Information processing systems, information processing methods

Abstract

To provide a novel information processing system that is superior in convenience, usefulness, or reliability. [Solution] An information processing system consisting of a first component, a second component, and a third component. The first component has the function of receiving documents and sending them to the third component, and receiving and providing criticisms. The second component has the function of generating questions, answers, and criticisms using a large-scale language model. The third component has the function of receiving documents, questions, answers, and criticisms, sharing them internally, and sending instructional texts to the second component. It also has the function of creating instructional texts. This makes it possible to ask questions, receive answers, and provide criticisms about documents.

Description

Technical Field

[0001] One embodiment of the present invention relates to an information processing system, an information processing method, or a semiconductor device.

[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, a manufacture, or a composition of matter. Therefore, more specific examples of the technical field of one embodiment of the present invention disclosed in this specification can include an information processing device, a semiconductor device, a storage device, driving methods thereof, or manufacturing methods thereof.

Background Art

[0003] In recent years, development of language models using neural networks has been actively carried out, and in particular, large language models (LLMs) have been attracting attention. A large language model is a natural language processing model trained using a large amount of data. With a large language model, for example, a dialogue model that gives responses to user instructions can be realized. Non-Patent Document 1 discloses GPT-4 (Generative Pre-trained Transformer 4) (registered trademark) as a large language model, and also discloses ChatGPT as a dialogue model.

[0004] The capabilities of natural language processing models have been significantly increased by using large language models. On the other hand, due to the increasing size of language models, it is difficult to incorporate and operate a language model in-house in terms of facilities and costs. For this reason, the use of external services that provide language models has become one mode of utilizing language models.

[0005] Furthermore, Patent Document 1 discloses an information processing device that receives document data, decomposes the text contained in the document data into multiple sentences, and extracts from the multiple sentences sentences that belong to a first classification in which the structure is described but the effect is not described, and sentences that belong to a second classification in which the effect is described but the structure is not described. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Summary of ChatGPT / GPT-4 Research and Perspective Towards the Future of Large Language Models, Yiheng Liu et al. (Submitted on 4 Apr 2023), [online], Internet<URL:https: / / arxiv.org / abs / 2304.01852> [Patent Documents]

[0007] [Patent Document 1] WO2024 / 176060 publication [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] One aspect of the present invention aims to provide a novel information processing system that is superior in convenience, usefulness, or reliability. Alternatively, it aims to provide a novel information processing method that is superior in convenience, usefulness, or reliability. Alternatively, it aims to provide a novel information processing system, a novel information processing method, or a novel semiconductor device.

[0009] Furthermore, the description of these problems does not preclude the existence of other problems. Moreover, one aspect of the present invention does not need to solve all of these problems. Other problems will naturally become apparent from the description in the specification, drawings, and claims, and it is possible to extract other problems from the description in the specification, drawings, and claims. [Means for solving the problem]

[0010] (1) One aspect of the present invention is an information processing system that receives a document and provides a critique of the description of the document relating to a question. The first response to the question may include "unable to answer," and the question may relate to the first response or document that does not include "unable to answer."

[0011] (2) One aspect of the present invention is an information processing system having a function for receiving a document and providing criticism, a function for transmitting a first instruction to a large-scale language model and receiving questions, a function for transmitting a second instruction to a large-scale language model and receiving responses, and a function for transmitting a third instruction to a large-scale language model and receiving criticism.

[0012] The first instruction includes the first instruction, the subject, and the conversation history, and the first instruction includes a procedure for generating questions related to the subject based on the conversation history. The subject is a document or a response. The conversation history includes questions and responses.

[0013] The second instruction document includes the second instruction, a question, and a document, and the second instruction includes a procedure for generating a response to the question from the description in the document. The response may include "unanswerable."

[0014] The third instruction document includes the third instruction, questions, answers, and documents, and the third instruction includes a procedure for generating critiques regarding the description of the documents relating to the answers to the questions.

[0015] (3) One aspect of the present invention is an information processing system having a first component, a second component, and a third component.

[0016] The first component has the functionality to receive documents and send them to the third component, and the functionality to receive and provide feedback.

[0017] The second component includes the functions of receiving a first instruction and sending a question to the third component, receiving a second instruction and sending a first response to the third component, receiving a third instruction and sending a critique to the third component, and processing using a large-scale language model. The large-scale language model includes the functions of generating a question according to the first instruction, generating a first response according to the second instruction, and generating a critique according to the third instruction.

[0018] The third component includes the functionality to receive and share documents, questions, first responses, and criticisms within the third component, the functionality to send the first, second, and third instructions to the second component, and the functionality to send criticisms to the first component. The third component also includes the first subcomponent.

[0019] The first subcomponent has the function of creating the first instruction, the second instruction, and the third instruction.

[0020] The first instruction includes the first instruction, the subject, and the conversation history, and the first instruction includes a procedure for generating questions related to the subject based on the conversation history. The subject is a document or the first response, and the conversation history includes questions and the first response.

[0021] The second instruction document includes a second instruction, a question, and a document, and the second instruction includes a procedure for generating a first response to the question from the description in the document. The first response may include "unanswerable."

[0022] The third instruction text includes a third instruction, a question and answer, a first response, and a document, and the third instruction includes a procedure for generating a critique with respect to the description of the document relating to the first response to the question and answer.

[0023] Thereby, it is possible to ask questions about the description of the document. In addition, it is possible to give the first response to the question and answer related to the description of the document. Further, it is possible to repeat the question and answer and the first response with respect to the description of the document. Furthermore, it is possible to examine the description of the document in detail. Moreover, it is possible to provide a critique for the description of the document relating to the first response to the question and answer. As a result, it is possible to provide a novel information processing system that is excellent in convenience, usefulness, or reliability.

[0024] (4) Further, one aspect of the present invention is the above information processing system, wherein the first sub-component comprises a function of creating a first instruction text when a first response that does not include unanswerable content or the document is shared, a function of creating a second instruction text when a question and answer is shared, and a function of creating a third instruction text when a first response that includes unanswerable content is shared.

[0025] Thereby, when a document or a first response that does not include unanswerable content is shared, a question and answer can be generated. In addition, when a first response that includes unanswerable content is shared, a critique can be generated. Further, it is possible to provide a critique for the description of the document relating to the first response to the question and answer. Furthermore, using the critique, for example, a user of the information processing system can be prompted to refine the document. As a result, it is possible to provide a novel information processing system that is excellent in convenience, usefulness, or reliability.

[0026] (5) Further, one aspect of the present invention is the above information processing system, wherein the first component has a function of receiving and providing records.

[0027] Furthermore, the third component has the function of creating a record and sending it to the first component. The record includes questions and answers and the first response.

[0028] This allows, for example, the use of records to encourage users of the information processing system to revise documents. As a result, it becomes possible to provide a novel information processing system that is superior in convenience, usefulness, or reliability.

[0029] (6) Another aspect of the present invention is the above-described information processing system, wherein the first component has the function of receiving question material and transmitting it to the third component. The question material includes either or both of the idea document and the frequently asked questions. The document is a document written from the idea document.

[0030] Furthermore, the third component has the function of receiving and sharing question materials within the third component. The first instruction statement includes the first instruction, the subject, the conversation history, and the question materials. The first instruction also includes a procedure for generating questions about the subject, taking into account the question materials and the conversation history.

[0031] This allows for the creation of questions and answers based on the idea document. It also allows for a thorough examination of whether the content of the idea document is completely and comprehensively included in the document. Furthermore, it allows for the creation of questions and answers based on frequently asked questions. Finally, it allows for a thorough examination of whether the document is written in a way that answers frequently asked questions. As a result, it becomes possible to provide a novel information processing system that is superior in terms of convenience, usefulness, and reliability.

[0032] (7) Another aspect of the present invention is the above-described information processing system, wherein the first component has the function of receiving and providing supplementary information.

[0033] The second component has the function of receiving the fourth and fifth directives and sending supplementary information to the third component. The large-scale language model has the function of generating a second response to the question according to the fourth directive, and the function of generating supplementary information regarding the second response according to the fifth directive.

[0034] The third component includes the functionality to accept supplementary information and send it to the first component, the functionality to send the fourth and fifth instructions to the second component, the functionality to share the search result set within the third component, and a second subcomponent.

[0035] The first subcomponent provides the functionality to generate queries, fourth directives, and fifth directives. A query includes a request to collect relevant prior art from a database and create a set of search results. The prior art falls within the scope of common knowledge in the field to which the document belongs. The fourth directive includes the fourth directive, the question, and the set of search results. The fourth directive includes a procedure for generating a second response to the question from the prior art. The second response includes bibliographic information identifying the prior art. The fifth directive includes the fifth directive, the second response, and the document. The fifth directive includes a procedure for generating supplementary information from the second response that complements the description of the document. The second response may include "unanswerable."

