system

Abstract

To provide a system that supports online purchase.SOLUTION: A system includes: means for answering to an inquiry input by a user about product purchase in an online site by an AI chatbot using a generative AI; means for receiving the inquiry of the user by a form post which can be input in a plurality of languages, translating the inquiry into a specific language, and presenting it to an operator when it is determined that the AI chatbot cannot answer; and means for receiving an answer to the inquiry from the operator, translating the answer into an original language, and presenting it to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Given the difficulty in securing operator resources, the challenge is to provide English-speaking users living in Japan with support for inquiries and online purchases in English. [Means for solving the problem]

[0005] To solve this problem, an AI chatbot is used to answer users' questions and support online purchases, with the logs used to improve the site. Furthermore, if the bot is unable to resolve the issue, a form can be submitted in English, with an operator confirming the answer in Japanese using AI translation, and the AI ​​then translates the answer into English and sends it back. Furthermore, the system is multilingual. [Brief explanation of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 2 is a sequence diagram showing a flow of processing in the data processing system according to the first embodiment of the first form example. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Embodiment 1. [Figure 13] FIG. 10 is a sequence diagram showing a processing flow of a data processing system in a second embodiment of the second form example. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Embodiment Example 2. [Figure 15] FIG. 10 is a sequence diagram showing the flow of processing in a data processing system according to a third embodiment of the third embodiment. [Figure 16]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Embodiment 3. [Figure 17] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when an emotion engine is combined. [Figure 19] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the second embodiment of the second form example when an emotion engine is combined. [Figure 20] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when an emotion engine is combined. [Figure 21] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the third embodiment of the third form example when an emotion engine is combined. [Figure 22] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0007] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, the processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0012] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0018] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0019] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0020] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0023] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0024] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0025] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] This invention is a system that uses an AI chatbot to answer user questions, support online purchases, and use the logs to improve the website. Specifically, it accepts user inquiries and purchase requests and provides appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[0029] "Example 2"

[0030] Furthermore, this embodiment of the present invention can also make the above system multilingual. Specifically, by changing the settings of the AI ​​chatbot or translation AI, it can also support languages ​​other than English. For example, it can also handle inquiries from users in French, Spanish, and other languages.

[0031] "Example 3"

[0032] Furthermore, this embodiment of the present invention is designed to function even when it is difficult to secure human resources. Specifically, an AI chatbot automatically processes inquiries and escalates them to a human operator as necessary. This reduces the burden on operators while allowing them to continue providing support to users.

[0033] The processing flow of each embodiment will be described below.

[0034] "Example 1"

[0035] Step 1: The AI ​​chatbot accepts user inquiries and purchase requests.

[0036] Step 2: The AI ​​chatbot provides the appropriate answer or action.

[0037] Step 3: If the bot can't resolve the issue, prompt the user to submit the form in English.

[0038] Step 4: The AI ​​translates the post into Japanese and an operator verifies and replies.

[0039] Step 5: The AI ​​translates the answer back into English and sends it to the user.

[0040] "Example 2"

[0041] Step 1: Change the settings of your AI chatbot and translation AI to support languages ​​other than English.

[0042] Step 2: For example, respond to inquiries from users in French or Spanish.

[0043] "Example 3"

[0044] Step 1: The AI ​​chatbot automatically handles the inquiry.

[0045] Step 2: Escalate to an operator if necessary.

[0046] Step 3: This reduces the burden on operators while still providing support to users.

[0047] Example 1

[0048] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0049] Conventional online purchasing support systems sometimes fail to provide prompt and appropriate answers to user questions, and a lack of operator resources is a problem, especially when multilingual support is required. Additionally, there was a lack of a way to effectively log user inquiries and use them to improve the site.

[0050] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0051] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming and responding in AI-translated Japanese, and the AI ​​translating the answer into English and sending it, a means for the AI ​​chatbot to generate answers to users' inquiries using a generative AI model, a means for generating prompts for the generative AI model to obtain appropriate answers, and a means for making the flow multilingual. This enables quick and appropriate answers to users' questions, allows the system to function even when operator resources are insufficient, and enables the site to be improved by effectively utilizing the logs.

[0052] An "artificial intelligence chatbot" is a program that automatically generates answers to users' questions and supports online purchases.

[0053] A "generative artificial intelligence model" is an artificial intelligence model used to generate appropriate responses to user queries.

[0054] A "prompt sentence" is text that is input to a generative artificial intelligence model to generate an appropriate answer.

[0055] "Form submission" is a way for users to enter and submit questions in English that the bot cannot solve.

[0056] An "operator" is a human representative who checks the Japanese content translated by artificial intelligence in response to user inquiries and provides appropriate answers.

[0057] "Artificial intelligence translation" is a technology that automatically translates user inquiries and operator responses into different languages.

[0058] "Multilingual support" means that the system has the ability to respond to user inquiries in multiple languages.

[0059] A "log" is data that records user inquiries and system responses and is used for later analysis and site improvement.

[0060] This invention is a system that answers users' questions, supports online purchases, and uses logs to improve the website. Specifically, it uses an AI chatbot to accept inquiries and purchase requests from users and provide appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[0061] Hardware and software used

[0062] Server: Hosts the data processing and AI models, specifically using virtual servers from a cloud service provider.

[0063] Device: The device from which the user accesses the site (e.g., PC, smartphone, tablet, etc.).

[0064] Generative AI model: An artificial intelligence model used to generate answers to user queries, specifically using natural language processing models.

[0065] Translation software: Software for translating between English and Japanese, specifically using a cloud-based translation API.

[0066] Explanation of program processing

[0067] 1. Receiving user inquiries

[0068] A user uses a device to enter a query into a website chatbot.

[0069] The terminal sends this input to the server.

[0070] The server receives the query and records it as a log in the database.

[0071] Specific operation: A user types "Do you have this item in stock?" into the chatbot. The device sends this message to the server, which records it in the database as "User ID: 1234, Inquiry: Do you have this item in stock?"

[0072] 2. Answer generation by AI chatbot

[0073] The server sends the received query to the generative AI model.

[0074] The generative AI model generates an appropriate answer and sends it back to the server.

[0075] The server generates the answer and sends it to the user's device.

[0076] Specific operation: The server sends the query "Do you have this item in stock?" to the generative AI model. The generative AI model generates the answer "In stock" and sends it back to the server. The server sends this answer to the user's device, and the user sees "In stock" on the chatbot's screen.

[0077] 3. Processing Purchase Requests

[0078] The user enters a purchase request.

[0079] The terminal sends this request to the server.

[0080] The server receives the request, checks inventory, and processes the order.

[0081] If necessary, the generative AI model provides additional information to the user.

[0082] How it works: The user types, "I want to buy this product." The device sends this request to the server, which checks the inventory. If the item is in stock, the server notifies the user via the generative AI model, saying, "Your order has been accepted."

[0083] 4. Escalation of Inquiries

[0084] If a query arises that the bot cannot resolve, the server prompts the user to submit a form in English.

[0085] A user submits a form in English.

[0086] The server sends the content to translation software, which translates it into Japanese.

[0087] What happens: The bot tells the user, "We can't solve this problem. Please submit the form in English." The user submits the form in English, saying, "I need help with my order." The server sends this to translation software, which translates it into Japanese as, "I need help with my order."

[0088] 5. Confirmation by the operator

[0089] The operator will review the translated inquiry and enter the appropriate response in Japanese.

[0090] The server sends this response back to the translation software, which translates it into English.

[0091] The server sends the translated response to the user's device.

[0092] What happens: The operator sees the inquiry "I need help with my order" and types "We have it in stock. We will process your order" in Japanese. The server sends this response to translation software, which translates it into English as "The item is in stock. We will process your order." The server then sends this translated response to the user's device.

[0093] 6. Leveraging logs

[0094] The server keeps a log of all queries and responses.

[0095] The server analyzes these logs and generates insights for improving the site.

[0096] Specific operation: The server saves logs such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock." These logs are periodically analyzed to generate insights, such as "There are many inquiries about a particular product," which can be used to improve the site.

[0097] Prompt Sentence Examples

[0098] User asks: "Do you have this item in stock?"

[0099] Prompt for generative AI model: "A user is asking, 'Do you have this item in stock?' Please generate an appropriate answer."

[0100] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0101] Step 1: Receiving user inquiries

[0102] Input: A user uses a device to type a query into a website chatbot.

[0103] What happens: A user types into the chatbot, "Do you have this item in stock?"

[0104] Data processing: The device sends this input to the server.

[0105] Output: The server receives the query and logs it in the database.

[0106] Specific operation: The server records "User ID: 1234, Enquiry: Is this item in stock?"

[0107] Step 2: Generate answers with an AI chatbot

[0108] Input: The query received by the server.

[0109] What happens: The server sends the query "Do you have this item in stock?" to the generative AI model.

[0110] Data computation: Generative AI models generate appropriate answers.

[0111] Output: The generative AI model generates the answer "In stock" and sends it back to the server.

[0112] Specific behavior: The server sends this response to the user's device, and the user sees the message "In stock" on the chatbot's screen.

[0113] Step 3: Processing the purchase request

[0114] Input: The user enters a purchase request.

[0115] What happens: The user types, "I want to buy this product."

[0116] Data processing: The device sends this request to the server.

[0117] Output: The server receives the request, checks inventory, and processes the order.

[0118] Specific operation: The server checks the inventory, and if the item is in stock, it notifies the user through the generative AI model that "your order has been accepted."

[0119] Step 4: Escalate the case

[0120] Input: The query the bot cannot resolve.

[0121] What happens: The bot informs the user, "We can't solve this problem. Please submit the form in English."

[0122] Data processing: User submits form in English.

[0123] Output: The server sends the content to translation software, which translates it into Japanese.

[0124] What happens: The server translates "I need help with my order" to "I need help with my order."

[0125] Step 5: Confirmation by operator

[0126] Input: The translated query.

[0127] Specific operation: The operator confirms the inquiry "I need help with my order" and types in Japanese "We have the item in stock. We will take your order."

[0128] Data calculation: The server sends this response back to the translation software, which translates it into English.

[0129] Output: Sends the translated answer to the user's device.

[0130] Specific behavior: The server translates the message into English as "The item is in stock. We will process your order" and sends it to the user.

[0131] Step 6: Leverage the logs

[0132] Input: A log of all queries and responses.

[0133] Specific operation: The server saves a log such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock."

[0134] Data calculations: Our servers analyze these logs and generate insights for improving the site.

[0135] Output: Insights for improving your site.

[0136] Specific operation: The server generates insights such as "there are many inquiries about a particular product" and uses this insight to improve the site.

[0137] (Application example 1)

[0138] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0139] Conventional online purchasing support systems sometimes failed to provide prompt and appropriate answers to user questions, and operator resources were often insufficient, especially when multilingual support was required. Furthermore, there was a lack of a way to effectively utilize user inquiry logs to improve the site. This resulted in a poor user experience and reduced willingness to purchase.

[0140] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0141] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming the response in Japanese translated by the AI, and the AI ​​translating the response into English and sending it; a means for making the above flow multilingual; a means for users to input questions about products as an application installed on a smartphone and the AI ​​chatbot providing an immediate response; a means for prompting users to submit a form in English for questions that cannot be resolved and translating the form into Japanese by the AI; and a means for an operator confirming the response and the AI ​​translating the response back into English and sending it to the user. This enables quick and appropriate responses to users' questions, facilitates multilingual support, compensates for a lack of operator resources, and improves the user experience.

[0142] "User" refers to the consumer or user making an online purchase.

[0143] "Questions" refer to questions or concerns users have about products or services.

[0144] "Online purchasing" refers to the act of purchasing goods or services over the Internet.

[0145] An "artificial intelligence chatbot" is a program that automatically provides answers to users' questions.

[0146] "Logs" refer to records of user inquiries and interactions with chatbots.

[0147] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0148] "Bot unresolved" refers to a situation where the chatbot is unable to provide an appropriate answer to the user's question.

[0149] "Form submission in English" refers to a form that allows users to enter questions or requests in English.

[0150] "Operator" refers to a human representative who responds to user inquiries.

[0151] "Artificial intelligence translation" refers to the process of using artificial intelligence to translate text into different languages.

[0152] "Multilingual" refers to providing services in multiple languages.

[0153] "Applications installed on a smartphone" refers to software programs that run on a smartphone.

[0154] "Providing immediate answers" refers to providing a quick response to a user's question.

[0155] "Encouraging form submission in English" means instructing users to enter their questions in English.

[0156] "Translating into Japanese" refers to converting English text into Japanese.

[0157] "Translate back into English and send" means translating the Japanese response into English and sending it to the user.

[0158] A system for implementing this invention includes an AI chatbot that answers users' questions, supports online purchases, and uses the results to improve the website. The system also includes a means for accepting form submissions in English when the bot cannot resolve the issue, for an operator to confirm and respond in Japanese using AI translation, and for the AI ​​to retranslate the response into English and send it back. The system also includes a means for making the above flow multilingual.

[0159] Hardware and software used

[0160] Hardware:

[0161] Smartphone

[0162] software:

[0163] Python

[0164] OpenAI(R) API

[0165] Google® trans library

[0166] Data processing and calculation

[0167] Accepting user input:

[0168] Users input questions about products through a smartphone application, which processes the input as text data within the application.

[0169] Answered by an AI chatbot:

[0170] The server uses the OpenAI API to generate answers to the user's questions, which are then returned to the user as text data.

[0171] Translation features:

[0172] The server uses the GoogleTrans library to translate between English and Japanese. When a user submits a form in English, the text is translated into Japanese and sent to the operator. The operator's response is also translated from Japanese to English and sent to the user.

[0173] Operator Intervention:

[0174] Any questions that the chatbot cannot solve are forwarded to an operator, who types the answer in Japanese, which is then translated back into English and sent to the user.

[0175] Specific examples

[0176] Example 1:

[0177] A user asks, "What size is this item?" The AI ​​chatbot immediately replies, "This item is a size medium."

[0178] Example 2:

[0179] If a user asks, "Can this product be shipped internationally?" and the chatbot cannot resolve the question, it prompts the user to submit a form in English. The user types, "Can this product be shipped internationally?", and the AI ​​translates it into Japanese as, "Can this product be shipped internationally?" The operator replies, "Yes, it can be shipped internationally," and the AI ​​translates it into English as, "Yes, it can be shipped internationally," and sends it to the user.

[0180] Prompt Sentence Examples

[0181] User: What size is this item?

[0182] AI Chatbot: This product is size M.

[0183] User: Can this item be shipped internationally?

[0184] AI Chatbot: Unable to resolve. Please submit the form in English.

[0185] User: Can this product be shipped internationally?

[0186] AI Translation: Can this item be shipped internationally?

[0187] Operator: Yes, you can.

[0188] AI Translation: Yes, it can be shipped internationally.

[0189] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0190] Step 1:

[0191] A user inputs a question about a product through a smartphone application. The input text data is processed within the application and sent to the server.

[0192] Input: User question text

[0193] Output: Text data sent to the server

[0194] Step 2:

[0195] The server uses the OpenAI API to generate answers to user questions, and the generated answers are stored as text data on the server.

[0196] Input: User question text

[0197] Data processing: Generate answers to questions using OpenAI APIs

[0198] Output: Generated answer text

[0199] Step 3:

[0200] The server sends the generated answer to the user's smartphone application, through which the user can confirm the answer.

[0201] Input: Generated answer text

[0202] Output: Answer text displayed in the user's smartphone application

[0203] Step 4:

[0204] If the user is not satisfied with the chatbot's answer, a message prompting them to submit a form in English is displayed. The user can then enter their question in English and submit it to the server.

[0205] Input: User's question text in English

[0206] Output: English question text sent to the server

[0207] Step 5:

[0208] The server uses the GoogleTrans library to translate the user's English question into Japanese, and the translated text is sent to the operator.

[0209] Input: User's question text in English

[0210] Data processing: Translated from English to Japanese using the Googletrans library

[0211] Output: Question text in Japanese sent to the operator

[0212] Step 6:

[0213] The operator inputs the answer in Japanese and sends it to the server, which receives the answer.

[0214] Input: Operator's response text in Japanese

[0215] Output: Answer text in Japanese sent to the server

[0216] Step 7:

[0217] The server uses the GoogleTrans library to translate the operator's Japanese response into English, and the translated text is sent to the user's smartphone application.

[0218] Input: Operator's Japanese response text

[0219] Data processing: Translate from Japanese to English using the Googletrans library

[0220] Output: Answer text in English that will be displayed on the user's smartphone application

[0221] Step 8:

[0222] The user can check the English response from the operator through the smartphone application.

[0223] Input: Answer text in English

[0224] Output: Answer text in English that will be displayed on the user's smartphone application

[0225] Example 2

[0226] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0227] Conventional AI chatbot systems only support certain languages, making it difficult to support multiple languages. Furthermore, the process for generating appropriate responses to user inquiries is complicated, making it difficult to respond when there are insufficient operator resources. This leads to a poor user experience and insufficient log collection for site improvement.

[0228] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0229] In this invention, the server includes a means for receiving user input data and determining the language, a means for changing the AI ​​chatbot's settings based on the determined language, and a means for generating an appropriate response using a generative AI model based on the set language. This enables multilingual support and provides prompt and appropriate responses to user inquiries. Furthermore, even if there are insufficient operator resources, the system automatically responds, enabling log collection for improving the user experience and site improvement.

[0230] An "artificial intelligence chatbot" is a program that automatically generates responses to users' questions and assists them in online purchases.

[0231] "Artificial intelligence translation" is a technology that automatically translates text between different languages.

[0232] A "generative AI model" is an artificial intelligence algorithm that generates appropriate responses based on user input.

[0233] "Language determination" is the process of automatically identifying the language of text entered by a user.

[0234] "Reconfiguration" is the process of adjusting the system's operating modes and parameters based on the determined language.

[0235] "Response generation" is the process of generating an appropriate reply to a user's input.

[0236] "Multilingual" refers to the system's ability to respond to user inquiries in several different languages.

[0237] An "operator" is a person whose role is to respond to user inquiries.

[0238] "Resource allocation" refers to preparing the personnel and equipment necessary for the system to operate normally.

[0239] "User experience" is a general term for the satisfaction and ease of use that users feel when using a system.

[0240] "Site Improvement" is the process of improving the functionality and design of a website based on user usage data and feedback.

[0241] MODE FOR CARRYING OUT THE INVENTION

[0242] This invention relates to a multilingual AI chatbot system. A specific embodiment of this system is described below.

[0243] System configuration

[0244] The system consists of three main components: the user, the terminal, and the server. The user makes a query through the terminal, and the terminal sends the data to the server. The server processes the received data and generates an appropriate response to send to the terminal.

[0245] Hardware and software used

[0246] Hardware: Terminals include devices such as PCs, smartphones, and tablets. Servers are high-performance computer systems, including cloud servers.

[0247] Software: The server uses a generative AI model (e.g., GPT-4®) to generate the response, and translation software such as Google Cloud Translation API to determine the language.

[0248] Program processing

[0249] When the server receives a user inquiry, it first determines the language. It uses the Google Cloud Translation API to determine the language. Based on the determined language, the server changes the AI ​​chatbot's settings. For example, if French is detected, the settings are changed to support French.

[0250] The server then uses the generative AI model to generate an appropriate response, which is translated into the user's language using translation software if further translation is required, and then sends the response to the device, which displays it to the user.

[0251] Specific examples

[0252] For example, consider a case where a user makes a query in French: "Bonjour, comment puis-je vous aider?" The device sends this input to the server. The server uses the Google Cloud Translation API to determine that the input is in French. The server then changes the AI ​​chatbot's settings to support French and uses the generative AI model to generate the response: "Bonjour! Comment puis-je vous aider today?" The generated response is sent to the device, which displays it to the user.

[0253] Prompt Sentence Examples

[0254] "Generate an appropriate response to a query in French. The user's input is 'Bonjour, comment puis-je vous aider?'"

[0255] In this way, the system can process queries in multiple languages, improving the user experience.

[0256] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0257] Step 1:

[0258] The user enters a query.

[0259] The user inputs a query into the input field of the terminal. For example, the user inputs "Bonjour, comment puis-je vous aider?" The input data is saved in text format on the terminal.

[0260] Step 2:

[0261] The terminal sends the input data to the server.

[0262] The terminal converts the data entered by the user into JSON format and sends it to the server using the HTTPS protocol. The input data is the text-formatted query, and the output is the JSON-formatted data sent to the server.

