Program, information processing method, and information processing apparatus.

Abstract

To provide a program or similar tool that can output a business plan. [Solution] The program causes the computer to execute a process that involves acquiring analytical information necessary for creating a business plan, providing the acquired analytical information to a language model to output multiple analytical items and corresponding analytical content for each analytical item, acquiring multiple source pieces of information related to the output analytical content, and providing the acquired multiple source pieces of information to the language model to output a business plan.

Description

Technical Field

[0001] The present disclosure relates to a program, an information processing method, and an information processing apparatus.

Background Art

[0002] In Patent Document 1, a system for supporting the creation of management diagrams and the like necessary for budget requests and subsidy application in public sewerage business is disclosed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the invention according to Patent Document 1 has a problem that it cannot output a business plan.

[0005] In one aspect, an object is to provide a program or the like that can output a business plan.

Means for Solving the Problems

[0006] A program according to one aspect acquires analysis information necessary for creating a business plan, gives the acquired analysis information to a language model, outputs a plurality of analysis items and analysis contents corresponding to each analysis item, acquires a plurality of source information related to the output analysis contents, and causes a computer to execute a process of outputting a business plan by giving the acquired plurality of source information to a language model.

Effects of the Invention

[0007] In one aspect, it becomes possible to output a business plan.

Brief Description of the Drawings

[0008] [Figure 1] This is an explanatory diagram showing an overview of the information processing system. [Figure 2] This is a block diagram showing an example server configuration. [Figure 3] This is a block diagram showing an example of a computer configuration. [Figure 4] A flowchart illustrating an example of the processing procedure for an information processing system. [Figure 5] A flowchart illustrating an example of the processing procedure for an information processing system. [Figure 6] A flowchart illustrating an example of the processing procedure for an information processing system. [Figure 7] This is an explanatory diagram showing an example of an input screen. [Figure 8] This is an explanatory diagram showing an example of the first prompt. [Figure 9] This is an explanatory diagram showing an example of the second prompt. [Figure 10] This is an explanatory diagram showing an example of an input screen. [Figure 11] This is an explanatory diagram showing an example of an input screen. [Figure 12] This is an explanatory diagram showing an example of the third prompt. [Figure 13] This is an explanatory diagram showing an example of the fourth prompt. [Figure 14] This is an explanatory diagram showing an example of an input screen. [Figure 15] This is an explanatory diagram showing an example of an input screen. [Figure 16] This is an explanatory diagram showing an example of a reference screen. [Figure 17] This is an explanatory diagram showing an example of the fifth prompt. [Figure 18] This is an explanatory diagram showing an example of a screen. [Figure 19] This is an explanatory diagram showing an example of a record layout for a formatted database. [Figure 20] A flowchart illustrating an example of the processing procedure for an information processing system. [Figure 21] It is an explanatory diagram showing an example of a reference screen. [Figure 22] It is an explanatory diagram showing an example of a sixth prompt. [Figure 23] It is an explanatory diagram showing an example of a fifth prompt. [Figure 24] It is a flowchart showing an example of the processing procedure of an information processing system. [Figure 25] It is an explanatory diagram showing an example of a screen example. [Figure 26] It is an explanatory diagram showing an example of a selection screen. [Figure 27] It is an explanatory diagram showing an example of a seventh prompt. [Figure 28] It is an explanatory diagram showing an example of an eighth prompt. [Figure 29] It is a flowchart showing an example of the processing procedure of an information processing system. [Figure 30] It is a flowchart showing an example of the processing procedure of an information processing system. [Figure 31] It is an explanatory diagram showing an example of a ninth prompt.

MODE FOR CARRYING OUT THE INVENTION

[0009] (Embodiment 1) FIG. 1 is an explanatory diagram showing an overview of an information processing system. The information processing system acquires a plurality of analysis items and analysis contents corresponding to each analysis item from analysis information necessary for creating a business plan, and generates a business plan based on a plurality of source information related to the acquired analysis contents. Examples of the types of business plans include, for example, applications for subsidies and grants, startup of a new business, corporate acquisitions and mergers, business succession, formulation of business strategies, and BCP plans. In this embodiment, among the business plans, the "business plan for subsidy application" will be described. The information processing system includes an information processing device 10 and an information processing device 20. The information processing system transmits and receives information via the network N.

[0010] The information processing device 10 is an information processing device that processes, stores, and transmits various types of information. The information processing device 10 is, for example, a server device, a personal computer, or a general-purpose tablet PC (personal computer). The information processing device 10 may also be a cloud server device that provides the functions included in the information processing device 10 as a cloud service. In this embodiment, the information processing device 10 will be described as a server 10.

[0011] The information processing device 20 is, for example, a server device, a smartphone, a tablet, a personal computer (hereinafter referred to as "computer"), or a general-purpose tablet PC. In this embodiment, the information processing device 20 will be described as a computer 20.

[0012] Figure 2 is a block diagram showing an example of a server configuration. Server 10 includes a control unit 11, a storage unit 12, a communication unit 13, a large-capacity storage unit 14, and a read unit 15. Each of the above-mentioned units is interconnected via a bus. The control unit 11 is configured using one or more processors such as a CPU (Central Processing Unit), MPU (Micro-Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-purpose computing on graphics processing units), TPU (Tensor Processing Unit), or AI (Artificial Intelligence) chip. The control unit 11 performs various information processing and control processing related to Server 10 by appropriately executing a control program 12P (program product) stored in the storage unit 12.

[0013] The storage unit 12 includes RAM (Random Access Memory) or ROM (Read Only Memory), etc. The storage unit 12 stores various data necessary for the control program 12P executed by the control unit 11. The storage unit 12 temporarily stores data generated when the control unit 11 executes the control program 12P.

[0014] Furthermore, the memory unit 12 stores the language model 121. The language model 121 is a general-purpose large language model (LLM: Large Language Model) constructed by, for example, performing unsupervised pre-training using a large set of texts and images. Alternatively, a model that has been fine-tuned specifically for tasks such as business plan creation support or business document creation support may be used. Examples of language models 121 include GPT (Generative Pre-trained Transformer)-3, GPT-3.5, GPT-4, RWKV (Receptance Weighted Key Value), PaLM2, or LLaMa (Large Language Model Meta AI).

[0015] When analytical information necessary for creating a business plan (hereinafter referred to as "analytical information") is input to the language model 121, it performs calculations to generate multiple analytical items and corresponding analytical content (hereinafter referred to as "analytical content") for each analytical item, and outputs the generated multiple analytical items and corresponding analytical content for each analytical item.

[0016] The analysis information includes an overview of the new business, keywords related to the new business, and the results of interviews regarding the new business. The analysis items include major and minor analysis items. Major analysis items are the broad framework analysis items essential for creating a business plan, such as market trend analysis and customer analysis. Minor analysis items are specific analysis items that subdivide each major analysis item. The analysis content consists of answers to each minor analysis item.

[0017] Furthermore, when source information related to the analysis content (hereinafter referred to as "source information") is input to the language model 121, it performs calculations to generate a business plan and outputs the generated business plan. The source information is the information that the language model 121 referenced in order to output the analysis content, and includes information held by the user and information on the WEB (World Wide Web), etc.

[0018] The language model 121 may be constructed by combining multiple algorithms. In this embodiment, as an example, an example in which the language model 121 is stored in the storage unit 12 is described, but it is not limited to this. The control unit 11 may access an external server (not shown) that stores the language model 121 and use the language model 121. The storage unit 12 also stores a source information file 122. The source information file 122 stores source information saved by the user, and source information obtained from the web by the information processing system through processing described later.

[0019] The communication unit 13 is a communication module that sends and receives information to and from the computer 20 via the network N. The large-capacity storage unit 14 includes RAM or ROM, etc. The large-capacity storage unit 14 stores the format DB 141. The format DB 141 stores the format of the business plan specified by the user. The format of the business plan specified by the user includes, for example, the format of the business plan according to the type of subsidy application. The format DB 141 will be described in detail in Embodiment 2.

[0020] In this embodiment, the storage unit 12 and the large-capacity storage unit 14 may be configured as a single storage device. The large-capacity storage unit 14 may be composed of multiple storage devices. The large-capacity storage unit 14 may also be an external storage device connected to the server 10.

[0021] The reading unit 15 reads information stored in the portable storage medium 1a. The portable storage medium 1a is, for example, a CD (Compact Disc), DVD (Digital Versatile Disc), USB (Universal Serial Bus) memory, or SD (Secure Digital). The reading unit 15 reads the control program 12P from the portable storage medium 1a. The control unit 11 stores the read control program 12P in the storage unit 12.

[0022] The control unit 11 may download the control program 12P from another computer via the network N. In that case, the control unit 11 stores the downloaded control program 12P in the storage unit 12. Alternatively, the control unit 11 may store the read control program 12P in the large-capacity storage unit 14.