[0036] The second subcomponent comprises a database and a management system for performing processing. The database has the function of storing prior art, and the management system has the function of creating search result sets from the database in response to queries.

[0037] This allows, for example, the collection of prior literature from the scope of ordinary knowledge in the field to which the document belongs. Furthermore, it enables the generation of a second response to a question from the prior literature. It also allows the use of literature information to provide evidence for the second response to the question. Additionally, supplementary information can be used to provide the second response to the question. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0038] (8) Another aspect of the present invention is the above-described information processing system, wherein the first subcomponent has the following functions: a function to create a query when it shares a first response that includes an unanswerable response; a function to create a fourth instruction when it shares a set of search results; a function to create a fifth instruction when it shares a second response that does not include an unanswerable response; and a function to create a third instruction when it shares a second response that includes an unanswerable response.

[0039] This allows for the creation of a query and a search result set, including prior art, from the database when a first response, including an "unanswerable" response, is shared. Furthermore, when the search result set is shared, a second response to the question can be generated from the prior art. If the second response does not include an "unanswerable" response, supplementary information can be generated from the second response regarding the document description of the second response to the question. This supplementary information can then be used, for example, to encourage users of the information processing system to revise the document. Additionally, if the second response includes an "unanswerable" response, criticism can be provided regarding the document description of the question. This criticism can then be used, for example, to encourage users of the information processing system to revise the document. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0040] (9) One aspect of the present invention is an information processing method having the first to thirteenth steps.

[0041] In the first step, the first component receives the document and sends it to the second component.

[0042] In the second step, the second component receives the document and shares it within itself. The second component comprises the first subcomponent.

[0043] In the third step, the first subcomponent creates the first instruction and sends it to the third component. The first instruction includes the first instruction, the target, and the conversation history, and the first instruction includes a procedure to generate questions related to the target based on the conversation history. The target is a document the first time, and the first response from the second time onward, and the conversation history includes questions and the first response from the second time onward.

[0044] In the fourth step, the third component receives the first instruction, uses a large-scale language model to generate a question, and sends it to the second component.

[0045] In the fifth step, the second component receives questions and shares them internally.

[0046] In the sixth step, the first subcomponent creates a second instruction document and sends it to the third component. The second instruction document includes a second instruction, a question, and a document, and the second instruction includes a procedure for generating a first response to the question from the document's description. The first response may include "unanswerable."

[0047] In the seventh step, the third component receives the second instruction, uses a large-scale language model to generate the first response, and sends it to the second component.

[0048] In the eighth step, the second component receives the first response and shares it internally.

[0049] In step 9, when the second component shares a first response that does not contain an unanswerable response, proceed to step 3; and when the second component shares a first response that contains an unanswerable response, proceed to step 10.

[0050] In the tenth step, the first subcomponent creates a third set of instructions and sends them to the third component. The third set of instructions includes a third set of instructions, questions, and a document, and the third set of instructions includes a procedure for generating criticism regarding the description of the document relating to the first response to the questions.

[0051] In the 11th step, the third component receives the third instruction, uses a large-scale language model to generate a critique, and sends it to the second component.

[0052] In step 12, the second component receives the feedback and sends it to the first component.

[0053] In the 13th step, the first component accepts and provides feedback.

[0054] This allows for questions to be asked about the document's description. Furthermore, a first response can be provided to questions related to the document's description. The questioning and first response can be repeated regarding the document's description. The document's description can also be examined in detail. Critiques can be provided regarding the document's description in relation to the first response to the question. Additionally, when a first response that does not contain a document or an unanswerable response is shared, a question can be generated. When a first response that includes an unanswerable response is shared, a critique can be generated. Critiques can also be provided regarding the document's description in relation to the first response to the question. Furthermore, the critique can be used, for example, to encourage users of the information processing system to revise the document. As a result, a novel information processing method superior in convenience, usefulness, or reliability can be provided.

[0055] (10) Another aspect of the present invention is the above-described information processing method, wherein the ninth step is replaced by steps 14 through 24.

[0056] In step 14, when the second component shares a first response that does not contain an unanswerable response, proceed to step 3; and when the second component shares a first response that contains an unanswerable response, proceed to step 15.

[0057] In step 15, the first subcomponent creates a query and shares it within the second component. The query includes a request to collect relevant prior art from a database and create a set of search results. The prior art falls within the scope of common knowledge in the field to which the document belongs. The second component also comprises a second subcomponent, which has the functionality to perform processing using a database and a management system. The database has the functionality to store prior art, and the management system has the functionality to create a set of search results from the database in response to the query.

[0058] In step 16, the second subcomponent creates a set of search results and shares them within the second component.

[0059] In step 17, the first subcomponent creates a fourth instruction and sends it to the third component. The fourth instruction includes a fourth instruction, a question, and a set of search results, and the fourth instruction includes a procedure for generating a second response to the question from prior art. The second response includes bibliographic information identifying the prior art, and the second response may include "unanswerable".

[0060] In step 18, the third component receives the fourth instruction, uses a large-scale language model to generate a second response, and sends it to the second component.

[0061] In step 19, the second component receives the second response and shares it internally.

[0062] In step 20, when the second component shares a second response that includes an unanswerable response, proceed to step 10; when the second component shares a second response that does not include an unanswerable response, proceed to step 21.

[0063] In step 21, the first subcomponent creates a fifth instruction and sends it to the third component. The fifth instruction includes the fifth instruction, the second response, and the document, and the fifth instruction includes a procedure to generate supplementary information from the second response to complement the description of the document.

[0064] In step 22, the third component receives the fifth instruction, uses a large-scale language model to generate a supplement, and sends it to the second component.

[0065] In step 23, the second component receives the supplement and sends it to the first component.

[0066] In step 24, the first component accepts and provides supplements.

[0067] This allows for the collection of prior art from the scope of ordinary knowledge in the field to which the document belongs. Furthermore, a second response to a question can be generated from the prior art. The basis for the second response can be shown using the bibliographic information. Supplementary information can also be used to provide the second response to the question. When a first response, including an unanswerable answer, is shared, a query can be created to generate a search result set from the database that includes prior art. When the search result set is shared, a second response to the question can be generated from the prior art. When the second response does not include an unanswerable answer, supplementary information can be generated from the second response regarding the document's description of the question. Supplementary information can also be used to encourage users of the information processing system to revise the document. When the second response includes an unanswerable answer, criticism can be provided regarding the document's description of the question. Criticism can also be used to encourage users of the information processing system to revise the document. As a result, a novel information processing method with superior convenience, usefulness, or reliability can be provided. [Effects of the Invention]

[0068] One aspect of the present invention can provide a novel information processing system that is superior in convenience, usefulness, or reliability. Alternatively, it can provide a novel information processing method that is superior in convenience, usefulness, or reliability. Alternatively, it can provide a novel information processing system, a novel information processing method, or a novel semiconductor device.

[0069] Furthermore, the description of these effects does not preclude the existence of other effects. Moreover, one aspect of the present invention does not necessarily have to possess all of these effects. Other effects will naturally become apparent from the description in the specification, drawings, and claims, and it is possible to extract other effects from the description in the specification, drawings, and claims. [Brief explanation of the drawing]

[0070] [Figure 1] Figure 1 is a diagram illustrating the configuration of an information processing system according to an embodiment. [Figure 2] Figure 2(A) is a diagram illustrating the configuration of components used in the information processing system according to the embodiment, and Figure 2(B) is a diagram illustrating the configuration of the conversation history provided by the information processing system according to the embodiment. [Figure 3] Figure 3 is a diagram illustrating the configuration of components used in the information processing system according to the embodiment. [Figure 4] Figures 4(A) and 4(B) illustrate the structure of instruction statements used in the information processing system according to the embodiment. [Figure 5] Figures 5(A) and 5(B) illustrate the structure of instruction statements used in the information processing system according to the embodiment. [Figure 6] Figure 6 is a diagram illustrating the configuration of an information processing system according to an embodiment. [Figure 7] Figure 7(A) is a diagram illustrating the configuration of components used in the information processing system according to the embodiment, and Figure 7(B) is a diagram illustrating the configuration of the conversation history provided by the information processing system according to the embodiment. [Figure 8] Figure 8 is a diagram illustrating the configuration of components used in the information processing system according to the embodiment. [Figure 9] Figures 9(A) to 9(C) illustrate the configuration of a search result set or instruction statement used in the information processing system according to the embodiment. [Figure 10]Figure 10 is a diagram illustrating the configuration of an information processing device used in an information processing system according to an embodiment. [Figure 11] Figure 11 is a diagram illustrating an information processing method according to an embodiment. [Figure 12] Figure 12 is a diagram illustrating an information processing method according to an embodiment. [Figure 13] Figure 13 is a diagram illustrating an information processing method according to an embodiment. [Figure 14] Figure 14 is a diagram illustrating an information processing method according to an embodiment. [Modes for carrying out the invention]

[0071] An information processing system according to one aspect of the present invention comprises a first component, a second component, and a third component.