[0263] Step 3:

[0264] The server determines the language.

[0265] The server calls the Google Cloud Translation API to determine the language of the received data. The API parses the input data and returns a language code (e.g., "fr" for French). The input is the user's query, and the output is the determined language code.

[0266] Step 4:

[0267] The server changes the settings of the AI ​​chatbot.

[0268] The server changes the AI ​​chatbot's settings based on the determined language code. For example, if the language code is "fr", it changes the settings to support French. The input is the language code, and the output is the changed chatbot settings.

[0269] Step 5:

[0270] The server generates a response.

[0271] The server generates an appropriate response using a generative AI model (e.g., GPT-4) based on the set language. The prompt text is "Generate an appropriate response to a question in French. The user's input is 'Bonjour, comment puis-je vous aider?'." The input is the prompt text and the user's question, and the output is the generated response.

[0272] Step 6:

[0273] The server generates a response and sends it to the terminal.

[0274] The server converts the generated response into JSON format and sends it to the terminal using the HTTPS protocol. The input is the generated response, and the output is the JSON format data sent to the terminal.

[0275] Step 7:

[0276] The terminal displays the response to the user.

[0277] The terminal displays the response received from the server to the user, for example, "Bonjour! Comment puis-je vous aider aujourd'hui?" The input is the response data received from the server, and the output is the text displayed to the user.

[0278] (Application example 2)

[0279] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0280] Conventional online purchasing support systems have limited support for user questions and insufficient multilingual support, making it difficult to provide adequate service to international users. Furthermore, limited operator resources can result in delayed responses, which can lead to a decline in user satisfaction. Furthermore, real-time translation and response generation are difficult, creating a need for an improved user experience.

[0281] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0282] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, having an operator confirm and respond in an AI-translated language, and having the AI ​​translate and send the response; a means for making the above flow multilingual; a means for translating user inquiries in real time, generating responses using a generative AI model, and translating the responses back into the original language; and a means including an application to be installed on a smartphone. This enables quick and appropriate responses even to international users, and is expected to improve user satisfaction.

[0283] "User" means any person or legal entity making an online purchase.

[0284] "Questions" are questions or concerns users have when making an online purchase.

[0285] "Online purchasing" is the act of purchasing goods or services over the Internet.

[0286] An "artificial intelligence chatbot" is a program that automatically responds to users' questions.

[0287] "Logs" are data that record user activities and inquiries.

[0288] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0289] "Bot Unresolved" refers to when an AI chatbot is unable to provide an appropriate answer to a user's question.

[0290] "Multilingual" refers to multiple different languages, especially languages ​​other than English.

[0291] "Form submission" means the act of a user submitting an inquiry or information through a web form.

[0292] An "operator" is a human representative who responds to user inquiries.

[0293] "Artificial intelligence translation" is the process of converting text from one language to another using artificial intelligence.

[0294] A "verification response" is a formal response provided by an operator to a user's inquiry.

[0295] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user inquiries.

[0296] "Real-time" refers to processing occurring immediately without delay.

[0297] A "smartphone" is a mobile phone that can connect to the Internet and install applications.

[0298] An "application" is a software program designed to perform a specific function.

[0299] A system for implementing this invention includes an artificial intelligence chatbot that answers user questions and assists with online purchasing. The system translates user inquiries in real time, generates responses using a generative AI model, and translates the responses back into the user's native language. The system also includes an application installed on a smartphone.

[0300] Hardware and Software Configuration

[0301] Hardware: Smartphone (iOS or ANDROID (registered trademark))

[0302] Software: OpenAI API (generative AI model), GoogleTrans (translation library)

[0303] Data processing and calculation

[0304] 1. Receive user inquiries:

[0305] A user makes a query through a smartphone application. For example, the user asks in French, "Bonjour, pouvez-vous me recommander un bon livre?"

[0306] 2. Enquiry Translation:

[0307] The server uses GoogleTrans to translate the user's query into English, for example, "Bonjour, pouvez-vous me recommander un bon livre?" to "Hello, can you recommend a good book?"

[0308] 3. Generate a response:

[0309] The server uses OpenAI's generative AI model to generate responses to English queries, such as "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[0310] 4. Retranslating responses:

[0311] The server translates the generated response into the original language (French) using Google Translate. For example, "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald." is translated to "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[0312] 5. Response to the user:

[0313] The translated response is returned to the user, who receives the response in French through a smartphone application.

[0314] Specific examples

[0315] User inquiry: “Bonjour, pouvez-vous me recommander un bon livre?”

[0316] Prompt the generative AI model: "Hello, can you recommend a good book?"

[0317] Generated response: "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[0318] Translated response: "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[0319] In this way, the system can provide multilingual online purchasing support, enabling fast and appropriate responses for international users.

[0320] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0321] Step 1:

[0322] The user makes an inquiry through a smartphone application.

[0323] Input: User's question (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[0324] Output: User query text

[0325] Specific behavior: The user enters a question into the application's input field and presses the submit button.

[0326] Step 2:

[0327] The server uses GoogleTrans to translate the user's query into English.

[0328] Input: User's query text (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[0329] Output: Text translated into English (e.g. "Hello, can you recommend a good book?")

[0330] What happens: The server calls the GoogleTrans API to translate the French text into English.

[0331] Step 3:

[0332] The server uses OpenAI's generative AI model to generate responses to English queries.

[0333] Input: Text translated into English (e.g. "Hello, can you recommend a good book?")

[0334] Output: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[0335] What happens: The server calls the OpenAI API, sends a prompt to the generative AI model, and generates a response.

[0336] Step 4:

[0337] The server translates the generated response back into the original language (French) using Google Translate.

[0338] Input: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[0339] Output: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[0340] What happens: The server calls the GoogleTrans API and translates the English response text into French.

[0341] Step 5:

[0342] The server returns the translated response to the user.

[0343] Input: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[0344] Output: Response text displayed on the user's smartphone

[0345] What happens: The server sends the translated response text to the application, which displays it on the user's smartphone.

[0346] Example 3

[0347] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0348] The previous system had issues such as delayed responses to user inquiries and difficulty in responding when operator resources were insufficient. It also lacked multilingual support, limiting support for global users. This resulted in issues such as a decline in user satisfaction and insufficient log collection for site improvement.

[0349] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0350] In this invention, the server includes a means for the automated response device to generate a response to an inquiry using a generative AI model, a means for the automated response device to determine whether escalation is necessary based on the content and complexity of the inquiry, and a means for sending a notification to an operator when it is determined that escalation is necessary. This allows for a prompt and appropriate response to inquiries from users, enables the system to function even when operator resources are insufficient, and enables multilingual support.

[0351] An "automatic answering machine" is a device that automatically generates and provides responses to inquiries from users.

[0352] A "generative AI model" is a model that uses artificial intelligence to generate appropriate responses from text data.

[0353] An "inquiry" is a question or request that a user enters into the system.

[0354] "Escalation" is the process of handing over a response to an operator when it is determined that a more advanced response is required based on the content and complexity of the inquiry.

[0355] "Operator" means a person who directly responds to inquiries from users and provides necessary support.

[0356] "Notification" is information sent from the system to the operator to inform them when escalation is necessary.

[0357] "Multilingual support" refers to the ability of the system to respond to user inquiries in multiple languages.

[0358] The present invention is a system that automatically generates a response to an inquiry from a user and escalates the inquiry to an operator as necessary. A specific embodiment of this system is described below.

[0359] First, a user inputs a query into the system using a device. The device then sends the query to a server. The server then passes the received query to a generative AI model, which generates an appropriate response. For example, OpenAI's GPT-4 can be used as this generative AI model.

[0360] The server evaluates the generated response and determines whether escalation is necessary based on the nature and complexity of the inquiry. If escalation is necessary, the server notifies an operator, who then responds directly to the user's inquiry and provides the necessary support.

[0361] As a concrete example, the following prompt sentence can be input to a generative AI model:

[0362] Example prompt sentence:

[0363] User: Forgot password.

[0364] AI Chatbot: We've sent you a link to reset your password. Please check your email.

[0365] Example prompt sentence:

[0366] User: I think my account has been hacked.

[0367] AI Chatbot: This issue has been escalated to an operator. Please wait a moment.

[0368] The system is able to respond to user inquiries promptly and appropriately, and is designed to continue functioning even when operator resources are insufficient. It also supports multiple languages, enabling it to provide support to users globally.

[0369] The hardware used includes the user's device (PC, smartphone, etc.) and the server, while the software used includes a generative AI model (e.g., GPT-4) and an algorithm for evaluating the content of the query.

[0370] In this way, a system is realized in which the user, the terminal, and the server work together to process inquiries efficiently. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0371] Step 1:

[0372] A user uses a terminal to enter a query into the system, which then transmits the query to the server.

[0373] Type: The user types "I forgot my password" into the chat window on the device.

[0374] Data processing: The terminal formats this text data for transmission to the server.

[0375] Output: The formatted query data is sent to the server.

[0376] Specific operation: The user enters text into the chat window on the device and presses the send button. The device sends the entered text to the server.

[0377] Step 2:

[0378] The server passes the received query to a generative AI model, such as OpenAI's GPT-4, to generate an appropriate response.

[0379] Input: The query data received by the server.

[0380] Data processing: The server converts the query data into the format required to input it into the generative AI model.

[0381] Output: Response data from the generative AI model.

[0382] Specific operation: The server inputs the user's query "I forgot my password" into GPT-4 and generates a response. GPT-4 then generates a response such as "We've sent you a link to reset your password. Please check your email."

[0383] Step 3:

[0384] The server evaluates the generated response and determines whether escalation is necessary based on the nature and complexity of the query.

[0385] Input: Response data from the generative AI model and the original query data.

[0386] Data processing: The server compares the response data with the query data and applies an algorithm to determine if escalation is necessary.

[0387] Output: The result of the decision as to whether escalation is required.

[0388] What happens: The server compares the generated response with the query and determines whether an escalation is necessary. For example, if the query is "I think my account has been hacked," the server determines that an escalation is necessary.

[0389] Step 4:

[0390] The server will send a notification to the operator if it determines that escalation is necessary.

[0391] Input: The result of the decision that escalation is necessary.

[0392] Data processing: The server generates a message to send a notification to the operator.

[0393] Output: Informational message sent to the operator.

[0394] Specific operation: The server displays a notification on the operator's management screen. The operator checks this notification and prepares to respond to the user.

[0395] Step 5:

[0396] Operators respond directly to user inquiries and provide the necessary support.

[0397] Input: Notification message sent to operator and original inquiry data.

[0398] Data processing: The operator collects information to respond appropriately to user inquiries and then carries out the response.

[0399] Output: The support provided to the user.

[0400] Specific Actions: The operator will call the user and provide further assistance, for example, checking the security settings of the user's account and taking any necessary measures.

[0401] (Application example 3)

[0402] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0403] Conventional online purchasing support systems have difficulty in quickly and accurately responding to user inquiries, and there have been problems with the quality of user support declining, especially when it is difficult to provide multilingual support or secure operator resources. Furthermore, there is a lack of a mechanism for appropriately escalating inquiries depending on the content of the inquiry, making it difficult to improve user satisfaction.

[0404] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0405] In this invention, the server includes an AI chatbot means for answering user questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, having an operator confirm and respond in an AI-translated language, and having the AI ​​translate and send the response; a means for the AI ​​chatbot to receive inquiries from users, generate prompt sentences using a generative AI model, and generate responses; a means for determining whether escalation is necessary based on the content of the responses and escalating to an operator if necessary; and a means for making the above flow multilingual. This allows for quick and accurate responses to user questions, enables the system to function even when operator resources are difficult to secure, and improves the quality of user support.

[0406] "User" refers to any person or entity making an online purchase.

[0407] "Questions" refer to questions or concerns users have about online purchases.

[0408] "Online purchasing" refers to the act of purchasing goods or services over the Internet.

[0409] "Support" means assisting users in making a successful online purchase.

[0410] "Log" refers to records of user inquiries, system responses, etc.

[0411] "Site Improvement" refers to improving the functionality and design of a website to enhance user convenience and satisfaction.

[0412] An "AI chatbot" is a program that uses artificial intelligence to automatically respond to user inquiries.

[0413] "Bot unresolved" refers to when an AI chatbot is unable to generate an appropriate response to a user's inquiry.

[0414] "Multilingual" refers to several different languages.

[0415] "Form submission" refers to a web form that allows users to enter and submit their inquiry.

[0416] "Operator" refers to a human representative who responds to user inquiries.

[0417] "AI translation" refers to the use of artificial intelligence to translate text between different languages.

[0418] A "generative AI model" refers to an artificial intelligence model that generates appropriate responses to user inquiries.

[0419] A "prompt sentence" refers to the text that is input to a generative AI model.

[0420] "Response" refers to the reply given by an AI chatbot or operator to a user's inquiry.

[0421] "Escalation" refers to handing over inquiries that the AI ​​chatbot cannot handle to an operator.

[0422] "Flow" refers to a series of processing steps from a user's inquiry to a response.

[0423] A system for implementing this invention includes an AI chatbot that answers user questions, supports online purchases, and uses the results to improve the site. The system includes a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, for an operator to confirm and respond in an AI-translated language, and for the AI ​​to translate and send the response. The system also includes a means for the AI ​​chatbot to receive inquiries from users, generate prompts using a generative AI model, and generate responses. The system also includes a means for determining whether escalation is necessary based on the response content and escalating to an operator if necessary. The above flow supports multiple languages.

[0424] Hardware and software used

[0425] Hardware:

[0426] Server (cloud server is also acceptable)

[0427] User device (smartphone, PC, etc.)

[0428] software:

[0429] Python

[0430] Flask

[0431] OpenAI API

[0432] Data processing and calculation

[0433] server:

[0434] The server receives inquiries from users. The inquiries are sent in JSON format, which the server parses. It then uses the OpenAI API to send prompts to a generative AI model to generate a response. The generated response determines whether escalation is necessary before being returned to the user. If escalation is necessary, the inquiry is handed over to an operator. The operator enters a response, which is then translated again by the AI ​​and sent to the user.

[0435] User device:

[0436] The user's device inputs a query and sends it to the server. The user receives a response from the AI ​​chatbot and can make additional queries if necessary.

[0437] Specific examples

[0438] When a user asks, "Do you have this item in stock?", the server receives the query and sends a prompt to the generative AI model. The generative AI model generates a response, "We currently have it in stock," and the server returns this to the user.

[0439] When a user inquires, "I would like to return an item," the server similarly sends a prompt to the generative AI model, which generates a response saying, "We will begin the return process."

[0440] When a user inquires, "This problem cannot be solved," the server determines that escalation is necessary and hands the inquiry over to an operator. The operator responds, and the AI ​​translates and responds to the user.

[0441] Example prompt sentence:

[0442] User: Do you have this item in stock?

[0443] AI Chatbot: Currently in stock.

[0444] User: I'd like to return it.

[0445] AI Chatbot: Initiate the return process.

[0446] User: I can't solve this problem.

[0447] AI Chatbot: Escalate to a human.

[0448] In this way, users' questions can be responded to quickly and accurately, the system can function even when it is difficult to secure operator resources, and the quality of user support can be improved.

[0449] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0450] Step 1:

[0451] The user inputs a query from the terminal and sends it to the server.

[0452] Input: User's query text (e.g., "Do you have this item in stock?")

[0453] Output: Query data sent to the server (JSON format)

[0454] Specific actions: The user uses a smartphone or computer to enter a query into the chatbot's interface and press the send button.

[0455] Step 2:

[0456] The server receives the user's query and sends a prompt to the generative AI model.

[0457] Input: User's query data (JSON format)

[0458] Output: The prompt sent to the generative AI model

[0459] Specific operation: The server analyzes the received query data, generates an appropriate prompt, and sends it to the OpenAI API.

[0460] Step 3:

[0461] The generative AI model generates a response based on the prompt sentence and sends it back to the server.

[0462] Input: prompt statement

[0463] Output: The generated response text (e.g., "Currently in stock")

[0464] Specific operation: The OpenAI API receives the prompt sentence, generates a response using the generative AI model, and sends the result back to the server.

[0465] Step 4:

[0466] The server receives the generated response and determines whether escalation is necessary.

[0467] Input: Generated response text

[0468] Output: Judgment result on whether escalation is necessary (e.g., no escalation required)

[0469] What happens: The server parses the generated response text and determines if escalation is necessary based on specific keywords or conditions.

[0470] Step 5:

[0471] If no escalation is required, the server sends the response back to the user.

[0472] Input: Generated response text

[0473] Output: The response text sent back to the user

[0474] Specific operation: The server sends the generated response text to the user's device, and the user confirms the response in the chatbot's interface.

[0475] Step 6:

[0476] If escalation is required, the server will hand over the query to an operator.

[0477] Input: Generated response text and user query data

[0478] Output: Query data sent to the operator

[0479] Specific behavior: If the server determines that escalation is necessary, it sends the user's inquiry data and the generated response text to an operator.

[0480] Step 7:

[0481] The operator enters a response to the inquiry and sends it to the server.

[0482] Input: Operator response text (e.g., "This item is currently out of stock")

[0483] Output: Operator response data sent to the server

[0484] Specific operation: The operator uses a dedicated interface to enter answers to user inquiries and send them to the server.

[0485] Step 8:

[0486] The server receives the operator's response, performs AI translation if necessary, and returns it to the user.

[0487] Input: Operator response data

[0488] Output: The translated answer text that is sent back to the user

[0489] Specific operation: The server receives the operator's answer data, performs AI translation if necessary, and sends the translated answer text to the user's device.

[0490] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0491] "Example 1"

[0492] In one embodiment of the present invention, an AI chatbot includes an emotion engine that recognizes a user's emotions. The emotion engine analyzes emotions from the user's text and voice input and provides feedback to the AI ​​chatbot. For example, if the user uses words that indicate anger or frustration, the emotion engine recognizes this and notifies the AI ​​chatbot of the user's emotional state. This allows the AI ​​chatbot to adjust its response based on the user's emotions.

[0493] "Example 2"

[0494] The emotion engine also provides information for site improvement based on user sentiment. For example, if many users express dissatisfaction with a particular process or function, the emotion engine will detect this and provide feedback to the site operator for improvement. This will enable a better user experience and improve the site.

[0495] "Example 3"

[0496] Furthermore, the emotion engine can track user emotions over time and analyze their changes. This allows us to understand the emotional changes experienced by users as they use the site and improve our services accordingly. For example, if a user's emotions suddenly worsen during a purchase process, we can identify the cause and take measures to resolve the issue.

[0497] The processing flow of each embodiment will be described below.

[0498] "Example 1"

[0499] Step 1: User sends a message to the AI ​​chatbot via text or voice.

[0500] Step 2: The emotion engine analyzes the user's message and recognizes its emotion.

[0501] Step 3: The emotion engine provides feedback on the emotional information it recognizes to the AI ​​chatbot.

[0502] Step 4: The AI ​​chatbot responds according to the user's emotions based on feedback from the emotion engine.

[0503] "Example 2"

[0504] Step 1: The emotion engine analyzes the user's emotions and provides feedback to the site operator.

[0505] Step 2: The site owner improves the site based on feedback from the emotion engine.

[0506] "Example 3"

[0507] Step 1: The emotion engine tracks the user's emotions over time.

[0508] Step 2: The emotion engine analyzes changes in the user's emotions.

[0509] Step 3: The emotion engine suggests service improvements based on changes in emotions.

[0510] Example 1

[0511] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0512] Traditional online purchasing support systems often fail to provide appropriate answers to users' questions or respond appropriately to their emotions. Furthermore, insufficient multilingual support and a lack of operator resources can lead to poor system functionality. This leads to a poor user experience and delays in site improvements.

[0513] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0514] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming and responding in Japanese using AI translation, and the AI ​​translating the response into English and sending it, a means for making the above flow multilingual, an emotion analysis means for analyzing emotions from the user's text input or voice input and feeding the results back to the AI ​​chatbot, and a means for adjusting responses based on the emotion analysis results. This enables appropriate answers to users' questions and responses based on their emotions, and multilingual support is also possible, so system functionality can be maintained even when operator resources are insufficient, improving user experience and improving the site.

[0515] An "artificial intelligence chatbot" is a program that automatically generates answers to users' questions and supports online purchases.

[0516] "Artificial intelligence translation" is a technology that automatically translates text entered in one language into another language.

