Product Information Generation Method, Product Information Generation Device, and Product Information Generation Program

Abstract

Support the improvement of product development efficiency. 【Solution means】 Based on the technical information indicating the manufacturing technology and the dialogue history information generated based on the dialogue with the user, the computer generates product information indicating a product that is manufactured using the manufacturing technology and contributes to the demand indicated by the dialogue history information. The computer outputs the product information.

Description

Technical Field

[0001] The present invention relates to product information generation technology.

Background Art

[0002] There are many small and medium-sized manufacturers (makers) in various regions of Japan that have been continuously producing excellent products with reliable technical capabilities. However, due to changes in the external environment such as labor shortages caused by the declining birthrate and aging population, diversification of sales channels accompanying digitalization, and diversification of consumers' values, it has become difficult to continue operations with only the same products and sales channels as before. In order for such manufacturers to continue their operations, it is desirable to develop new products. There are many manufacturers, not only among small and medium-sized enterprises in each region but also among large enterprises, that are focusing on and struggling with the development of new products.

[0003] Regarding the development of new products, a product development method is known that can form an appropriate corporate chain and develop products according to consumers' needs (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Among manufacturers, there are many small and medium-sized enterprises that have been solely focused on honing their technical capabilities without being involved in marketing or sales strategies as subcontractors for large enterprises. Such manufacturers often do not know where to start even if they want to challenge the development and sales of new products. The same problem exists not only among small and medium-sized enterprises but also among large enterprises.

[0006] On one side, the present invention aims to assist in improving the efficiency of product development.

Means for Solving the Problem

[0007] According to one embodiment, a computer executes the following processes.

[0008] The computer generates product information indicating a product that is manufactured using manufacturing technology and contributes to the demand indicated by the conversation history information, based on the technical information indicating the manufacturing technology and the conversation history information generated based on the interaction with the user. The computer outputs the product information.

Effects of the Invention

[0009] On one side, it is possible to assist in improving the efficiency of product development.

Brief Explanation of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10A

Figure 10B

Figure 11

Embodiments for Carrying Out the Invention

[0011] Hereinafter, embodiments will be described in detail with reference to the drawings.

[0012] FIG. 1 shows a functional configuration example of the product information generation apparatus according to the embodiment. The product information generation apparatus 101 in FIG. 1 includes a generation unit 111 and an output unit 112.

[0013] FIG. 2 is a flowchart showing an example of the first product information generation process performed by the product information generation apparatus 101 in FIG. 1. The generation unit 111 generates product information indicating a product that is manufactured using manufacturing technology and contributes to the demand indicated by the dialogue history information based on the technical information indicating the manufacturing technology and the dialogue history information generated based on the dialogue with the user (step 201). The output unit 112 outputs the product information (step 202).

[0014] According to the product information generation apparatus 101 in FIG. 1, the efficiency of product development can be supported.

[0015] FIG. 3 shows a configuration example of a product information generation system including the product information generation apparatus 101 in FIG. 1. The product information generation system in FIG. 3 includes a management server 311, a generation server 312, a text generation system 313, an image generation system 314, a manufacturer terminal device 315, and user terminal devices 316-1 to 316-N (N is an integer of 1 or more). The generation server 312 corresponds to the product information generation apparatus 101 in FIG. 1.

[0016] The manufacturer terminal device 315 is an information processing device (computer) of the manufacturer that manufactures the product, and the user terminal device 316-i (i = 1 to N) is an information processing device of the user. The user is a consumer who is a candidate to purchase the product. The manufacturer that manufactures the product is an example of a specific manufacturer. The product information generation system may include the manufacturer terminal devices 315 of each of the plurality of manufacturers.

[0017] The manufacturer terminal device 315 and the user terminal device 316-i may be mobile terminal devices or personal computers. The mobile terminal device may be a smartphone, a tablet, or a notebook personal computer.

[0018] The management server 311 communicates with the generation server 312 and the manufacturer terminal device 315 via the communication network 317. The generation server 312 communicates with the management server 311, the text generation system 313, the image generation system 314, and the user terminal devices 316-1 to 316-N via the communication network 317. The communication network 317 is, for example, a WAN (Wide Area Network).

[0019] The manufacturer terminal device 315 transmits, according to the instructions of the manufacturer's employees, technical information indicating the manufacturing technology of the manufacturer to the management server 311. The manufacturing technology represents the technical strength of the manufacturer. The technical information is, for example, text described in natural language.

[0020] The management server 311 receives the technical information from the manufacturer terminal device 315 and stores the received technical information. The management server 311 transmits the stored technical information to the generation server 312.

[0021] The user of each user terminal device 316-i interacts with the generation server 312 using the chat function or the like of the user terminal device 316-i.

[0022] The generation server 312 receives technical information from the management server 311. Also, the generation server 312 generates interaction history information by interacting with the users of each user terminal device 316-i. Next, the generation server 312 uses the technical information and the interaction history information to generate product information indicating a new product, and uses the generated product information to generate an image of the new product. Then, the generation server 312 transmits the product information and the image to the management server 311.

[0023] The new product is manufactured using the manufacturing technology indicated by the technical information and contributes to the demand indicated by the interaction history information.

[0024] When generating the interaction history information and the product information, the generation server 312 instructs the text generation system 313 to generate a desired text and uses the text generated by the text generation system 313.

[0025] The text generation system 313 includes a text generation AI (Artificial Intelligence). The text generation AI is a text generation model trained by machine learning. The text generation AI may be a large language model such as GPT (Generative Pre-trained Transformer).

[0026] The text generation system 313 generates an output sentence using the text generation AI in response to an instruction from the generation server 312 and transmits the generated output sentence to the generation server 312. The output sentence is, for example, text described in natural language.

[0027] The generation server 312 instructs the image generation system 314 to generate a desired image based on the product information and uses the image generated by the image generation system 314 as the image of the product.

[0028] The image generation system 314 includes an image generation AI. The image generation AI is an image generation model trained by machine learning. The image generation system 314 generates an image using the image generation AI in response to an instruction from the generation server 312, and transmits the generated image to the generation server 312.

[0029] Furthermore, the generation server 312 transmits product information and images to the user terminal devices 316-1 to 316-N to request each user for an evaluation of the new product. Then, the generation server 312 receives evaluation information indicating the user's evaluation from each user terminal device 316-i, and transmits evaluation result information indicating the evaluation result of the new product to the management server 311.