[0072] The first component has the functionality to receive documents and send them to the third component, and the functionality to receive and provide feedback.

[0073] The second component includes the functions of receiving a first instruction and sending a question to the third component, receiving a second instruction and sending a first response to the third component, receiving a third instruction and sending a critique to the third component, and processing using a large-scale language model. The large-scale language model includes the functions of generating a question according to the first instruction, generating a first response according to the second instruction, and generating a critique according to the third instruction.

[0074] The third component includes the functionality to receive and share documents, questions, first responses, and criticisms within the third component, the functionality to send the first, second, and third instructions to the second component, and the functionality to send criticisms to the first component. The third component also includes the first subcomponent.

[0075] The first subcomponent has the function of creating a first instruction, a second instruction, and a third instruction. The first instruction includes a first instruction and a subject, the first instruction including a procedure for generating a question related to the subject, the subject being a document or a first response. The second instruction includes a second instruction, a question, and a document, the second instruction including a procedure for generating a first response to the question from the description in the document. The first response may include "unanswerable". The third instruction includes a third instruction, a question, a first response, and a document, the third instruction including a procedure for generating a critique of the description in the document relating to the first response to the question.

[0076] This allows for questions to be asked about the document's description. Furthermore, an initial response can be provided to questions related to the document's description. The questioning and initial response process can be repeated. The document's description can be examined in detail. Critiques can also be provided regarding the document's description in relation to the initial response to the question. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0077] Embodiments will be described in detail with reference to the drawings. However, it will be readily apparent to those skilled in the art that the present invention is not limited to the following description, and that its form and details can be modified in various ways without departing from the spirit and scope of the present invention. Accordingly, the present invention is not to be interpreted as being limited to the contents of the embodiments shown below. In the configuration of the invention described below, the same reference numerals are used in common across different drawings for the same parts or parts having similar functions, and repeated descriptions are omitted.

[0078] In this specification, ordinal numbers such as "first," "second," etc., are used to avoid confusion of components and do not limit the number of components or the order of components (e.g., process order or layering order). Furthermore, even if a term does not have an ordinal number in this specification, an ordinal number may be added in the claims to avoid confusion of components. Even if a term has an ordinal number in this specification, a different ordinal number may be added in the claims. Even if a term has an ordinal number in this specification, an ordinal number may be omitted in the claims.

[0079] In the drawings attached to this specification, components are classified by function and shown as independent blocks in block diagrams. However, in reality, it is difficult to completely separate components by function, and a single component may be involved in multiple functions.

[0080] (Embodiment 1) In this embodiment, an information processing system according to one aspect of the present invention will be described with reference to Figures 1 to 10.

[0081] Figure 1 is a diagram illustrating the configuration of an information processing system according to one embodiment of the present invention.

[0082] Figure 2(A) is a diagram illustrating the configuration of components that can be used in an information processing system according to one embodiment of the present invention, and Figure 2(B) is a diagram illustrating the configuration of records provided by an information processing system according to one embodiment of the present invention.

[0083] Figure 3 illustrates the configuration of a component that can be used in an information processing system according to one embodiment of the present invention.

[0084] Figures 4(A) and 4(B) illustrate the structure of instruction statements transmitted and received within an information processing system according to one embodiment of the present invention.

[0085] Figures 5(A) and 5(B) illustrate the structure of instruction statements transmitted and received within an information processing system according to one embodiment of the present invention.

[0086] Figure 6 is a diagram illustrating the configuration of an information processing system according to one embodiment of the present invention.

[0087] Figure 7(A) is a diagram illustrating the configuration of components that can be used in an information processing system according to one embodiment of the present invention, and Figure 7(B) is a diagram illustrating the configuration of records provided by an information processing system according to one embodiment of the present invention.

[0088] Figure 8 illustrates the configuration of components that can be used in an information processing system according to one embodiment of the present invention.

[0089] Figure 9(A) is a diagram illustrating the structure of a search result set created by an information processing system according to one embodiment of the present invention, and Figures 9(B) and 9(C) are diagrams illustrating the structure of instruction statements transmitted and received within the information processing system according to one embodiment of the present invention.

[0090] Figure 10 is a block diagram illustrating the configuration of an information processing device that can be used in an information processing system according to one embodiment of the present invention.

[0091] <Example of an information processing system configuration 1> An information processing system according to one aspect of the present invention includes component 110, component 130, and component 120 (see Figure 1).

[0092] For example, the information processing devices that perform the functions of component 110, component 130, and component 120 each include a computing device and a communication device. Furthermore, these communication devices can be connected to each other via the network 51 to constitute an information processing system according to one embodiment of the present invention.

[0093] <Component 110 Configuration Example 1> Component 110 has the function of receiving a document Doc and sending it to component 120.

[0094] For example, a document showing how to implement the invention can be used as Document Doc. A document written in a manner that enables a person with ordinary skill in the art to which the invention for which a patent is sought can implement the invention can also be used as Document Doc. Furthermore, a document that specifically demonstrates how to implement the invention can also be used as Document Doc. In this specification, Document Doc includes a document describing embodiments of the invention, and preferably includes a document specifically illustrating embodiments of the invention.

[0095] For example, a user 99 of the information processing system inputs a document Doc into component 110. Or, for example, inputs a command into component 110 to select and send a document Doc stored in various storage devices. Specifically, the user 99 of the information processing system inputs into component 110 using input devices such as a keyboard, mouse, facsimile machine, or eye-tracking device.

[0096] Furthermore, component 110 has a function to accept criticism (Cmnt) regarding the description of document Doc and provide it to, for example, the user 99 of the information processing system. Specifically, it provides it to the user 99 of the information processing system using output devices such as display devices, speakers, printers, facsimile machines, and various storage devices.

[0097] Component 110 includes, for example, a display device Dsp and an input device CID (see Figure 2(A)). The display device Dsp has the function of displaying a document Doc or a critique Cmnt. The display device Dsp also has the function of displaying a conversation history CH (see Figure 2(B)).

[0098] <Component 130 Configuration Example 1> Component 130 has the function of receiving instruction Pt1 and sending question Q to component 120, receiving instruction Pt2 and sending response A1 to component 120, and receiving instruction Pt3 and sending criticism Cmnt to component 120 (see Figure 1). Component 130 also has the function of performing processing using the large-scale language model LLM.

[0099] 《Example Configuration of a Large-Scale Language Model (LLM) 1》 The large-scale language model LLM includes a function to generate a question Q according to instruction Pt1, a function to generate a response A1 according to instruction Pt2, and a function to generate a critique Cmnt according to instruction Pt3.

[0100] For example, large-scale language models such as GPT-3®, GPT-3.5, GPT-4®, LaMDA, Llama2, or Llama3 can be used in large-scale language model LLMs.

[0101] <Component 120 Configuration Example 1> Component 120 has the functionality to accept documents (Doc), questions (Q), answers (A1), and critiques (Cmnt), and to share them within Component 120.

[0102] Furthermore, component 120 includes the function of sending instruction statements Pt1, Pt2, and Pt3 to component 130, and the function of sending criticism Cmnt to component 110.

[0103] Furthermore, component 120 includes subcomponent 120A (see Figure 3). For the sake of clarity, in this specification, a configuration having one or more functions is referred to as a component or subcomponent. Also, when a component has the function of sharing information internally and includes multiple subcomponents, the multiple subcomponents can utilize the same information.

[0104] 《Example Configuration of Subcomponent 120A》 Subcomponent 120A has the function of creating instruction statement Pt1, instruction statement Pt2, and instruction statement Pt3.

[0105] [Example 1 of the structure of instruction statement Pt1] Instruction Pt1 includes instruction g1, the target object (Obj), and the conversation history CH (see Figure 4(A)). Instruction g1 includes a procedure for generating a question Q related to the target object, based on the conversation history CH. The target object is either document Doc or response A1, and the conversation history CH includes the question Q and response A1.