[0023] In this embodiment, server 10 may be composed of multiple servers. Server 10 may be a virtual machine virtually constructed by software within a single device. Server 10 may be a local server installed in the facility where server 10 is located. Server 10 may be a cloud server connected via network N. Furthermore, the control program 12P may run on a single server, or it may be distributed and run on multiple servers interconnected via network N.

[0024] Figure 3 is a block diagram showing an example of a computer configuration. The computer 20 includes a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, a display unit 25, and a large-capacity storage unit 26. The above-mentioned units are interconnected via a bus. The control unit 21 is configured using one or more processors such as a CPU, MPU, or GPU. The storage unit 22 includes RAM or ROM.

[0025] The memory unit 22 stores various data necessary for the control program 22P (program product) executed by the control unit 21. The memory unit 22 temporarily stores data generated when the control program 22P is executed. The control unit 21 performs information processing and control processing related to the computer 20 by appropriately executing the control program 22P stored in the memory unit 22.

[0026] Furthermore, the storage unit 22 stores the source information file 122. The computer 20 appropriately synchronizes the source information file 122 stored in the storage unit 22 with the source information file 122 stored in the server 10. File synchronization via the network N has been performed conventionally, so a description is omitted.

[0027] The communication unit 23 is a communication module that sends and receives information to and from the server 10 via the network N. The input unit 24 is, for example, a keyboard or mouse. The display unit 25 is, for example, a liquid crystal panel or an organic EL (electro-luminescence) panel. The large-capacity storage unit 26 includes RAM or ROM. The large-capacity storage unit 26 stores the format DB 141. The computer 20 appropriately synchronizes the format DB 141 stored in the large-capacity storage unit 26 with the format DB 141 stored in the server 10.

[0028] In this embodiment, the storage unit 22 and the large-capacity storage unit 26 may be configured as a single storage device. The large-capacity storage unit 26 may be composed of multiple storage devices. The large-capacity storage unit 26 may also be an external storage device connected to the computer 20.

[0029] The following describes the process by which the information processing system of this embodiment outputs a business plan. Figures 4 to 6 are flowcharts showing an example of the processing procedure of the information processing system. When a user executes the program of this embodiment, the control unit 21 displays a screen for receiving analysis information input by the user (hereinafter referred to as the input screen) on the display unit 25 (step S101).

[0030] Figure 7 is an explanatory diagram showing an example of an input screen. The input screen shown in Figure 7 includes a menu field T0, a first field R1, a second field R2, and a third field R3. The first field R1, the second field R2, and the third field R3 are displayed in parallel from left to right. Menu field T0 is displayed above the first field R1, the second field R2, and the third field R3. Menu field T0 displays multiple selectable menus, including a business plan field T1, a business analysis field T2, and a data field T3.

[0031] Figure 7 shows the screen corresponding to the business analysis section T2 among several menus. The first section R1 accepts user input of analysis information. The first section R1 includes the input field R11, text button b1, and generate button b2. The input field R11 accepts analysis information input via the input unit 24. In the input field R11 of Figure 7, an example of analysis information entered by the user is displayed: "A micro-hydropower generation service that utilizes unused urban water resources, taking SDGs into consideration."

[0032] The text button b1 is a button for accepting input of analysis information in file format. When the text button b1 is selected, the control unit 21 obtains analysis information by performing appropriate processing according to the type of file entered. Specifically, for image files or PDF files containing images, OCR (Optical Character Recognition) processing is performed, and for plain text files, PDF files without images, HTML files, spreadsheet files, or Markdown files, etc., analysis information is obtained by performing text parsing processing according to each file format. The generate button b2 is a button for sending the input information obtained via the input field R11 to the server 10. The second field R2 displays multiple major analysis items output by the language model 121, minor analysis items within the major analysis items, and the analysis content corresponding to each analysis item. The third field R3 displays the analysis items and analysis content selected by the user, as well as an object for selecting source information. In Figure 7, since the input of analysis information is being accepted in the first field R1, the second field R2 and the third field R3 are blank.

[0033] The control unit 11 acquires analysis information via the first column R1 (step S102). The control unit 11 displays the acquired analysis information in the first column R1 (step S103). The control unit 21 determines whether the generate button b2 has been selected (step S104). If the control unit 21 determines that the generate button b2 has not been selected (step S104: NO), it returns to the process in step S102 and continues to accept input of analysis information for the first column R1. If the communication unit 23 determines that the generate button b2 has been selected (step S104: YES), it sends the analysis information to the server 10 (step S105).

[0034] The communication unit 13 receives analysis information transmitted from the computer 20 (step S201). The control unit 11 generates a first prompt 100 (hereinafter referred to as the first prompt 100) which includes the received analysis information and a command to generate major analysis items (step S202). The command to generate major analysis items is pre-stored in the storage unit 12 or the large-capacity storage unit 14. The command to generate major analysis items can be modified as appropriate according to the embodiment.

[0035] Figure 8 is an explanatory diagram showing an example of the first prompt. The first prompt 100 includes a role information field 110, an analysis information field 120, a constraint field 130, an example field 140, and a first generation command field 150. The role information field 110 describes the position and role of the language model 121 necessary when generating major analysis items. The contents described in the role information field 110 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the role information field 110 can be appropriately changed according to the embodiment.

[0036] In Figure 8, the role information column 110 contains: "You are a business planning expert and a professional assistant who provides deep insights into new business launches, business continuity, and business restructuring." Your role is to generate several major analytical items necessary for creating a business plan, based on the given analytical information. It is stated.

[0037] The analysis information column 120 contains the analysis information transmitted from computer 20. In the analysis information column 120 of Figure 8, "A micro-hydropower generation service that utilizes unused urban water resources, taking the SDGs into consideration." It is stated.

[0038] The constraints section 130 contains notes to improve the consistency of the major analysis items generated by the language model 121. The contents of the constraints section 130 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the constraints section 130 can be modified as appropriate to suit the embodiment.

[0039] In the constraints column 130 of Figure 8, 1. Each major analysis item should be limited to perspectives from which information can be gathered through web searches. 2. For each major analysis item, avoid overly general expressions, including specific business names and proper nouns. 3. ..." It is stated.

[0040] Example section 140 contains examples to assist in the generation of major analysis items by the language model 121. The examples include both good and bad examples. The contents described in example section 140 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in example section 140 can be modified as appropriate to suit the embodiment.

[0041] In the example section 140 of Figure 8, "Good example (concrete and context-independent):" 1. Key players and competitive landscape in the Japanese electric vehicle market ... Bad examples (too general or context-dependent): 1...." It is stated.

[0042] The first generation instruction field 150 contains instructions for the language model 121 to output major analysis items. The contents of the first generation instruction field 150 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the first generation instruction field 150 can be changed as appropriate to suit the embodiment.

[0043] In Figure 8, the first generation instruction column 150 contains: "Please generate several key major analysis items related to the analysis information, following the instructions above. When generating these major analysis items, please consider the following points." 1. The main analysis items must be specific perspectives that allow for information gathering through web searches in subsequent investigations. 2. The major analysis items must express important analytical perspectives related to the success factors of the business and have a self-contained meaning. It is stated.

[0044] The control unit 11 inputs the generated first prompt 100 to the language model 121, thereby extracting multiple major analysis items generated by the language model 121 (step S203). The control unit 11 generates a second prompt 200 (hereinafter referred to as the second prompt 200) which includes the analysis information, multiple major analysis items, and the generation command for the minor analysis items within the major analysis items (step S204).

[0045] Figure 9 is an explanatory diagram showing an example of a second prompt. The second prompt 200 includes a second role information field 210, an analysis information field 120, a major analysis item field 220, a second constraint field 230, a second example field 240, and a second generation instruction field 250. The second role information field 210 describes the position and role of the language model 121 necessary when generating sub-analysis items within the major analysis item. The contents described in the second role information field 210 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the second role information field 210 can be appropriately changed according to the embodiment.

[0046] In Figure 9, the second role information section 210 states: "You are a business planning expert and a professional assistant who provides deep insights into new business launches, business continuity, and business restructuring." Your role is to generate the necessary sub-analysis items for creating a business plan, based on the given analytical information and major analysis items. It is stated.

[0047] The analysis information field 120 contains the analysis information transmitted from the computer 20. The contents of the analysis information field 120 in Figure 9 are the same as those in Figure 8. The major analysis item column 220 contains the major analysis items generated by the language model 121. In Figure 9, the major analysis item column 220 contains: The report includes "market trend analysis, customer analysis, competitor analysis, product / service analysis, risk analysis, and marketing strategy analysis."

[0048] In this embodiment, an example is shown in which major analysis items extracted by the language model 121 are listed in the major analysis item column 220, but this is not limited to this. Instead of major analysis items extracted by the language model 121, major analysis items pre-stored in the storage unit 12 or the large-capacity storage unit 14 may be listed in the major analysis item column 220.