[0517] "Emotion analysis" is a technology that analyzes emotions from a user's text or voice input and provides feedback on the results.

[0518] "Multilingual support" refers to having functionality that supports multiple languages ​​and providing appropriate services to users who speak different languages.

[0519] An "operator" is a human representative who responds to inquiries from users.

[0520] A "log" is a record of a user's operation history and inquiries recorded by the system.

[0521] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0522] "Text input" refers to the act of a user entering text information using a keyboard or touchscreen.

[0523] "Voice input" refers to the act of a user inputting voice information using a microphone.

[0524] "Feedback" refers to returning information to the system or user based on analysis results and evaluations.

[0525] MODE FOR CARRYING OUT THE INVENTION

[0526] This invention is a system that answers users' questions, supports online purchases, and uses the logs to improve the website. A specific embodiment of this system is described below.

[0527] System configuration

[0528] The system consists of the following main components:

[0529] 1. Artificial Intelligence Chatbots

[0530] 2. Artificial Intelligence Translation Engine

[0531] 3. Sentiment Analysis Engine

[0532] 4. Operator Interface

[0533] 5. User Device

[0534] Hardware and software used

[0535] Server: Hardware that manages the entire system and runs various engines. For example, a cloud-based server or an on-premise server is used.

[0536] User terminal: A device through which a user accesses the system, including a PC, smartphone, tablet, etc.

[0537] Artificial Intelligence Chatbot: Software that automatically generates answers to user inquiries.

[0538] Artificial intelligence translation engine: Software that translates user inquiries and operator responses into multiple languages.

[0539] Sentiment analysis engine: Software that analyzes emotions from a user's text or voice input.

[0540] Operator interface: An interface through which an operator can provide confirmation and response to a user's inquiry.

[0541] System Operation

[0542] The user uses a device to make a text or voice inquiry to the chatbot. The device then sends the user's input to a server. The server then passes the received user's inquiry to an AI chatbot, which then analyzes the inquiry and generates an appropriate answer. The generated answer is then sent to the device via the server and displayed to the user.

[0543] The emotion analysis engine analyzes the user's text and voice input to recognize the user's emotional state. For example, if a user types, "This service is really terrible!", the emotion analysis engine will recognize anger and feed the results back to the AI ​​chatbot. The chatbot will then adjust its response based on the emotion analysis results and provide an appropriate response to the user.

[0544] The AI ​​translation engine translates user inquiries and operator responses into multiple languages. For example, if a user makes an inquiry in English and the chatbot is unable to resolve the issue, it will prompt the user to submit a form in English. The content entered by the user into the form is sent via the server to the AI ​​translation engine and translated into Japanese. The operator checks the translated Japanese post and enters an appropriate response. The operator's response is sent again to the AI ​​translation engine, translated into English, and sent to the user.

[0545] Specific examples

[0546] For example, if a user asks, "How do I return this product?", an AI chatbot can search for information about the return process and generate a response such as, "Please see this link for the return process." If a user types, "This service is terrible!", a sentiment analysis engine will recognize anger and the chatbot will respond with, "Sorry. What was the problem?"

[0547] Prompt Sentence Examples

[0548] "If a user asks about a product, check availability and respond. And if a user expresses anger, use your emotion engine to respond appropriately."

[0549] The above is a specific embodiment for carrying out the present invention.

[0550] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0551] Step 1:

[0552] Users use their device to send text or voice queries to the chatbot.

[0553] Input: User text or voice input

[0554] Output: Query data from the terminal to the server

[0555] Specific operation: The user types "Do you have this item in stock?" The device sends this text data to the server.

[0556] Step 2:

[0557] The server passes the received user inquiry to an artificial intelligence chatbot.

[0558] Input: Inquiry data sent from the terminal

[0559] Output: Data transfer to an artificial intelligence chatbot

[0560] Specific operation: The server analyzes the user's inquiry and passes it to the artificial intelligence chatbot.

[0561] Step 3:

[0562] The AI ​​chatbot analyzes the inquiry and generates an appropriate response.

[0563] Input: Query data sent from the server

[0564] Output: Generated response data

[0565] Specific behavior: The AI ​​chatbot references the inventory database and generates the answer, "Yes, we have it in stock."

[0566] Step 4:

[0567] The server sends the generated response to the terminal.

[0568] Input: Answer data from an AI chatbot

[0569] Output: Response data to the terminal

[0570] Specific operation: The server sends the generated answer to the terminal and displays it to the user.

[0571] Step 5:

[0572] The emotion analysis engine analyzes the user's text and voice input to recognize their emotional state.

[0573] Input: User text or voice input

[0574] Output: Sentiment analysis result data

[0575] How it works: If a user types, "This service is so terrible!", the sentiment analysis engine will recognize the anger and feed it back to the AI ​​chatbot.

[0576] Step 6:

[0577] The AI ​​chatbot adjusts its response based on the results of sentiment analysis.

[0578] Input: Sentiment analysis result data

[0579] Output: Adjusted correspondence data

[0580] What it does: The AI ​​chatbot responds with something like, "Sorry. What was the problem?"

[0581] Step 7:

[0582] The server sends the adjusted response to the terminal.

[0583] Input: Tailored response data from an artificial intelligence chatbot

[0584] Output: Adjusted corresponding data to the device

[0585] Specific operation: The server sends the adjusted response to the terminal and displays it to the user.

[0586] Step 8:

[0587] If the server detects an inquiry that the AI ​​chatbot cannot resolve, it prompts the user to submit a form in English.

[0588] Input: Unresolved inquiry data from an artificial intelligence chatbot

[0589] Output: Data prompting the user to submit the form

[0590] What happens: The server prompts the user to "fill in the form with your details in English."

[0591] Step 9:

[0592] The terminal sends the information the user has entered into the form to the server.

[0593] Input: User form-entered data

[0594] Output: Form input data to the server

[0595] Specific operation: The user enters "Please tell me about the return procedure" into the form, and the device sends the data to the server.

[0596] Step 10:

[0597] The server sends the received English posts to an artificial intelligence translation engine, which translates them into Japanese.

[0598] Input: User's English post data

[0599] Output: Data translated into Japanese

[0600] Specific operation: The server sends the user's English post to an artificial intelligence translation engine, which translates "Please tell me about the return procedure" into Japanese.

[0601] Step 11:

[0602] An operator will review the translated Japanese post and enter an appropriate response.

[0603] Input: Post data translated into Japanese

[0604] Output: Operator response data

[0605] Specific actions: The operator responds by saying, "Please refer to this link for return procedures."

[0606] Step 12:

[0607] The server then sends the operator's response back to the AI ​​translation engine, where it is translated into English.

[0608] Input: Operator response data

[0609] Output: Answer data translated into English

[0610] Specific operation: The server sends the operator's response to an artificial intelligence translation engine, which translates it into English as "Please refer to this link for the return procedure."

[0611] Step 13:

[0612] The device will then send the translated English response to the user.

[0613] Input: Response data translated into English

[0614] Output: User response data

[0615] What happens: The device displays the translated English answer to the user.

[0616] (Application example 1)

[0617] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0618] Conventional online purchasing support systems struggled to respond appropriately to user questions, particularly due to insufficient multilingual support and sentiment analysis. Furthermore, when operator resources were limited, it was difficult to provide a prompt response, resulting in a poor user experience. Furthermore, the system lacked the convenience of a smartphone application.

[0619] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0620] In this invention, the server includes an AI chatbot means for answering user questions, supporting online purchases, and using the results to improve the website; a means for accepting form submissions in English when the bot is unable to resolve the issue, having an operator confirm and respond in Japanese using AI translation, and then having the AI ​​translate the response into English and send it; a means for making the above flow multilingual; a means including an emotion analysis engine for analyzing user emotions and adjusting responses accordingly; and a means for functioning as an application installed on a smartphone. This enables prompt and appropriate responses to user questions and improves the user experience through multilingual support and emotion analysis. The system also allows for continued functioning even when operator resources are limited.

[0621] An "artificial intelligence chatbot" is an automated response system designed to answer user questions and assist with online purchases.

[0622] "Artificial intelligence translation" is a technology that automatically translates text between different languages.

[0623] An "emotion analysis engine" is a system that analyzes emotions from the user's text or voice input and provides feedback on the results.

[0624] "Multilingual" refers to the ability of a system to function in multiple languages.

[0625] An "application installed on a smartphone" is a software program that runs on a smartphone.

[0626] An "operator" is a human representative who responds to inquiries from users.

[0627] "Log" refers to data on user operation history and inquiry details recorded by the system.

[0628] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0629] "Form submission" is the act of using a web form to allow a user to enter and submit certain information.

[0630] "English translation" is the act of translating Japanese text into English.

[0631] To implement the present invention, the following system configuration and program are required.

[0632] System Configuration

[0633] The server contains an AI chatbot that answers users' questions and supports online purchases. When users input questions using their smartphones, the chatbot generates responses in real time. It uses a sentiment analysis engine to analyze the user's emotions and responds accordingly. If the bot cannot resolve the issue, it prompts the user to submit a form in English, which is then translated into Japanese using AI translation. An operator verifies the response, which is then translated back into English and sent to the user. These functions are multilingual and are provided as an application to be installed on smartphones.

[0634] Hardware and software used

[0635] Hardware: Smartphone

[0636] Software: OpenAI API, Google Translate API, TextBlob

[0637] Data processing and calculation

[0638] The server receives text input from the user and generates an appropriate response using the OpenAI API. It uses TextBlob as a sentiment analysis engine to analyze the user's emotions. If the sentiment score is below a certain threshold, it determines that the user is angry and takes appropriate action. If the bot cannot resolve the issue, it uses the Google Translate API to translate the English form submission into Japanese and send it to an operator. The operator's confirmation response is then translated back into English and sent to the user.

[0639] Specific examples

[0640] For example, if a user asks, "How do I return this item?", the following process occurs:

[0641] 1. The user uses their smartphone to type, "How do I return this item?"

[0642] 2. The server parses the sentiment score using TextBlob to ensure the user is not angry.

[0643] 3. Use the OpenAI API to generate a chatbot response to "How do I return this item?"

[0644] 4. If the response does not contain "cannot be resolved", display the generated response to the user.

[0645] Prompt Sentence Examples

[0646] If the user types "How do I return this item?", the following is an example of a prompt:

[0647] User Question: How do I return this item?

[0648] Chatbot response:

[0649] This allows for quick and appropriate responses to user questions, improves the user experience through multilingual support and sentiment analysis, and keeps the system functioning even when operator resources are limited.

[0650] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0651] Step 1:

[0652] The user uses a smartphone to enter a question.

[0653] Input: User question text (e.g., "How do I return this item?")

[0654] Output: The user's question text is sent to the server.

[0655] Specific actions: The user opens the application on their smartphone, enters a question, and presses the send button.

[0656] Step 2:

[0657] The server receives the user's question text and analyzes the sentiment score using a sentiment analysis engine.

[0658] Input: User question text

[0659] Output: Sentiment score (e.g. -0.2)

[0660] What happens: The server uses TextBlob to parse the user's question text and calculates a sentiment score.

[0661] Step 3:

[0662] The server evaluates the emotion score and determines whether the user is angry.

[0663] Input: Sentiment score

[0664] Output: User's emotional state (e.g., not angry)

[0665] Specific behavior: The server checks whether the emotion score is below a certain threshold (e.g., -0.5) and determines that the user is not angry.

[0666] Step 4:

[0667] The server uses the OpenAI API to generate chatbot responses to user questions.

[0668] Input: User question text

[0669] Output: Chatbot response text (e.g., "Here's how to return your order...")

[0670] Specific operation: The server sends the user's question text to the OpenAI API and receives the generated response text.

[0671] Step 5:

[0672] The server sends the chatbot's response text to the user.

[0673] Input: Chatbot response text

[0674] Output: Response text displayed on the user's smartphone

[0675] What happens: The server generates a response text and sends it to the user's smartphone, where the application displays it.

[0676] Step 6:

[0677] The server checks whether the chatbot's response contains the message "cannot be resolved."

[0678] Input: Chatbot response text

[0679] Output: Unresolvable flag (e.g., not included)

[0680] What happens: The server parses the response text and checks whether it contains the keyword "unresolved."

[0681] Step 7:

[0682] If the server determines that it cannot resolve the issue, it prompts the user to submit the form in English.

[0683] Input: Unresolvable flag

[0684] Output: Form submission request in English

[0685] Specific operation: The server sends a message in English to the user's smartphone prompting them to submit the form.

[0686] Step 8:

[0687] The user submits a form in English and the server receives it.

[0688] Input: User's form submission text in English

[0689] Output: Form submission text in English is sent to the server.

[0690] Specific behavior: A user enters a form submission in English and presses the submit button.

[0691] Step 9:

[0692] The server translates the English form submission into Japanese using the Google Translate API.

[0693] Input: Form post text in English

[0694] Output: Japanese translated text

[0695] Specific behavior: The server sends the English form post text to the Google Translate API and receives the translated Japanese text.

[0696] Step 10:

[0697] The server sends the translated text to the operator.

[0698] Input: Japanese translated text

[0699] Output: Text displayed on the operator's terminal

[0700] Specific operation: The server sends the translated Japanese text to the operator's terminal, and the operator checks it.

[0701] Step 11:

[0702] The operator provides a confirmation response, which is received by the server.

[0703] Input: Operator confirmation answer text

[0704] Output: The confirmation response text is sent to the server.

[0705] Specific operation: The operator enters a confirmation answer and presses the send button.

[0706] Step 12:

[0707] The server translates the confirmation response into English using the Google Translate API.

[0708] Input: Confirmation answer text

[0709] Output: Text translated into English

[0710] Specific operation: The server sends the confirmation response text to the Google Translate API and receives the translated English text.

[0711] Step 13:

[0712] The server sends the user a confirmation response translated into English.

[0713] Input: Confirmation answer text translated into English

[0714] Output: Confirmation answer text displayed on the user's smartphone

[0715] What it does: The server sends the translated English text to the user's smartphone, where the application displays it.

[0716] Example 2

[0717] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0718] Previous online support systems often failed to provide appropriate answers to users' questions and lacked multilingual support. Furthermore, they lacked a mechanism for analyzing user sentiment and using it to improve the site, making it difficult to improve the user experience. Furthermore, when operator resources were limited, the system's functionality was often reduced.

[0719] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0720] In this invention, the server includes an AI chatbot that answers users' questions, supports online purchases, and uses the results to improve the website; a chatbot that accepts form submissions in multiple languages ​​when the chatbot is unable to resolve the issue, has an operator confirm and respond in a language translated by the AI, and then translates and sends the response; an emotion analysis engine that analyzes users' emotions and provides feedback for website improvement; and a multilingual support system for the above flow. This allows for appropriate answers to users' questions, strengthens multilingual support, and analyzes user emotions to improve the website. Furthermore, even when operator resources are limited, the system's functionality is not compromised, improving the user experience.

[0721] "User" means an individual or legal entity that uses the System to make inquiries or purchases.

[0722] "Questions" are questions or uncertainties that users have about the system.

[0723] "Online purchasing" is the act of purchasing goods or services over the Internet.

[0724] An "artificial intelligence chatbot" is a program that automatically responds to user inquiries.

[0725] "Log" refers to the user's operation history and inquiry details recorded by the system.

[0726] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0727] "Multilingual" refers to multiple different languages ​​and means that the system supports these languages.

[0728] "Form submission" means the act of a user submitting an inquiry or information through a web form.

[0729] An "operator" is a human representative who responds to inquiries from users.

[0730] "Artificial intelligence translation" is the process of translating text into different languages ​​using artificial intelligence techniques.

[0731] An "emotion analysis engine" is a program that analyzes emotions from a user's text and provides the results.

[0732] "Feedback" means any opinion or information provided to improve a system or service.

[0733] "Flow" refers to a series of processing steps or procedures within a system.

[0734] This invention is a system that answers users' questions, supports online purchases, and uses the logs to improve the website. The system is composed of multiple components, including an AI chatbot, AI translation, and a sentiment analysis engine.

[0735] Hardware and software used

[0736] The server should be a high-performance server (e.g., a high-performance server). The following software should be used:

[0737] Artificial Intelligence Chatbots: Popular Chatbot Platforms

[0738] AI Translation: General Translation API

[0739] Sentiment Analysis Engine: A general sentiment analysis tool

[0740] Program processing

[0741] When the server receives a user inquiry, it first analyzes the inquiry using an AI chatbot, then invokes an AI translation to translate the inquiry into the appropriate language based on the analysis results, and the translated content is passed back to the AI ​​chatbot, which generates an appropriate response for the user.

[0742] Furthermore, the sentiment analysis engine analyzes the emotions contained in user inquiries and responses. For example, if many users express dissatisfaction with a particular process or function, the sentiment analysis engine will detect this and provide feedback to the site operator for improvement.

[0743] Specific examples

[0744] If a user asks in French, "I find the search function on this site difficult to use," the server performs the following process:

[0745] 1. User: Enter "La fonction de recherche de ce site est difficile a utiliser".

[0746] 2. Terminal: Sends the entered query to the server.

[0747] 3. Server: Receives the inquiry and passes it to the AI ​​chatbot.

[0748] 4. Artificial intelligence chatbot: Analyzes the inquiry content and extracts the intent that "the search function is difficult to use."

[0749] 5. Server: Passes the analysis results to an AI translator and requests it to translate from French to English.

[0750] 6. AI translation: Translates as "The search function of this site is difficult to use."

[0751] 7. Server: Passes the translation results to the AI ​​chatbot.

[0752] 8. Artificial intelligence chatbot: Generates the response, "We are sorry to hear that you are having trouble with the search function. We will look into this issue."

[0753] 9. Server: Passes the response to AI translation and requests it to be translated from English to French.

[0754] 10. Artificial Intelligence Translation: Translate "Nous sommes desoles d'apprendre que vous rencontrez des difficulties avec la fonction de recherche.

[0755] 11. Server: Sends the translated response to the device.

[0756] 12. Terminal: displays the response to the user.

[0757] 13. Server: Passes the query content to the sentiment analysis engine and performs sentiment analysis.

[0758] 14. Sentiment analysis engine: Analyzes the content of inquiries and detects feelings of dissatisfaction.

[0759] 15. Server: Feedback the results of the sentiment analysis to the site operator, reporting that "many users are dissatisfied with the search function."

[0760] Prompt Sentence Examples

[0761] "Please explain how a sentiment analysis engine provides feedback when a user is unhappy with a particular feature."

[0762] "Please provide a concrete example of how an AI chatbot would respond to a query in French."

[0763] In this way, the system achieves multilingual support and an improved user experience.

[0764] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0765] Step 1:

[0766] The user enters a query.

[0767] Users use their devices to enter questions and opinions into the system's inquiry form, and the entered data is saved in text format on the device.

[0768] Step 2:

[0769] The terminal sends a query to the server.

[0770] The terminal sends the query entered by the user to the server via the Internet. The input data is sent to the server in text format.

[0771] Step 3:

[0772] The server receives the query and passes it to an artificial intelligence chatbot.

[0773] The server passes the received inquiry to an AI chatbot, which receives the input data in text format.

[0774] Step 4:

[0775] An artificial intelligence chatbot analyzes the inquiry.

[0776] The AI ​​chatbot analyzes the content of the inquiry and extracts the intent and important keywords. The input data is in text format, and the output data is returned to the server in a structured data format as the analysis results.

[0777] Step 5:

[0778] The server passes the analysis results to the artificial intelligence translation system.

[0779] The server then passes the analysis results to an AI translator and requests that they be translated into the specified language. The input data is in structured data format, and the output data is returned in translated text format.

[0780] Step 6:

[0781] Artificial intelligence translation will translate the inquiry into the specified language.

[0782] AI translation translates the query into the specified language, with the input data in a structured data format and the output data returned to the server in a translated text format.

[0783] Step 7:

[0784] The server passes the translation results to an artificial intelligence chatbot.

[0785] The server then passes the translated query back to the AI ​​chatbot, with the input data being in the form of translated text and the output data being returned in the form of structured data as the chatbot's response.

[0786] Step 8:

[0787] An artificial intelligence chatbot generates appropriate responses.

[0788] The AI ​​chatbot generates an appropriate response based on the translated query, with the input data in the form of translated text and the output data in the form of structured data as a response, which is returned to the server.

[0789] Step 9:

[0790] The server passes the response to an artificial intelligence translator to translate it into the user's language.