[0106] For example, a document (Doc) can be divided into one or more semantic paragraphs, and each semantic paragraph can be used as the target object (Obj). Also, when creating an instruction sentence (Pt1) for the first time, the conversation history (CH) is empty, but from the second time onward, the conversation history (CH) includes a question (Q) and an answer (A1).

[0107] For example, when using document Doc as the target Object, the document in the following paragraph can be used as instruction statement Pt1.

[0108] Please create several questions (Q) regarding the use of a document (Doc) as the target object (Obj).

[0109] Furthermore, for example, when response A1 is the target object, the document in the following paragraph can be used as instruction statement Pt1.

[0110] "Regarding response A1, please create a new question Q based on the conversation history CH."

[0111] [Example of the structure of instruction statement Pt2] Instruction Pt2 includes instruction g2, question Q, and document Doc (see Figure 5(A)). Instruction g2 includes a procedure for generating response A1 to question Q from the description in document Doc. Response A1 may include an ITA (Information Item A) that cannot be answered.

[0112] For example, the document in the following paragraph can be used in instruction Pt2. If not specified in document Doc, a response indicating that it is an unanswerable ITA will be sent. Specifically, response A1 will include "Not specified".

[0113] "Please prepare your answer to Question Q from the document Doc. If the answer is not found in the document Doc, please write 'Not found'."

[0114] [Example of the structure of instruction statement Pt3] Instruction Pt3 includes instruction g3, question Q, response A1, and document Doc (see Figure 5(B)). Instruction g3 includes a procedure for generating a critique Cmnt regarding the description in document Doc concerning response A1 to question Q.

[0115] For example, the document in the following paragraph can be used as instruction statement Pt3.

[0116] "Regarding Question Q, it is not possible to answer it from the document Doc. Please tell us what information should be added to the document Doc in order to answer Question Q."

[0117] This allows for questions (Q) to be asked about the description in document Doc. It also allows for responses (A1) to questions (Q) related to the description in document Doc. Furthermore, the questioning and responses (A1) regarding the description in document Doc can be repeated. Additionally, the description in document Doc can be examined in detail. Furthermore, critical comments (Cmnt) can be provided regarding the description in document Doc related to responses (A1) to questions (Q). As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0118] 《Example Configuration of Subcomponent 120A 2》 Furthermore, subcomponent 120A has the function of creating instruction statement Pt1 when response A1 or document Doc that does not contain an unanswerable ITA is shared. Subcomponent 120A also has the function of creating instruction statement Pt2 when question Q is shared. This allows for repeated question and answer sessions.

[0119] Furthermore, subcomponent 120A has the function of creating instruction statement Pt3 when it shares response A1 which includes an unanswerable ITA. In other words, when the large-scale language model LLM responds that it cannot answer question Q from the description in document Doc, subcomponent 120A creates instruction statement Pt3.

[0120] This allows for the generation of a question / question Q when a response A1 or document Doc is shared that does not contain an unanswerable ITA. It also allows for the generation of a critique Cmnt when a response A1 containing an unanswerable ITA is shared. Furthermore, a critique Cmnt can be provided regarding the description in document Doc concerning the response A1 to the question / question Q. Additionally, the critique Cmnt can be used, for example, to encourage users of the information processing system to revise document Doc. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0121] <Component 110 Configuration Example 2> Component 110 has the function of receiving a conversation history CH and providing it, for example, to a user 99 of the information processing system (see Figure 1).

[0122] <Component 120 Configuration Example 2> Component 120 has the function of creating a conversation history CH and sending it to component 110. The conversation history CH includes question Q and answer A1 (see Figure 2(B)).

[0123] This allows, for example, the use of the conversation history (CH) to prompt users of the information processing system to revise documents (Doc). As a result, it becomes possible to provide a novel information processing system that is superior in terms of convenience, usefulness, and reliability.

[0124] <Component 110 Configuration Example 3> Component 110 has the function of receiving question material (QM) and sending it to component 120 (see Figure 1).

[0125] [Example of Questionnaire QM Structure] Question Material QM includes points of focus or items to consider when generating questions. For example, Question Material QM may include either an Idea Document CN or a Frequently Asked Questions (FAQ), or both. Note that the Frequently Asked Questions (FAQ) can be rephrased as a collection of anticipated questions.

[0126] Document Doc is a document written from the Idea CN. For example, when using Document Doc as a document showing how to implement the invention, a document written so that a person with ordinary skill in the art to which the invention for which a patent is sought can implement it, or a document that shows how to implement it in detail, bullet points, memos, or notes showing how to implement the invention can be used in the Idea CN.

[0127] Furthermore, frequently asked questions (FAQs) can be used to address questions such as: "Are the operation and effects of the invention described?", "Are the reasons for the preferred configuration described?", "Are the definitions of technical terms provided?", and "Are the detailed conditions described for the experimental data?".

[0128] <Component 120 Configuration Example 3> Component 120 has the function of receiving question material (QM) and sharing it internally within Component 120.

[0129] [Example of the structure of instruction statement Pt1 2] Instruction Pt1 includes instruction g1, the target object (Obj), the conversation history CH, and the question material QM (see Figure 4(B)). Instruction g1 includes a procedure for generating a question Q about the target object (Obj) based on the conversation history CH and referring to the question material QM.

[0130] For example, when using document Doc as the target Object, the document in the following paragraph can be used as instruction statement Pt1.

[0131] "Create multiple questions (Q) related to the question material (QM) from the document (Doc)."

[0132] Furthermore, for example, when response A1 is the target object, the document in the following paragraph can be used as instruction statement Pt1.

[0133] "Regarding response A1, please create a new question Q related to question material QM, taking into account the conversation history CH."

[0134] This allows for the creation of questions and answers (Q) by referring to the Idea Document (CN). It also allows for a detailed examination of whether the contents of the Idea Document (CN) are completely included in the Document (Doc). Furthermore, it allows for the creation of questions and answers (Q) by referring to the Frequently Asked Questions (FAQ). Finally, it allows for a detailed examination of whether the Document (Doc) is written in a way that answers the Frequently Asked Questions (FAQ). As a result, it becomes possible to provide a novel information processing system that is superior in terms of convenience, usefulness, and reliability.

[0135] <Component 110 Configuration Example 4> Component 110 has the function of accepting supplementary Apx for the description of document Doc and providing it, for example, to the user 99 of the information processing system (see Figure 6).

[0136] Component 110 includes, for example, a display device Dsp and an input device CID (see Figure 7(A)). The display device Dsp has the function of displaying a document Doc or supplement Apx. The display device Dsp also has the function of displaying a conversation history CH (see Figure 7(B)).

[0137] <Component 130 Configuration Example 2> Component 130 has the function of receiving instruction statements Pt4 and Pt5 and sending supplementary Apx to component 120 (see Figure 6).

[0138] 《Example Configuration of a Large-Scale Language Model (LLM) 2》 The large-scale language model LLM has a function to generate a response A2 to a question Q according to instruction Pt4, and a function to generate supplementary Apx related to response A2 according to instruction Pt5.

[0139] <Component 120 Configuration Example 4> Component 120 includes the function of receiving supplementary Apx and sending it to component 110, the function of sending instruction Pt4 and instruction Pt5 to component 130, and the function of sharing the search result set SRS within component 120.

[0140] 《Example Configuration of Subcomponent 120A 3》 Subcomponent 120A has the functionality to create a query Qu, instruction Pt4, and instruction Pt5.

[0141] [Example of Query Qu configuration] A query (Qu) includes a request to collect relevant prior art references (Ref) from the database (DB) and create a search result set (SRS). Note that the prior art references (Ref) fall within the scope of common knowledge in the field to which the document (Doc) belongs.

[0142] For example, the search result set SRS includes a prior art reference (Ref) and a reference information RID that identifies the prior art reference (see Figure 9(A)).

[0143] [Example of the structure of instruction sheet Pt4] Instruction Pt4 includes instruction g4, question Q, and search result set SRS (see Figure 9(B)). Instruction g4 includes a procedure for generating response A2 to question Q from prior art reference Ref. Response A2 includes the bibliographic information RID that identifies the prior art reference Ref.

[0144] For example, the document in the following paragraph can be used in instruction Pt4. If there is no prior art reference, respond with an "ITA" indicating that the answer is unavailable. Specifically, response A2 includes "Not listed".

[0145] "For question Q, please prepare your answer based on the prior art references (Ref). If the information is not found in the prior art references, please answer 'Not found'."

[0146] [Example of the structure of instruction sheet Pt5] Instruction Pt5 includes instruction g5, response A2, and document Doc (see Figure 9(C)). Instruction g5 includes a procedure for generating supplementary Apx from response A2 to supplement the description in document Doc.