[0049] The second constraints section 230 contains notes to improve the consistency of the sub-analysis items generated by the language model 121. The contents of the second constraints section 230 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the second constraints section 230 can be modified as appropriate to suit the embodiment. In Figure 9, under the second constraint condition column 230, "1. Sub-analysis items should be limited to those related to the major analysis items." 2. Each sub-analysis item should be limited to aspects for which information can be gathered through web searches. 3. ..." It is stated.

[0050] The second example section 240 contains examples to assist in the generation of sub-analysis items by the language model 121. The examples include both good and bad examples. The contents described in the second example section 240 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the second example section 240 can be modified as appropriate to suit the embodiment.

[0051] In Figure 9, in the second example column 240, "Good example (concrete and context-independent):" 1. What were the top 3 best-selling electric vehicle models in Japan in 2023, and what were the sales figures for each? 2.... Bad examples (too general or context-dependent): 1. What is the market size of this industry? (This question lacks specificity, and "this industry" is unclear.) 2. ..." It is stated.

[0052] The second generation instruction field 250 contains instructions for the language model 121 to extract sub-analysis items. The contents of the second generation instruction field 250 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the second generation instruction field 250 can be modified as appropriate to suit the embodiment.

[0053] In Figure 9, the second generation instruction field 250 contains: "Please generate the analysis information and important sub-analysis items related to the major analysis items according to the above instructions. When generating the sub-analysis items, please consider the following points." 1. Please limit the sub-analysis items to those that can be answered through a web search. 2. Sub-analysis items should be fully understandable on their own, independent of the context of the analysis information and the major analysis items. It is stated.

[0054] The control unit 11 inputs the generated second prompt 200 to the language model 121, thereby extracting the sub-analysis items output by the language model 121 (step S205). The control unit 11 stores the extracted major and minor analysis items in the large-capacity storage unit 14 (step S206). The communication unit 13 transmits the extracted major and minor analysis items to the computer 20 (step S207).

[0055] The communication unit 23 receives the major analysis items and minor analysis items transmitted from the server 10 (step S106). The control unit 21 displays the received major analysis items and minor analysis items on the input screen (step S107).

[0056] Figure 10 is an explanatory diagram showing an example of an input screen. The second column R2 displays multiple item fields R21 and the first add button b4. The number of item fields R21 is displayed according to the number of major analysis items. Item field R21 displays the major analysis item MA, the sub-analysis items ma within the major analysis item, the answer button b3, and the second add button b5. Multiple sub-analysis items ma are displayed in association with one major analysis item MA. The first add button b4 is displayed at the bottom of the second column R2.

[0057] In Figure 10, the sub-analysis items ma, "How are the criteria for environmental impact assessments of power generation systems using building wastewater set?" and "What are the legal regulations and permit acquisition processes related to urban micro-hydropower generation?" are displayed in correspondence with the major analysis item MA, "Risk Analysis."

[0058] The answer button b3 is displayed below each sub-analysis item ma. The answer button b3 is a button to instruct the generation of analysis content. The first add button b4 is a button to accept additional input from the user for major analysis items. The second add button b5 is a button to accept additional input from the user for sub-analysis items.

[0059] The control unit 21 determines whether or not it has received the selection of the first additional button b4 (step S108). If the control unit 21 determines that it has received the selection of the first additional button b4 (step S108: YES), it receives additional input of major analysis items from the user via the input unit 24 (step S109). The communication unit 23 transmits the additional major analysis items received to the server 10 (step S110).

[0060] The communication unit 13 receives the major analysis items transmitted from the computer 20 (step S208). The control unit 11 generates a second prompt (see Figure 9) by combining the analysis information received in step S201, the major analysis items received in step S208, and the generation commands for the minor analysis items within the major analysis items (step S209). The major analysis item field 220 of the second prompt 200 contains the major analysis items additionally entered by the user. The control unit 11 inputs the generated second prompt 200 into the language model 121 to extract the minor analysis items output by the language model 121 (step S210). The control unit 11 stores the major analysis items and minor analysis items in the large-capacity storage unit 14 (step S211). The communication unit 13 transmits the extracted minor analysis items to the computer 20 (step S212).

[0061] The communication unit 23 receives the sub-analysis items sent from the server 10 (step S111). The control unit 21 displays the major analysis items additionally entered by the user and the sub-analysis items sent from the server 10 on the input screen (step S112). In this case, a new item field R21 is added to the second field R2 of the input screen. The new item field R21 displays the major analysis item MA additionally entered by the user, the sub-analysis item ma within the major analysis item output by the language model 121, the answer button b3, and the second add button b5.

[0062] After step S112 is completed, or if the control unit 21 determines that it has not received a selection for the first add button b4 (step S108: NO), it determines whether or not it has received a selection for the second add button b5 (step S113). If the control unit 21 determines that it has received a selection for the second add button b5 (step S113: YES), it accepts additional input of sub-analysis items from the user via the input unit 24 (step S114). The control unit 21 displays the received sub-analysis items in the item field R21 (step S115).

[0063] Figure 11 is an explanatory diagram showing an example of an input screen. Above the second add button b5 in item field R21, the user-added sub-analysis item ma, "What are some examples of troubles and disputes in electricity sharing in local communities?", and the corresponding answer button b3 are displayed.

[0064] After step S115 is completed, or if the control unit 21 determines that it has not received a selection for the second additional button b5 (step S113: NO), it determines whether or not it has received a selection for the answer button b3 (step S116). If the control unit 21 determines that it has not received a selection for the answer button b3 (step S116: NO), it returns to step S108 and waits until it receives a selection for the first additional button b4, the second additional button b5, or the answer button b3. If the control unit 21 determines that it has received a selection for the answer button b3 (step S116: YES), it sends a command to generate the analysis content (hereinafter referred to as the generation command) and the major and minor analysis items corresponding to the answer button b3 to the server 10 via the communication unit 23 (step S117).

[0065] The communication unit 13 receives a generation command transmitted from the computer 20, along with the major and minor analysis items corresponding to the answer button b3 (step S213). The control unit 11 generates a third prompt 300 (hereinafter referred to as the third prompt 300) which includes a generation command for the minor analysis items and a query for web search (step S214).

[0066] Figure 12 is an explanatory diagram showing an example of a third prompt. The third prompt 300 includes a third role information field 310, a sub-analysis item field 320, and a third generation command field 330. The third role information field 310 describes the position and role of the language model 121 necessary for generating a web search query (hereinafter referred to as a web search query) that derives the analysis content. The contents described in the third role information field 310 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the third role information field 310 can be appropriately changed according to the embodiment.

[0067] In Figure 12, the third role information section 310 states: "You are a professional in information gathering and research using web searches. Considering the context of the conversation, devise web search queries to derive analysis results for the given sub-analysis items, and generate a list of up to three web search queries. However, as an exception, if the response is simple, such as "Hello," and clearly does not require a web search, return an empty list. If it is not clear, assume that a web search will be performed." It is stated.

[0068] The sub-analysis item column 320 contains the sub-analysis items. In Figure 12, item 320 in the sub-analysis section contains: "How are the environmental impact assessment criteria set for power generation systems that utilize building wastewater?" It is stated.

[0069] The third generation instruction field 330 contains instructions for the language model 121 to output a web search query. The contents of the third generation instruction field 330 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the third generation instruction field 330 can be modified as appropriate to suit the embodiment.

[0070] In the third generation instruction column 330 of Figure 12, Please generate a web search query according to the above instructions. When generating the web search query, please consider the following points. 1. Carefully analyze the sub-analysis items to identify key topics, keywords, and concepts. 2. Create web search queries that are concise and clear, so that they function effectively in search engines. 3. Include specific keywords such as proper nouns, dates, and technical terms whenever possible. 4. Generate web search queries using the language used in the sub-analysis items. 5. Include the year as needed to obtain the most up-to-date information. 6. Industry-specific jargon and abbreviations should be used in combination with general terminology. 7. Understand the intent of the sub-analysis items and create web search queries that provide not only direct answers but also relevant background information. 8. Generate diverse web search queries to obtain information from multiple perspectives and different sources as needed. 9. Each query focuses on one main concept or sub-analysis item. 10. Web search queries should be specific and focused, avoiding ambiguity. 11...." It is stated.

[0071] The control unit 11 inputs the generated third prompt 300 to the language model 121 to obtain the web search query generated by the language model 121 (step S215). The control unit 11 obtains web search information related to the obtained search query from the web (step S216). The control unit 11 generates a fourth prompt 400 (hereinafter referred to as the fourth prompt 400) which includes major analysis items, minor analysis items, web search information, and a command to generate the analysis content (step S217).

[0072] Figure 13 is an explanatory diagram showing an example of the fourth prompt. The fourth prompt 400 includes a fourth role information field 410, a sub-analysis item field 320, a web search information field 420, a citation method field 430, and a fourth generation command field 440. The fourth role information field 410 describes the position and role of the language model 121 necessary for generating the analysis content for the sub-analysis items. The contents described in the fourth role information field 410 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the fourth role information field 410 can be appropriately changed according to the embodiment.