[0791] The server then passes the generated response to the AI ​​translator again, requesting it to be translated into the user's language. The input data is in structured data format, and the output data is returned in translated text format.

[0792] Step 10:

[0793] The server sends the translated response to the terminal.

[0794] The server sends the translated response to the terminal, where the input data is in translated text format and the output data is displayed on the terminal.

[0795] Step 11:

[0796] The terminal displays the response to the user.

[0797] The terminal displays the response received from the server to the user. The input data is in the translated text form, and the output data is in the text form displayed to the user.

[0798] Step 12:

[0799] The server passes the query content to the emotion analysis engine, which performs emotion analysis.

[0800] The server passes the user's inquiry to the emotion analysis engine for emotion analysis. The input data is in text format, and the output data is returned in structured data format as the emotion analysis results.

[0801] Step 13:

[0802] The sentiment analysis engine returns the analysis results to the server.

[0803] The sentiment analysis engine returns the analysis results to the server. The input data is in text format, and the output data is in structured data format as the sentiment analysis results.

[0804] Step 14:

[0805] The server provides the analysis results as feedback to the website operator.

[0806] The server feeds back the results of the sentiment analysis to the website operator and provides information for improving the website. The input data is in a structured data format as the result of the sentiment analysis, and the output data is feedback information provided to the website operator.

[0807] (Application example 2)

[0808] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0809] Conventional online purchasing support systems lacked multilingual support for user questions, making it difficult to respond appropriately to inquiries in languages ​​other than English. Furthermore, there was a lack of feedback to improve the site based on user sentiment, and the user experience was not sufficiently improved. Furthermore, the system needed to function smoothly even when operator resources were limited.

[0810] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0811] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, an operator confirming and responding in an AI-translated language, and the AI ​​translating and sending the response, a means including an emotion engine for analyzing user emotions and providing feedback for improvements to the site operator, and a means for making the above flow multilingual. This enables multilingual support for users, provides feedback for site improvements based on emotions, and enables the system to function smoothly even when operator resources are limited.

[0812] An "AI chatbot" is a dialogue system that uses artificial intelligence to automatically respond to users' questions and support online purchases.

[0813] "Multilingual support" refers to the ability to support multiple different languages ​​and provide appropriate responses based on the language used by the user.

[0814] "Form submission" refers to the act of a user submitting an inquiry or information using an online input form.

[0815] An "operator" is a human representative who responds to inquiries from users.

[0816] "AI translation" is a technology that uses artificial intelligence to convert text from one language to another.

[0817] The "Emotion Engine" is a system that analyzes user emotions and provides feedback to site operators for improvements based on the results.

[0818] "Feedback" is information for improvement provided based on user experiences and opinions.

[0819] "Website Operator" means the person or organization that manages and operates an online website.

[0820] A "log" is a record of a user's operation history and inquiries recorded by the system.

[0821] The embodiments of the present invention will be described in detail below.

[0822] System Program

[0823] The system consists of a server, a user terminal, and an operator terminal. The server runs programs including an AI chatbot that answers user questions, a multilingual form submission system, and an emotion engine.

[0824] Hardware and software used

[0825] Hardware: Servers, smartphones, operator terminals

[0826] Software: Python, TENSORFLOW (registered trademark), Google Cloud Translation API, Sentiment Analysis API

[0827] Data processing and calculation

[0828] 1. Receiving User Enquiries:

[0829] An inquiry is made from a user terminal (such as a smartphone).

[0830] The query is sent to the server.

[0831] 2. Multilingual support:

[0832] The server uses the Google Cloud Translation API to translate the query into English.

[0833] The translated content is passed to an AI chatbot.

[0834] 3. Chatbot response:

[0835] AI chatbots generate appropriate responses to user queries.

[0836] Retranslate the response into the user's language.

[0837] 4. Sentiment analysis:

[0838] The server uses the Sentiment Analysis API to analyze the user's query and detect their emotions.

[0839] Based on the sentiment data, we provide feedback to site operators for improvement.

[0840] Specific examples

[0841] User inquiry: The user asks in French, "When will this item arrive?"

[0842] Translation: "Quand ce produit arrivera-t-il ?" translated into English using the Google Cloud Translation API.

[0843] Chatbot response: The AI ​​chatbot replies, "The product will arrive in 3 days."

[0844] Retranslate: Retranslate the response into French and tell the user, "Le produit arrivera dans 3 jours."

[0845] Sentiment Analysis: Analyze user sentiment using the Sentiment Analysis API and determine whether it is positive.

[0846] Prompt Sentence Examples

[0847] If a user asks in French, "When will this item arrive?", use the Google Cloud Translation API to translate it into English and pass it to an AI chatbot. Translate the chatbot's response back into French and reply to the user. Also, use the Sentiment Analysis API to analyze user sentiment and provide feedback to the publisher.

[0848] In this way, multilingual support for users is possible, sentiment-based feedback for site improvement is provided, and the system can function smoothly even when operator resources are limited.

[0849] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0850] Step 1:

[0851] The user device receives the inquiry. The user inputs the inquiry into a device such as a smartphone and presses the send button. The input is in the user's language, and the inquiry content is sent to the server.

[0852] Step 2:

[0853] The server receives the query content. The server receives the query content sent from the user terminal and stores it in a database. The input is the user's query content, and the output is the stored query data.

[0854] Step 3:

[0855] The server translates the query into English using the Google Cloud Translation API. The server then sends the received query to the Google Cloud Translation API and receives the translated English text. The input is the user's query, and the output is the translated English text.

[0856] Step 4:

[0857] The server passes the translated content to the AI ​​chatbot. The server sends the translated English text to the AI ​​chatbot, which generates an appropriate response. The input is the translated English text, and the output is the response generated by the AI ​​chatbot.

[0858] Step 5:

[0859] The server retranslates the AI ​​chatbot's response into the user's language. The server then sends the response received from the AI ​​chatbot back to the Google Cloud Translation API and receives the translated text into the user's language. The input is the AI ​​chatbot's response, and the output is the translated text into the user's language.

[0860] Step 6:

[0861] The server sends the translated response to the user terminal. The server sends the translated response to the user's terminal in the user's language and displays it to the user. The input is the text translated into the user's language, and the output is the response displayed on the user terminal.

[0862] Step 7:

[0863] The server analyzes the user's sentiment using the Sentiment Analysis API. The server sends the user's query to the Sentiment Analysis API and receives the sentiment analysis results. The input is the user's query, and the output is the sentiment analysis results.

[0864] Step 8:

[0865] The server provides feedback to the website operator based on the results of the sentiment analysis. The server stores the results of the sentiment analysis in a database and generates feedback for improvement to the website operator. The input is the result of the sentiment analysis, and the output is the feedback provided to the website operator.

[0866] Example 3

[0867] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0868] Traditional online support systems often had slow responses to user inquiries and lacked operator resources. They also lacked the means to track changes in user sentiment and use it to improve services. This resulted in poor user experience and decreased site usage.

[0869] The identification process by the identification processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means. In this invention, the server includes an automatic response device means, a means for receiving an inquiry, an operator responding with a confirmation, and the automatic response device translating and sending the response, a means for supporting multiple languages, a means for tracking user emotions over time and analyzing changes therein, and a means for escalating to an operator as necessary. This enables prompt and appropriate support to users while reducing the burden on operators, improving the user experience and increasing the site usage rate.

[0870] An "automatic answering machine" is a device that automatically generates a response to a user's inquiry and provides an answer.

[0871] "Means for accepting an inquiry, having an operator confirm and respond, and having an automatic answering device translate and send the response" refers to a means for carrying out a series of processes in which an inquiry from a user is accepted, an operator confirms the content of the inquiry, creates a response, and the automatic answering device translates the response and sends it to the user.

[0872] "Means for supporting multiple languages" refers to means for enabling the system to respond to user inquiries in multiple languages.

[0873] "Means for tracking a user's emotions over time and analyzing changes therein" refers to means for continuously monitoring a user's emotions and analyzing changes therein.

[0874] "Means for escalating to an operator as necessary" refers to a means for transferring an inquiry to an operator when the inquiry cannot be handled by the automated answering device or when the inquiry meets certain conditions.

[0875] The present invention provides a system that provides a prompt and appropriate response to user inquiries, reduces the burden on operators, and improves the user experience. Specific embodiments of this system will be described below.

[0876] Hardware and software used

[0877] Hardware: Servers, devices (users' PCs and smartphones)

[0878] Software: Automated response devices (e.g., Dialogflow), sentiment analysis engines (e.g., IBM Watson® Tone Analyzer)

[0879] System Overview

[0880] When a user sends a query using a terminal, the server receives the query and forwards it to an automated answering machine, which analyzes the query and generates an appropriate answer, which is then sent to the user via the server.

[0881] If the automated answering machine cannot process the query, the server escalates the query to a human operator, who creates a confirmation response that the automated answering machine translates and sends to the user.

[0882] The server then sends the user's query to a sentiment analysis engine, tracks the user's sentiment over time, and analyzes changes in that sentiment. This data is then stored in a database and used to improve the service.

[0883] Specific examples

[0884] User Question: "How do I return an item?"

[0885] The automated response: "Here's how to return it..."

[0886] Sentiment analysis engine analysis result: "User sentiment is neutral"

[0887] Prompt Sentence Examples

[0888] Prompt: "A user has inquired about how to return an item. Your automated response system generated an appropriate response, and your sentiment analysis engine determined the user's sentiment to be neutral. Based on these results, please suggest ways to improve our service."

[0889] This system enables quick and appropriate responses to user inquiries, reducing the burden on operators and is expected to improve the user experience and increase the site usage rate. The flow of the identification process in Example 3 will be explained using Figure 21.

[0890] Step 1:

[0891] A user submits an inquiry. Using a device (PC or smartphone), the user enters a question into the inquiry form and clicks the send button. The input is the user's inquiry, and the output is the inquiry data sent to the server. Specifically, the user enters "Please tell me how to return an item" and clicks the send button.

[0892] Step 2:

[0893] The server receives the inquiry. The server receives the inquiry data from the user and records it in a log. The input is the inquiry data sent by the user, and the output is the inquiry data recorded in the log. In concrete terms, the server receives the inquiry "Please tell me how to return an item" and records it in a log.

[0894] Step 3:

[0895] The server forwards the inquiry to an automatic response device. The server forwards the inquiry data it received to the automatic response device (e.g., Dialogflow). The input is the inquiry data recorded on the server, and the output is the inquiry data sent to the automatic response device. Specifically, the server sends the data "Please tell me how to return the product" to Dialogflow.

[0896] Step 4:

[0897] The automated response device analyzes the inquiry and generates a response. The automated response device analyzes the inquiry content using natural language processing technology and generates an appropriate response. The input is the inquiry data sent to the automated response device, and the output is the generated response data. Specifically, Dialogflow generates the response, "The return method is as follows..."

[0898] Step 5:

[0899] The server sends the answer to the user. The server sends the answer it received from the automatic answering machine to the user. The input is the answer data received from the automatic answering machine, and the output is the answer data sent to the user. In concrete terms, the server sends the answer "The return method is as follows..." to the user's terminal.

[0900] Step 6:

[0901] The server sends the query to the sentiment analysis engine. The server sends the user's query to the sentiment analysis engine (e.g. IBM Watson Tone Analyzer). The input is the query data recorded on the server, and the output is the query data sent to the sentiment analysis engine. Specifically, the server sends the data "Please tell me how to return the product" to IBM Watson Tone Analyzer.

[0902] Step 7:

[0903] The sentiment analysis engine analyzes the user's emotions. The sentiment analysis engine analyzes the content of the inquiry and analyzes the user's emotions. The input is the inquiry data sent to the sentiment analysis engine, and the output is the analysis result data. Specifically, IBM Watson Tone Analyzer returns the analysis result that "the user's emotions are neutral."

[0904] Step 8:

[0905] The server tracks changes in emotions and stores them in a database. The server receives the analysis results from the emotion analysis engine, tracks changes in the user's emotions, and stores them in a database. The input is the analysis result data received from the emotion analysis engine, and the output is the emotion change data stored in the database. Specifically, the server stores the data "user's emotion is neutral" in the database.

[0906] Step 9:

[0907] Escalate to an operator if necessary. If the server detects a query that the automated answering machine cannot process, it escalates to an operator. The input is the query data that the automated answering machine cannot process, and the output is the query data forwarded to the operator. Specifically, the server forwards the query "A technical problem has occurred" to the operator.

[0908] (Application example 3)

[0909] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0910] Traditional online purchasing support systems are inadequate in responding to user questions, particularly in the areas of multilingual support and emotional response. Furthermore, when operator resources are insufficient, the system may not function smoothly. This can lead to lower user satisfaction and reduced site usage.

[0911] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0912] In this invention, the server includes: a means for using an AI chatbot with generative AI to respond to inquiries entered by users regarding product purchases on online sites; a means for accepting the user's inquiry via a form submission that can be entered in multiple languages, translating the inquiry into a specific language and presenting it to an operator if the AI ​​chatbot determines that it cannot provide an answer; and a means for accepting an answer to the inquiry from the operator, translating the answer back into the original language, and presenting it to the user. The server also includes a means for using an emotion engine to estimate an emotion score indicating the user's emotion and tracking changes in the emotion score, and a means for notifying the operator if the emotion score falls below a certain threshold. The server also includes a prompt sentence for determining whether the inquiry needs to be escalated to the operator based on the content or complexity of the inquiry entered by the user, and a means for using the generative AI to determine whether the AI ​​chatbot cannot provide an answer. This enables prompt and appropriate responses to user questions and enables service improvements based on changes in emotion. Furthermore, even when operator resources are insufficient, the system functions smoothly and user satisfaction can be improved.

[0913] An "AI chatbot" is a dialogue system that uses artificial intelligence to automatically respond to users' questions and support online purchases.

[0914] "Multilingual support" refers to the ability to process user inquiries and responses in multiple languages.

[0915] The "emotion engine" is a system that tracks users' emotions over time and analyzes their changes.

[0916] The "emotion score" is a numerical indicator of the user's emotional state.

[0917] "Escalate" is the process of transferring an inquiry to an operator when the AI ​​chatbot is unable to respond.

[0918] An "operator" is a person who responds to user inquiries.

[0919] "Log" refers to historical data of user inquiries and responses.

[0920] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[0921] A system for implementing this invention is configured as follows: The server includes an AI chatbot means for answering user questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, an operator confirming and replying in an AI-translated language, and the AI ​​translating and sending the reply, an emotion engine means for tracking user emotions over time and analyzing changes, a means for escalating to an operator if the emotion score falls below a certain threshold, and a means for making the above flow multilingual.

[0922] Program processing explanation

[0923] When the server receives a user inquiry, an AI chatbot using generative AI first automatically generates a response. This response is generated using OpenAI's API. The user's inquiry is analyzed by an emotion engine, and an emotion score for the user is calculated. If this emotion score falls below a certain threshold, the inquiry is escalated to an operator. The operator then replies in the AI-translated language, which is then translated again by AI and sent to the user.

[0924] The hardware used is mobile devices such as smartphones and tablets, and the software uses OpenAI's API and sentiment analysis engine (e.g., SentimentAnalyzer).

[0925] Specific examples

[0926] For example, if a user inquires, "My order hasn't arrived yet. What's going on?", the AI ​​chatbot will respond, "We will check the status of your order. Please wait a moment." After that, an emotion score indicating the user's emotions is estimated based on the user's inquiry and the AI ​​chatbot's response, and if the emotion score falls below a threshold, the inquiry is escalated to an operator. The operator will check the user's inquiry and the chatbot's response and take appropriate action.

[0927] Example prompts to input to a generative AI model:

[0928] User asks: "I haven't received my order, what's going on?"

[0929] The AI ​​chatbot responds: "Please wait a moment while we check the status of your order."

[0930] Sentiment score: -0.6 Escalate: True

[0931] Furthermore, when the server receives an inquiry from a user, it uses a prompt sentence that determines whether the inquiry needs to be escalated to an operator based on the content or complexity of the inquiry entered by the user and generative AI to determine whether the AI ​​chatbot can provide an answer. If the server determines that the AI ​​chatbot cannot provide an answer, it accepts the user's inquiry through a form submission that can be entered in multiple languages, translates the inquiry into a specific language, and presents it to the operator. It also accepts an answer to the inquiry from the operator, translates the answer into the original language, and presents it to the user.

[0932] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0933] Step 1:

[0934] A user inputs an inquiry from a device such as a smartphone or tablet.

[0935] Input: User's inquiry (e.g. "My order hasn't arrived yet. What's going on?")

[0936] Output: The query is sent to the server.

[0937] Step 2:

[0938] The server passes the received inquiry to the AI ​​chatbot and generates a response.

[0939] Input: User's inquiry

[0940] Data processing: Use OpenAI's API to generate responses to inquiries.

[0941] Output: The AI ​​chatbot's response (e.g., "We'll check the status of your order. Please wait a moment.")

[0942] Step 3:

[0943] The server passes the user query and the generated response to an emotion engine to calculate an emotion score for the user.

[0944] Input: AI chatbot response

[0945] Data calculation: Use a sentiment analysis engine (e.g., SentimentAnalyzer) to calculate a sentiment score for the response text.

[0946] Output: Sentiment score (e.g. -0.6)

[0947] Step 4:

[0948] The server determines whether the emotion score is below a certain threshold.

[0949] Input: Sentiment score

[0950] Data calculation: Compare whether the sentiment score is below a threshold (e.g., -0.5).

[0951] Output: Escalation required (True or False)

[0952] Step 5:

[0953] If the sentiment score is below the threshold, the server escalates the query to an operator.

[0954] Input: Escalation Required (True)

[0955] Specific operation: The server notifies the operator of the user's inquiry and the AI ​​chatbot's response.

[0956] Output: Inquiry escalated to operator and response

[0957] Step 6:

[0958] The operator reviews the escalated inquiry and prepares an appropriate response.

[0959] Input: Escalated inquiry and response

[0960] Specific operation: The operator creates a confirmation response in the AI-translated language.

[0961] Output: Operator response

[0962] Step 7:

[0963] The server then uses AI to translate the operator's response again and sends it to the user.

[0964] Input: Operator's answer

[0965] Data processing: Using AI translation, the operator's response is translated into the user's language.

[0966] Output: The translated answer is sent to the user.

[0967] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0968] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0969] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0970] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0971] [Second embodiment]

[0972] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0973] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0974] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0975] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0976] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0977] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0978] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0979] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0980] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0981] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0982] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0983] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0984] "Example 1"

[0985] This invention is a system that uses an AI chatbot to answer user questions, support online purchases, and use the logs to improve the website. Specifically, it accepts user inquiries and purchase requests and provides appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[0986] "Example 2"

[0987] Furthermore, this embodiment of the present invention can also make the above system multilingual. Specifically, by changing the settings of the AI ​​chatbot or translation AI, it can also support languages ​​other than English. For example, it can also handle inquiries from users in French, Spanish, and other languages.

[0988] "Example 3"

[0989] Furthermore, this embodiment of the present invention is designed to function even when it is difficult to secure human resources. Specifically, an AI chatbot automatically processes inquiries and escalates them to a human operator as necessary. This reduces the burden on operators while allowing them to continue providing support to users.

[0990] The processing flow of each embodiment will be described below.

[0991] "Example 1"

[0992] Step 1: The AI ​​chatbot accepts user inquiries and purchase requests.

[0993] Step 2: The AI ​​chatbot provides the appropriate answer or action.

[0994] Step 3: If the bot can't resolve the issue, prompt the user to submit the form in English.

[0995] Step 4: The AI ​​translates the post into Japanese and an operator verifies and replies.

[0996] Step 5: The AI ​​translates the answer back into English and sends it to the user.

[0997] "Example 2"

[0998] Step 1: Change the settings of your AI chatbot and translation AI to support languages ​​other than English.

[0999] Step 2: For example, respond to inquiries from users in French or Spanish.

[1000] "Example 3"

[1001] Step 1: The AI ​​chatbot automatically handles the inquiry.

[1002] Step 2: Escalate to an operator if necessary.

[1003] Step 3: This reduces the burden on operators while still providing support to users.

[1004] Example 1

[1005] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1006] Conventional online purchasing support systems sometimes fail to provide prompt and appropriate answers to user questions, and a lack of operator resources is a problem, especially when multilingual support is required. Additionally, there was a lack of a way to effectively log user inquiries and use them to improve the site.