[0147] For example, the document in the following paragraph can be used as instruction Pt5.

[0148] "Questions regarding the target Objective (Obj) cannot be answered from the document (Doc). To enable this, please create a proposal to supplement the document (Doc) based on response A2."

[0149] <Component 120 Configuration Example 5> Furthermore, component 120 includes subcomponent 120B (see Figure 8).

[0150] 《Example configuration of subcomponent 120B》 Subcomponent 120B includes the functionality to perform processing using a database DB and a management system DBMS.

[0151] Furthermore, the database (DB) has the function of storing prior art references (Ref). For example, archives of public documents, private documents, and publicly known documents can be used in the database (DB). Specifically, published patent gazettes, patent publications, books, and other documents can be included in the search targets. In addition, the management system (DBMS) has the function of creating a search result set (SRS) from the database (DB) according to the query (Qu).

[0152] This allows, for example, the collection of prior art references (Ref) from the scope of common knowledge in the field to which the document (Doc) belongs. Furthermore, it enables the generation of answers (A2) to questions (Q) from these prior art references. Additionally, the basis for the answers (A2) can be shown using the literature information (RID). Supplementary information (Apx) can also be used to provide the answers (A2) to questions. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0153] 《Example Configuration of Subcomponent 120A 4》 Furthermore, subcomponent 120A has the function of creating query Qu when it shares response A1 which includes an unanswerable ITA. In other words, when the large-scale language model LLM responds that it cannot answer question Q from the description in document Doc, subcomponent 120A creates query Qu.

[0154] Furthermore, subcomponent 120A has the function of creating instruction statement Pt4 when the search result set SRS is shared.

[0155] Furthermore, subcomponent 120A has the function of creating instruction statement Pt5 when it shares response A2 that does not contain an unanswerable ITA. In other words, when the large-scale language model LLM answers question Q from the description in prior art Ref, subcomponent 120A creates instruction statement Pt5.

[0156] Furthermore, subcomponent 120A has the function of creating instruction statement Pt3 when it shares response A2 which includes an unanswerable ITA. In other words, when the large-scale language model LLM responds that it cannot answer question Q not only from the document Doc but also from the search result set SRS, subcomponent 120A creates instruction statement Pt3.

[0157] This allows for the creation of a query Qu and a search result set SRS containing prior art references Ref from the database DB when a response A1 containing an unanswerable ITA is shared. Furthermore, when the search result set SRS is shared, a response A2 to question Q can be generated from the prior art references Ref. If response A2 does not contain an unanswerable ITA, a supplementary Apx can be generated from response A2 regarding the description of the document Doc related to question Q. The supplementary Apx can then be used, for example, to encourage users of the information processing system to revise the document Doc. Additionally, if response A2 contains an unanswerable ITA, a critique Cmnt can be provided regarding the description of the document Doc related to question Q. The critique Cmnt can then be used, for example, to encourage users of the information processing system to revise the document Doc. As a result, a novel information processing system with superior convenience, usefulness, and reliability can be provided.

[0158] <Example of Information Processing System Configuration 2> The information processing system described in this embodiment includes component 110, component 120, and component 130 (see Figure 1).

[0159] For example, an information processing system according to one aspect of the present invention can be configured with an information processing device that performs the functions of component 110, an information processing device that performs the functions of component 120, and an information processing device that performs the functions of component 130. The number of information processing devices constituting the information processing system according to one aspect of the present invention is one or more. Furthermore, for example, an information processing system according to one aspect of the present invention can be configured by connecting multiple information processing devices using a network 51.

[0160] By configuring an information processing system according to one aspect of the present invention using multiple information processing devices, the load related to information processing can be distributed.

[0161] <Example of Information Processing Device Configuration 1> The configuration example 1 of the information processing device described in this embodiment can be used for component 110. The configuration example 1 of the information processing device can also be called a client computer or the like. For example, a desktop computer or a mobile computer can be used for component 110.

[0162] Configuration Example 1 of the information processing device can receive data input by a user of an information processing system according to one aspect of the present invention. Furthermore, Configuration Example 1 of the information processing device can provide the user with data output by an information processing system according to one aspect of the present invention.

[0163] In component 110, for example, dedicated application software, a web browser, etc., are operated. Users of an information processing system according to one embodiment of the present invention can access the information processing system through either of these. As a result, they can enjoy services using the information processing system according to one embodiment of the present invention.

[0164] <Example of Information Processing Device Configuration 2> The configuration example 2 of the information processing device described in this embodiment can be used for component 120. For example, a workstation, server computer, supercomputer, etc., can be used for component 120.

[0165] Furthermore, it is preferable that the second example of the information processing device configuration has the functionality of a parallel computer. By using it as a parallel computer, it is possible to perform large-scale calculations necessary for, for example, learning and inference of artificial intelligence (AI).

[0166] Furthermore, the second example of the information processing device configuration can perform processing using a natural language model with AI.

[0167] For example, it is preferable to be able to perform processing using natural language models such as GPT-3®, GPT-3.5, GPT-4®, LaMDA, Llama2, and Llama3.

[0168] <Example of Information Processing Device Configuration 3> For example, the configuration example 3 of the information processing device described in this embodiment can be used for component 130. Note that component 130 is larger in scale and has higher computing power than component 120. For example, a large computer such as a server computer or supercomputer can be used for component 130.

[0169] Furthermore, it is preferable that the information processing device configuration example 3 has the functionality of a parallel computer. By using it as a parallel computer, for example, it can perform large-scale calculations necessary for AI training and inference.

[0170] Furthermore, the third example of the information processing device configuration can perform processing using an AI-based natural language model. In particular, it can execute processing using a general-purpose language model capable of performing various natural language processing tasks.

[0171] For example, processing can be performed using natural language models such as GPT-3(registered trademark), GPT-3.5, GPT-4(registered trademark), LaMDA, Llama2, and Llama3. Processing using GPT-4(registered trademark) is particularly preferable. For instance, being able to perform processing using large-scale language models, which are significantly larger than conventional natural language models, can enable more natural document generation or dialogue.

[0172] Furthermore, those who provide services using an information processing system according to one embodiment of the present invention do not necessarily need to own the configuration example 3 of the information processing device themselves. For example, a service provider can utilize a portion of the services provided by other businesses, etc., using the configuration example 3 of the information processing device.

[0173] <Example of Network 51 Configuration> A network 51 that can be used in an information processing system according to one embodiment of the present invention can connect multiple information processing devices. This allows the connected multiple information processing devices to transmit and receive data from each other. Furthermore, it allows for the distribution of the information processing load.

[0174] When performing wireless communication, communication protocols or technologies such as 4th generation mobile communication systems (4G), 5th generation mobile communication systems (5G), 6th generation mobile communication systems (6G), or specifications standardized by IEEE such as Wi-Fi® and Bluetooth® can be used.

[0175] For example, a local network can be used as network 51. An intranet or extranet can also be used as network 51. Furthermore, PAN (Personal Area Network), LAN (Local Area Network), CAN (Campus Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), GAN (Global Area Network), etc., can be used as network 51.

[0176] Furthermore, for example, a global network can be used for network 51. Specifically, the internet, which is the foundation of the World Wide Web (WWW), can be used.

[0177] Furthermore, a person who provides services using an information processing system according to one aspect of the present invention can, for example, provide services using an information processing method according to one aspect of the present invention via a network 51.

[0178] Furthermore, if an information processing system according to one embodiment of the present invention is built within a local network, the possibility of confidential information leakage can be reduced compared to using the Internet.

[0179] <Example of Information Processing Device Configuration 4> An information processing device 20 that can be used in an information processing system according to one embodiment of the present invention includes, for example, an input unit 21, a storage unit 22, a processing unit 23, an output unit 24, and a transmission line 25 (see Figure 10).

[0180] In the drawings attached to this specification, the components are classified by function and shown as independent blocks in the block diagram. However, in reality, it is difficult to completely separate the components by function, and one component may be involved in multiple functions. For example, a part of the processing unit 23 may function as the input unit 21. Also, one function may be involved in multiple components. For example, the processing performed by the processing unit 23 may be executed by different information processing devices depending on the processing.

[0181] <Input section 21> The input unit 21 can receive data from outside the information processing device. For example, the input unit 21 can receive data via the network 51. Specifically, a device such as a personal computer equipped with a communication port or communication function can be used.

[0182] The input unit 21 supplies the received data to either or both of the storage unit 22 and the processing unit 23 via the transmission line 25.

[0183] <Storage section 22> The memory unit 22 has the function of storing the program executed by the processing unit 23. The memory unit 22 may also have the function of storing data generated by the processing unit 23 (for example, calculation results, analysis results, inference results), data received by the input unit 21, etc.