[0073] In Figure 13, the fourth role information section 410 states: "You are an AI assistant. In addition to general conversational abilities, you are an expert in business research, providing deep insights into diverse business themes such as new business launches, business succession, and business restructuring. You are currently conducting research on the following sub-analysis items. Please use the following web search results to generate analysis content for these sub-analysis items." It is stated.

[0074] The sub-analysis item column 320 contains the sub-analysis items. The contents of the sub-analysis item column 320 in Figure 13 are the same as those in Figure 12. The web search information field 420 contains web search information related to the search query. The web search information is presented in either text or link format. In Figure 13, the web search information field 420 contains web search information in link format.

[0075] The citation method section 430 describes how to cite web search information when generating analysis content based on web search information. The contents described in the citation method section 430 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents described in the citation method section 430 can be modified as appropriate to suit the embodiment.

[0076] In the citation method section 430 of Figure 13, 1. When using web search results, include a link immediately after the relevant sentence or paragraph. 2. Cite only the most relevant web search results, and if there are multiple relevant results, list them side by side as shown in (1)(2). 3. Do not include a list of citations at the end of the entire answer. 4. ..." It is stated.

[0077] The fourth generation instruction field 440 contains instructions for the language model 121 to output the analysis content. The contents of the fourth generation instruction field 440 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the fourth generation instruction field 440 can be modified as appropriate to suit the embodiment.

[0078] In Figure 13, the fourth generation instruction column 440 contains: Please generate the analysis results according to the above instructions. When generating the analysis results, please consider the following points. 1. Use a fair and journalistic tone. 2. Integrate search results and create a consistent answer. 3. Citations should be made appropriately to clearly indicate the source of information. 4. If there are any points you are unsure about, please state that fact without making assumptions. 5. Aim for informative, logical, and positive answers. It is stated.

[0079] The control unit 11 inputs the generated fourth prompt 400 to the language model 121, thereby acquiring the analysis content generated by the language model 121 (step S218). The control unit 11 acquires multiple source information referenced by the language model 121 to generate the analysis content (step S219). The control unit 11 stores the acquired analysis content in the large-capacity storage unit 14, associating it with major analysis items and minor analysis items (step S220). The communication unit 13 transmits the acquired analysis content and multiple source information to the computer 20 (step S221).

[0080] The communication unit 23 receives the analysis content and multiple source information transmitted from the server 10 (step S118). The control unit 21 displays the received analysis content on the input screen (step S119).

[0081] Figure 14 is an explanatory diagram showing an example of an input screen. The item field R21 displays the analysis content C corresponding to the sub-analysis item ma, and a confirmation button b6 is displayed below the analysis content C. The confirmation button b6 is a button that displays the source information corresponding to the analysis item and analysis content in the third field R3.

[0082] In Figure 14, the analysis content C, which corresponds to the sub-analysis item ma, "How are the standards for environmental impact assessment of power generation systems that utilize building wastewater set?", is displayed, and below analysis content C, the confirmation button b6 is displayed.

[0083] The control unit 21 determines whether or not another answer button b3 displayed in item field R21 has been selected (step S120). If the control unit 21 determines that another answer button b3 has been selected (step S120: YES), it returns to step S117. If the control unit 21 determines that no other answer button b3 has been selected (step S120: NO), it determines whether or not the selection of the confirmation button b6 has been accepted (step S121).

[0084] If the control unit 21 determines that the confirmation button b6 is not selected (step S121: NO), it returns to step S120. If the control unit 21 determines that the confirmation button b6 is selected (step S121: YES), it displays the corresponding sub-analysis items and analysis content, and the object for selecting source information in the third column R3 (step S122).

[0085] Figure 15 is an explanatory diagram showing an example of an input screen. The third column, R3, displays the selected sub-analysis item ma and analysis content C, as well as object O for selecting source information. Both the analysis content C and object O in the third column, R3, are displayed as selectable. In Figure 15, the third column, R3, displays three objects O. Object O displays selectable source information, including web pages or document files related to the text in the analysis content.

[0086] When the control unit 21 receives a selection of source information displayed in object O, it stores the source information in the source information file 122 of the storage unit 12 (step S123). When the control unit 21 receives a selection of source information displayed in object O in the third column R3, it may overlay a screen (not shown) for detailed display of the entire content of the source information onto the third column R3. Also, when the control unit 21 receives a selection of analysis content C in the third column R3, it may overlay a screen (not shown) for displaying a portion of the content of the source information referenced to output the analysis content C (for example, the portion of the entire source information that was actually referenced, between 300 and 500 characters).

[0087] The control unit 21 determines whether or not it has received the selection of another confirmation button b6 (step S124). If the control unit 21 determines that it has received the selection of another confirmation button b6 (step S124: YES), it returns to step S122. If the control unit 21 determines that it has not received the selection of another confirmation button b6 (step S124: NO), it determines whether or not it has received the selection of data field T3 (step S125).

[0088] If the control unit 21 determines that it has not accepted the selection of data field T3 (step S125: NO), it returns to step S124. If the control unit 21 determines that it has accepted the selection of data field T3 (step S125: YES), it displays a screen for referencing the source information stored in the source information file 122 (hereinafter referred to as the reference screen) on the display unit 25 (step S126).

[0089] Figure 16 is an explanatory diagram showing an example of a reference screen. The reference screen has a data field T3 selected and includes a source information section T31, a chat field T32 for assisting in the creation of a business plan, and a create button b7. The source information section T31 displays multiple source information stored in the source information file 122. The source information section T31 also displays information such as the name, storage date, file size, and saver of each source information. The create button b7 is a button for instructing the creation of a business plan. Queries are entered into the chat field T32 to assist in the creation of a business plan. The user enters queries to assist in the creation of a business plan into the chat field T32 via the input section 24.

[0090] In the chat field T32 of Figure 16, the following queries are provided to assist in creating a business plan: "1. Section name: Project plan for this grant program 2. Instructions: (1) Describe the market trends and the background for developing the business plan, and briefly describe the outline of the plan for this grant project. (2) Describe the problems and challenges that arose before the equipment was installed. (3) Write a piece of about 400 characters. (4) Write the numbers clearly. This is displayed. In this embodiment, an example is shown where a query entered by the user is displayed in the chat field T32, but this is not limited to this. The chat field T32 may be left blank.

[0091] The control unit 21 retrieves the query entered by the user in the chat field T32 (step S127). The control unit 21 determines whether or not the selection of the create button b7 has been accepted (step S128). If the control unit 21 determines that the selection of the create button b7 has not been accepted (step S128: NO), it returns to step S126 and continues to display the reference screen.

[0092] If the control unit 21 determines that it has received the selection of the create button b7 (step S128: YES), it sends multiple source information, instructions for creating a business plan, and the query entered in the chat field T32 to the server 10 via the communication unit 23 (step S129). If the chat field T32 is blank, the control unit 21 may send multiple source information and instructions for creating a business plan to the server 10. Alternatively, based on the user's instructions, the control unit 21 may obtain multiple source information stored in the source information file 122 or multiple source information from the web via the chat field T32. In that case, the control unit 21 sends the obtained multiple source information, instructions for creating a business plan, and the query entered in the chat field T32 to the server 10. The following explanation will describe an example in which the control unit 21 sends multiple source information, instructions for creating a business plan, and the query entered in the chat field T32 to the server 10.

[0093] The communication unit 13 receives multiple source information, instructions for creating a business plan, and queries entered in the chat field T32 transmitted from the computer 20 (step S222). The control unit 11 generates a fifth prompt 500 (hereinafter referred to as the fifth prompt 500) which includes the multiple source information, queries entered in the chat field T32, and instructions for generating a business plan (step S223).

[0094] Figure 17 is an explanatory diagram showing an example of a fifth prompt. The fifth prompt 500 includes a fifth role information field 510, a source information field 520, a query field 530, a citation format field 540, and a fifth generation instruction field 550. The fifth role information field 510 describes the position and role of the language model 121 necessary for generating the business plan. The contents described in the fifth role information field 510 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents described in the fifth role information field 510 can be appropriately changed according to the embodiment.

[0095] In Figure 17, the fifth role information section 510 states: "You are a business planning expert and a professional assistant who provides deep insights into new business launches, business continuity, and business restructuring. Your role is to create business plans based on information from multiple sources." It is stated.

[0096] The source information section 520 contains multiple source information entries stored in the source information file 122. The source information is presented in either link or text format. The query entered in the chat field T32 will be displayed in the query field 530.

[0097] In the query column 530 of Figure 17, "1. Section name: Project plan for this grant program 2. Instructions: (1) Describe the market trends and the background for developing the business plan, and briefly describe the outline of the plan for this grant project. (2) Describe the problems and challenges that arose before the equipment was installed. (3)...." It is stated.