[0030] Figure 4 shows a functional configuration example of the management server 311 in Figure 3. The management server 311 in Figure 4 includes a communication unit 411, an access control unit 412, a campaign management unit 413, a product management unit 414, an evaluation result acquisition unit 415, and a storage unit 416.

[0031] The communication unit 411 communicates with the generation server 312 and the manufacturer terminal device 315 via the communication network 317.

[0032] The storage unit 416 stores manufacturer information 421. The manufacturer information 421 includes the manufacturer's account information.

[0033] The access control unit 412 controls access from the manufacturer terminal device 315. The access control unit 412 determines, for example, whether to permit login from the manufacturer terminal device 315 using the manufacturer information 421.

[0034] The campaign management unit 413 receives technical information 422 indicating the manufacturing technology of the manufacturer from the manufacturer terminal device 315 via the communication unit 411, and stores the received technical information 422 in the storage unit 416. The campaign management unit 413 transmits the technical information 422 to the generation server 312 via the communication unit 411. The technical information 422 is text indicating the manufacturing technology.

[0035] For example, the technical information 422 may include texts describing molding techniques for various materials, processing techniques for various materials, a variety of decoration techniques, production systems, design techniques for production equipment, accumulation of technical experience, measurement techniques for product development, etc.

[0036] Furthermore, the campaign management unit 413 receives, via the communication unit 411, campaign information 423 indicating a campaign set by the manufacturer from the manufacturer terminal device 315, and stores the received campaign information 423 in the storage unit 416. The campaign management unit 413 transmits the campaign information 423 to the generation server 312 via the communication unit 411.

[0037] A campaign represents a concept of a new product. For example, a product theme, manufacturing techniques that can be used in manufacturing the product, etc. are set as the campaign. A campaign is an example of a predetermined matter.

[0038] The campaign information 423 is a text indicating the campaign. The manufacturer may use part or all of the technical information 422 as the campaign information 423. When the product information generation system includes the manufacturer terminal devices 315 of a plurality of manufacturers respectively, the campaign management unit 413 receives the technical information 422 and the campaign information 423 of each of those manufacturers and transmits them to the generation server 312.

[0039] The manufacturer can also change or delete the campaign information 423. In this case, the manufacturer terminal device 315 transmits, according to the instructions of the employee, a change instruction or a deletion instruction for the campaign information 423 to the management server 311.

[0040] The campaign management unit 413 receives the change instruction or the deletion instruction from the manufacturer terminal device 315 via the communication unit 411. Then, the campaign management unit 413 changes the campaign information 423 according to the change instruction and deletes the campaign information 423 according to the deletion instruction.

[0041] The Product Management Department 414 receives product information and images from the Generation Server 312 via the Communication Department 411 and transmits the product information and images to the Manufacturer Terminal Device 315.

[0042] The Manufacturer Terminal Device 315 displays the product information and images received from the Management Server 311 on the screen, and the Manufacturer Terminal Device 315 transmits a change instruction or deletion instruction for the product information and images to the Management Server 311 according to the instructions of the employee.

[0043] The Product Management Department 414 receives a change instruction or deletion instruction from the Manufacturer Terminal Device 315 via the Communication Department 411 and transmits the change instruction or deletion instruction to the Generation Server 312.

[0044] The Evaluation Result Acquisition Department 415 receives evaluation result information from the Generation Server 312 via the Communication Department 411 and transmits the evaluation result information to the Manufacturer Terminal Device 315.

[0045] The Manufacturer Terminal Device 315 displays the evaluation result information received from the Management Server 311 on the screen. Thereby, the employee can confirm the user's evaluation of the new product.

[0046] Figure 5 shows a functional configuration example of the Generation Server 312 in Figure 3. The Generation Server 312 in Figure 5 includes a Communication Department 511, an Access Control Department 512, a Campaign Selection Department 513, a Dialogue Department 514, a Product Information Generation Department 515, an Evaluation Department 516, and a Storage Department 517. The Communication Department 511 and the Product Information Generation Department 515 respectively correspond to the Output Department 112 and the Generation Department 111 in Figure 1.

[0047] The Communication Department 511 communicates with the Management Server 311, the Text Generation System 313, the Image Generation System 314, and the User Terminal Devices 316-1 to 316-N via the Communication Network 317.

[0048] The memory unit 517 stores user information 521 and dialogue history information 522. The user information 521 includes the account information of each user. The dialogue history information 522 is text indicating the history of the dialogue between the generation server 312 and each user.

[0049] The access control unit 512 controls access from each user terminal device 316-i. For example, the access control unit 512 uses the user information 521 to determine whether to permit login from the user terminal device 316-i, and selects the hearing target person and the evaluator from the logged-in users. The hearing target person is the user who is the target of the hearing regarding the campaign, and the evaluator is the user who evaluates the new product.

[0050] The campaign selection unit 513 receives the technical information 422 and the campaign information 423 from the management server 311 via the communication unit 511, and stores the received technical information 422 and campaign information 423 in the memory unit 517. When the product information generation system includes the manufacturer terminal devices 315 of a plurality of manufacturers respectively, the campaign selection unit 513 receives the technical information 422 and the campaign information 423 of each of those manufacturers, and stores them in the memory unit 517.

[0051] Next, the campaign selection unit 513 selects the campaign information 423 to be used for the hearing from the campaign information 423 stored in the memory unit 517.

[0052] The dialogue unit 514 communicates with the user terminal device 316-i of the selected hearing target person via the communication unit 511, and conducts a hearing regarding the campaign indicated by the selected campaign information 423.

[0053] In the hearing, the dialogue unit 514 dialogues with the user of the user terminal device 316-i, acquires the text representing the user's speech, and records it in the dialogue history information 522. The dialogue between the dialogue unit 514 and the user is conducted, for example, in a chat format using text.

[0054] As an example of hearing, the dialogue unit 514 instructs the text generation system 313, via the communication unit 511, to generate a question Q1 that asks about the problems or demands the user is aware of regarding the campaign. For example, the dialogue unit 514 sends the text of the campaign information 423 to the text generation system 313, thereby inputting the text indicating the campaign into the text generation AI of the text generation system 313 and instructing the generation of the question Q1.