[1007] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1008] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming and responding in AI-translated Japanese, and the AI ​​translating the answer into English and sending it, a means for the AI ​​chatbot to generate answers to users' inquiries using a generative AI model, a means for generating prompts for the generative AI model to obtain appropriate answers, and a means for making the flow multilingual. This enables quick and appropriate answers to users' questions, allows the system to function even when operator resources are insufficient, and enables the site to be improved by effectively utilizing the logs.

[1009] An "artificial intelligence chatbot" is a program that automatically generates answers to users' questions and supports online purchases.

[1010] A "generative artificial intelligence model" is an artificial intelligence model used to generate appropriate responses to user queries.

[1011] A "prompt sentence" is text that is input to a generative artificial intelligence model to generate an appropriate answer.

[1012] "Form submission" is a way for users to enter and submit questions in English that the bot cannot solve.

[1013] An "operator" is a human representative who checks the Japanese content translated by artificial intelligence in response to user inquiries and provides appropriate answers.

[1014] "Artificial intelligence translation" is a technology that automatically translates user inquiries and operator responses into different languages.

[1015] "Multilingual support" means that the system has the ability to respond to user inquiries in multiple languages.

[1016] A "log" is data that records user inquiries and system responses and is used for later analysis and site improvement.

[1017] This invention is a system that answers users' questions, supports online purchases, and uses logs to improve the website. Specifically, it uses an AI chatbot to accept inquiries and purchase requests from users and provide appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[1018] Hardware and software used

[1019] Server: Hosts the data processing and AI models, specifically using virtual servers from a cloud service provider.

[1020] Device: The device from which the user accesses the site (e.g., PC, smartphone, tablet, etc.).

[1021] Generative AI model: An artificial intelligence model used to generate answers to user queries, specifically using natural language processing models.

[1022] Translation software: Software for translating between English and Japanese, specifically using a cloud-based translation API.

[1023] Explanation of program processing

[1024] 1. Receiving user inquiries

[1025] A user uses a device to enter a query into a website chatbot.

[1026] The terminal sends this input to the server.

[1027] The server receives the query and records it as a log in the database.

[1028] Specific operation: A user types "Do you have this item in stock?" into the chatbot. The device sends this message to the server, which records it in the database as "User ID: 1234, Inquiry: Do you have this item in stock?"

[1029] 2. Answer generation by AI chatbot

[1030] The server sends the received query to the generative AI model.

[1031] The generative AI model generates an appropriate answer and sends it back to the server.

[1032] The server generates the answer and sends it to the user's device.

[1033] Specific operation: The server sends the query "Do you have this item in stock?" to the generative AI model. The generative AI model generates the answer "In stock" and sends it back to the server. The server sends this answer to the user's device, and the user sees "In stock" on the chatbot's screen.

[1034] 3. Processing Purchase Requests

[1035] The user enters a purchase request.

[1036] The terminal sends this request to the server.

[1037] The server receives the request, checks inventory, and processes the order.

[1038] If necessary, the generative AI model provides additional information to the user.

[1039] How it works: The user types, "I want to buy this product." The device sends this request to the server, which checks the inventory. If the item is in stock, the server notifies the user via the generative AI model, saying, "Your order has been accepted."

[1040] 4. Escalation of Inquiries

[1041] If a query arises that the bot cannot resolve, the server prompts the user to submit a form in English.

[1042] A user submits a form in English.

[1043] The server sends the content to translation software, which translates it into Japanese.

[1044] What happens: The bot tells the user, "We can't solve this problem. Please submit the form in English." The user submits the form in English, saying, "I need help with my order." The server sends this to translation software, which translates it into Japanese as, "I need help with my order."

[1045] 5. Confirmation by the operator

[1046] The operator will review the translated inquiry and enter the appropriate response in Japanese.

[1047] The server sends this response back to the translation software, which translates it into English.

[1048] The server sends the translated response to the user's device.

[1049] What happens: The operator sees the inquiry "I need help with my order" and types "We have it in stock. We will process your order" in Japanese. The server sends this response to translation software, which translates it into English as "The item is in stock. We will process your order." The server then sends this translated response to the user's device.

[1050] 6. Leveraging logs

[1051] The server keeps a log of all queries and responses.

[1052] The server analyzes these logs and generates insights for improving the site.

[1053] Specific operation: The server saves logs such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock." These logs are periodically analyzed to generate insights, such as "There are many inquiries about a particular product," which can be used to improve the site.

[1054] Prompt Sentence Examples

[1055] User asks: "Do you have this item in stock?"

[1056] Prompt for generative AI model: "A user is asking, 'Do you have this item in stock?' Please generate an appropriate answer."

[1057] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1058] Step 1: Receiving user inquiries

[1059] Input: A user uses a device to type a query into a website chatbot.

[1060] What happens: A user types into the chatbot, "Do you have this item in stock?"

[1061] Data processing: The device sends this input to the server.

[1062] Output: The server receives the query and logs it in the database.

[1063] Specific operation: The server records "User ID: 1234, Enquiry: Is this item in stock?"

[1064] Step 2: Generate answers with an AI chatbot

[1065] Input: The query received by the server.

[1066] What happens: The server sends the query "Do you have this item in stock?" to the generative AI model.

[1067] Data computation: Generative AI models generate appropriate answers.

[1068] Output: The generative AI model generates the answer "In stock" and sends it back to the server.

[1069] Specific behavior: The server sends this response to the user's device, and the user sees the message "In stock" on the chatbot's screen.

[1070] Step 3: Processing the purchase request

[1071] Input: The user enters a purchase request.

[1072] What happens: The user types, "I want to buy this product."

[1073] Data processing: The device sends this request to the server.

[1074] Output: The server receives the request, checks inventory, and processes the order.

[1075] Specific operation: The server checks the inventory, and if the item is in stock, it notifies the user through the generative AI model that "your order has been accepted."

[1076] Step 4: Escalate the case

[1077] Input: The query the bot cannot resolve.

[1078] What happens: The bot informs the user, "We can't solve this problem. Please submit the form in English."

[1079] Data processing: User submits form in English.

[1080] Output: The server sends the content to translation software, which translates it into Japanese.

[1081] What happens: The server translates "I need help with my order" to "I need help with my order."

[1082] Step 5: Confirmation by operator

[1083] Input: The translated query.

[1084] Specific operation: The operator confirms the inquiry "I need help with my order" and types in Japanese "We have the item in stock. We will take your order."

[1085] Data calculation: The server sends this response back to the translation software, which translates it into English.

[1086] Output: Sends the translated answer to the user's device.

[1087] Specific behavior: The server translates the message into English as "The item is in stock. We will process your order" and sends it to the user.

[1088] Step 6: Leverage the logs

[1089] Input: A log of all queries and responses.

[1090] Specific operation: The server saves a log such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock."

[1091] Data calculations: Our servers analyze these logs and generate insights for improving the site.

[1092] Output: Insights for improving your site.

[1093] Specific operation: The server generates insights such as "there are many inquiries about a particular product" and uses this insight to improve the site.

[1094] (Application example 1)

[1095] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1096] Conventional online purchasing support systems sometimes failed to provide prompt and appropriate answers to user questions, and operator resources were often insufficient, especially when multilingual support was required. Furthermore, there was a lack of a way to effectively utilize user inquiry logs to improve the site. This resulted in a poor user experience and reduced willingness to purchase.

[1097] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1098] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming the response in Japanese translated by the AI, and the AI ​​translating the response into English and sending it; a means for making the above flow multilingual; a means for users to input questions about products as an application installed on a smartphone and the AI ​​chatbot providing an immediate response; a means for prompting users to submit a form in English for questions that cannot be resolved and translating the form into Japanese by the AI; and a means for an operator confirming the response and the AI ​​translating the response back into English and sending it to the user. This enables quick and appropriate responses to users' questions, facilitates multilingual support, compensates for a lack of operator resources, and improves the user experience.

[1099] "User" refers to the consumer or user making an online purchase.

[1100] "Questions" refer to questions or concerns users have about products or services.

[1101] "Online purchasing" refers to the act of purchasing goods or services over the Internet.

[1102] An "artificial intelligence chatbot" is a program that automatically provides answers to users' questions.

[1103] "Logs" refer to records of user inquiries and interactions with chatbots.

[1104] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1105] "Bot unresolved" refers to a situation where the chatbot is unable to provide an appropriate answer to the user's question.

[1106] "Form submission in English" refers to a form that allows users to enter questions or requests in English.

[1107] "Operator" refers to a human representative who responds to user inquiries.

[1108] "Artificial intelligence translation" refers to the process of using artificial intelligence to translate text into different languages.

[1109] "Multilingual" refers to providing services in multiple languages.

[1110] "Applications installed on a smartphone" refers to software programs that run on a smartphone.

[1111] "Providing immediate answers" refers to providing a quick response to a user's question.

[1112] "Encouraging form submission in English" means instructing users to enter their questions in English.

[1113] "Translating into Japanese" refers to converting English text into Japanese.

[1114] "Translate back into English and send" means translating the Japanese response into English and sending it to the user.

[1115] A system for implementing this invention includes an AI chatbot that answers users' questions, supports online purchases, and uses the results to improve the website. The system also includes a means for accepting form submissions in English when the bot cannot resolve the issue, for an operator to confirm and respond in Japanese using AI translation, and for the AI ​​to retranslate the response into English and send it back. The system also includes a means for making the above flow multilingual.

[1116] Hardware and software used

[1117] Hardware:

[1118] Smartphone

[1119] software:

[1120] Python

[1121] OpenAI API

[1122] Googletrans library

[1123] Data processing and calculation

[1124] Accepting user input:

[1125] Users input questions about products through a smartphone application, which processes the input as text data within the application.

[1126] Answered by an AI chatbot:

[1127] The server uses the OpenAI API to generate answers to the user's questions, which are then returned to the user as text data.

[1128] Translation features:

[1129] The server uses the GoogleTrans library to translate between English and Japanese. When a user submits a form in English, the text is translated into Japanese and sent to the operator. The operator's response is also translated from Japanese to English and sent to the user.

[1130] Operator Intervention:

[1131] Any questions that the chatbot cannot solve are forwarded to an operator, who types the answer in Japanese, which is then translated back into English and sent to the user.

[1132] Specific examples

[1133] Example 1:

[1134] A user asks, "What size is this item?" The AI ​​chatbot immediately replies, "This item is a size medium."

[1135] Example 2:

[1136] If a user asks, "Can this product be shipped internationally?" and the chatbot cannot resolve the question, it prompts the user to submit a form in English. The user types, "Can this product be shipped internationally?", and the AI ​​translates it into Japanese as, "Can this product be shipped internationally?" The operator replies, "Yes, it can be shipped internationally," and the AI ​​translates it into English as, "Yes, it can be shipped internationally," and sends it to the user.

[1137] Prompt Sentence Examples

[1138] User: What size is this item?

[1139] AI Chatbot: This product is size M.

[1140] User: Can this item be shipped internationally?

[1141] AI Chatbot: Unable to resolve. Please submit the form in English.

[1142] User: Can this product be shipped internationally?

[1143] AI Translation: Can this item be shipped internationally?

[1144] Operator: Yes, you can.

[1145] AI Translation: Yes, it can be shipped internationally.

[1146] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1147] Step 1:

[1148] A user inputs a question about a product through a smartphone application. The input text data is processed within the application and sent to the server.

[1149] Input: User question text

[1150] Output: Text data sent to the server

[1151] Step 2:

[1152] The server uses the OpenAI API to generate answers to user questions, and the generated answers are stored as text data on the server.

[1153] Input: User question text

[1154] Data processing: Generate answers to questions using OpenAI APIs

[1155] Output: Generated answer text

[1156] Step 3:

[1157] The server sends the generated answer to the user's smartphone application, through which the user can confirm the answer.

[1158] Input: Generated answer text

[1159] Output: Answer text displayed in the user's smartphone application

[1160] Step 4:

[1161] If the user is not satisfied with the chatbot's answer, a message prompting them to submit a form in English is displayed. The user can then enter their question in English and submit it to the server.

[1162] Input: User's question text in English

[1163] Output: English question text sent to the server

[1164] Step 5:

[1165] The server uses the GoogleTrans library to translate the user's English question into Japanese, and the translated text is sent to the operator.

[1166] Input: User's question text in English

[1167] Data processing: Translated from English to Japanese using the Googletrans library

[1168] Output: Question text in Japanese sent to the operator

[1169] Step 6:

[1170] The operator inputs the answer in Japanese and sends it to the server, which receives the answer.

[1171] Input: Operator's response text in Japanese

[1172] Output: Answer text in Japanese sent to the server

[1173] Step 7:

[1174] The server uses the GoogleTrans library to translate the operator's Japanese response into English, and the translated text is sent to the user's smartphone application.

[1175] Input: Operator's Japanese response text

[1176] Data processing: Translate from Japanese to English using the Googletrans library

[1177] Output: Answer text in English that will be displayed on the user's smartphone application

[1178] Step 8:

[1179] The user can check the English response from the operator through the smartphone application.

[1180] Input: Answer text in English

[1181] Output: Answer text in English that will be displayed on the user's smartphone application

[1182] Example 2

[1183] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1184] Conventional AI chatbot systems only support certain languages, making it difficult to support multiple languages. Furthermore, the process for generating appropriate responses to user inquiries is complicated, making it difficult to respond when there are insufficient operator resources. This leads to a poor user experience and insufficient log collection for site improvement.

[1185] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1186] In this invention, the server includes a means for receiving user input data and determining the language, a means for changing the AI ​​chatbot's settings based on the determined language, and a means for generating an appropriate response using a generative AI model based on the set language. This enables multilingual support and provides prompt and appropriate responses to user inquiries. Furthermore, even if there are insufficient operator resources, the system automatically responds, enabling log collection for improving the user experience and site improvement.

[1187] An "artificial intelligence chatbot" is a program that automatically generates responses to users' questions and assists them in online purchases.

[1188] "Artificial intelligence translation" is a technology that automatically translates text between different languages.

[1189] A "generative AI model" is an artificial intelligence algorithm that generates appropriate responses based on user input.

[1190] "Language determination" is the process of automatically identifying the language of text entered by a user.

[1191] "Reconfiguration" is the process of adjusting the system's operating modes and parameters based on the determined language.

[1192] "Response generation" is the process of generating an appropriate reply to a user's input.

[1193] "Multilingual" refers to the system's ability to respond to user inquiries in several different languages.

[1194] An "operator" is a person whose role is to respond to user inquiries.

[1195] "Resource allocation" refers to preparing the personnel and equipment necessary for the system to operate normally.

[1196] "User experience" is a general term for the satisfaction and ease of use that users feel when using a system.

[1197] "Site Improvement" is the process of improving the functionality and design of a website based on user usage data and feedback.

[1198] MODE FOR CARRYING OUT THE INVENTION

[1199] This invention relates to a multilingual AI chatbot system. A specific embodiment of this system is described below.

[1200] System configuration

[1201] The system consists of three main components: the user, the terminal, and the server. The user makes a query through the terminal, and the terminal sends the data to the server. The server processes the received data and generates an appropriate response to send to the terminal.

[1202] Hardware and software used

[1203] Hardware: Terminals include devices such as PCs, smartphones, and tablets. Servers are high-performance computer systems, including cloud servers.

[1204] Software: The server uses a generative AI model (e.g., GPT-4) to generate the response, and translation software such as Google Cloud Translation API to determine the language.

[1205] Program processing

[1206] When the server receives a user inquiry, it first determines the language. It uses the Google Cloud Translation API to determine the language. Based on the determined language, the server changes the AI ​​chatbot's settings. For example, if French is detected, the settings are changed to support French.

[1207] The server then uses the generative AI model to generate an appropriate response, which is translated into the user's language using translation software if further translation is required, and then sends the response to the device, which displays it to the user.

[1208] Specific examples

[1209] For example, consider a case where a user makes a query in French: "Bonjour, comment puis-je vous aider?" The device sends this input to the server. The server uses the Google Cloud Translation API to determine that the input is in French. The server then changes the AI ​​chatbot's settings to support French and uses the generative AI model to generate the response: "Bonjour! Comment puis-je vous aider today?" The generated response is sent to the device, which displays it to the user.

[1210] Prompt Sentence Examples

[1211] "Generate an appropriate response to a query in French. The user's input is 'Bonjour, comment puis-je vous aider?'"

[1212] In this way, the system can process queries in multiple languages, improving the user experience.

[1213] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1214] Step 1:

[1215] The user enters a query.

[1216] The user inputs a query into the input field of the terminal. For example, the user inputs "Bonjour, comment puis-je vous aider?" The input data is saved in text format on the terminal.

[1217] Step 2:

[1218] The terminal sends the input data to the server.

[1219] The terminal converts the data entered by the user into JSON format and sends it to the server using the HTTPS protocol. The input data is the text-formatted query, and the output is the JSON-formatted data sent to the server.

[1220] Step 3:

[1221] The server determines the language.

[1222] The server calls the Google Cloud Translation API to determine the language of the received data. The API parses the input data and returns a language code (e.g., "fr" for French). The input is the user's query, and the output is the determined language code.

[1223] Step 4:

[1224] The server changes the settings of the AI ​​chatbot.

[1225] The server changes the AI ​​chatbot's settings based on the determined language code. For example, if the language code is "fr", it changes the settings to support French. The input is the language code, and the output is the changed chatbot settings.

[1226] Step 5:

[1227] The server generates a response.

[1228] The server generates an appropriate response using a generative AI model (e.g., GPT-4) based on the set language. The prompt text is "Generate an appropriate response to a question in French. The user's input is 'Bonjour, comment puis-je vous aider?'." The input is the prompt text and the user's question, and the output is the generated response.

[1229] Step 6:

[1230] The server generates a response and sends it to the terminal.

[1231] The server converts the generated response into JSON format and sends it to the terminal using the HTTPS protocol. The input is the generated response, and the output is the JSON format data sent to the terminal.

[1232] Step 7:

[1233] The terminal displays the response to the user.

[1234] The terminal displays the response received from the server to the user, for example, "Bonjour! Comment puis-je vous aider aujourd'hui?" The input is the response data received from the server, and the output is the text displayed to the user.

[1235] (Application example 2)

[1236] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1237] Conventional online purchasing support systems have limited support for user questions and insufficient multilingual support, making it difficult to provide adequate service to international users. Furthermore, limited operator resources can result in delayed responses, which can lead to a decline in user satisfaction. Furthermore, real-time translation and response generation are difficult, creating a need for an improved user experience.

[1238] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1239] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, having an operator confirm and respond in an AI-translated language, and having the AI ​​translate and send the response; a means for making the above flow multilingual; a means for translating user inquiries in real time, generating responses using a generative AI model, and translating the responses back into the original language; and a means including an application to be installed on a smartphone. This enables quick and appropriate responses even to international users, and is expected to improve user satisfaction.

[1240] "User" means any person or legal entity making an online purchase.

[1241] "Questions" are questions or concerns users have when making an online purchase.

[1242] "Online purchasing" is the act of purchasing goods or services over the Internet.

[1243] An "artificial intelligence chatbot" is a program that automatically responds to users' questions.

[1244] "Logs" are data that record user activities and inquiries.

[1245] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1246] "Bot Unresolved" refers to when an AI chatbot is unable to provide an appropriate answer to a user's question.

[1247] "Multilingual" refers to multiple different languages, especially languages ​​other than English.

[1248] "Form submission" means the act of a user submitting an inquiry or information through a web form.

[1249] An "operator" is a human representative who responds to user inquiries.

[1250] "Artificial intelligence translation" is the process of converting text from one language to another using artificial intelligence.

[1251] A "verification response" is a formal response provided by an operator to a user's inquiry.

[1252] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user inquiries.

[1253] "Real-time" refers to processing occurring immediately without delay.

[1254] A "smartphone" is a mobile phone that can connect to the Internet and install applications.

[1255] An "application" is a software program designed to perform a specific function.

[1256] A system for implementing this invention includes an artificial intelligence chatbot that answers user questions and assists with online purchasing. The system translates user inquiries in real time, generates responses using a generative AI model, and translates the responses back into the user's native language. The system also includes an application installed on a smartphone.

[1257] Hardware and Software Configuration

[1258] Hardware: Smartphone (iOS or Android)

[1259] Software: OpenAI API (generative AI model), GoogleTrans (translation library)

[1260] Data processing and calculation

[1261] 1. Receive user inquiries:

[1262] A user makes a query through a smartphone application. For example, the user asks in French, "Bonjour, pouvez-vous me recommander un bon livre?"