[0098] The citation format section 540 contains the methods for citing multiple source information used when generating the business plan, correct examples of citation of said source information, and incorrect examples of citation of said source information. The contents of the citation format section 540 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents of the citation format section 540 can be modified as appropriate to suit the embodiment.

[0099] In the citation format column 540 of Figure 17, 1. Format of citation: • Please include the sources of the information used in the generated text. ...." 2. Correct citation example: The market size is projected to grow rapidly at a compound annual growth rate (CAGR) of 15% from 2020 to 2025. (1)... 3. Examples of incorrect citations: Cases where quotations are grouped together at the end of a text, even though they could be cited at the sentence or paragraph level: The market size is growing rapidly. Furthermore, the market share of competitors is also expanding. (1)(2)(3)... It is stated.

[0100] The fifth generation instruction field 550 contains instructions for the language model 121 to output a business plan. The contents of the fifth generation instruction field 550 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the fifth generation instruction field 550 can be changed as appropriate to suit the embodiment.

[0101] In Figure 17, the fifth generation instruction column 550 contains: Please generate a business plan according to the above instructions. When generating the business plan, please consider the following points. 1. Maintain content and scope appropriate to the section name. 2. Use the source information as the primary source of information, and describe related information and examples. 3. ..." It is stated.

[0102] The control unit 11 inputs the generated fifth prompt 500 to the language model 121 to acquire the business plan generated by the language model 121 (step S224). The control unit 11 stores the acquired business plan in the large-capacity storage unit 14 (step S225). The communication unit 13 transmits the acquired business plan to the computer 20 (step S226).

[0103] The communication unit 23 receives the business plan transmitted from the server 10 (step S130). The control unit 21 displays the received business plan on the display unit 25 (step S131). The business plan transmitted from the server 10 is displayed on the screen corresponding to the business plan field T1. Figure 18 is an explanatory diagram showing an example of a screen. In the screen shown in Figure 18, the business plan sent from server 10 is displayed in the business plan field T1.

[0104] According to Embodiment 1, the information processing system can acquire analytical information necessary for creating a business plan, and by providing the acquired analytical information to a language model, it can output multiple analytical items and the corresponding analytical content for each analytical item.

[0105] According to Embodiment 1, the information processing system can output a business plan by acquiring multiple source pieces of information related to the analysis content output by the language model and providing the acquired multiple source pieces of information to the language model.

[0106] According to Embodiment 1, the information processing system can extract multiple major analysis items and minor analysis items within the major analysis items using a language model based on the acquired analysis information, and output analysis content corresponding to each minor analysis item using the language model.

[0107] According to Embodiment 1, the information processing system can additionally accept input from the user for major or minor analysis items.

[0108] According to Embodiment 1, the information processing system can receive the selection of source information, including web pages or document files related to the text in the analysis content, and store the received source information in the storage unit.

[0109] According to Embodiment 1, the information processing system can output an input field for analysis information in the first column, output analysis items and analysis content in the second column, and output an object for selecting the selected analysis items and analysis content and source information in the third column.

[0110] According to Embodiment 1, the information processing system can output a chat field to assist in the creation of a business plan, and can output a business plan by providing the queries and source information entered in the chat field to a language model.

[0111] (Embodiment 2) Embodiment 2 describes a method in which an information processing system acquires the format of a business plan to be output and outputs a business plan based on the acquired format.

[0112] Methods for obtaining the business plan format include reading from the format DB141, uploading the business plan format by the user, and manual input by the user. Embodiment 2 describes an example of how to obtain the business plan format by reading the business plan format from the format DB141.

[0113] Figure 19 is an explanatory diagram showing an example of the record layout of the format database. Format DB 141 stores the type of business plan and the format of the business plan corresponding to that type. Format DB 141 includes a business plan column and a format column. The business plan column stores the type of business plan. The format column stores the format of the business plan corresponding to that type of business plan. The top record in Figure 19 stores the type of business plan "Business plan for subsidy application" and the format of the business plan corresponding to that type of business plan "Format A".

[0114] Figure 20 is a flowchart showing an example of the processing procedure of an information processing system. In the flowchart shown in Figure 20, steps S126 to S129 and steps S222 to S224 are replaced with steps S301 to S305 and steps S401 to S407 in the processing shown in Figures 4 to 6. Steps similar to those in Figures 4 to 6 are omitted from explanation. When the control unit 21 determines that it has received a selection in data field T3 (step S125: YES), it displays the reference screen on the display unit 25 (step S301).

[0115] Figure 21 is an explanatory diagram showing an example of a reference screen. The reference screen further includes a Type field T33. The Type field T33 is a pull-down menu button. The Type field T33 accepts the user's selection of the type of business plan. The type of business plan displayed in the Type field T33 corresponds to the type of business plan stored in the format DB141. In the Type field T33 of Figure 21, "Business plan for subsidy application" is selected. The reference screen may also include an upload field (not shown) for the user to upload the business plan format. Furthermore, the reference screen may also include a manual input field (not shown) for the user to manually enter the business plan format.

[0116] The control unit 21 retrieves the query entered by the user in the chat field T32 (step S302). The control unit 21 accepts the user's selection of the type of business plan (step S303). The control unit 21 determines whether or not the selection of the create button b7 has been accepted (step S304). If the control unit 21 determines that the selection of the create button b7 has not been accepted (step S304: NO), it returns to step S301 and continues to display the reference screen. If the control unit 21 determines that the selection of the create button b7 has been accepted (step S304: YES), it sends multiple source information, instructions for creating a business plan, the query entered in the chat field T32, and the type of business plan that was accepted to the server 10 via the communication unit 23 (step S305).

[0117] The communication unit 13 receives multiple source information transmitted from the computer 20, instructions for creating a business plan, queries entered in the chat field T32, and the type of business plan for which a selection has been accepted (step S401). The control unit 11 reads the format of the business plan corresponding to the received type of business plan from the format DB 141 (step S402).

[0118] The control unit 11 generates a sixth prompt 600 (hereinafter referred to as the sixth prompt) which includes the format of the read business plan and commands for generating multiple sections that constitute the business plan format and guidelines for each section (step S403). The multiple sections that constitute the business plan format are multiple elements that constitute the business plan format, and include, for example, "Vision" and "Mission". The guidelines for each section are, for example, points to note when writing each section, such as character limits and constraints.

[0119] Figure 22 is an explanatory diagram showing an example of the sixth prompt. The sixth prompt 600 includes a sixth role information field 610, a format field 620, a generation method field 630, and a sixth generation command field 640. The sixth role information field 610 describes the position and role of the language model 121 necessary for generating the section structure and guidelines from the business plan format. The contents described in the sixth role information field 610 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents described in the sixth role information field 610 can be appropriately changed according to the embodiment.

[0120] In Figure 22, the sixth role information section 610 states: "You are an expert who analyzes business plan formats and generates multiple sections and guidelines for each section." Your role will be as follows: 1. Identify all sections within the business plan format and clarify their hierarchical structure. 2. Generate guidelines and requirements related to each section. It is stated.

[0121] The format field 620 displays the format of the business plan obtained by reading from the format DB 141, uploading by the user, or manually entering by the user. In Figure 22, the format field 620 displays "Format A" which was read from the format DB 141.

[0122] The generation method section 630 describes how to generate the multiple sections that make up the business plan format, and the guidelines for each section. The contents described in the generation method section 630 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents described in the generation method section 630 can be modified as appropriate to suit the embodiment.

[0123] In the generation method section 630 of Figure 22, 1. Read the entire document and identify the main sections. 2. Find the subsections within each section and understand the hierarchical relationships. 3. Collect the following information for each section: - Section name - Guidelines for disclosure (if any) - Restrictions (such as character limits, if any)" It is stated.

[0124] The sixth generation instruction field 640 contains instructions for the language model 121 to output a business plan. The contents of the sixth generation instruction field 640 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the sixth generation instruction field 640 can be changed as appropriate to suit the embodiment.

[0125] In Figure 22, the sixth generation instruction column 640 contains: "Please generate multiple sections and guidelines for each section from the business plan format, following the instructions above. When generating the multiple sections and guidelines for each section, please consider the following points." 1. If no guidelines or restrictions exist, those fields can be omitted. 2. The section hierarchy can have any depth. 3. ..." It is stated.

[0126] The control unit 11 inputs the generated sixth prompt 600 to the language model 121, thereby extracting multiple sections and guidelines for each section generated by the language model 121 (step S404). The control unit 11 combines multiple source information, queries entered in the chat field T32, each section and guidelines for each section, and the business plan generation command to generate the fifth prompt 500 (step S405).

[0127] Figure 23 is an explanatory diagram showing an example of the fifth prompt. The fifth prompt 500 includes the fifth role information field 510, source information field 520, query field 530, citation format field 540, and fifth generation instruction field 550, as well as the section content field 560 and created section field 570. The section content field 560 contains the name of each section output by the language model 121 and the guidelines for each section. In the section content column 560 of Figure 23, 1. Section name: "Vision" 2. Guidelines: "Write within 300 characters..." ...." It is stated.