[0055] The text generation system 313 sends the text of the question Q1 generated from the campaign by the text generation AI to the generation server 312.

[0056] The dialogue unit 514 obtains the question Q1 by receiving the text of the question Q1 from the text generation system 313 via the communication unit 511. Then, the dialogue unit 514 presents the question Q1 to the user by sending the text of the question Q1 to the user terminal device 316-i via the communication unit 511, and records the question Q1 in the dialogue history information 522.

[0057] The user inputs the text of the answer A1 to the question Q1 into the user terminal device 316-i, and the user terminal device 316-i sends the text of the answer A1 to the generation server 312.

[0058] The dialogue unit 514 obtains the answer A1 by receiving the text of the answer A1 from the user terminal device 316-i via the communication unit 511, and records it in the dialogue history information 522. Thereby, the problems or demands the user is aware of regarding the campaign can be included in the dialogue history information 522.

[0059] Next, the dialogue unit 514 sends the text of the answer A1 to the text generation system 313 via the communication unit 511, thereby inputting the answer A1 into the text generation AI and instructing the generation of a question Q2 related to the problems or demands the user is aware of.

[0060] The text generation system 313 transmits the text of question Q2 generated from answer A1 by the text generation AI to the generation server 312.

[0061] The dialogue unit 514 obtains question Q2 by receiving the text of question Q2 from the text generation system 313 via the communication unit 511. Then, the dialogue unit 514 presents question Q2 to the user by transmitting the text of question Q2 to the user terminal device 316-i via the communication unit 511, and records question Q2 in the dialogue history information 522.

[0062] The user inputs the text of answer A2 to question Q2 into the user terminal device 316-i, and the user terminal device 316-i transmits the text of answer A2 to the generation server 312.

[0063] The dialogue unit 514 obtains answer A2 by receiving the text of answer A2 from the user terminal device 316-i via the communication unit 511, and records it in the dialogue history information 522. Thereby, detailed information recognized by the user in daily life related to the problem or request can be included in the dialogue history information 522.

[0064] Next, the product information generation unit 515 generates product information 523 indicating a new product using the technical information 422 and the dialogue history information 522. The new product is manufactured using the manufacturing technology indicated by the technical information 422 and contributes to the demand indicated by the dialogue history information 522.

[0065] The product information generation unit 515 transmits the texts of the technical information 422 and the dialogue history information 522 to the text generation system 313 via the communication unit 511, for example. Thereby, the product information generation unit 515 inputs the technical information 422 and the dialogue history information 522 to the text generation AI and instructs the generation of the product information 523.

[0066] The text generation system 313 transmits the text of the product information 523 generated from the technical information 422 and the dialogue history information 522 by the text generation AI to the generation server 312.

[0067] The product information generation unit 515 obtains the product information 523 by receiving the text of the product information 523 from the text generation system 313 via the communication unit 511, and stores it in the storage unit 517.

[0068] By inputting the technical information 422 into the text generation AI, it is possible to generate the product information 523 of products that can utilize the technical capabilities of the manufacturer. Also, by inputting the dialogue history information 522 into the text generation AI, it is possible to generate the product information 523 of products that contribute to the user's needs, which is derived from the problems or demands that the user is aware of and the detailed information related thereto. By using not only the problems or demands but also the detailed information related thereto together, the accuracy of the product information 523 is improved.

[0069] Next, the product information generation unit 515 generates an image 524 of a new product using the product information 523. For example, the product information generation unit 515 inputs the product information 523 into the text generation AI by transmitting the text of the product information 523 to the text generation system 313 via the communication unit 511, and instructs the generation of the image generation instruction information. The image generation instruction information is information that instructs the image generation system 314 to generate the image 524.

[0070] The text generation system 313 transmits the text of the image generation instruction information generated from the product information 523 by the text generation AI to the generation server 312.

[0071] The product information generation unit 515 obtains the image generation instruction information by receiving the text of the image generation instruction information from the text generation system 313 via the communication unit 511. Then, the product information generation unit 515 inputs the image generation instruction information into the image generation AI of the image generation system 314 by transmitting the image generation instruction information to the image generation system 314 via the communication unit 511, and instructs the generation of the image 524.

[0072] The image generation system 314 transmits the image 524 generated by the image generation AI to the generation server 312.

[0073] The product information generation unit 515 obtains the image 524 by receiving the image 524 from the image generation system 314 via the communication unit 511, and stores it in the storage unit 517.

[0074] Next, the dialogue unit 514 presents the product information 523 and the image 524 to the user by transmitting the product information 523 and the image 524 to the user terminal device 316-i via the communication unit 511.

[0075] The product information generation unit 515 transmits the product information 523 and the image 524 to the management server 311 via the communication unit 511, and receives a change instruction or a deletion instruction from the manufacturer for the product information 523 and the image 524 from the management server 311. Then, the product information generation unit 515 changes the product information 523 and the image 524 according to the change instruction, and deletes the product information 523 and the image 524 according to the deletion instruction.

[0076] By presenting the image 524 together with the product information 523, the user and the employees of the manufacturer can easily confirm the appearance of the unknown product indicated by the product information 523.

[0077] The evaluation unit 516 presents the product information 523 and the image 524 to the users selected as evaluators by transmitting the product information 523 and the image 524 to each user terminal device 316-i of a plurality of evaluators via the communication unit 511.

[0078] The user of each user terminal device 316-i inputs an evaluation of the product indicated by the product information 523 and the image 524 into the user terminal device 316-i, and the user terminal device 316-i transmits evaluation information 525 indicating the input evaluation to the generation server 312. As an example, the user can input a positive evaluation of the product by clicking the "Like" button displayed together with the product information 523 and the image 524.

[0079] As another example, assume that product information 523 and an image 524 of a new product are generated from the dialogue history information 522 of each of a plurality of users for a campaign set by a manufacturer. In this case, the manufacturer selects a plurality of products from the generated new products, and the evaluation unit 516 presents the product information 523 and the image 524 of each of the selected plurality of products to the user. The user selects the most desired product among the plurality of products based on the presented product information 523 and the image 524, and votes with comments.