[1263] 2. Enquiry Translation:

[1264] The server uses GoogleTrans to translate the user's query into English, for example, "Bonjour, pouvez-vous me recommander un bon livre?" to "Hello, can you recommend a good book?"

[1265] 3. Generate a response:

[1266] The server uses OpenAI's generative AI model to generate responses to English queries, such as "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[1267] 4. Retranslating responses:

[1268] The server translates the generated response into the original language (French) using Google Translate. For example, "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald." is translated to "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[1269] 5. Response to the user:

[1270] The translated response is returned to the user, who receives the response in French through a smartphone application.

[1271] Specific examples

[1272] User inquiry: “Bonjour, pouvez-vous me recommander un bon livre?”

[1273] Prompt the generative AI model: "Hello, can you recommend a good book?"

[1274] Generated response: "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[1275] Translated response: "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[1276] In this way, the system can provide multilingual online purchasing support, enabling fast and appropriate responses for international users.

[1277] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1278] Step 1:

[1279] The user makes an inquiry through a smartphone application.

[1280] Input: User's question (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[1281] Output: User query text

[1282] Specific behavior: The user enters a question into the application's input field and presses the submit button.

[1283] Step 2:

[1284] The server uses GoogleTrans to translate the user's query into English.

[1285] Input: User's query text (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[1286] Output: Text translated into English (e.g. "Hello, can you recommend a good book?")

[1287] What happens: The server calls the GoogleTrans API to translate the French text into English.

[1288] Step 3:

[1289] The server uses OpenAI's generative AI model to generate responses to English queries.

[1290] Input: Text translated into English (e.g. "Hello, can you recommend a good book?")

[1291] Output: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[1292] What happens: The server calls the OpenAI API, sends a prompt to the generative AI model, and generates a response.

[1293] Step 4:

[1294] The server translates the generated response back into the original language (French) using Google Translate.

[1295] Input: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[1296] Output: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[1297] What happens: The server calls the GoogleTrans API and translates the English response text into French.

[1298] Step 5:

[1299] The server returns the translated response to the user.

[1300] Input: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[1301] Output: Response text displayed on the user's smartphone

[1302] What happens: The server sends the translated response text to the application, which displays it on the user's smartphone.

[1303] Example 3

[1304] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1305] The previous system had issues such as delayed responses to user inquiries and difficulty in responding when operator resources were insufficient. It also lacked multilingual support, limiting support for global users. This resulted in issues such as a decline in user satisfaction and insufficient log collection for site improvement.

[1306] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1307] In this invention, the server includes a means for the automated response device to generate a response to an inquiry using a generative AI model, a means for the automated response device to determine whether escalation is necessary based on the content and complexity of the inquiry, and a means for sending a notification to an operator when it is determined that escalation is necessary. This allows for a prompt and appropriate response to inquiries from users, enables the system to function even when operator resources are insufficient, and enables multilingual support.

[1308] An "automatic answering machine" is a device that automatically generates and provides responses to inquiries from users.

[1309] A "generative AI model" is a model that uses artificial intelligence to generate appropriate responses from text data.

[1310] An "inquiry" is a question or request that a user enters into the system.

[1311] "Escalation" is the process of handing over a response to an operator when it is determined that a more advanced response is required based on the content and complexity of the inquiry.

[1312] "Operator" means a person who directly responds to inquiries from users and provides necessary support.

[1313] "Notification" is information sent from the system to the operator to inform them when escalation is necessary.

[1314] "Multilingual support" refers to the ability of the system to respond to user inquiries in multiple languages.

[1315] The present invention is a system that automatically generates a response to an inquiry from a user and escalates the inquiry to an operator as necessary. A specific embodiment of this system is described below.

[1316] First, a user inputs a query into the system using a device. The device then sends the query to a server. The server then passes the received query to a generative AI model, which generates an appropriate response. For example, OpenAI's GPT-4 can be used as this generative AI model.

[1317] The server evaluates the generated response and determines whether escalation is necessary based on the nature and complexity of the inquiry. If escalation is necessary, the server notifies an operator, who then responds directly to the user's inquiry and provides the necessary support.

[1318] As a concrete example, the following prompt sentence can be input to a generative AI model:

[1319] Example prompt sentence:

[1320] User: Forgot password.

[1321] AI Chatbot: We've sent you a link to reset your password. Please check your email.

[1322] Example prompt sentence:

[1323] User: I think my account has been hacked.

[1324] AI Chatbot: This issue has been escalated to an operator. Please wait a moment.

[1325] The system is able to respond to user inquiries promptly and appropriately, and is designed to continue functioning even when operator resources are insufficient. It also supports multiple languages, enabling it to provide support to users globally.

[1326] The hardware used includes the user's device (PC, smartphone, etc.) and the server, while the software used includes a generative AI model (e.g., GPT-4) and an algorithm for evaluating the content of the query.

[1327] In this way, a system is realized in which the user, the terminal, and the server work together to process inquiries efficiently. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1328] Step 1:

[1329] A user uses a terminal to enter a query into the system, which then transmits the query to the server.

[1330] Type: The user types "I forgot my password" into the chat window on the device.

[1331] Data processing: The terminal formats this text data for transmission to the server.

[1332] Output: The formatted query data is sent to the server.

[1333] Specific operation: The user enters text into the chat window on the device and presses the send button. The device sends the entered text to the server.

[1334] Step 2:

[1335] The server passes the received query to a generative AI model, such as OpenAI's GPT-4, to generate an appropriate response.

[1336] Input: The query data received by the server.

[1337] Data processing: The server converts the query data into the format required to input it into the generative AI model.

[1338] Output: Response data from the generative AI model.

[1339] Specific operation: The server inputs the user's query "I forgot my password" into GPT-4 and generates a response. GPT-4 then generates a response such as "We've sent you a link to reset your password. Please check your email."

[1340] Step 3:

[1341] The server evaluates the generated response and determines whether escalation is necessary based on the nature and complexity of the query.

[1342] Input: Response data from the generative AI model and the original query data.

[1343] Data processing: The server compares the response data with the query data and applies an algorithm to determine if escalation is necessary.

[1344] Output: The result of the decision as to whether escalation is required.

[1345] What happens: The server compares the generated response with the query and determines whether an escalation is necessary. For example, if the query is "I think my account has been hacked," the server determines that an escalation is necessary.

[1346] Step 4:

[1347] The server will send a notification to the operator if it determines that escalation is necessary.

[1348] Input: The result of the decision that escalation is necessary.

[1349] Data processing: The server generates a message to send a notification to the operator.

[1350] Output: Informational message sent to the operator.

[1351] Specific operation: The server displays a notification on the operator's management screen. The operator checks this notification and prepares to respond to the user.

[1352] Step 5:

[1353] Operators respond directly to user inquiries and provide the necessary support.

[1354] Input: Notification message sent to operator and original inquiry data.

[1355] Data processing: The operator collects information to respond appropriately to user inquiries and then carries out the response.

[1356] Output: The support provided to the user.

[1357] Specific Actions: The operator will call the user and provide further assistance, for example, checking the security settings of the user's account and taking any necessary measures.

[1358] (Application example 3)

[1359] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1360] Conventional online purchasing support systems have difficulty in quickly and accurately responding to user inquiries, and there have been problems with the quality of user support declining, especially when it is difficult to provide multilingual support or secure operator resources. Furthermore, there is a lack of a mechanism for appropriately escalating inquiries depending on the content of the inquiry, making it difficult to improve user satisfaction.

[1361] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1362] In this invention, the server includes an AI chatbot means for answering user questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, having an operator confirm and respond in an AI-translated language, and having the AI ​​translate and send the response; a means for the AI ​​chatbot to receive inquiries from users, generate prompt sentences using a generative AI model, and generate responses; a means for determining whether escalation is necessary based on the content of the responses and escalating to an operator if necessary; and a means for making the above flow multilingual. This allows for quick and accurate responses to user questions, enables the system to function even when operator resources are difficult to secure, and improves the quality of user support.

[1363] "User" refers to any person or entity making an online purchase.

[1364] "Questions" refer to questions or concerns users have about online purchases.

[1365] "Online purchasing" refers to the act of purchasing goods or services over the Internet.

[1366] "Support" means assisting users in making a successful online purchase.

[1367] "Log" refers to records of user inquiries, system responses, etc.

[1368] "Site Improvement" refers to improving the functionality and design of a website to enhance user convenience and satisfaction.

[1369] An "AI chatbot" is a program that uses artificial intelligence to automatically respond to user inquiries.

[1370] "Bot unresolved" refers to when an AI chatbot is unable to generate an appropriate response to a user's inquiry.

[1371] "Multilingual" refers to several different languages.

[1372] "Form submission" refers to a web form that allows users to enter and submit their inquiry.

[1373] "Operator" refers to a human representative who responds to user inquiries.

[1374] "AI translation" refers to the use of artificial intelligence to translate text between different languages.

[1375] A "generative AI model" refers to an artificial intelligence model that generates appropriate responses to user inquiries.

[1376] A "prompt sentence" refers to the text that is input to a generative AI model.

[1377] "Response" refers to the reply given by an AI chatbot or operator to a user's inquiry.

[1378] "Escalation" refers to handing over inquiries that the AI ​​chatbot cannot handle to an operator.

[1379] "Flow" refers to a series of processing steps from a user's inquiry to a response.

[1380] A system for implementing this invention includes an AI chatbot that answers user questions, supports online purchases, and uses the results to improve the site. The system includes a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, for an operator to confirm and respond in an AI-translated language, and for the AI ​​to translate and send the response. The system also includes a means for the AI ​​chatbot to receive inquiries from users, generate prompts using a generative AI model, and generate responses. The system also includes a means for determining whether escalation is necessary based on the response content and escalating to an operator if necessary. The above flow supports multiple languages.

[1381] Hardware and software used

[1382] Hardware:

[1383] Server (cloud server is also acceptable)

[1384] User device (smartphone, PC, etc.)

[1385] software:

[1386] Python

[1387] Flask

[1388] OpenAI API

[1389] Data processing and calculation

[1390] server:

[1391] The server receives inquiries from users. The inquiries are sent in JSON format, which the server parses. It then uses the OpenAI API to send prompts to a generative AI model to generate a response. The generated response determines whether escalation is necessary before being returned to the user. If escalation is necessary, the inquiry is handed over to an operator. The operator enters a response, which is then translated again by the AI ​​and sent to the user.

[1392] User device:

[1393] The user's device inputs a query and sends it to the server. The user receives a response from the AI ​​chatbot and can make additional queries if necessary.

[1394] Specific examples

[1395] When a user asks, "Do you have this item in stock?", the server receives the query and sends a prompt to the generative AI model. The generative AI model generates a response, "We currently have it in stock," and the server returns this to the user.

[1396] When a user inquires, "I would like to return an item," the server similarly sends a prompt to the generative AI model, which generates a response saying, "We will begin the return process."

[1397] When a user inquires, "This problem cannot be solved," the server determines that escalation is necessary and hands the inquiry over to an operator. The operator responds, and the AI ​​translates and responds to the user.

[1398] Example prompt sentence:

[1399] User: Do you have this item in stock?

[1400] AI Chatbot: Currently in stock.

[1401] User: I'd like to return it.

[1402] AI Chatbot: Initiate the return process.

[1403] User: I can't solve this problem.

[1404] AI Chatbot: Escalate to a human.

[1405] In this way, users' questions can be responded to quickly and accurately, the system can function even when it is difficult to secure operator resources, and the quality of user support can be improved.

[1406] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1407] Step 1:

[1408] The user inputs a query from the terminal and sends it to the server.

[1409] Input: User's query text (e.g., "Do you have this item in stock?")

[1410] Output: Query data sent to the server (JSON format)

[1411] Specific actions: The user uses a smartphone or computer to enter a query into the chatbot's interface and press the send button.

[1412] Step 2:

[1413] The server receives the user's query and sends a prompt to the generative AI model.

[1414] Input: User's query data (JSON format)

[1415] Output: The prompt sent to the generative AI model

[1416] Specific operation: The server analyzes the received query data, generates an appropriate prompt, and sends it to the OpenAI API.

[1417] Step 3:

[1418] The generative AI model generates a response based on the prompt sentence and sends it back to the server.

[1419] Input: prompt statement

[1420] Output: The generated response text (e.g., "Currently in stock")

[1421] Specific operation: The OpenAI API receives the prompt sentence, generates a response using the generative AI model, and sends the result back to the server.

[1422] Step 4:

[1423] The server receives the generated response and determines whether escalation is necessary.

[1424] Input: Generated response text

[1425] Output: Judgment result on whether escalation is necessary (e.g., no escalation required)

[1426] What happens: The server parses the generated response text and determines if escalation is necessary based on specific keywords or conditions.

[1427] Step 5:

[1428] If no escalation is required, the server sends the response back to the user.

[1429] Input: Generated response text

[1430] Output: The response text sent back to the user

[1431] Specific operation: The server sends the generated response text to the user's device, and the user confirms the response in the chatbot's interface.

[1432] Step 6:

[1433] If escalation is required, the server will hand over the query to an operator.

[1434] Input: Generated response text and user query data

[1435] Output: Query data sent to the operator

[1436] Specific behavior: If the server determines that escalation is necessary, it sends the user's inquiry data and the generated response text to an operator.

[1437] Step 7:

[1438] The operator enters a response to the inquiry and sends it to the server.

[1439] Input: Operator response text (e.g., "This item is currently out of stock")

[1440] Output: Operator response data sent to the server

[1441] Specific operation: The operator uses a dedicated interface to enter answers to user inquiries and send them to the server.

[1442] Step 8:

[1443] The server receives the operator's response, performs AI translation if necessary, and returns it to the user.

[1444] Input: Operator response data

[1445] Output: The translated answer text that is sent back to the user

[1446] Specific operation: The server receives the operator's answer data, performs AI translation if necessary, and sends the translated answer text to the user's device.

[1447] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1448] "Example 1"

[1449] In one embodiment of the present invention, an AI chatbot includes an emotion engine that recognizes a user's emotions. The emotion engine analyzes emotions from the user's text and voice input and provides feedback to the AI ​​chatbot. For example, if the user uses words that indicate anger or frustration, the emotion engine recognizes this and notifies the AI ​​chatbot of the user's emotional state. This allows the AI ​​chatbot to adjust its response based on the user's emotions.

[1450] "Example 2"

[1451] The emotion engine also provides information for site improvement based on user sentiment. For example, if many users express dissatisfaction with a particular process or function, the emotion engine will detect this and provide feedback to the site operator for improvement. This will enable a better user experience and improve the site.

[1452] "Example 3"

[1453] Furthermore, the emotion engine can track user emotions over time and analyze their changes. This allows us to understand the emotional changes experienced by users as they use the site and improve our services accordingly. For example, if a user's emotions suddenly worsen during a purchase process, we can identify the cause and take measures to resolve the issue.

[1454] The processing flow of each embodiment will be described below.

[1455] "Example 1"

[1456] Step 1: User sends a message to the AI ​​chatbot via text or voice.

[1457] Step 2: The emotion engine analyzes the user's message and recognizes its emotion.

[1458] Step 3: The emotion engine provides feedback on the emotional information it recognizes to the AI ​​chatbot.

[1459] Step 4: The AI ​​chatbot responds according to the user's emotions based on feedback from the emotion engine.

[1460] "Example 2"

[1461] Step 1: The emotion engine analyzes the user's emotions and provides feedback to the site operator.

[1462] Step 2: The site owner improves the site based on feedback from the emotion engine.

[1463] "Example 3"

[1464] Step 1: The emotion engine tracks the user's emotions over time.

[1465] Step 2: The emotion engine analyzes changes in the user's emotions.

[1466] Step 3: The emotion engine suggests service improvements based on changes in emotions.

[1467] Example 1

[1468] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1469] Traditional online purchasing support systems often fail to provide appropriate answers to users' questions or respond appropriately to their emotions. Furthermore, insufficient multilingual support and a lack of operator resources can lead to poor system functionality. This leads to a poor user experience and delays in site improvements.

[1470] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1471] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming and responding in Japanese using AI translation, and the AI ​​translating the response into English and sending it, a means for making the above flow multilingual, an emotion analysis means for analyzing emotions from the user's text input or voice input and feeding the results back to the AI ​​chatbot, and a means for adjusting responses based on the emotion analysis results. This enables appropriate answers to users' questions and responses based on their emotions, and multilingual support is also possible, so system functionality can be maintained even when operator resources are insufficient, improving user experience and improving the site.

[1472] An "artificial intelligence chatbot" is a program that automatically generates answers to users' questions and supports online purchases.

[1473] "Artificial intelligence translation" is a technology that automatically translates text entered in one language into another language.

[1474] "Emotion analysis" is a technology that analyzes emotions from a user's text or voice input and provides feedback on the results.

[1475] "Multilingual support" refers to having functionality that supports multiple languages ​​and providing appropriate services to users who speak different languages.

[1476] An "operator" is a human representative who responds to inquiries from users.

[1477] A "log" is a record of a user's operation history and inquiries recorded by the system.

[1478] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1479] "Text input" refers to the act of a user entering text information using a keyboard or touchscreen.

[1480] "Voice input" refers to the act of a user inputting voice information using a microphone.

[1481] "Feedback" refers to returning information to the system or user based on analysis results and evaluations.

[1482] MODE FOR CARRYING OUT THE INVENTION

[1483] This invention is a system that answers users' questions, supports online purchases, and uses the logs to improve the website. A specific embodiment of this system is described below.

[1484] System configuration

[1485] The system consists of the following main components:

[1486] 1. Artificial Intelligence Chatbots

[1487] 2. Artificial Intelligence Translation Engine

[1488] 3. Sentiment Analysis Engine

[1489] 4. Operator Interface

[1490] 5. User Device

[1491] Hardware and software used

[1492] Server: Hardware that manages the entire system and runs various engines. For example, a cloud-based server or an on-premise server is used.

[1493] User terminal: A device through which a user accesses the system, including a PC, smartphone, tablet, etc.

[1494] Artificial Intelligence Chatbot: Software that automatically generates answers to user inquiries.

[1495] Artificial intelligence translation engine: Software that translates user inquiries and operator responses into multiple languages.

[1496] Sentiment analysis engine: Software that analyzes emotions from a user's text or voice input.

[1497] Operator interface: An interface through which an operator can provide confirmation and response to a user's inquiry.

[1498] System Operation

[1499] The user uses a device to make a text or voice inquiry to the chatbot. The device then sends the user's input to a server. The server then passes the received user's inquiry to an AI chatbot, which then analyzes the inquiry and generates an appropriate answer. The generated answer is then sent to the device via the server and displayed to the user.

[1500] The emotion analysis engine analyzes the user's text and voice input to recognize the user's emotional state. For example, if a user types, "This service is really terrible!", the emotion analysis engine will recognize anger and feed the results back to the AI ​​chatbot. The chatbot will then adjust its response based on the emotion analysis results and provide an appropriate response to the user.

[1501] The AI ​​translation engine translates user inquiries and operator responses into multiple languages. For example, if a user makes an inquiry in English and the chatbot is unable to resolve the issue, it will prompt the user to submit a form in English. The content entered by the user into the form is sent via the server to the AI ​​translation engine and translated into Japanese. The operator checks the translated Japanese post and enters an appropriate response. The operator's response is sent again to the AI ​​translation engine, translated into English, and sent to the user.

[1502] Specific examples

[1503] For example, if a user asks, "How do I return this product?", an AI chatbot can search for information about the return process and generate a response such as, "Please see this link for the return process." If a user types, "This service is terrible!", a sentiment analysis engine will recognize anger and the chatbot will respond with, "Sorry. What was the problem?"

[1504] Prompt Sentence Examples

[1505] "If a user asks about a product, check availability and respond. And if a user expresses anger, use your emotion engine to respond appropriately."

[1506] The above is a specific embodiment for carrying out the present invention.

[1507] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1508] Step 1:

[1509] Users use their device to send text or voice queries to the chatbot.

[1510] Input: User text or voice input

[1511] Output: Query data from the terminal to the server

[1512] Specific operation: The user types "Do you have this item in stock?" The device sends this text data to the server.

[1513] Step 2:

[1514] The server passes the received user inquiry to an artificial intelligence chatbot.