[0184] The storage unit 22 may have a database. The information processing device may also have a database separate from the storage unit 22. The information processing device may have the function to retrieve data from a database located outside the storage unit 22, outside the information processing device itself, or outside the information processing system. Furthermore, the information processing device may have the function to retrieve data from both its own database and an external database.

[0185] Either or both of the storage and / or file server can be used in the storage unit 22. Furthermore, a database recording the paths of files stored on the file server can be used in the storage unit 22.

[0186] The storage unit 22 includes at least one of volatile memory and non-volatile memory. Examples of volatile memory include DRAM (Dynamic Random Access Memory) and SRAM (Static Random Access Memory). Examples of non-volatile memory include ReRAM (Resistive Random Access Memory), PRAM (Phase Change Random Access Memory), FeRAM (Ferroelectric Random Access Memory), MRAM (Magnetoresistive Random Access Memory), and flash memory. The storage unit 22 may also include at least one of NOSRAM (registered trademark) and DOSRAM (registered trademark). The storage unit 22 may also include a recording media drive. Examples of recording media drives include hard disk drives (HDD) and solid state drives (SSD).

[0187] NOSRAM is an abbreviation for "Nonvolatile Oxide Semiconductor Random Access Memory (RAM)". NOSRAM is a type of memory where the memory cell is a 2-transistor (2T) or 3-transistor (3T) gain cell, and the transistors are transistors that use metal oxide in the channel formation region (also called OS transistors). OS transistors have an extremely small current flowing between the source and drain when off, i.e., a leakage current. By utilizing the characteristic of extremely low leakage current, NOSRAM can be used as a non-volatile memory by holding charge corresponding to the data within the memory cell. In particular, NOSRAM can read the stored data without destroying it (non-destructive read), making it suitable for computational processing that involves repeating data read operations a large amount. Because the data capacity of NOSRAM can be increased by stacking it, it can be used as a large-scale cache memory, main memory, or storage memory to improve the performance of semiconductor devices.

[0188] DOSRAM is an abbreviation for "Dynamic Oxide Semiconductor RAM," and refers to RAM with a 1T (transistor) 1C (capacitance) type memory cell. DOSRAM is a type of DRAM formed using OS transistors, and it is a memory that temporarily stores information sent from an external source. DOSRAM is a memory that takes advantage of the low off-current of OS transistors.

[0189] In this specification, "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 called oxide semiconductors or simply OS), etc. For example, when a metal oxide is used in the semiconductor layer of a transistor, that metal oxide may be referred to as an oxide semiconductor.

[0190] The metal oxide in the channel-forming region preferably contains indium (In). When the metal oxide in the channel-forming region contains indium, the carrier mobility (electron mobility) of the OS transistor is increased. For example, indium oxide (InOx) or indium gallium zinc oxide (In-Ga-Zn oxide, also written as "IGZO") can be used in the channel-forming region. Furthermore, the metal oxide in the channel-forming region is preferably an oxide semiconductor containing element M. Element M is preferably at least one of aluminum (Al), gallium (Ga), and tin (Sn). Other elements applicable to element M include boron (B), silicon (Si), titanium (Ti), iron (Fe), nickel (Ni), germanium (Ge), yttrium (Y), zirconium (Zr), molybdenum (Mo), lanthanum (La), cerium (Ce), neodymium (Nd), hafnium (Hf), tantalum (Ta), and tungsten (W). However, in some cases, element M may be a combination of multiple elements as mentioned above. Element M is, for example, an element with a high bond energy with oxygen. For example, an element with a higher bond energy with oxygen than indium. Furthermore, the metal oxide containing the channel-forming region is preferably a metal oxide containing zinc (Zn). Metal oxides containing zinc may be more prone to crystallization.

[0191] The metal oxides present in the channel-forming regions are not limited to indium-containing metal oxides. For example, the metal oxides present in the channel-forming regions may be zinc-tin oxides, gallium-tin oxides, or other metal oxides that do not contain indium but contain zinc, gallium, or tin.

[0192] <Processing 23> The processing unit 23 has the function of performing calculations, analyses, and inferences using data supplied from either or both of the input unit 21 and the storage unit 22. The processing unit 23 can supply the generated data (e.g., calculation results, analysis results, inference results) to either or both of the storage unit 22 and the output unit 24.

[0193] The processing unit 23 has the function of acquiring data from the storage unit 22. The processing unit 23 may also have the function of recording or registering data in the storage unit 22.

[0194] The processing unit 23 may, for example, have an arithmetic circuit. The processing unit 23 may, for example, have a central processing unit (CPU). The processing unit 23 may also have a graphics processing unit (GPU). The processing unit 23 may also have a neural processing unit (NPU / neural network processing unit).

[0195] The processing unit 23 may have a microprocessor such as a DSP (Digital Signal Processor). The microprocessor can be implemented using a PLD (Programmable Logic Device) such as an FPGA (Field Programmable Gate Array) or FPAA (Field Programmable Analog Array). The processing unit 23 may also have a quantum processor. The processing unit 23 can perform various data processing and program control by interpreting and executing instructions from various programs via the processor. Programs that can be executed by the processor are stored in at least one of the processor's memory area and the storage unit 22.

[0196] The processing unit 23 may have main memory. The main memory may include at least one of volatile memory such as RAM and non-volatile memory such as ROM (Read Only Memory). Furthermore, the main memory may include at least one of the above-mentioned NOSRAM and DOSRAM.

[0197] For RAM, for example, DRAM or SRAM is used, and a virtual memory space is allocated and used as the workspace for the processing unit 23. The operating system, application programs, program modules, program data, and lookup tables stored in the storage unit 22 are loaded into RAM for execution. These data, programs, and program modules loaded into RAM are each directly accessed and manipulated by the processing unit 23.

[0198] ROM can store BIOS (Basic Input / Output System) and firmware, etc., which do not require rewriting. Examples of ROM include mask ROM, OTPROM (One Time Programmable Read Only Memory), and EPROM (Erasable Programmable Read Only Memory). Examples of EPROM include UV-EPROM (Ultra-Violet Erasable Programmable Read Only Memory), which allows data to be erased by ultraviolet irradiation, EEPROM (Electrically Erasable Programmable Read Only Memory), and flash memory.

[0199] The processing unit 23 may have either or both an OS transistor and a transistor having silicon in its channel formation region (Si transistor).

[0200] The processing unit 23 preferably has an OS transistor. Because the OS transistor has an extremely small off-current, using the OS transistor as a switch to hold the charge (data) that has flowed into a capacitive element that functions as a memory element ensures that the data can be retained for a long period of time. By using this characteristic in at least one of the registers and cache memory of the processing unit, the processing unit can be operated only when necessary, and at other times the information of the previous processing is saved to the memory element, thereby turning off the processing unit. In other words, normally-off computing becomes possible, and the power consumption of the information processing system can be reduced.

[0201] It is preferable for information processing devices to use AI for at least some of their processing.

[0202] Information processing devices preferably utilize artificial neural networks (ANNs, also simply referred to as neural networks). Neural networks are implemented using circuits (hardware) or programs (software).

[0203] In this specification, the term "neural network" refers to any model that mimics the neural network of living organisms, determines the strength of connections between neurons through learning, and possesses problem-solving capabilities. A neural network has an input layer, an intermediate layer (hidden layer), and an output layer.

[0204] In this specification and other documents, when discussing neural networks, the process of determining the connection strength (also called weight coefficient) between neurons from existing information is sometimes referred to as "learning."

[0205] In this specification and other documents, the process of constructing a neural network using connection strengths obtained through learning and deriving new conclusions from it may be referred to as "inference."

[0206] <Output section 24> The output unit 24 can output at least one of the calculation results, analysis results, and inference results from the processing unit 23 to the outside of the information processing device. For example, the output unit 24 can transmit data via the network 51. Specifically, a device such as a personal computer equipped with a communication port or communication function can be used. Alternatively, a device equipped with a communication function may be used for both the input unit 21 and the output unit 24.

[0207] <Transmission Line 25> The transmission line 25 has the function of transmitting data. Data can be transmitted and received between the input unit 21, the storage unit 22, the processing unit 23, and the output unit 24 via the transmission line 25. Specifically, an external bus, LAN, or the internet can be used as the transmission line 25.

[0208] This embodiment can be appropriately combined with other embodiments shown in this specification.

[0209] (Embodiment 2) In this embodiment, an information processing method according to one aspect of the present invention will be described with reference to Figures 11 to 14.

[0210] Figure 11 is a flowchart illustrating an information processing method according to one embodiment of the present invention.