[0128] The "Created Sections" column 570 displays the business plans for each section that the language model has output in the past. In Figure 23, the "Created Sections" column 570 is empty because it is the first time that each section of the business plan is being output.

[0129] Furthermore, in the fifth generation instruction column 550 of Figure 23, Please generate a business plan according to the above instructions. When generating the business plan, please consider the following points. 1. Maintain content and scope appropriate to the section name. 2. Use the source information as the primary source of information, and describe related information and examples. 3. The added information should be cited appropriately according to the "citation format". 4. Do not make speculative statements that are not based on source information. 5. Be mindful of the overall flow of the business plan and maintain consistency and coherence with the content of the previous sections in the "Prepared Sections". 6..." It is stated.

[0130] The control unit 11 inputs the generated fifth prompt 500 to the language model 121 to obtain the business plan generated by the language model 121 (step S406). The control unit 11 determines whether or not a business plan has been obtained for all sections (step S407). If the control unit 11 determines that a business plan has not been obtained for all sections (step S407: NO), it returns to step S405. If the control unit 11 determines that a business plan has been obtained for all sections (step S407: YES), it proceeds to step S225.

[0131] Embodiment 2 describes an example in which an information processing system extracts multiple sections and guidelines for each section from a business plan format and obtains a business plan using the extracted number of sections and guidelines for each section, but is not limited to this. The information processing system may obtain a business plan without extracting multiple sections and guidelines for each section from the business plan format.

[0132] In that case, the control unit 11 generates the fifth prompt 500 without generating the sixth prompt 600. The fifth prompt 500 includes a format field (not shown) instead of the section content field 560 and the created section field 570. The format field displays the format of the acquired business plan. The control unit 11 outputs the business plan by inputting the generated fifth prompt 500 into the language model 121.

[0133] According to Embodiment 2, the information processing system can obtain a business plan format corresponding to the type of business plan, and based on the obtained business plan format, it can extract multiple sections that constitute the business plan format and guidelines for each section.

[0134] According to Embodiment 2, the information processing system can output a business plan by providing a language model with the extracted multiple sections, the guidelines for each section, and multiple source information.

[0135] According to Embodiment 2, the information processing system can generate a consistent business plan by generating the business plan section by section.

[0136] (Embodiment 3) Embodiment 3 describes the processing that occurs when an information processing system receives a selection of the first object from among the first, second, and third objects displayed together with a screen containing analytical information or a business plan.

[0137] Figure 24 is a flowchart showing an example of the processing procedure of an information processing system. The flowchart in Figure 24 adds steps S501-S510 and S601-S608 from step S130 onwards in the process shown in Figures 4-6. Steps similar to those in Figures 4-6 are omitted from explanation. The control unit 21 displays the business plan received from the server 10 on the display unit 25 (step S501).

[0138] Figure 25 is an explanatory diagram showing an example of a screen. The business plan sent from server 10 is displayed on the screen corresponding to the business plan field T1. The screen shown in Figure 25 displays the original text reception field T11 and the editing field T12. The original text reception field T11 displays the business plan sent from server 10 or the analysis information entered by the user. The original text reception field T11 accepts edits from the user via the input unit 24. The user can freely edit the content written in the original text reception field T11 via the input unit 24. The following explanation will use an example where the business plan is displayed in the original text reception field T11.

[0139] The editing field T12 displays the first object O1, the second object O2, and the third object O3 as selectable options. The first object O1 is used by the user to select multiple source information stored in the source information file 122 of the storage unit 12. The second object O2 is used to select the said source information for the language model 121. The third object O3 is used to retrieve source information for the web. The following describes the processing when the selection of the first object O1 is received. The processing when the selection of the second object O2 is received is described in Embodiment 4. The processing when the selection of the third object O3 is received is described in Embodiment 5.

[0140] The control unit 21 receives edits to the business plan displayed in the source text reception field T11 via the input unit 24 (step S502). The control unit 21 determines whether or not it has accepted the selection of the first object O1 (step S503). If the control unit 21 determines that it has not accepted the selection of the first object O1 (step S503: NO), it returns to step S502. If the control unit 21 determines that it has accepted the selection of the first object O1 (step S503: YES), it displays a screen for selecting source information stored in the storage unit 12 (hereinafter referred to as the selection screen) on the display unit 25 (step S504).

[0141] Figure 26 is an explanatory diagram showing an example of a selection screen. The selection screen d1 includes multiple checkboxes, a source information section T31, and an extension button b8. The source information section T31 displays multiple source information stored in the source information file 122. Each checkbox is displayed in a list corresponding to each source information. In Figure 26, each checkbox is located near each source information (for example, on the left side). The user can select the corresponding source information by checking each checkbox. The extension button b8 is a button for sending the source information selected by the user to the server 10.

[0142] The control unit 21 accepts the user's selection of source information via multiple checkboxes (step S505). The control unit 21 determines whether or not the extension button b8 has been selected (step S506). If the control unit 21 determines that the extension button b8 has not been selected (step S506: NO), it returns to step S505.

[0143] If the control unit 21 determines that it has received a selection from the extension button b8 (step S506: YES), it reads the multiple source information selected via the checkboxes from the source information file 122 (step S507). The communication unit 23 sends the business plan edited in the original text reception field T11, the multiple source information read out, and an instruction to extract search keywords from the business plan (hereinafter referred to as the search keyword extraction instruction) to the server 10 (step S508).

[0144] The communication unit 13 receives the business plan, multiple source information, and search keyword extraction instructions transmitted from the computer 20 (step S601). The control unit 11 generates a seventh prompt 700 (hereinafter referred to as the seventh prompt 700) which includes the business plan transmitted from the computer 20 and a search keyword generation instruction (step S602).

[0145] Figure 27 is an explanatory diagram showing an example of the seventh prompt. The seventh prompt 700 includes a seventh role information field 710, a source text field 720, and a seventh generation command field 730. The seventh role information field 710 describes the position and role of the language model 121 necessary for generating search keywords. The contents described in the seventh role information field 710 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents described in the seventh role information field 710 can be appropriately changed according to the embodiment.

[0146] In Figure 27, in the seventh role information section 710, "You are a specialist in generating natural language search keywords for vector-based searches of relevant information to improve text." Your task is to generate up to three optimal search keywords to effectively find relevant information that will help improve the text, based on the original text below. It is stated.

[0147] The original text field 720 contains the business plan transmitted from computer 20. If analysis information is transmitted from computer 20, the original text field 720 contains the analysis information.

[0148] In Figure 27, under the original text column 720, "Against the backdrop of growing environmental awareness in recent years, there is increasing social interest in the use of sustainable energy sources and renewable energy (1). In this context, the utilization of untapped water resources in urban areas holds the potential for creating new renewable energy..." It is stated.

[0149] The seventh generation instruction field 730 contains instructions for causing the language model 121 to output search keywords. The contents of the seventh generation instruction field 730 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the seventh generation instruction field 730 can be modified as appropriate to suit the embodiment.

[0150] In Figure 27, in the seventh generation instruction column 730, Please generate search keywords according to the above instructions. Please consider the following points when generating search keywords. 1. Identify the main topics, concepts, and information needs of the provided source text, and create search keywords to find relevant information. 2. Search keywords are generated in natural language sentence format for use in vector searches. 3. Understand the main point and intent of the original text, and generate search keywords to obtain information that supports it. 4. Create search keywords that can be used to obtain information useful for detailing and concretizing texts such as industry trends, statistical data, and related information. 5. ..." It is stated.

[0151] The control unit 11 inputs the generated seventh prompt 700 to the language model 121, thereby extracting the search keywords generated by the language model 121 (step S603). Specifically, the control unit 11 extracts "environment," "sustainable," and "renewable energy" as examples of search keywords. The control unit 11 extracts source information related to the search keywords (hereinafter referred to as related source information) from the source information transmitted from the computer 20 (source information selected by the user) (step S604).

[0152] The method for extracting relevant source information in Embodiment 3 will now be described in detail. The control unit 11 identifies a portion of the text data related to the search keyword from the text data contained in the source information transmitted from the computer 20 (source information selected by the user). Specifically, the control unit 11 identifies a portion of the text data related to the search keyword "environment," a portion of the text data related to the search keyword "sustainable," and a portion of the text data related to the search keyword "promotion of renewable energy," respectively, from the text data contained in each piece of source information.

[0153] The control unit 11 identifies a portion of the text data related to each search keyword from the text data by, for example, calculating the similarity between the vectors of each search keyword and each text data contained in each source information. If there are multiple sources of information transmitted from the computer 20 (source information selected by the user), the control unit 11 performs the above processing on all of the source information.

[0154] The control unit 11 extracts a portion of the text data identified by the above process as related source information. Through the above process, the control unit 11 can extract a portion of the text data related to the search keyword from the text data contained in the source information transmitted from the computer 20 (source information selected by the user) as related source information.