[0080] The evaluation unit 516 receives evaluation information 525 from each user terminal device 316-i of a plurality of evaluators via the communication unit 511, and stores it in the storage unit 517. Then, the evaluation unit 516 generates evaluation result information indicating an evaluation result by aggregating the evaluation information 525, and transmits the evaluation result information to the management server 311 via the communication unit 511. The evaluation result information may include the number or ratio of users who input positive evaluations, and may also include the number of votes obtained by voting with comments and the comments of the users.

[0081] According to the product information generation system of FIG. 3, product information 523 and an image 524 of an unknown product that contributes to the user's needs are automatically generated from the manufacturer's technical information 422 and the user's dialogue history information 522, and provided to the manufacturer. Thereby, it is possible to support the efficiency of product development by the manufacturer.

[0082] Even a manufacturer who is not familiar with marketing or sales strategies does not need to investigate the needs of consumers by himself / herself, and can easily obtain ideas for new products that can be expected to be commercialized by simply providing the technical information 422 showing the company's technical capabilities.

[0083] Instead of providing the management server 311 and the generation server 312 separately, the generation server 312 may also have the functions of the management server 311.

[0084] Either one or both of the functions of the product management department 414 and the evaluation result acquisition department 415 of the management server 311 may be implemented on one or two other servers. Either one or both of the functions of the campaign selection unit 513 and the evaluation unit 516 of the generation server 312 may be implemented on one or two other servers.

[0085] FIG. 6 shows an example of a chat screen of the hearing conducted by the dialogue unit 514 in FIG. 5. Messages 601-1 to 601-13 represent the statements of the dialogue unit 514, and messages 602-1 to 602-10 represent the statements of the user.

[0086] First, the dialogue unit 514 sends message 601-1 to the user. Message 601-1 includes text representing a question asking about the user's age and gender, and text representing a question asking about the problems or demands the user is aware of regarding the campaign. The question in message 601-1 asking about the problems or demands the user is aware of regarding the campaign is an example of question Q1.

[0087] Next, the user sends message 602-1 to the dialogue unit 514. Message 602-1 includes text representing the user's age and gender.

[0088] Next, the dialogue unit 514 sends message 601-2 to the user. Message 601-2 includes the following text.

[0089] "Thank you for your reply! Example 1: ····· Example 2: ····· Example 3: ····· Example 4: ····· Example 5: ····· Please let us know your concerns with reference to the above examples.

[0090] "·····" represents specific text. Message 601-2 is also an example of question Q1.

[0091] Next, the user sends message 602-2 to the dialogue unit 514. Message 602-2 contains text representing a problem or desire that the user is aware of regarding the campaign. Message 602-2 is an example of an answer A1 to question Q1.

[0092] Next, the dialogue unit 514 sends message 601-3 to the user. Message 601-3 contains text representing consent for message 602-2 and text representing a question for message 602-2. Message 601-3 is an example of a question Q2 related to the problem or desire that the user is aware of.

[0093] Next, the user sends message 602-3 to the dialogue unit 514. Message 602-3 contains text representing an answer to message 601-3. Message 602-3 is an example of an answer A2 to question Q2.

[0094] Next, the dialogue unit 514 sends message 601-4 to the user. Message 601-4 contains text representing confirmation of the content of message 602-3 and text representing a question for message 602-3. Message 601-4 is also an example of a question Q2 related to the problem or desire that the user is aware of.

[0095] Next, the user sends message 602-4 to the dialogue unit 514. Message 602-4 contains text representing an answer to message 601-4. Message 602-4 is also an example of an answer A2 to question Q2.

[0096] Next, the dialogue unit 514 sends message 601-5 to the user. Message 601-5 contains the following text.

[0097] "Summary of the hearing: · User attributes: ····· · Common things to improve in daily life: ····· · Detailed information: ····· Idea Product Title ····· Summary of the Idea Target Audience ····· Concerns ····· Concept ····· Features ····· Usage Method ····· (Convey requests)(Re-propose ideas) (Generate images)』

[0098] Among the text of Message 601-5, the first part of the text from "Summary of the Hearing:" to "·Detailed Information:·····" represents a summary of the information elicited from the user through the hearing. Message 601-1 to Message 601-4, Message 602-1 to Message 602-4, and the first part of Message 601-5 are recorded in the dialogue history information 522.

[0099] Among the text of Message 601-5, the second part of the text from "Idea" to "Usage Method ·····" represents the product information 523 generated from the technical information 422 and the dialogue history information 522.

[0100] The last (Convey requests), (Re-propose ideas), and (Generate images) in Message 601-5 represent candidates for responses to Message 601-5.

[0101] The product information generation unit 515 inputs the technical information 422 and the dialogue history information 522 into the text generation AI and instructs the generation of the product information 523 to obtain the product information 523. The dialogue unit 514 uses the obtained product information 523 as the second part of Message 601-5.

[0102] Instructions from the product information generation unit 515 to the text generation AI include, for example, prompt P1, prompt P2, and prompt P3. Prompt P1 includes, for example, the following text.

[0103] 『#Purpose You are a professional inventor. For TEXT1, to solve TEXT2, please output multiple ideas for fictional products using {specific technology} (either single or in combinations). #Constraints ····· #Output format: - Idea - User usage scenario of the idea - Details of the scientific and logical reasons for which {specific technology} specifically solves TEXT2 - {Specific technology} to be solved #Specific technology: TEXT3』

[0104] "TEXT1" represents the text of the first part of message 601-5, "TEXT2" represents the text of message 602-2, and "TEXT3" represents the text of technical information 422.

[0105] Prompt P2 includes, for example, the following text.

[0106] 『Please output any idea that meets the conditions from TEXT4. #Conditions: ····· #Output format: ·····』

[0107] "TEXT4" represents the text of the output sentence for prompt P1. The output sentence for prompt P1 includes text representing multiple ideas.

[0108] Prompt P3 includes, for example, the following text.

[0109] Instruction: You are a [Product Description Writer]. Based on the {Idea} Please output in Markdown format a product using the following technology to highlight the main features and advantages and improve TEXT5 in order to attract potential buyers based on the following constraints and input text. # Idea TEXT6 # Constraints ····· # Output Example: ## Product Title ····· ## Idea Summary Target Audience ····· Concerns ····· Concept ····· ## Features ····· ## Usage ····· · Input Text: TEXT1 · Output Text: Markdown format ## Product Title: ## Idea Summary Target Audience Concerns Concept ## Features ## Usage』

[0110] "TEXT5" represents the text of the campaign, and "TEXT6" represents the text of the output sentence for Prompt P2. The output sentence for Prompt P2 contains the text representing one idea. The text of the output sentence for Prompt P3 is used as the second part of Message 601-5.