[0211] Figure 12 is a flowchart illustrating an information processing method according to one embodiment of the present invention.

[0212] Figure 13 is a flowchart illustrating an information processing method according to one embodiment of the present invention.

[0213] Figure 14 is a flowchart illustrating an information processing method according to one embodiment of the present invention.

[0214] <Example of information processing method 1> An information processing method according to one aspect of the present invention comprises steps S1 to S13 (see Figures 11 and 12).

[0215] <Step S1> In step S1, component 110 receives the document Doc and sends it to component 120 (see Figure 11).

[0216] For example, a user 99 of the information processing system inputs a document Doc into component 110. Or, for example, inputs a command into component 110 to select and send a document Doc stored in a storage device.

[0217] <Step S2> In step S2, component 120 receives the document Doc and shares it internally. Component 120 includes subcomponent 120A.

[0218] <Step S3> In step S3, subcomponent 120A creates instruction statement Pt1 and sends it to component 130.

[0219] Instruction Pt1 includes instruction g1, target object (Obj), and conversation history CH. Instruction g1 includes a procedure for generating a question (Q) related to the target object (Obj) based on the conversation history CH. The target object (Obj) is document (Doc) the first time, and response (A1) from the second time onward. When instruction Pt1 is created for the first time, the conversation history CH is empty, but from the second time onward, the conversation history CH includes question (Q) and response (A1).

[0220] <Step S4> In step S4, component 130 receives instruction Pt1. Component 130 also generates question Q using the large-scale language model LLM and sends it to component 120.

[0221] <Step S5> In step S5, component 120 receives the question Q and shares it internally within component 120.

[0222] <Step S6> In step S6, subcomponent 120A creates instruction statement Pt2 and sends it to component 130.

[0223] Instruction Pt2 includes instruction g2, question Q, and document Doc. Instruction g2 includes a procedure for generating response A1 to question Q from the description in document Doc. Response A1 may include an ITA (Information Item A) that cannot be answered.

[0224] <Step S7> In step S7, component 130 receives instruction Pt2. Component 130 also generates response A1 using the large-scale language model LLM and sends it to component 120.

[0225] <Step S8> In step S8, component 120 receives response A1 and shares it internally.

[0226] <Step S9> In step S9, if component 120 shares a response A1 that does not include an unanswerable ITA (as shown in Figure 11 for PTA), the process proceeds to step S3. Alternatively, if component 120 shares a response A1 that includes an unanswerable ITA (as shown in Figure 11 for ITA), the process proceeds to step S10 (see Figure 12).

[0227] <Step S10> In step S10, subcomponent 120A creates instruction statement Pt3 and sends it to component 130.

[0228] Instruction Pt3 includes instruction g3, question Q, and document Doc, and instruction g3 includes a procedure for generating a critique Cmnt regarding the description of document Doc relating to response A1 to question Q.

[0229] <Step S11> In step S11, component 130 receives instruction Pt3, generates a critique Cmnt using the large-scale language model LLM, and sends it to component 120.

[0230] <Step S12> In step S12, component 120 receives the criticism Cmnt and sends it to component 110.

[0231] <Step S13> In step S13, component 110 receives a critique Cmnt and provides it, for example, to a user 99 of the information processing system.

[0232] This allows for questions (Q) to be asked about the description in document Doc. It also allows for responses (A1) to questions (Q) related to the description in document Doc. Furthermore, the process of asking questions (Q) and providing responses (A1) to the description in document Doc can be repeated. The description in document Doc can also be examined in detail. Critical comments (Cmnt) can be provided regarding the description in document Doc related to responses (A1) to questions (Q). Additionally, questions (Q) can be generated when a response (A1) that does not contain an unanswerable ITA (Information Task) is shared. Critical comments (Cmnt) can also be generated when a response (A1) that contains an unanswerable ITA (Information Task) is shared. Critical comments (Cmnt) can also be provided regarding the description in document Doc related to responses (A1) to questions (Q). Furthermore, critical comments (Cmnt) can be used, for example, to encourage users of the information processing system to revise the document Doc. As a result, a novel information processing method with superior convenience, usefulness, or reliability can be provided.

[0233] <Example of information processing method 2> An information processing method according to one aspect of the present invention has steps S14 to S24 instead of step S9 (see Figures 13 and 14).

[0234] <Step S14> In step S14, if component 120 shares a response A1 that does not include an unanswerable ITA (as shown in Figure 13 for PTA), the process proceeds to step S3. If component 120 shares a response A1 that includes an unanswerable ITA (as shown in Figure 13 for ITA), the process proceeds to step S15 (see Figure 13).

[0235] <Step S15> In step S15, subcomponent 120A creates query Qu and shares it within component 120 (see Figure 14).

[0236] A query (Qu) includes a request to collect relevant prior art references (Ref) from the database (DB) and create a search result set (SRS). Note that the prior art references (Ref) fall within the scope of common knowledge in the field to which the document (Doc) belongs.

[0237] Furthermore, component 120 includes subcomponent 120B. Subcomponent 120B includes functions for processing using a database DB and a management system DBMS. The database DB has a function for storing prior art references (Ref), and the management system DBMS has a function for creating a search result set SRS from the database DB in response to a query Qu.

[0238] <Step S16> In step S16, subcomponent 120B creates a search result set SRS and shares it within component 120.

[0239] <Step S17> In step S17, subcomponent 120A creates instruction statement Pt4 and sends it to component 130.

[0240] Instruction Pt4 includes instruction g4, question Q, and search result set SRS. Instruction g4 includes a procedure for generating response A2 to question Q from prior art reference Ref. Response A2 includes bibliographic information RID that identifies the prior art reference Ref. Response A2 may also include an ITA (Information Item A) indicating that the response is unanswerable.

[0241] <Step S18> In step S18, component 130 receives instruction Pt4, generates response A2 using the large-scale language model LLM, and sends it to component 120.

[0242] <Step S19> In step S19, component 120 receives response A2 and shares it internally.

[0243] <Step S20> In step S20, if component 120 shares a response A2 that includes an unanswerable ITA (as shown in Figure 14), the process proceeds to step S10. Alternatively, if component 120 shares a response A2 that does not include an unanswerable ITA (as shown in Figure 14), the process proceeds to step S21.

[0244] <Step S21> In step S21, subcomponent 120A creates instruction statement Pt5 and sends it to component 130.

[0245] Instruction Pt5 includes instruction g5, response A2, and document Doc, and instruction g5 includes a procedure for generating supplementary Apx from response A2 to supplement the description in document Doc.

[0246] <Step S22> In step S22, component 130 receives instruction Pt5. Component 130 also generates supplement Apx using the large-scale language model LLM and sends it to component 120.

[0247] <Step S23> In step S23, component 120 receives the supplement Apx and sends it to component 110.

[0248] <Step S24> In step S24, component 110 receives supplement Apx and provides it, for example, to user 99 of the information processing system.

[0249] This allows for the collection of prior art references (Ref) from the scope of ordinary knowledge in the field to which the document Doc belongs. Furthermore, it is possible to generate a response A2 to a question (Q) from the prior art references. The basis for the response A2 to the question (Q) can be shown using the literature information RID. Additionally, a supplementary Apx can be used to provide the response A2 to the question (Q). When a response A1 containing an unanswerable ITA is shared, a query (Qu) can be created to generate a search result set (SRS) from the database (DB) containing the prior art references. When the search result set (SRS) is shared, a response A2 to the question (Q) can be generated from the prior art references. Furthermore, when response A2 does not contain an unanswerable ITA, a supplementary Apx can be generated from response A2 regarding the description of the document Doc related to the question (Q). The supplementary Apx can also be used to encourage users of the information processing system to revise the document Doc, for example. Furthermore, when response A2 contains an unanswerable ITA, a critical comment (Cmnt) can be provided regarding the description of the document Doc related to the question (Q). Furthermore, by using the Critical Commentary (Cmnt), for example, users of the information processing system can be encouraged to revise their documents (Doc). As a result, a novel information processing method with superior convenience, usefulness, or reliability can be provided.

[0250] This embodiment can be appropriately combined with other embodiments shown in this specification. [Explanation of Symbols]

[0251] Apx supplement CH Conversation History CID input device Cmnt comment DB database DBMS management system Doc document Dsp display device ITA unanswered LLM large language model Obj object QM query material Qu query Ref prior document RID document information SRS search result set 20 information processing apparatus 21 input unit 22 storage unit 23 processing unit 24 output unit 25 transmission path 51 network 99 user 110 component 120 component 120A subcomponent 120B subcomponent 130 component

Claims

1. It has the function of receiving documents and providing critiques of the descriptions in documents related to questions, The responses to the above questions include the inability to answer, The aforementioned questions and answers, including responses that do not contain unanswerable questions, or information processing systems related to the aforementioned documents.