[0155] In the example described above, the control unit 11 extracts a portion of the text data related to the search keyword from the text data included in the source information as related source information, but it is not limited to this. The control unit 11 may also extract a portion of the image, content, and video included in the source information that is related to the search keyword as related source information.

[0156] The control unit 11 generates an eighth prompt 800 (hereinafter referred to as the eighth prompt 800) which includes the business plan transmitted from the computer 20, the extracted related source information, and the business plan generation command (step S605). Figure 28 is an explanatory diagram showing an example of the eighth prompt. The eighth prompt 800 includes the eighth role information field 810, the source text field 720, the citation format field 820, the related source information field 830, and the eighth generation instruction field 840. The eighth role information field 810 contains the position and role of the language model 121 necessary when expanding the source text (business plan). The contents of the eighth role information field 810 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents of the eighth role information field 810 can be appropriately changed according to the embodiment.

[0157] In Figure 28, in the eighth role information section 810, "You are a professional assistant for expanding, refining, and writing business plans." Your task is to expand on the original document (business plan) based on relevant source information. It is stated.

[0158] The original text field 720 contains the business plan transmitted from computer 20. If analysis information is transmitted from computer 20, the original text field 720 contains the analysis information. The contents of the original text field 720 in Figure 28 are the same as those in Figure 27.

[0159] The citation format section 820 contains the citation format of related source information used when expanding the original text (business plan), correct examples of citations of said related source information, and incorrect examples of citations of said source information. The contents of the citation format section 820 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents of the citation format section 820 can be modified as appropriate to suit the embodiment.

[0160] In the citation format column 820 of Figure 28, "Format of quotation: 1. Please include the source of the newly added information in the expanded text. 2.... Correct citation example: The market size is projected to grow rapidly at a compound annual growth rate (CAGR) of 15% from 2020 to 2025. Incorrect citation example: 1. Cases where quotations are listed together at the end of the answer, even though they could be cited at the sentence or paragraph level: ...." It is stated.

[0161] The related source information section 830 contains the extracted related source information. The related source information is provided in either link or text format. The eighth generation instruction field 840 contains instructions for the language model 121 necessary to expand the source text. The contents of the eighth generation instruction field 840 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the eighth generation instruction field 840 can be modified as appropriate to suit the embodiment.

[0162] In Figure 28, in the 8th generation instruction column 840, Please expand on the original document (business plan) according to the above guidelines. When expanding on the original document (business plan), please consider the following points. 1. To elaborate and specify the content while maintaining the main point and intent of the original text. 2. Add new relevant information and examples, using "Related Source Information" as the primary source of information. 3. The added information should be cited appropriately according to the "citation format". 4. Do not make speculative additions that are not based on "related source information." ...." It is stated.

[0163] The control unit 11 inputs the generated eighth prompt 800 to the language model 121 to acquire the business plan generated by the language model 121 (step S606). The control unit 11 stores the acquired business plan in the large-capacity storage unit 14 (step S607). The communication unit 13 transmits the acquired business plan to the computer 20 (step S608).

[0164] The communication unit 23 receives the business plan transmitted from the server 10 (step S509). The control unit 21 displays the received business plan on the display unit 25 (step S510).

[0165] Embodiment 3 describes an example in which an information processing system extracts a portion of text data related to a search keyword from the text data contained in the source information (source information selected by the user) transmitted from the computer 20, and uses the extracted related source information to obtain a business plan, but is not limited to this. The information processing system may also obtain a business plan using the source information (source information selected by the user) transmitted from the computer 20.

[0166] In that case, the control unit 11 generates the eighth prompt 800 without generating the seventh prompt 700. The related source information field 830 of the eighth prompt 800 contains the source information sent from the computer 20 (source information selected by the user). The control unit 11 outputs the business plan by inputting the generated eighth prompt 800 into the language model 121.

[0167] According to Embodiment 3, the information processing system can output a first object for the user to select multiple source information stored in the memory unit, a second object for the language model to select source information stored in the memory unit, and a third object for obtaining source information from the web.

[0168] According to Embodiment 3, when the information processing system receives a selection of a first object, it outputs a screen for selecting multiple source information, accepts the selection of source information, and can extract search keywords using a language model based on the analysis information or business plan.

[0169] According to Embodiment 3, the information processing system can output a business plan by extracting source information related to the search keyword from the selected source information and providing the extracted source information to a language model.

[0170] (Embodiment 4) Embodiment 4 describes the processing that occurs when an information processing system receives a selection of the second object from among the first, second, and third objects displayed together with a screen containing analytical information or a business plan.

[0171] Figure 29 is a flowchart showing an example of the processing procedure of an information processing system. The flowchart in Figure 29 replaces steps S503-S508 and S601-S606 with steps S701-S702 and S801-S806 in the process shown in Figure 24. Steps similar to those in Figure 24 are omitted from explanation.

[0172] The control unit 21 determines whether or not it has accepted the selection of the second object O2 (step S701). If the control unit 21 determines that it has not accepted the selection of the second object O2 (step S701: NO), it returns to step S502. If the control unit 21 determines that it has accepted the selection of the second object O2 (step S701: YES), it sends the business plan edited in the original text reception field T11 and the search keyword extraction instruction to the server 10 via the communication unit 23 (step S702).

[0173] The communication unit 13 receives the business plan and search keyword extraction instructions transmitted from the computer 20 (step S801). The control unit 11 generates the seventh prompt 700 (see Figure 27) by combining the business plan and search keyword generation instructions transmitted from the computer 20 (step S802).

[0174] The control unit 11 inputs the generated seventh prompt 700 to the language model 121, thereby extracting the search keywords generated by the language model 121 (step S803). Specifically, the control unit 11 extracts "environment," "sustainable," and "renewable energy" as examples of search keywords. The control unit 11 then extracts relevant source information from the source information contained in the source information file 122 (step S804).

[0175] The method for extracting relevant source information in Embodiment 4 will now be described in detail. The control unit 11 identifies a portion of the text data related to the search keyword from all the text data contained in all the source information in the source information file 122. Specifically, the control unit 11 identifies a portion of the text data related to the search keyword "environment," a portion of the text data related to the search keyword "sustainable," and a portion of the text data related to the search keyword "promotion of renewable energy," respectively, from each piece of text data contained in each piece of source information.

[0176] The control unit 11 identifies a portion of the text data related to each search keyword from the text data by, for example, calculating the similarity between the vectors of each search keyword and each text data contained in each source information. The control unit 11 performs the above processing on all source information contained in the source information file 122.

[0177] The control unit 11 extracts a portion of the text data identified by the above process as related source information. Through this process, the control unit 11 can extract a portion of the text data related to the search keyword from all the source information text data contained in the source information file 122 as related source information.

[0178] The control unit 11 generates an eighth prompt 800 (see Figure 28) by combining the business plan, related source information, and business plan generation command transmitted from the computer 20 (step S805). The control unit 11 inputs the generated eighth prompt 800 to the language model 121 to obtain the business plan generated by the language model 121 (step S806). The control unit 11 proceeds to step S607.

[0179] Embodiment 4 describes an example in which an information processing system extracts a portion of the text data related to a search keyword from the text data of all source information contained in the source information file 122 as related source information, and uses the extracted related source information to obtain a business plan, but is not limited to this. The information processing system may also obtain a business plan using all the source information contained in the source information file 122.

[0180] In that case, the control unit 11 generates the eighth prompt 800 without generating the seventh prompt 700. The related source information field 830 of the eighth prompt 800 contains all the source information included in the source information file 122. The control unit 11 outputs the business plan by inputting the generated eighth prompt 800 into the language model 121.

[0181] According to Embodiment 4, when the information processing system receives a selection of a second object, it can output a business plan by extracting search keywords using a language model based on analysis information or a business plan, extracting source information related to the search keywords from the source information stored in the memory unit, and providing the extracted source information to the language model.

[0182] (Embodiment 5) Embodiment 5 describes the processing that occurs when an information processing system receives a selection of the third object from among the first, second, and third objects displayed together with a screen containing analytical information or a business plan.

[0183] Figure 30 is a flowchart showing an example of a processing procedure for an information processing system. The flowchart in Figure 30 replaces steps S503-S508 and S601-S606 with steps S901-S902 and S1001-S1006 in the process shown in Figure 24. Steps similar to those in Figure 24 are omitted from explanation.

[0184] The control unit 21 determines whether or not it has accepted the selection of the third object O3 (step S901). If the control unit 21 determines that it has not accepted the selection of the third object O3 (step S901: NO), it returns to step S502. If the control unit 21 determines that it has accepted the selection of the third object O3 (step S901: YES), it sends the business plan edited in the original text reception field T11 and the search keyword extraction instruction to the server 10 via the communication unit 23 (step S902).