[0111] Next, the user sends Message 602-5 to the dialogue unit 514. Message 602-5 contains the following text.

[0112] "Redefine Ideas"

[0113] In this case, the product information generation unit 515 re-enters the technical information 422 and the dialogue history information 522 into the text generation AI, and re-instructs the generation of the product information 523, thereby obtaining the newly generated product information 523.

[0114] Next, the dialogue unit 514 sends the message 601-6 to the user. The message 601-6 contains the following text.

[0115] "Summary of the Hearing: · User attributes: ····· · Common things to improve in daily life: ····· · Detailed information: ····· Idea Product title ····· Idea summary Intended recipient ····· Concerns ····· Concept ····· Features ····· Usage method ····· (Convey requests) (Redefine ideas) (Generate images)""

[0116] Among the text of the message 601-6, the first part of the text from "Summary of the Hearing:" to "· Detailed information: ·····" is the same as the first part of the text of the message 601-5.

[0117] Among the text of the message 601-6, the second part of the text from "Idea" to "Usage method ·····" represents the newly generated product information 523.

[0118] The last ones of "Convey a request", "Re-propose an idea", and "Generate an image" in Message 601-6 represent candidates for responses to Message 601-6.

[0119] Note that instead of sending Message 601-5 containing one "idea" selected from a plurality of "ideas" to the user using Prompt P2, the dialogue unit 514 may send a message containing a plurality of "ideas" to the user. In this case, the user can select a desired "idea" from among the plurality of "ideas" included in the received message.

[0120] Next, the user sends Message 602-6 to the dialogue unit 514. Message 602-6 contains the following text.

[0121] "Convey a request"

[0122] Next, the dialogue unit 514 sends Message 601-7 to the user. Message 601-7 contains the following text.

[0123] "Please write your request for the idea in the chat input field and send it."

[0124] Next, the user sends Message 602-7 to the dialogue unit 514. Message 602-7 contains text for a correction instruction for the idea.

[0125] Next, the dialogue unit 514 sends Messages 601-8 to 601-10 to the user. Message 601-8 contains the following text.

[0126] "Idea" Product title ····· Summary of the idea Recipient ····· Concerns ····· Concept ····· Features ····· Usage method ····· (Convey wishes)(Re-propose ideas) (Generate an image)』

[0127] The last (Convey wishes), (Re-propose ideas), and (Generate an image) in Message 601-8 represent candidates for responses to Message 601-8.

[0128] The product information generation unit 515 obtains the text of Message 601-8 by inputting the text of the idea and Message 602-7 into the text generation AI and instructing the correction of the product information 523. The prompt P4 for the correction instruction to the text generation AI includes, for example, the following text.

[0129] 『# Instruction: You are a [product description writer]. Please modify {idea} according to the user's correction instruction and output it in Markdown format. # Idea TEXT7 # Correction instruction [User input] # Output example: ## Product title ····· ## Idea summary Person to whom it is to be delivered ····· Concerns ····· Concept ····· ## Features ····· ## Usage method ····· ·Output text: Markdown format ## Product title: ## Idea summary Person to whom it is to be delivered Concerns Concept ## Features ## Usage

[0130] "TEXT7" represents the text of the idea before modification, and "[User Input]" represents the text of Message 602-7. The text of the output sentence for Prompt P4 is used as Message 601-8. Message 601-9 includes the following text.

[0131] "Generating an image. Once the generation is complete, please select the image to be used for the product idea."

[0132] Figure 7 shows an example of Message 601-10. Message 601-10 in Figure 7 includes four images numbered "1" to "4". Each image represents a candidate for the product image for the campaign. Each image and "Propose the image again" represent candidates for the response to Message 601-10.

[0133] The product information generation unit 515 obtains the image generation instruction information by inputting Message 601-8 to the text generation AI and instructing the generation of the image generation instruction information. The instruction to generate the image generation instruction information includes, for example, Prompt P5, Prompt P6, Prompt P7, and Prompt P8. Prompt P5 includes, for example, the following text.

[0134] "Please create a prompt for the image generation AI to generate high-quality images with reference to the following # elements, # requirements, and # guidelines for prompt creation. # Elements: TEXT8 # Requirements: ····· # Guidelines for prompt creation: ····· Output format: ·····"

[0135] 「TEXT8」 represents the text of the output sentence for prompt P3 or prompt P4. The output sentence for prompt P5 is used as image generation instruction information.

[0136] Prompt P6 includes, for example, the following text.

[0137] 'Please create a prompt for an image generation AI to generate high-quality images using the following elements. # Elements: TEXT8 # Requirements: ····· # Output format: ·····』

[0138] The output sentence for prompt P6 is used as image generation instruction information.

[0139] Prompt P7 includes, for example, the following text.

[0140] 'Please create a prompt for an image generation AI to generate high-quality images using the following elements. # Elements: TEXT8 # Requirements: ····· # Output format: ·····』

[0141] As the "# Requirements" included in prompt P7, text different from that included in prompt P6 is used. The output sentence for prompt P7 is used as image generation instruction information.

[0142] Prompt P8 includes, for example, the following text.

[0143] 'Please create a prompt for an image generation AI to generate high-quality images using the following elements. # Elements: TEXT8 # Requirements: ····· # Output format: ·····

[0144] As the "# request" included in prompt P8, text different from the "# requests" included in prompts P6 and P7 is used. The output sentence for prompt P8 is used as image generation instruction information.

[0145] The product information generation unit 515 obtains the image 524 by inputting the image generation instruction information into the image generation AI and instructing the generation of the image 524. The dialogue unit 514 uses the obtained image 524 as each image included in the message 601-10. The images numbered "1" to "number 4" in Figure 7 are obtained from the image generation AI by inputting the output sentences of prompts P5 to P8 into the image generation AI, respectively.

[0146] Next, the user sends messages 602-8 and 602-9 to the dialogue unit 514.

[0147] Figure 8 shows an example of message 602-8. The message 602-8 in Figure 8 includes the image numbered "1" selected by the user from the four images in Figure 7. "Modify slightly by myself" and "Complete image generation" in Figure 8 represent candidates for instructions for the image in message 602-8. Message 602-9 includes the following text.