[1515] Input: Inquiry data sent from the terminal

[1516] Output: Data transfer to an artificial intelligence chatbot

[1517] Specific operation: The server analyzes the user's inquiry and passes it to the artificial intelligence chatbot.

[1518] Step 3:

[1519] The AI ​​chatbot analyzes the inquiry and generates an appropriate response.

[1520] Input: Query data sent from the server

[1521] Output: Generated response data

[1522] Specific behavior: The AI ​​chatbot references the inventory database and generates the answer, "Yes, we have it in stock."

[1523] Step 4:

[1524] The server sends the generated response to the terminal.

[1525] Input: Answer data from an AI chatbot

[1526] Output: Response data to the terminal

[1527] Specific operation: The server sends the generated answer to the terminal and displays it to the user.

[1528] Step 5:

[1529] The emotion analysis engine analyzes the user's text and voice input to recognize their emotional state.

[1530] Input: User text or voice input

[1531] Output: Sentiment analysis result data

[1532] How it works: If a user types, "This service is so terrible!", the sentiment analysis engine will recognize the anger and feed it back to the AI ​​chatbot.

[1533] Step 6:

[1534] The AI ​​chatbot adjusts its response based on the results of sentiment analysis.

[1535] Input: Sentiment analysis result data

[1536] Output: Adjusted correspondence data

[1537] What it does: The AI ​​chatbot responds with something like, "Sorry. What was the problem?"

[1538] Step 7:

[1539] The server sends the adjusted response to the terminal.

[1540] Input: Tailored response data from an artificial intelligence chatbot

[1541] Output: Adjusted corresponding data to the device

[1542] Specific operation: The server sends the adjusted response to the terminal and displays it to the user.

[1543] Step 8:

[1544] If the server detects an inquiry that the AI ​​chatbot cannot resolve, it prompts the user to submit a form in English.

[1545] Input: Unresolved inquiry data from an artificial intelligence chatbot

[1546] Output: Data prompting the user to submit the form

[1547] What happens: The server prompts the user to "fill in the form with your details in English."

[1548] Step 9:

[1549] The terminal sends the information the user has entered into the form to the server.

[1550] Input: User form-entered data

[1551] Output: Form input data to the server

[1552] Specific operation: The user enters "Please tell me about the return procedure" into the form, and the device sends the data to the server.

[1553] Step 10:

[1554] The server sends the received English posts to an artificial intelligence translation engine, which translates them into Japanese.

[1555] Input: User's English post data

[1556] Output: Data translated into Japanese

[1557] Specific operation: The server sends the user's English post to an artificial intelligence translation engine, which translates "Please tell me about the return procedure" into Japanese.

[1558] Step 11:

[1559] An operator will review the translated Japanese post and enter an appropriate response.

[1560] Input: Post data translated into Japanese

[1561] Output: Operator response data

[1562] Specific actions: The operator responds by saying, "Please refer to this link for return procedures."

[1563] Step 12:

[1564] The server then sends the operator's response back to the AI ​​translation engine, where it is translated into English.

[1565] Input: Operator response data

[1566] Output: Answer data translated into English

[1567] Specific operation: The server sends the operator's response to an artificial intelligence translation engine, which translates it into English as "Please refer to this link for the return procedure."

[1568] Step 13:

[1569] The device will then send the translated English response to the user.

[1570] Input: Response data translated into English

[1571] Output: User response data

[1572] What happens: The device displays the translated English answer to the user.

[1573] (Application example 1)

[1574] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1575] Conventional online purchasing support systems struggled to respond appropriately to user questions, particularly due to insufficient multilingual support and sentiment analysis. Furthermore, when operator resources were limited, it was difficult to provide a prompt response, resulting in a poor user experience. Furthermore, the system lacked the convenience of a smartphone application.

[1576] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1577] In this invention, the server includes an AI chatbot means for answering user questions, supporting online purchases, and using the results to improve the website; a means for accepting form submissions in English when the bot is unable to resolve the issue, having an operator confirm and respond in Japanese using AI translation, and then having the AI ​​translate the response into English and send it; a means for making the above flow multilingual; a means including an emotion analysis engine for analyzing user emotions and adjusting responses accordingly; and a means for functioning as an application installed on a smartphone. This enables prompt and appropriate responses to user questions and improves the user experience through multilingual support and emotion analysis. The system also allows for continued functioning even when operator resources are limited.

[1578] An "artificial intelligence chatbot" is an automated response system designed to answer user questions and assist with online purchases.

[1579] "Artificial intelligence translation" is a technology that automatically translates text between different languages.

[1580] An "emotion analysis engine" is a system that analyzes emotions from the user's text or voice input and provides feedback on the results.

[1581] "Multilingual" refers to the ability of a system to function in multiple languages.

[1582] An "application installed on a smartphone" is a software program that runs on a smartphone.

[1583] An "operator" is a human representative who responds to inquiries from users.

[1584] "Log" refers to data on user operation history and inquiry details recorded by the system.

[1585] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1586] "Form submission" is the act of using a web form to allow a user to enter and submit certain information.

[1587] "English translation" is the act of translating Japanese text into English.

[1588] To implement the present invention, the following system configuration and program are required.

[1589] System Configuration

[1590] The server contains an AI chatbot that answers users' questions and supports online purchases. When users input questions using their smartphones, the chatbot generates responses in real time. It uses a sentiment analysis engine to analyze the user's emotions and responds accordingly. If the bot cannot resolve the issue, it prompts the user to submit a form in English, which is then translated into Japanese using AI translation. An operator verifies the response, which is then translated back into English and sent to the user. These functions are multilingual and are provided as an application to be installed on smartphones.

[1591] Hardware and software used

[1592] Hardware: Smartphone

[1593] Software: OpenAI API, Google Translate API, TextBlob

[1594] Data processing and calculation

[1595] The server receives text input from the user and generates an appropriate response using the OpenAI API. It uses TextBlob as a sentiment analysis engine to analyze the user's emotions. If the sentiment score is below a certain threshold, it determines that the user is angry and takes appropriate action. If the bot cannot resolve the issue, it uses the Google Translate API to translate the English form submission into Japanese and send it to an operator. The operator's confirmation response is then translated back into English and sent to the user.

[1596] Specific examples

[1597] For example, if a user asks, "How do I return this item?", the following process occurs:

[1598] 1. The user uses their smartphone to type, "How do I return this item?"

[1599] 2. The server parses the sentiment score using TextBlob to ensure the user is not angry.

[1600] 3. Use the OpenAI API to generate a chatbot response to "How do I return this item?"

[1601] 4. If the response does not contain "cannot be resolved", display the generated response to the user.

[1602] Prompt Sentence Examples

[1603] If the user types "How do I return this item?", the following is an example of a prompt:

[1604] User Question: How do I return this item?

[1605] Chatbot response:

[1606] This allows for quick and appropriate responses to user questions, improves the user experience through multilingual support and sentiment analysis, and keeps the system functioning even when operator resources are limited.

[1607] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1608] Step 1:

[1609] The user uses a smartphone to enter a question.

[1610] Input: User question text (e.g., "How do I return this item?")

[1611] Output: The user's question text is sent to the server.

[1612] Specific actions: The user opens the application on their smartphone, enters a question, and presses the send button.

[1613] Step 2:

[1614] The server receives the user's question text and analyzes the sentiment score using a sentiment analysis engine.

[1615] Input: User question text

[1616] Output: Sentiment score (e.g. -0.2)

[1617] What happens: The server uses TextBlob to parse the user's question text and calculates a sentiment score.

[1618] Step 3:

[1619] The server evaluates the emotion score and determines whether the user is angry.

[1620] Input: Sentiment score

[1621] Output: User's emotional state (e.g., not angry)

[1622] Specific behavior: The server checks whether the emotion score is below a certain threshold (e.g., -0.5) and determines that the user is not angry.

[1623] Step 4:

[1624] The server uses the OpenAI API to generate chatbot responses to user questions.

[1625] Input: User question text

[1626] Output: Chatbot response text (e.g., "Here's how to return your order...")

[1627] Specific operation: The server sends the user's question text to the OpenAI API and receives the generated response text.

[1628] Step 5:

[1629] The server sends the chatbot's response text to the user.

[1630] Input: Chatbot response text

[1631] Output: Response text displayed on the user's smartphone

[1632] What happens: The server generates a response text and sends it to the user's smartphone, where the application displays it.

[1633] Step 6:

[1634] The server checks whether the chatbot's response contains the message "cannot be resolved."

[1635] Input: Chatbot response text

[1636] Output: Unresolvable flag (e.g., not included)

[1637] What happens: The server parses the response text and checks whether it contains the keyword "unresolved."

[1638] Step 7:

[1639] If the server determines that it cannot resolve the issue, it prompts the user to submit the form in English.

[1640] Input: Unresolvable flag

[1641] Output: Form submission request in English

[1642] Specific operation: The server sends a message in English to the user's smartphone prompting them to submit the form.

[1643] Step 8:

[1644] The user submits a form in English and the server receives it.

[1645] Input: User's form submission text in English

[1646] Output: Form submission text in English is sent to the server.

[1647] Specific behavior: A user enters a form submission in English and presses the submit button.

[1648] Step 9:

[1649] The server translates the English form submission into Japanese using the Google Translate API.

[1650] Input: Form post text in English

[1651] Output: Japanese translated text

[1652] Specific behavior: The server sends the English form post text to the Google Translate API and receives the translated Japanese text.

[1653] Step 10:

[1654] The server sends the translated text to the operator.

[1655] Input: Japanese translated text

[1656] Output: Text displayed on the operator's terminal

[1657] Specific operation: The server sends the translated Japanese text to the operator's terminal, and the operator checks it.

[1658] Step 11:

[1659] The operator provides a confirmation response, which is received by the server.

[1660] Input: Operator confirmation answer text

[1661] Output: The confirmation response text is sent to the server.

[1662] Specific operation: The operator enters a confirmation answer and presses the send button.

[1663] Step 12:

[1664] The server translates the confirmation response into English using the Google Translate API.

[1665] Input: Confirmation answer text

[1666] Output: Text translated into English

[1667] Specific operation: The server sends the confirmation response text to the Google Translate API and receives the translated English text.

[1668] Step 13:

[1669] The server sends the user a confirmation response translated into English.

[1670] Input: Confirmation answer text translated into English

[1671] Output: Confirmation answer text displayed on the user's smartphone

[1672] What it does: The server sends the translated English text to the user's smartphone, where the application displays it.

[1673] Example 2

[1674] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1675] Previous online support systems often failed to provide appropriate answers to users' questions and lacked multilingual support. Furthermore, they lacked a mechanism for analyzing user sentiment and using it to improve the site, making it difficult to improve the user experience. Furthermore, when operator resources were limited, the system's functionality was often reduced.

[1676] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1677] In this invention, the server includes an AI chatbot that answers users' questions, supports online purchases, and uses the results to improve the website; a chatbot that accepts form submissions in multiple languages ​​when the chatbot is unable to resolve the issue, has an operator confirm and respond in a language translated by the AI, and then translates and sends the response; an emotion analysis engine that analyzes users' emotions and provides feedback for website improvement; and a multilingual support system for the above flow. This allows for appropriate answers to users' questions, strengthens multilingual support, and analyzes user emotions to improve the website. Furthermore, even when operator resources are limited, the system's functionality is not compromised, improving the user experience.

[1678] "User" means an individual or legal entity that uses the System to make inquiries or purchases.

[1679] "Questions" are questions or uncertainties that users have about the system.

[1680] "Online purchasing" is the act of purchasing goods or services over the Internet.

[1681] An "artificial intelligence chatbot" is a program that automatically responds to user inquiries.

[1682] "Log" refers to the user's operation history and inquiry details recorded by the system.

[1683] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1684] "Multilingual" refers to multiple different languages ​​and means that the system supports these languages.

[1685] "Form submission" means the act of a user submitting an inquiry or information through a web form.

[1686] An "operator" is a human representative who responds to inquiries from users.

[1687] "Artificial intelligence translation" is the process of translating text into different languages ​​using artificial intelligence techniques.

[1688] An "emotion analysis engine" is a program that analyzes emotions from a user's text and provides the results.

[1689] "Feedback" means any opinion or information provided to improve a system or service.

[1690] "Flow" refers to a series of processing steps or procedures within a system.

[1691] This invention is a system that answers users' questions, supports online purchases, and uses the logs to improve the website. The system is composed of multiple components, including an AI chatbot, AI translation, and a sentiment analysis engine.

[1692] Hardware and software used

[1693] The server should be a high-performance server (e.g., a high-performance server). The following software should be used:

[1694] Artificial Intelligence Chatbots: Popular Chatbot Platforms

[1695] AI Translation: General Translation API

[1696] Sentiment Analysis Engine: A general sentiment analysis tool

[1697] Program processing

[1698] When the server receives a user inquiry, it first analyzes the inquiry using an AI chatbot, then invokes an AI translation to translate the inquiry into the appropriate language based on the analysis results, and the translated content is passed back to the AI ​​chatbot, which generates an appropriate response for the user.

[1699] Furthermore, the sentiment analysis engine analyzes the emotions contained in user inquiries and responses. For example, if many users express dissatisfaction with a particular process or function, the sentiment analysis engine will detect this and provide feedback to the site operator for improvement.

[1700] Specific examples

[1701] If a user asks in French, "I find the search function on this site difficult to use," the server performs the following process:

[1702] 1. User: Enter "La fonction de recherche de ce site est difficile a utiliser".

[1703] 2. Terminal: Sends the entered query to the server.

[1704] 3. Server: Receives the inquiry and passes it to the AI ​​chatbot.

[1705] 4. Artificial intelligence chatbot: Analyzes the inquiry content and extracts the intent that "the search function is difficult to use."

[1706] 5. Server: Passes the analysis results to an AI translator and requests it to translate from French to English.

[1707] 6. AI translation: Translates as "The search function of this site is difficult to use."

[1708] 7. Server: Passes the translation results to the AI ​​chatbot.

[1709] 8. Artificial intelligence chatbot: Generates the response, "We are sorry to hear that you are having trouble with the search function. We will look into this issue."

[1710] 9. Server: Passes the response to AI translation and requests it to be translated from English to French.

[1711] 10. Artificial Intelligence Translation: Translate "Nous sommes desoles d'apprendre que vous rencontrez des difficulties avec la fonction de recherche.

[1712] 11. Server: Sends the translated response to the device.

[1713] 12. Terminal: displays the response to the user.

[1714] 13. Server: Passes the query content to the sentiment analysis engine and performs sentiment analysis.

[1715] 14. Sentiment analysis engine: Analyzes the content of inquiries and detects feelings of dissatisfaction.

[1716] 15. Server: Feedback the results of the sentiment analysis to the site operator, reporting that "many users are dissatisfied with the search function."

[1717] Prompt Sentence Examples

[1718] "Please explain how a sentiment analysis engine provides feedback when a user is unhappy with a particular feature."

[1719] "Please provide a concrete example of how an AI chatbot would respond to a query in French."

[1720] In this way, the system achieves multilingual support and an improved user experience.

[1721] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1722] Step 1:

[1723] The user enters a query.

[1724] Users use their devices to enter questions and opinions into the system's inquiry form, and the entered data is saved in text format on the device.

[1725] Step 2:

[1726] The terminal sends a query to the server.

[1727] The terminal sends the query entered by the user to the server via the Internet. The input data is sent to the server in text format.

[1728] Step 3:

[1729] The server receives the query and passes it to an artificial intelligence chatbot.

[1730] The server passes the received inquiry to an AI chatbot, which receives the input data in text format.

[1731] Step 4:

[1732] An artificial intelligence chatbot analyzes the inquiry.

[1733] The AI ​​chatbot analyzes the content of the inquiry and extracts the intent and important keywords. The input data is in text format, and the output data is returned to the server in a structured data format as the analysis results.

[1734] Step 5:

[1735] The server passes the analysis results to the artificial intelligence translation system.

[1736] The server then passes the analysis results to an AI translator and requests that they be translated into the specified language. The input data is in structured data format, and the output data is returned in translated text format.

[1737] Step 6:

[1738] Artificial intelligence translation will translate the inquiry into the specified language.

[1739] AI translation translates the query into the specified language, with the input data in a structured data format and the output data returned to the server in a translated text format.

[1740] Step 7:

[1741] The server passes the translation results to an artificial intelligence chatbot.

[1742] The server then passes the translated query back to the AI ​​chatbot, with the input data being in the form of translated text and the output data being returned in the form of structured data as the chatbot's response.

[1743] Step 8:

[1744] An artificial intelligence chatbot generates appropriate responses.

[1745] The AI ​​chatbot generates an appropriate response based on the translated query, with the input data in the form of translated text and the output data in the form of structured data as a response, which is returned to the server.

[1746] Step 9:

[1747] The server passes the response to an artificial intelligence translator to translate it into the user's language.

[1748] The server then passes the generated response to the AI ​​translator again, requesting it to be translated into the user's language. The input data is in structured data format, and the output data is returned in translated text format.

[1749] Step 10:

[1750] The server sends the translated response to the terminal.

[1751] The server sends the translated response to the terminal, where the input data is in translated text format and the output data is displayed on the terminal.

[1752] Step 11:

[1753] The terminal displays the response to the user.

[1754] The terminal displays the response received from the server to the user. The input data is in the translated text form, and the output data is in the text form displayed to the user.

[1755] Step 12:

[1756] The server passes the query content to the emotion analysis engine, which performs emotion analysis.

[1757] The server passes the user's inquiry to the emotion analysis engine for emotion analysis. The input data is in text format, and the output data is returned in structured data format as the emotion analysis results.

[1758] Step 13:

[1759] The sentiment analysis engine returns the analysis results to the server.

[1760] The sentiment analysis engine returns the analysis results to the server. The input data is in text format, and the output data is in structured data format as the sentiment analysis results.

[1761] Step 14:

[1762] The server provides the analysis results as feedback to the website operator.

[1763] The server feeds back the results of the sentiment analysis to the website operator and provides information for improving the website. The input data is in a structured data format as the result of the sentiment analysis, and the output data is feedback information provided to the website operator.

[1764] (Application example 2)

[1765] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1766] Conventional online purchasing support systems lacked multilingual support for user questions, making it difficult to respond appropriately to inquiries in languages ​​other than English. Furthermore, there was a lack of feedback to improve the site based on user sentiment, and the user experience was not sufficiently improved. Furthermore, the system needed to function smoothly even when operator resources were limited.

[1767] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1768] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, an operator confirming and responding in an AI-translated language, and the AI ​​translating and sending the response, a means including an emotion engine for analyzing user emotions and providing feedback for improvements to the site operator, and a means for making the above flow multilingual. This enables multilingual support for users, provides feedback for site improvements based on emotions, and enables the system to function smoothly even when operator resources are limited.

[1769] An "AI chatbot" is a dialogue system that uses artificial intelligence to automatically respond to users' questions and support online purchases.

[1770] "Multilingual support" refers to the ability to support multiple different languages ​​and provide appropriate responses based on the language used by the user.

[1771] "Form submission" refers to the act of a user submitting an inquiry or information using an online input form.

[1772] An "operator" is a human representative who responds to inquiries from users.

[1773] "AI translation" is a technology that uses artificial intelligence to convert text from one language to another.

[1774] The "Emotion Engine" is a system that analyzes user emotions and provides feedback to site operators for improvements based on the results.

[1775] "Feedback" is information for improvement provided based on user experiences and opinions.

[1776] "Website Operator" means the person or organization that manages and operates an online website.

[1777] A "log" is a record of a user's operation history and inquiries recorded by the system.

[1778] The embodiments of the present invention will be described in detail below.

[1779] System Program

[1780] The system consists of a server, a user terminal, and an operator terminal. The server runs programs including an AI chatbot that answers user questions, a multilingual form submission system, and an emotion engine.

[1781] Hardware and software used

[1782] Hardware: Servers, smartphones, operator terminals

[1783] Software: Python, TensorFlow, Google Cloud Translation API, Sentiment Analysis API

[1784] Data processing and calculation

[1785] 1. Receiving User Enquiries:

[1786] An inquiry is made from a user terminal (such as a smartphone).

[1787] The query is sent to the server.

[1788] 2. Multilingual support:

[1789] The server uses the Google Cloud Translation API to translate the query into English.

[1790] The translated content is passed to an AI chatbot.

[1791] 3. Chatbot response:

[1792] AI chatbots generate appropriate responses to user queries.

[1793] Retranslate the response into the user's language.