[0185] The communication unit 13 receives the business plan and the search keyword extraction instruction transmitted from the computer 20 (step S1001). The control unit 11 generates a ninth prompt 900 (hereinafter, ninth prompt 900) which includes the business plan and the search keyword generation instruction transmitted from the computer 20 (step S1002).

[0186] Figure 31 is an explanatory diagram showing an example of the ninth prompt. The ninth prompt 900 includes a ninth role information field 910, a source text field 720, a search keyword example field 920, and a ninth generation command field 930. The ninth role information field 910 describes the position and role of the language model 121 necessary for generating search keywords. The contents described in the ninth role information field 910 are pre-stored in the storage unit 12 or the large-capacity storage unit 14. The contents described in the ninth role information field 910 can be appropriately changed according to the embodiment.

[0187] In Figure 31, in the ninth role information section 910, "You are a specialist in generating natural language search keywords for vector-based searches of relevant information to improve text." Your task is to generate up to three optimal search keywords to effectively find relevant information that will help improve the text, based on the following source text (and any instructions provided). It is stated.

[0188] The original text field 720 contains the business plan transmitted from computer 20. If analysis information is transmitted from computer 20, the original text field 720 contains the analysis information. The contents of the original text field 720 in Figure 31 are the same as those in Figure 27.

[0189] The example search keyword field 920 contains specific examples of how the language model 121 can output appropriate search keywords. In the example search keyword section 920 of Figure 31, "1. Market Size Forecast 2025 AND (SaaS OR Cloud Services)" 2. "Competitive Analysis: Market Share Trends - Latest AND Artificial Intelligence" It is stated.

[0190] The ninth generation instruction field 930 contains instructions for causing the language model 121 to output search keywords. The contents of the ninth generation instruction field 930 are pre-stored in the memory unit 12 or the large-capacity memory unit 14. The contents of the ninth generation instruction field 930 can be modified as appropriate to suit the embodiment.

[0191] In Figure 31, the 9th generation instruction column 930 contains: Please generate search keywords according to the above instructions. Please consider the following points when generating search keywords. 1. Identify the main topics, keywords, and concepts in the original text, and create search keywords to delve deeper into them. 2. Understand the main point and intent of the original text, and generate search keywords to obtain information that supports it. 3. Create search keywords that can obtain useful information to make the text more detailed and specific, such as industry trends, market statistics, and competitor information. 4. Include search keywords to supplement technical terms and details. 5. ..." It is stated.

[0192] The control unit 11 inputs the generated ninth prompt 900 to the language model 121, thereby extracting the search keyword generated by the language model 121 (step S1003). The control unit 11 obtains web source information related to the search keyword from the web (step S1004).

[0193] Specifically, the control unit 11 obtains web source information related to the search keyword from the web based on the frequency of occurrence of the search keyword, the reliability of the website, etc., and the similarity between the search keyword and the website, etc. The similarity between the search keyword and the website, etc. can be obtained, for example, by calculating the cosine similarity between the vectorized search keyword and the content on the vectorized website.

[0194] The control unit 11 generates the eighth prompt 800 (see Figure 28) by combining the business plan transmitted from the computer 20, the extracted web source information, and the business plan generation command (step S1005). In this case, the related source information field 830 of the eighth prompt 800 contains the search keywords and related web source information. The control unit 11 inputs the generated eighth prompt 800 into the language model 121 to obtain the business plan generated by the language model 121 (step S1006). The control unit 11 then proceeds to step S607.

[0195] According to Embodiment 5, when the information processing system receives a selection of a third object, it can output a business plan by extracting search keywords using a language model based on analysis information or a business plan, extracting source information related to the search keywords from the source information stored in the memory unit, and providing the extracted source information to the language model.

[0196] The matters described in each of the embodiments described above can be combined with one another. Furthermore, the independent claims and dependent claims described in the claims can be combined with one another in any combination, regardless of the form of reference. Moreover, although the claims use a form in which claims referencing two or more other claims (multi-claim form), the claims are not limited to this. A form in which multi-claims referencing at least one multi-claim (multi-multi-claim) may also be used.

[0197] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the present invention is indicated by the claims, not in the sense described above, and all modifications within the sense and scope equivalent to the claims are intended. [Explanation of symbols]

[0198] 10. Information Processing Equipment (Server) 11 Control Unit 12 Storage section 12P control program 121 language model 122 source information file 13 communication section 14 mass storage section 15 reading section 1a portable storage medium 20 information processing device (computer) 21 control section 22 storage section 22P control program 23 communication section 24 input section 25 display section 26 mass storage section T0 menu bar T1 business plan column T2 business analysis column T3 data column R1 first column R11 input column b1 text button b2 generate button R2 second column R21 item column MA major analysis item ma minor analysis item C analysis content b3 answer button b4 first addition button b5 second addition button b6 confirmation button R3 third column O object T31 source information section T32 chat column T33 type column b7 create button 100 first prompt 110 role information column 120 analysis information column 130 constraint condition column 140 example column 150 first generation command column 200 second prompt 210 second role information column 220 Major analysis item column 230 Second constraint column 240 Second example column 250 2nd generated instruction column 300 Third Prompt 310 Third Role Information Section 320 Small analysis item column 330 3rd generation instruction column 400 Fourth Prompt 410 Fourth Role Information Section 420 Web Search Information Section 430 Citation method column 440 4th generation instruction column 500 Fifth Prompt 510 Fifth Role Information Section 520 Source Information Section 530 Query field 540 Citation format field 550 5th generation instruction column 560 Section Contents 570 Created section fields 600 6th prompt 610 Section 6: Role Information 620 Format field 630 Generation method column 640 6th generated instruction column T11 Original text submission section T12 Editing Section O1 First Object O2 Second Object O3 Third Object d1 Selection screen b8 Extended Button 700 Seventh Prompt 710 Seventh Role Information Section 720 Original text field 730 7th generation instruction column 800 8th Prompt 810 Section 8: Role Information 820 Citation format field 830 Related Source Information Section 840 8th generation instruction column 900 Ninth Prompt 910 Section 9: Role Information 720 Original text field 920 Example Search Keywords 930 9th generation instruction column

Claims

1. Obtain the analytical information necessary to create a business plan. By providing the acquired analysis information to the language model, multiple analysis items and the corresponding analysis content for each analysis item are output. Obtain multiple source information related to the outputted analysis results, By providing the language model with multiple source pieces of information obtained, a business plan is generated. A program that instructs a computer to perform a process.

2. Based on the acquired analysis information, a language model is used to extract multiple major analysis items and sub-analysis items within those major analysis items. The aforementioned language model outputs analysis content corresponding to each sub-analysis item. The program according to claim 1.

3. The system will accept additional input from users for major or minor analysis items. The program according to claim 2.

4. We accept the selection of source information, including web pages or document files related to the text in the analysis. The received source information is stored in the memory unit. The program according to claim 1.

5. Output the input field for analysis information in column 1. Output the analysis items and analysis details in the second column. The third column outputs the selected analysis items and analysis content, as well as an object for selecting source information. The program according to claim 1.

6. The system outputs a first object for the user to select multiple source information stored in the memory unit, a second object for the language model to select source information stored in the memory unit, and a third object for obtaining source information from the web. The program according to claim 4.

7. When the selection of the first object is accepted, a screen for selecting multiple source information is displayed, Accepting selection of source information, Based on the aforementioned analytical information or business plan, a language model is used to extract search keywords. From the received source information, extract source information related to the search keyword, By providing the extracted source information to the language model, a business plan is output. The program according to claim 6.

8. If the selection of the second object is accepted, the language model will extract search keywords based on the analysis information or business plan. Extract source information related to the search keyword from the source information stored in the memory unit. By providing the extracted source information to the language model, a business plan is output. The program according to claim 6.

9. If the selection of the third object is accepted, the language model will extract search keywords based on the analysis information or business plan. Obtain web source information related to the extracted search keywords, By providing the acquired source information to the language model, a business plan is output. The program according to claim 6.

10. A chat window will be displayed to assist in creating the business plan. The business plan is generated by providing the language model with queries entered in the chat field and information from multiple sources. The program according to claim 1.

11. Obtain a business plan format, Based on the acquired format, the language model extracts multiple sections that constitute the format, and the guidelines for each section. By providing the extracted sections, their respective guidelines, and multiple source information to a language model, a business plan is generated. The program according to claim 1.

12. Obtain the analytical information necessary to create a business plan. By providing the acquired analysis information to the language model, multiple analysis items and the corresponding analysis content for each analysis item are output. Obtain multiple source information related to the outputted analysis results, By providing the language model with multiple source pieces of information obtained, a business plan is generated. Information processing methods.

13. An information processing device having a control unit, The control unit, Obtain the analytical information necessary to create a business plan. By providing the acquired analysis information to the language model, multiple analysis items and the corresponding analysis content for each analysis item are output. Obtain multiple source information related to the outputted analysis results, By providing the language model with multiple source pieces of information obtained, a business plan is generated. Information processing device.

Images