[0148] 'Modify slightly by myself'

[0149] Next, the dialogue unit 514 sends message 601-11 to the user. Message 601-11 includes the following text.

[0150] 'Please enter the points to be slightly modified. (Modify the image)'

[0151] (Modify the image) at the end of message 601-11 represents a candidate for the answer to message 601-11.

[0152] Next, the user sends message 602-10 to the dialogue unit 514. Message 602-10 includes the text of the correction instruction for the image included in message 602-8.

[0153] Next, the dialogue unit 514 sends message 601-12 and message 601-13 to the user.

[0154] Figure 9 shows an example of message 601-12. Message 601-12 in Figure 9 includes the image generated by correcting the image in Figure 8 based on message 602-10. "Manually make fine corrections" and "Complete image generation" in Figure 9 represent candidates for instructions for the image in message 601-12.

[0155] The product information generation unit 515 inputs the image generation instruction information and message 602-10 into the text generation AI, and by instructing the correction of the image generation instruction information, obtains new image generation instruction information for generating the image in Figure 9. The prompt P9 for instructing the correction to the text generation AI includes, for example, the following text.

[0156] 'Please correct the following image generation AI prompt according to the specified instruction and generate a new image generation AI prompt. Instruction: TEXT9 Original image generation AI prompt: TEXT10 Please output the corrected prompt.'

[0157] "TEXT9" represents the text of message 602-10. "TEXT10" represents the text of the image generation instruction information for generating the image in Figure 8. The output sentence for prompt P9 is used as the corrected image generation instruction information.

[0158] The product information generation unit 515 inputs the corrected image generation instruction information into the image generation AI and instructs the regeneration of the image 524, thereby obtaining the corrected image 524. The dialogue unit 514 uses the obtained image 524 as the image included in the message 601-12. The message 601-13 includes the following text.

[0159] 'Thank you for your hard work. Here is the generated idea. Idea Product title ····· Summary of the idea Recipient ····· Concerns ····· Concept ····· Features ····· IMAGE1 (Enter product idea)』

[0160] 「IMAGE1」 represents the image included in the message 601-12 in FIG. 9. The last (Enter product idea) in the message 601-13 represents the candidate for the answer to the message 601-13.

[0161] FIGS. 10A and 10B are flowcharts showing an example of the second product information generation process performed by the generation server 312 in FIG. 5. First, the dialogue unit 514 communicates with the user terminal device 316-i to conduct a hearing with the user regarding the campaign indicated by the campaign information 423.

[0162] In the hearing, the dialogue unit 514 inputs the text of the campaign information 423 into the text generation AI of the text generation system 313 and instructs the generation of the question Q1 for asking about the problems or demands that the user is aware of regarding the campaign (step 1001).

[0163] The text generation system 313 transmits the question Q1 generated from the campaign by the text generation AI to the generation server 312.

[0164] Next, the dialogue unit 514 presents the question Q1 to the user (step 1002) and records the question Q1 in the dialogue history information 522 (step 1003).

[0165] The user terminal device 316-i transmits the user's answer A1 to the question Q1 to the generation server 312.

[0166] The dialogue unit 514 records the answer A1 in the dialogue history information 522 (step 1004). Then, the dialogue unit 514 inputs the answer A1 into the text generation AI and instructs the generation of a question Q2 related to the problem or request that the user is aware of (step 1005).

[0167] The text generation system 313 transmits the question Q2 generated from the answer A1 by the text generation AI to the generation server 312.

[0168] Next, the dialogue unit 514 presents the question Q2 to the user (step 1006) and records the question Q2 in the dialogue history information 522 (step 1007).

[0169] The user terminal device 316-i transmits the user's answer A2 to the question Q2 to the generation server 312.

[0170] The dialogue unit 514 records the answer A2 in the dialogue history information 522 (step 1008). The generation server 312 may repeat the processing of steps 1005 to 1008 a predetermined number of times.

[0171] Next, the product information generation unit 515 inputs the technical information 422 and the dialogue history information 522 into the text generation AI and instructs the generation of the product information 523 (step 1009).

[0172] The text generation system 313 transmits the product information 523 generated from the technical information 422 and the dialogue history information 522 by the text generation AI to the generation server 312.

[0173] Next, the dialogue unit 514 presents the product information 523 to the user (step 1010) and checks whether a re-proposal of the product information 523 is instructed by the user (step 1011).

[0174] If a re-proposal of the product information 523 is not instructed (step 1011, NO), the dialogue unit 514 checks whether a correction of the product information 523 is instructed by the user (step 1012).

[0175] If a correction of the product information 523 is not instructed (step 1012, NO), the product information generation unit 515 inputs the product information 523 into the text generation AI and instructs the generation of image generation instruction information (step 1013).

[0176] The text generation system 313 transmits the image generation instruction information generated from the product information 523 by the text generation AI to the generation server 312.

[0177] The product information generation unit 515 inputs the image generation instruction information into the image generation AI of the image generation system 314 and instructs the generation of a plurality of images 524 that are candidates for the product image (step 1014).

[0178] The image generation system 314 transmits the plurality of images 524 generated by the image generation AI to the generation server 312.

[0179] The dialogue unit 514 presents the plurality of images 524 to the user (step 1015), receives the identification information indicating the image 524 selected by the user (step 1016), and then the dialogue unit 514 checks whether a correction of the selected image 524 is instructed by the user (step 1017).

[0180] If no correction of the selected image 524 is instructed (step 1017, NO), the dialogue unit 514 presents the product information 523 and the selected image 524 to the user (step 1018).

[0181] If a re-proposal of the product information 523 is instructed (step 1011, YES), the generation server 312 repeats the processing from step 1009 onward.

[0182] If a correction of the product information 523 is instructed (step 1012, YES), the product information generation unit 515 inputs the product information 523 and the user's correction instruction into the text generation AI and instructs the correction of the product information 523 (step 1019).

[0183] The text generation system 313 transmits the corrected product information 523 generated from the product information 523 and the correction instruction by the text generation AI to the generation server 312. Then, the generation server 312 repeats the processing from step 1010 onward.