2. A function to accept documents and provide criticism, The first instruction is sent to a large-scale language model, and it has a function to accept questions, A function to send a second instruction to the large-scale language model and receive a response, It has the function of transmitting a third instruction to the large-scale language model and receiving the criticism, The first instruction sentence includes the first instruction, the subject, and the conversation history, The first instruction includes a procedure for generating the questions related to the subject based on the conversation history, The subject is the document or the response, The aforementioned conversation history includes the aforementioned questions and answers, The second instruction statement includes the second instruction, the question and answer, and the document, The second instruction includes a procedure for generating the response to the question from the description in the document, The aforementioned response may include an inability to answer. The third instruction statement includes the third instruction, the question, the response, and the document, The third instruction includes a procedure for generating the criticism of the description of the document relating to the response to the question.

3. The first component and The second component and It has a third component, The first component has the function of receiving documents and transmitting them to the third component, and the function of receiving and providing criticisms. The second component has the function of receiving a first instruction and sending a question to the third component, the function of receiving a second instruction and sending a first response to the third component, and the function of receiving a third instruction and sending the criticism to the third component. It has the capability to perform processing using a large-scale language model, The large-scale language model comprises a function to generate the question according to the first instruction, a function to generate the first response according to the second instruction, and a function to generate the criticism according to the third instruction. The third component includes a function to receive and share the document, the question, the first response, and the criticism within the third component, a function to transmit the first instruction, the second instruction, and the third instruction to the second component, and a function to transmit the criticism to the first component. The third component comprises a first subcomponent, The first subcomponent has the function of creating the first instruction, the second instruction, and the third instruction. The first instruction sentence includes the first instruction, the subject, and the conversation history, The first instruction includes a procedure for generating the questions related to the subject based on the conversation history, The subject is the document or the first response, The conversation history includes the question and answer and the first response, The second instruction statement includes the second instruction, the question and answer, and the document, The second instruction includes a procedure for generating the first response to the question from the description in the document, The first response may include "unable to answer". The third instruction statement includes the third instruction, the question, the first response, and the document. The third instruction includes a procedure for generating the criticism of the description of the document relating to the first response to the question.

4. The first subcomponent has the function of creating the first instruction when the first response or the document is shared, which does not include an unanswerable response. When the aforementioned questions and answers are shared, the function to create the second instruction statement is provided, The information processing system according to claim 3, comprising a function to create a third instruction when the first response, including "unable to answer," is shared.

5. The first component described above has the function of receiving and providing records, The third component has the function of creating the record and transmitting it to the first component. The information processing system according to claim 3, wherein the record includes the question and answer and the first response.

6. The first component has the function of receiving question material and transmitting it to the third component. The aforementioned questionnaire materials include either or both an idea form and / or frequently asked questions. The aforementioned document is a document written from the aforementioned idea paper, The third component has a function to receive the question material and share it within the third component, The first instruction statement includes the first instruction, the subject, the conversation history, and the question material. The information processing system according to claim 3, wherein the first instruction includes a procedure for generating the questions concerning the subject based on the question materials and the conversation history.

7. The first component has the functionality to accept and provide supplements, The second component has the function of receiving the fourth instruction and the fifth instruction and transmitting the supplement to the third component. The large-scale language model includes a function to generate a second response to the question in accordance with the fourth instruction, and a function to generate the supplementary information relating to the second response in accordance with the fifth instruction, The third component includes a function to receive the supplement and transmit it to the first component, a function to transmit the fourth instruction and the fifth instruction to the second component, and a function to share the search result set within the third component. The third component comprises a second subcomponent, The first subcomponent has the function of creating a query, the fourth instruction, and the fifth instruction. The query includes a request to collect prior art related to the question from a database and create the search result set. The aforementioned prior art is within the scope of ordinary knowledge in the field to which the aforementioned document belongs. The fourth instruction statement includes the fourth instruction, the question and answer, and the set of search results. The fourth instruction includes a procedure for generating the second response to the question from the prior art, The second response includes bibliographic information that identifies the prior art, The fifth instruction statement includes the fifth instruction, the second response, and the document, The fifth instruction includes a procedure for generating the supplement that complements the description in the document from the second response, The second response may include "unable to answer". The second subcomponent comprises a function for performing processing using the database and management system, The aforementioned database has a function for storing the aforementioned prior art, The information processing system according to claim 3, wherein the management system has a function to create the search result set from the database in response to the query.

8. The first subcomponent has the function of creating the query when it shares the first response, which includes an unanswerable response, When the aforementioned search result set is shared, a function is provided to create the fourth instruction statement, A function to create the fifth instruction when the second response, which does not include "unable to answer," is shared, The information processing system according to claim 7, further comprising a function to create a third instruction when the second response, including "unable to answer," is shared.

9. An information processing method comprising the first to thirteenth steps, In the first step described above, the first component receives a document and transmits it to the second component. In the second step, the second component receives the document and shares it within the second component. The second component comprises a first subcomponent, In the third step, the first subcomponent creates a first instruction statement and sends it to the third component. The first instruction sentence includes the first instruction, the subject, and the conversation history, The first instruction includes a procedure for generating questions related to the subject based on the conversation history, The subject in the first instance is the aforementioned document, and from the second instance onward, it is the first response. The conversation history, for the second and subsequent times, includes the question and answer and the first response. In the fourth step, the third component receives the first instruction, generates the question using a large-scale language model, and transmits it to the second component. In the fifth step, the second component receives the question and shares it internally within the second component. In the sixth step, the first subcomponent creates a second instruction and sends it to the third component. The second instruction statement includes the second instruction, the question and answer, and the document, The second instruction includes a procedure for generating the first response to the question from the description in the document, The first response may include "unable to answer". In the seventh step, the third component receives the second instruction, uses the large-scale language model to generate the first response, and transmits it to the second component. In the eighth step, the second component receives the first response and shares it internally within the second component. In the ninth step, when the second component has shared the first response which does not include an unanswerable response, the process proceeds to the third step. When the second component shares the first response, including an unanswerable response, the process proceeds to the tenth step. In the 10th step, the first subcomponent creates a third instruction and sends it to the third component. The third instruction statement includes the third instruction, the question and answer, and the document, The third instruction includes a procedure for generating a critique of the description in the document relating to the first response to the question, In the 11th step, the third component receives the third instruction, generates the critique using the large-scale language model, and transmits it to the second component. In the twelfth step, the second component receives the criticism and transmits it to the first component. In the 13th step, the first component is an information processing method for receiving and providing the criticism.

10. An information processing method comprising steps 14 to 24, instead of step 9, In the 14th step, when the second component has shared the first response which does not contain an unanswerable response, the process proceeds to the third step. When the second component shares the first response, including an unanswerable response, the process proceeds to step 15. In the 15th step, the first subcomponent creates a query and shares it within the second component, The aforementioned query includes a request to collect prior art related to the aforementioned question from a database and create a set of search results. The aforementioned prior art is within the scope of ordinary knowledge in the field to which the aforementioned document belongs. The second component comprises a second subcomponent, The second subcomponent comprises a function for performing processing using the database and management system, The aforementioned database has a function for storing the aforementioned prior art, The management system includes a function to create the search result set from the database in response to the query, In the sixteenth step, the second subcomponent creates the search result set and shares it within the second component. In step 17, the first subcomponent creates a fourth instruction and sends it to the third component. The fourth instruction statement includes the fourth instruction, the question and answer, and the set of search results. The fourth instruction includes a procedure for generating a second response to the question from the prior art, The second response includes bibliographic information that identifies the prior art, The second response may include "unable to answer". In step 18, the third component receives the fourth instruction, uses the large-scale language model to generate the second response, and transmits it to the second component. In step 19, the second component receives the second response and shares it internally within the second component. In step 20, when the second component shares the second response, including an unanswerable response, proceed to step 10. When the second component has shared the second response which does not contain an unanswerable response, proceed to the 21st step: In step 21 above, the first subcomponent creates a fifth instruction and sends it to the third component. The fifth instruction statement includes the fifth instruction, the second response, and the document, The fifth instruction includes a procedure for generating supplements that complement the description in the document from the second response, In step 22, the third component receives the fifth instruction statement, Using the aforementioned large-scale language model, the supplement is generated and transmitted to the second component. In step 23 above, the second component receives the supplement and transmits it to the first component. The information processing method according to claim 9, wherein in the 24th step, the first component receives and provides the supplement.

Images