Information, 237: Information, 24 1 [1]: Item, 241 [2]: Item, 241 [3]: Item, 241 [k]: Item, 241: Item, 243 [1]: Item, 243 [2]: Item, 243 [h]: Item, 243: Item, 244: Item

Claims

The system includes a reception unit, a document data extraction unit, a prompt data generation unit, an information extraction unit, and a display data generation unit, the receiving unit has a function of receiving intellectual property data, the intellectual property data includes first document data having a description, second document data having claims, third document data having a notice of reasons for refusal, and fourth document data having cited documents; the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating prompt data including instruction sentence data and extracted document data, and outputting the generated prompt data to a language model; the information extraction unit has a function of extracting information from response data generated by the language model based on the prompt data; The information processing device, wherein the display data generating unit has a function of generating display data based on the information and outputting the display data to an information terminal. 　 In claim 1, the document data extraction unit has a function of extracting the first document data and the fourth document data; The instruction sentence data includes an instruction sentence for causing the language model to present a feature that is disclosed in the first document data and is not disclosed in the fourth document data. 　 In claim 1, the document data extraction unit has a function of extracting the third document data and the fourth document data; An information processing device, wherein the instruction sentence data has an instruction sentence for causing the language model to determine whether the reason for rejection included in the third document data is appropriately stated taking into account the content of the fourth document data. 　 In claim 1, the intellectual property data includes a plurality of the third document data; the document data extraction unit has a function of extracting a plurality of the third document data; The instruction sentence data includes an instruction sentence for causing the language model to summarize and present the contents of the plurality of third document data. 　 The system includes a reception unit, a document data extraction unit, a prompt data generation unit, and an information extraction unit, the receiving unit has a function of receiving intellectual property data, the intellectual property data includes first document data having a description, second document data having claims, third document data having a notice of reasons for refusal, and fourth document data having cited documents; the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating first prompt data and second prompt data each including instruction sentence data and at least one of extracted document data, and outputting the first prompt data and second prompt data to a language model; the information extraction unit has a function of extracting first information from first response data generated by the language model based on the first prompt data; the information extraction unit has a function of extracting second information from second response data generated by the language model based on the second prompt data; The second prompt data includes the first information. 　 In claim 5, the document data extraction unit has a function of extracting the first document data and the fourth document data; the instruction sentence data included in the first prompt data includes an instruction sentence for causing the language model to present a feature that is disclosed in the first document data and is not disclosed in the fourth document data; An information processing device, wherein the instruction sentence data contained in the second prompt data has an instruction sentence for causing the language model to determine whether the first information is disclosed in the first document data and whether the first information is not disclosed in the fourth document data. 　 In claim 5 or claim 6, A display data generating unit is included, The information processing device, wherein the display data generation unit has a function of generating display data based on the second information and outputting the display data to an information terminal. 　 The system includes a reception unit, a document data extraction unit, a prompt data generation unit, and an information extraction unit, the receiving unit has a function of receiving intellectual property data, the intellectual property data includes first document data having a description, second document data having claims, third document data having a notice of reasons for refusal, and fourth document data having cited documents; the document data extraction unit has a function of extracting at least one of the first to fourth document data, the prompt data generation unit has a function of generating prompt data including instruction sentence data and extracted document data; the prompt data generation unit has a function of outputting the prompt data to a language model k times (k is an integer equal to or greater than 2); the information extraction unit has a function of extracting first to k-th pieces of information from first to k-th pieces of response data generated by the language model based on the prompt data, respectively; each of the first to k-th pieces of information has one or more items; The information processing device, wherein the information extraction unit has a function of further extracting the items having the same content that are included in two or more of the first to kth information. 　 In claim 8, The information processing apparatus, wherein the information extraction unit has a function of calculating a similarity for each of the items included in the first to k-th information, and extracting the items whose similarity is equal to or greater than a predetermined value. 　 In claim 8 or claim 9, A display data generating unit is included, The information processing device, wherein the display data generation unit has a function of generating display data based on the extracted items and outputting the display data to an information terminal.

Images