[0184] If a correction of the selected image 524 is instructed (step 1017, YES), the product information generation unit 515 inputs the image generation instruction information of the image 524 and the user's correction instruction into the text generation AI and instructs the correction of the image generation instruction information (step 1020).

[0185] The text generation system 313 transmits the corrected image generation instruction information generated from the image generation instruction information and the correction instruction by the text generation AI to the generation server 312.

[0186] The product information generation unit 515 inputs the corrected image generation instruction information into the image generation AI and instructs the regeneration of the image 524 (step 1021).

[0187] The image generation system 314 transmits the corrected image 524 generated from the corrected image generation instruction information by the image generation AI to the generation server 312.

[0188] The dialogue unit 514 presents the corrected image 524 to the user (step 1022), and the generation server 312 repeats the processing after step 1017.

[0189] The configuration of the product information generation device 101 in FIG. 1 is merely an example, and some components may be omitted or changed according to the use or conditions of the product information generation device 101.

[0190] The configuration of the product information generation system in FIG. 3 is merely an example, and some components may be omitted or changed according to the use or conditions of the product information generation system.

[0191] The configuration of the management server 311 in FIG. 4 is merely an example, and some components may be omitted or changed according to the configuration or conditions of the product information generation system. For example, when it is not necessary to evaluate a new product by the user, the evaluation result acquisition unit 415 can be omitted.

[0192] The configuration of the generation server 312 in FIG. 5 is merely an example, and some components may be omitted or changed according to the configuration or conditions of the product information generation system. For example, when it is not necessary to evaluate a new product by the user, the evaluation unit 516 can be omitted. The dialogue unit 514 may perform a hearing using a dialogue-type AI instead of the text generation AI.

[0193] The flowcharts shown in FIGS. 2, 10A, and 10B are merely examples, and some processes may be omitted or changed according to the configuration or conditions of the product information generation system. For example, when it is not necessary to present the image 524, the processes of steps 1013 to 1017 and steps 1020 to 1022 can be omitted.

[0194] The chat screen shown in FIG. 6 and the contents of the above-described messages 601-1 to 601-13 and messages 602-1 to 602-10 are merely examples. The chat screen and messages in the hearing vary according to the technical information 422, campaign information 423, and the user. The images 524 shown in FIGS. 7 to 9 are merely examples, and the images 524 vary according to the product information 523.

[0195] The text of the above-described technical information 422 is merely an example, and the technical information 422 varies according to the manufacturer. The text of the above-described prompts P1 to P9 is merely an example, and the generation server 312 may perform processing using another prompt.

[0196] FIG. 11 shows a hardware configuration example of an information processing apparatus used as the product information generation apparatus 101 in FIG. 1, the management server 311 in FIG. 4, and the generation server 312 in FIG. 5. The information processing apparatus in FIG. 11 includes a CPU (Central Processing Unit) 1111, a memory 1112, an input device 1113, an output device 1114, an auxiliary storage device 1115, a medium drive device 1116, and a network connection device 1117. These components are hardware and are connected to each other by a bus 1118.

[0197] The memory 1112 is, for example, a semiconductor memory such as a ROM (Read Only Memory) or a RAM (Random Access Memory), and stores programs and data used for processing. The memory 1112 may operate as the storage unit 416 in FIG. 4 or the storage unit 517 in FIG. 5.

[0198] The CPU 1111 (processor) operates as the generation unit 111 in FIG. 1, for example, by executing a program using the memory 1112.

[0199] By executing a program using the memory 1112, the CPU 1111 also operates as the access control unit 412, campaign management unit 413, product management unit 414, and evaluation result acquisition unit 415 of FIG. 4.

[0200] By executing a program using the memory 1112, the CPU 1111 also operates as the access control unit 512, campaign selection unit 513, dialogue unit 514, product information generation unit 515, and evaluation unit 516 of FIG. 5.

[0201] The input device 1113 is, for example, a keyboard, a pointing device, etc., and is used for inputting instructions or information from an operator. The output device 1114 is, for example, a display device, a printer, a speaker, etc., and is used for querying or instructing an operator and outputting a processing result. The output device 1114 may operate as the output unit 112 of FIG. 1. The processing result may be product information 523 and an image 524.

[0202] The auxiliary storage device 1115 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, a tape device, a flash memory, etc. The auxiliary storage device 1115 may be a hard disk drive or an SSD (Solid State Drive). The information processing device can store a program and data in the auxiliary storage device 1115 and load them into the memory 1112 for use. The auxiliary storage device 1115 may operate as the storage unit 416 of FIG. 4 or the storage unit 517 of FIG. 5.

[0203] The media drive device 1116 drives the portable recording medium 1119 and accesses the recorded content thereon. The portable recording medium 1119 is a memory device, a flexible disk, an optical disk, a magneto-optical disk, or the like. The portable recording medium 1119 may be a CD-ROM (Compact Disk Read Only Memory), a DVD (Digital Versatile Disk), a USB (Universal Serial Bus) memory, or the like. The operator can store programs and data in the portable recording medium 1119 and load them into the memory 1112 for use.

[0204] As described above, the computer-readable recording medium for storing the programs and data used in the processing is a physical (non-transitory) recording medium such as the memory 1112, the auxiliary storage device 1115, or the portable recording medium 1119.

[0205] The network connection device 1117 is a communication device connected to the communication network 317 and performing data conversion associated with communication. The information processing device can receive programs and data from an external device via the network connection device 1117 and load them into the memory 1112 for use. The network connection device 1117 may operate as the output unit 112 in FIG. 1, the communication unit 411 in FIG. 4, or the communication unit 511 in FIG. 5.

[0206] The information processing device may further include a device such as a GPU (Graphics Processing Unit).

[0207] Note that the information processing device does not necessarily include all the components in FIG. 11, and it is also possible to omit some components according to the application or conditions. For example, when an interface with the operator is not required, the input device 1113 and the output device 1114 may be omitted. When the portable recording medium 1119 is not used, the media drive device 1116 may be omitted.

[0208] As the text generation system 313, image generation system 314, manufacturer terminal device 315, and user terminal device 316-i in FIG. 3, an information processing device similar to that in FIG. 11 can be used.

[0209] Although the disclosed embodiments and their advantages have been described in detail, those skilled in the art will be able to make various changes, additions, and omissions without departing from the scope of the invention clearly described in the claims.

Explanation of Reference Numerals

[0210] 101 Product Information Generation Device 111 Generation Unit 112 Output Unit 311 Management Server 312 Generation Server 313 Text Generation System 314 Image Generation System 315 Manufacturer Terminal Device 316-1 to 316-N User Terminal Devices 317 Communication Network 411, 511 Communication Unit 412, 512 Access Control Unit 413 Campaign Management Unit 414 Product Management Unit 415 Evaluation Result Acquisition Unit 416, 517 Storage Unit 421 Manufacturer Information 422 Technical Information 423 Campaign Information 513 Campaign Selection Unit 514 Dialogue Unit 515 Product Information Generation Unit 516 Evaluation Unit 521 User Information 522 Dialogue History Information 523 Product Information 524 Image 525 Evaluation Information 601-1 to 601-13, 602-1 to 602-10 Messages 1111 CPU 1112 Memory 1113 Input device 1114 Output device 1115 Auxiliary storage device 1116 Media drive 1117 Network connection device 1118 Bus 1119 Portable recording medium

Claims

1. Output questions that elicit the problems that each of one or more users, who are potential consumers who purchase products, are aware of regarding a campaign set by a manufacturer with manufacturing technology, Obtain the answers of each of the users to the questions, Generate dialogue history information for each of the users based on the dialogue with each of the users, including the answers of each of the users, Generate first product information indicating a product that is manufactured using the manufacturing technology and that contributes to the needs of the one or more users indicated by the dialogue history information, based on the technical information indicating the manufacturing technology and the dialogue history information of each of the users, Generate revised second product information based on the first product information and the revision instructions of each of the users for the first product information, Output the second product information, A computer executes the process, The questions are texts generated by the text generation model based on information indicating the campaign input from the computer to the text generation model trained by machine learning, The process of generating the first product information includes: The process of inputting the technical information and the dialogue history information of each of the users into the text generation model, The process of obtaining the first product information generated by the text generation model from the technical information and the dialogue history information of each of the users, and includes, The process of generating the second product information includes: The process of inputting the first product information and the revision instructions of each of the users into the text generation model, The process of obtaining the second product information generated by the text generation model from the first product information and the revision instructions of each of the users, and includes, The product that contributes to the needs of the one or more users is a product in which the problems indicated by the answers of each of the users are solved by the manufacturing technology. A product information generation method characterized by this.

2. Output questions related to the problem based on the answers of each of the users to the questions that elicit the problem, Obtain the answers of each of the users to the questions related to the problem, The computer further executes the process, The process of generating the dialogue history information includes the process of generating the dialogue history information based on the answers of each of the users to the questions that elicit the problem and the answers of each of the users to the questions related to the problem. The product information generation method according to Claim 1, characterized by this.

3. Before the process of outputting questions for eliciting the said problems, input information indicating the said campaign into the said text generation model, Obtain questions for eliciting the said problems, which are generated by the said text generation model from the information indicating the said campaign, Before the process of outputting questions related to the said questions, input the responses of each of the said users to the questions for eliciting the said problems into the said text generation model, Obtain questions related to the said questions, which are generated by the said text generation model from the responses of each of the said users to the questions for eliciting the said problems, The method for generating product information according to claim 2, wherein the computer further executes the process.

4. The said manufacturing technology represents the molding technology of various materials, the processing technology of various materials, the decoration technology, the production system, the design technology of production equipment, the accumulation of technical experience, or the measurement technology for product development, which are possessed by a specific manufacturer. The method for generating product information according to any one of claims 1 to 3.

5. The method for generating product information according to any one of claims 1 to 3, wherein the computer further executes the process of outputting an image generated based on the said second product information.

6. Before the process of outputting the said image, generate image generation instruction information based on the said second product information, Input the image generation instruction information into an image generation model trained by machine learning, Obtain the said image generated by the said image generation model from the image generation instruction information, The method for generating product information according to claim 5, wherein the computer further executes the process.

7. Regarding a campaign set by a manufacturer having manufacturing technology, output questions for eliciting problems that each of one or more users, who are consumers and potential customers for purchasing products, are aware of, obtain the responses of each of the said users to the said questions, and based on the conversation with each of the said users including the responses of each of the said users, a dialogue unit that generates the dialogue history information of each of the said users; Based on the technical information indicating the said manufacturing technology and the dialogue history information of each of the said users, generate first product information indicating products that are manufactured using the said manufacturing technology and contribute to the needs of the one or more users indicated by the said dialogue history information, and based on the said first product information and the correction instructions of each of the said users for the said first product information, a generation unit that generates corrected second product information. an output unit that outputs the second product information; comprising; the question is text generated by the text generation model based on information indicating the campaign input from the computer to the text generation model trained by machine learning; the generation unit inputs the technical information and the dialogue history information of each user into the text generation model, obtains the product information generated by the text generation model from the technical information and the dialogue history information of each user, inputs the first product information and the correction instructions of each user into the text generation model, and obtains the second product information generated by the text generation model from the first product information and the correction instructions of each user; A product information generation device, wherein the product contributing to the needs of the one or more users is a product in which the problems indicated by the answers of each user are solved by the manufacturing technology.

8. For a campaign set by a manufacturer having a manufacturing technology, output a question for asking about the problems that each of one or more users, who are consumers who are candidates for customers who purchase products, are aware of; obtain the answers of each user to the question; generate dialogue history information of each user based on the dialogue with each user including the answers of each user; generate first product information indicating a product manufactured using the manufacturing technology and contributing to the needs of the one or more users indicated by the dialogue history information based on the technical information indicating the manufacturing technology and the dialogue history information of each user; generate corrected second product information based on the first product information and the correction instructions of each user for the first product information; output the second product information; cause a computer to execute the process; the question is text generated by the text generation model based on information indicating the campaign input from the computer to the text generation model trained by machine learning; the process of generating the product information; a process of inputting the technical information and the dialogue history information of each user into the text generation model; a process of obtaining the product information generated by the text generation model from the technical information and the dialogue history information of each user; including; the process of generating the second product information; A process of inputting the first product information and the correction instructions of each user into the text generation model; A process of obtaining the second product information generated by the text generation model from the first product information and the correction instructions of each user; including; A product information generation program, wherein the product that contributes to the needs of the one or more users is a product in which the problems indicated by the answers of each user are solved by the manufacturing technology.

Images