[1794] 4. Sentiment analysis:

[1795] The server uses the Sentiment Analysis API to analyze the user's query and detect their emotions.

[1796] Based on the sentiment data, we provide feedback to site operators for improvement.

[1797] Specific examples

[1798] User inquiry: The user asks in French, "When will this item arrive?"

[1799] Translation: "Quand ce produit arrivera-t-il ?" translated into English using the Google Cloud Translation API.

[1800] Chatbot response: The AI ​​chatbot replies, "The product will arrive in 3 days."

[1801] Retranslate: Retranslate the response into French and tell the user, "Le produit arrivera dans 3 jours."

[1802] Sentiment Analysis: Analyze user sentiment using the Sentiment Analysis API and determine whether it is positive.

[1803] Prompt Sentence Examples

[1804] If a user asks in French, "When will this item arrive?", use the Google Cloud Translation API to translate it into English and pass it to an AI chatbot. Translate the chatbot's response back into French and reply to the user. Also, use the Sentiment Analysis API to analyze user sentiment and provide feedback to the publisher.

[1805] In this way, multilingual support for users is possible, sentiment-based feedback for site improvement is provided, and the system can function smoothly even when operator resources are limited.

[1806] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1807] Step 1:

[1808] The user device receives the inquiry. The user inputs the inquiry into a device such as a smartphone and presses the send button. The input is in the user's language, and the inquiry content is sent to the server.

[1809] Step 2:

[1810] The server receives the query content. The server receives the query content sent from the user terminal and stores it in a database. The input is the user's query content, and the output is the stored query data.

[1811] Step 3:

[1812] The server translates the query into English using the Google Cloud Translation API. The server then sends the received query to the Google Cloud Translation API and receives the translated English text. The input is the user's query, and the output is the translated English text.

[1813] Step 4:

[1814] The server passes the translated content to the AI ​​chatbot. The server sends the translated English text to the AI ​​chatbot, which generates an appropriate response. The input is the translated English text, and the output is the response generated by the AI ​​chatbot.

[1815] Step 5:

[1816] The server retranslates the AI ​​chatbot's response into the user's language. The server then sends the response received from the AI ​​chatbot back to the Google Cloud Translation API and receives the translated text into the user's language. The input is the AI ​​chatbot's response, and the output is the translated text into the user's language.

[1817] Step 6:

[1818] The server sends the translated response to the user terminal. The server sends the translated response to the user's terminal in the user's language and displays it to the user. The input is the text translated into the user's language, and the output is the response displayed on the user terminal.

[1819] Step 7:

[1820] The server analyzes the user's sentiment using the Sentiment Analysis API. The server sends the user's query to the Sentiment Analysis API and receives the sentiment analysis results. The input is the user's query, and the output is the sentiment analysis results.

[1821] Step 8:

[1822] The server provides feedback to the website operator based on the results of the sentiment analysis. The server stores the results of the sentiment analysis in a database and generates feedback for improvement to the website operator. The input is the result of the sentiment analysis, and the output is the feedback provided to the website operator.

[1823] Example 3

[1824] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1825] Traditional online support systems often had slow responses to user inquiries and lacked operator resources. They also lacked the means to track changes in user sentiment and use it to improve services. This resulted in poor user experience and decreased site usage.

[1826] The identification process by the identification processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means. In this invention, the server includes an automatic response device means, a means for receiving an inquiry, an operator responding with a confirmation, and the automatic response device translating and sending the response, a means for supporting multiple languages, a means for tracking user emotions over time and analyzing changes therein, and a means for escalating to an operator as necessary. This enables prompt and appropriate support to users while reducing the burden on operators, improving the user experience and increasing the site usage rate.

[1827] An "automatic answering machine" is a device that automatically generates a response to a user's inquiry and provides an answer.

[1828] "Means for accepting an inquiry, having an operator confirm and respond, and having an automatic answering device translate and send the response" refers to a means for carrying out a series of processes in which an inquiry from a user is accepted, an operator confirms the content of the inquiry, creates a response, and the automatic answering device translates the response and sends it to the user.

[1829] "Means for supporting multiple languages" refers to means for enabling the system to respond to user inquiries in multiple languages.

[1830] "Means for tracking a user's emotions over time and analyzing changes therein" refers to means for continuously monitoring a user's emotions and analyzing changes therein.

[1831] "Means for escalating to an operator as necessary" refers to a means for transferring an inquiry to an operator when the inquiry cannot be handled by the automated answering device or when the inquiry meets certain conditions.

[1832] The present invention provides a system that provides a prompt and appropriate response to user inquiries, reduces the burden on operators, and improves the user experience. Specific embodiments of this system will be described below.

[1833] Hardware and software used

[1834] Hardware: Servers, devices (users' PCs and smartphones)

[1835] Software: Automated response devices (e.g., Dialogflow), sentiment analysis engines (e.g., IBM Watson Tone Analyzer)

[1836] System Overview

[1837] When a user sends a query using a terminal, the server receives the query and forwards it to an automated answering machine, which analyzes the query and generates an appropriate answer, which is then sent to the user via the server.

[1838] If the automated answering machine cannot process the query, the server escalates the query to a human operator, who creates a confirmation response that the automated answering machine translates and sends to the user.

[1839] The server then sends the user's query to a sentiment analysis engine, tracks the user's sentiment over time, and analyzes changes in that sentiment. This data is then stored in a database and used to improve the service.

[1840] Specific examples

[1841] User Question: "How do I return an item?"

[1842] The automated response: "Here's how to return it..."

[1843] Sentiment analysis engine analysis result: "User sentiment is neutral"

[1844] Prompt Sentence Examples

[1845] Prompt: "A user has inquired about how to return an item. Your automated response system generated an appropriate response, and your sentiment analysis engine determined the user's sentiment to be neutral. Based on these results, please suggest ways to improve our service."

[1846] This system enables quick and appropriate responses to user inquiries, reducing the burden on operators and is expected to improve the user experience and increase the site usage rate. The flow of the identification process in Example 3 will be explained using Figure 21.

[1847] Step 1:

[1848] A user submits an inquiry. Using a device (PC or smartphone), the user enters a question into the inquiry form and clicks the send button. The input is the user's inquiry, and the output is the inquiry data sent to the server. Specifically, the user enters "Please tell me how to return an item" and clicks the send button.

[1849] Step 2:

[1850] The server receives the inquiry. The server receives the inquiry data from the user and records it in a log. The input is the inquiry data sent by the user, and the output is the inquiry data recorded in the log. In concrete terms, the server receives the inquiry "Please tell me how to return an item" and records it in a log.

[1851] Step 3:

[1852] The server forwards the inquiry to an automatic response device. The server forwards the inquiry data it received to the automatic response device (e.g., Dialogflow). The input is the inquiry data recorded on the server, and the output is the inquiry data sent to the automatic response device. Specifically, the server sends the data "Please tell me how to return the product" to Dialogflow.

[1853] Step 4:

[1854] The automated response device analyzes the inquiry and generates a response. The automated response device analyzes the inquiry content using natural language processing technology and generates an appropriate response. The input is the inquiry data sent to the automated response device, and the output is the generated response data. Specifically, Dialogflow generates the response, "The return method is as follows..."

[1855] Step 5:

[1856] The server sends the answer to the user. The server sends the answer it received from the automatic answering machine to the user. The input is the answer data received from the automatic answering machine, and the output is the answer data sent to the user. In concrete terms, the server sends the answer "The return method is as follows..." to the user's terminal.

[1857] Step 6:

[1858] The server sends the query to the sentiment analysis engine. The server sends the user's query to the sentiment analysis engine (e.g. IBM Watson Tone Analyzer). The input is the query data recorded on the server, and the output is the query data sent to the sentiment analysis engine. Specifically, the server sends the data "Please tell me how to return the product" to IBM Watson Tone Analyzer.

[1859] Step 7:

[1860] The sentiment analysis engine analyzes the user's emotions. The sentiment analysis engine analyzes the content of the inquiry and analyzes the user's emotions. The input is the inquiry data sent to the sentiment analysis engine, and the output is the analysis result data. Specifically, IBM Watson Tone Analyzer returns the analysis result that "the user's emotions are neutral."

[1861] Step 8:

[1862] The server tracks changes in emotions and stores them in a database. The server receives the analysis results from the emotion analysis engine, tracks changes in the user's emotions, and stores them in a database. The input is the analysis result data received from the emotion analysis engine, and the output is the emotion change data stored in the database. Specifically, the server stores the data "user's emotion is neutral" in the database.

[1863] Step 9:

[1864] Escalate to an operator if necessary. If the server detects a query that the automated answering machine cannot process, it escalates to an operator. The input is the query data that the automated answering machine cannot process, and the output is the query data forwarded to the operator. Specifically, the server forwards the query "A technical problem has occurred" to the operator.

[1865] (Application example 3)

[1866] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1867] Traditional online purchasing support systems are inadequate in responding to user questions, particularly in the areas of multilingual support and emotional response. Furthermore, when operator resources are insufficient, the system may not function smoothly. This can lead to lower user satisfaction and reduced site usage.

[1868] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1869] In this invention, the server includes: a means for using an AI chatbot with generative AI to respond to inquiries entered by users regarding product purchases on online sites; a means for accepting the user's inquiry via a form submission that can be entered in multiple languages, translating the inquiry into a specific language and presenting it to an operator if the AI ​​chatbot determines that it cannot provide an answer; and a means for accepting an answer to the inquiry from the operator, translating the answer back into the original language, and presenting it to the user. The server also includes a means for using an emotion engine to estimate an emotion score indicating the user's emotion and tracking changes in the emotion score, and a means for notifying the operator if the emotion score falls below a certain threshold. The server also includes a prompt sentence for determining whether the inquiry needs to be escalated to the operator based on the content or complexity of the inquiry entered by the user, and a means for using the generative AI to determine whether the AI ​​chatbot cannot provide an answer. This enables prompt and appropriate responses to user questions and enables service improvements based on changes in emotion. Furthermore, even when operator resources are insufficient, the system functions smoothly and user satisfaction can be improved.

[1870] An "AI chatbot" is a dialogue system that uses artificial intelligence to automatically respond to users' questions and support online purchases.

[1871] "Multilingual support" refers to the ability to process user inquiries and responses in multiple languages.

[1872] The "emotion engine" is a system that tracks users' emotions over time and analyzes their changes.

[1873] The "emotion score" is a numerical indicator of the user's emotional state.

[1874] "Escalate" is the process of transferring an inquiry to an operator when the AI ​​chatbot is unable to respond.

[1875] An "operator" is a person who responds to user inquiries.

[1876] "Log" refers to historical data of user inquiries and responses.

[1877] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[1878] A system for implementing this invention is configured as follows: The server includes an AI chatbot means for answering user questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, an operator confirming and replying in an AI-translated language, and the AI ​​translating and sending the reply, an emotion engine means for tracking user emotions over time and analyzing changes, a means for escalating to an operator if the emotion score falls below a certain threshold, and a means for making the above flow multilingual.

[1879] Program processing explanation

[1880] When the server receives a user inquiry, an AI chatbot using generative AI first automatically generates a response. This response is generated using OpenAI's API. The user's inquiry is analyzed by an emotion engine, and an emotion score for the user is calculated. If this emotion score falls below a certain threshold, the inquiry is escalated to an operator. The operator then replies in the AI-translated language, which is then translated again by AI and sent to the user.

[1881] The hardware used is mobile devices such as smartphones and tablets, and the software uses OpenAI's API and sentiment analysis engine (e.g., SentimentAnalyzer).

[1882] Specific examples

[1883] For example, if a user inquires, "My order hasn't arrived yet. What's going on?", the AI ​​chatbot will respond, "We will check the status of your order. Please wait a moment." After that, an emotion score indicating the user's emotions is estimated based on the user's inquiry and the AI ​​chatbot's response, and if the emotion score falls below a threshold, the inquiry is escalated to an operator. The operator will check the user's inquiry and the chatbot's response and take appropriate action.

[1884] Example prompts to input to a generative AI model:

[1885] User asks: "I haven't received my order, what's going on?"

[1886] The AI ​​chatbot responds: "Please wait a moment while we check the status of your order."

[1887] Sentiment score: -0.6 Escalate: True

[1888] Furthermore, when the server receives an inquiry from a user, it uses a prompt sentence that determines whether the inquiry needs to be escalated to an operator based on the content or complexity of the inquiry entered by the user and generative AI to determine whether the AI ​​chatbot can provide an answer. If the server determines that the AI ​​chatbot cannot provide an answer, it accepts the user's inquiry through a form submission that can be entered in multiple languages, translates the inquiry into a specific language, and presents it to the operator. It also accepts an answer to the inquiry from the operator, translates the answer into the original language, and presents it to the user.

[1889] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1890] Step 1:

[1891] A user inputs an inquiry from a device such as a smartphone or tablet.

[1892] Input: User's inquiry (e.g. "My order hasn't arrived yet. What's going on?")

[1893] Output: The query is sent to the server.

[1894] Step 2:

[1895] The server passes the received inquiry to the AI ​​chatbot and generates a response.

[1896] Input: User's inquiry

[1897] Data processing: Use OpenAI's API to generate responses to inquiries.

[1898] Output: The AI ​​chatbot's response (e.g., "We'll check the status of your order. Please wait a moment.")

[1899] Step 3:

[1900] The server passes the user query and the generated response to an emotion engine to calculate an emotion score for the user.

[1901] Input: AI chatbot response

[1902] Data calculation: Use a sentiment analysis engine (e.g., SentimentAnalyzer) to calculate a sentiment score for the response text.

[1903] Output: Sentiment score (e.g. -0.6)

[1904] Step 4:

[1905] The server determines whether the emotion score is below a certain threshold.

[1906] Input: Sentiment score

[1907] Data calculation: Compare whether the sentiment score is below a threshold (e.g., -0.5).

[1908] Output: Escalation required (True or False)

[1909] Step 5:

[1910] If the sentiment score is below the threshold, the server escalates the query to an operator.

[1911] Input: Escalation Required (True)

[1912] Specific operation: The server notifies the operator of the user's inquiry and the AI ​​chatbot's response.

[1913] Output: Inquiry escalated to operator and response

[1914] Step 6:

[1915] The operator reviews the escalated inquiry and prepares an appropriate response.

[1916] Input: Escalated inquiry and response

[1917] Specific operation: The operator creates a confirmation response in the AI-translated language.

[1918] Output: Operator response

[1919] Step 7:

[1920] The server then uses AI to translate the operator's response again and sends it to the user.

[1921] Input: Operator's answer

[1922] Data processing: Using AI translation, the operator's response is translated into the user's language.

[1923] Output: The translated answer is sent to the user.

[1924] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1925] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1926] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[1927] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1928] [Third embodiment]

[1929] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1930] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1931] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1932] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1933] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1934] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1935] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1936] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1937] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1938] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1939] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1940] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1941] "Example 1"

[1942] This invention is a system that uses an AI chatbot to answer user questions, support online purchases, and use the logs to improve the website. Specifically, it accepts user inquiries and purchase requests and provides appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[1943] "Example 2"

[1944] Furthermore, this embodiment of the present invention can also make the above system multilingual. Specifically, by changing the settings of the AI ​​chatbot or translation AI, it can also support languages ​​other than English. For example, it can also handle inquiries from users in French, Spanish, and other languages.

[1945] "Example 3"

[1946] Furthermore, this embodiment of the present invention is designed to function even when it is difficult to secure human resources. Specifically, an AI chatbot automatically processes inquiries and escalates them to a human operator as necessary. This reduces the burden on operators while allowing them to continue providing support to users.

[1947] The processing flow of each embodiment will be described below.

[1948] "Example 1"

[1949] Step 1: The AI ​​chatbot accepts user inquiries and purchase requests.

[1950] Step 2: The AI ​​chatbot provides the appropriate answer or action.

[1951] Step 3: If the bot can't resolve the issue, prompt the user to submit the form in English.

[1952] Step 4: The AI ​​translates the post into Japanese and an operator verifies and replies.

[1953] Step 5: The AI ​​translates the answer back into English and sends it to the user.

[1954] "Example 2"

[1955] Step 1: Change the settings of your AI chatbot and translation AI to support languages ​​other than English.

[1956] Step 2: For example, respond to inquiries from users in French or Spanish.

[1957] "Example 3"

[1958] Step 1: The AI ​​chatbot automatically handles the inquiry.

[1959] Step 2: Escalate to an operator if necessary.

[1960] Step 3: This reduces the burden on operators while still providing support to users.

[1961] Example 1

[1962] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1963] Conventional online purchasing support systems sometimes fail to provide prompt and appropriate answers to user questions, and a lack of operator resources is a problem, especially when multilingual support is required. Additionally, there was a lack of a way to effectively log user inquiries and use them to improve the site.

[1964] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1965] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site, a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming and responding in AI-translated Japanese, and the AI ​​translating the answer into English and sending it, a means for the AI ​​chatbot to generate answers to users' inquiries using a generative AI model, a means for generating prompts for the generative AI model to obtain appropriate answers, and a means for making the flow multilingual. This enables quick and appropriate answers to users' questions, allows the system to function even when operator resources are insufficient, and enables the site to be improved by effectively utilizing the logs.

[1966] An "artificial intelligence chatbot" is a program that automatically generates answers to users' questions and supports online purchases.

[1967] A "generative artificial intelligence model" is an artificial intelligence model used to generate appropriate responses to user queries.

[1968] A "prompt sentence" is text that is input to a generative artificial intelligence model to generate an appropriate answer.

[1969] "Form submission" is a way for users to enter and submit questions in English that the bot cannot solve.

[1970] An "operator" is a human representative who checks the Japanese content translated by artificial intelligence in response to user inquiries and provides appropriate answers.

[1971] "Artificial intelligence translation" is a technology that automatically translates user inquiries and operator responses into different languages.

[1972] "Multilingual support" means that the system has the ability to respond to user inquiries in multiple languages.

[1973] A "log" is data that records user inquiries and system responses and is used for later analysis and site improvement.

[1974] This invention is a system that answers users' questions, supports online purchases, and uses logs to improve the website. Specifically, it uses an AI chatbot to accept inquiries and purchase requests from users and provide appropriate answers and actions. If the bot cannot resolve an inquiry, the user is prompted to submit a form in English, which the AI ​​translates into Japanese. An operator then verifies the response, which the AI ​​then translates back into English and sends to the user.

[1975] Hardware and software used

[1976] Server: Hosts the data processing and AI models, specifically using virtual servers from a cloud service provider.

[1977] Device: The device from which the user accesses the site (e.g., PC, smartphone, tablet, etc.).

[1978] Generative AI model: An artificial intelligence model used to generate answers to user queries, specifically using natural language processing models.

[1979] Translation software: Software for translating between English and Japanese, specifically using a cloud-based translation API.

[1980] Explanation of program processing

[1981] 1. Receiving user inquiries

[1982] A user uses a device to enter a query into a website chatbot.

[1983] The terminal sends this input to the server.

[1984] The server receives the query and records it as a log in the database.

[1985] Specific operation: A user types "Do you have this item in stock?" into the chatbot. The device sends this message to the server, which records it in the database as "User ID: 1234, Inquiry: Do you have this item in stock?"

[1986] 2. Answer generation by AI chatbot

[1987] The server sends the received query to the generative AI model.

[1988] The generative AI model generates an appropriate answer and sends it back to the server.

[1989] The server generates the answer and sends it to the user's device.

[1990] Specific operation: The server sends the query "Do you have this item in stock?" to the generative AI model. The generative AI model generates the answer "In stock" and sends it back to the server. The server sends this answer to the user's device, and the user sees "In stock" on the chatbot's screen.

[1991] 3. Processing Purchase Requests

[1992] The user enters a purchase request.

[1993] The terminal sends this request to the server.

[1994] The server receives the request, checks inventory, and processes the order.

[1995] If necessary, the generative AI model provides additional information to the user.

[1996] How it works: The user types, "I want to buy this product." The device sends this request to the server, which checks the inventory. If the item is in stock, the server notifies the user via the generative AI model, saying, "Your order has been accepted."

[1997] 4. Escalation of Inquiries

[1998] If a query arises that the bot cannot resolve, the server prompts the user to submit a form in English.

[1999] A user submits a form in English.

[2000] The server sends the content to translation software, which translates it into Japanese.

[2001] What happens: The bot tells the user, "We can't solve this problem. Please submit the form in English." The user submits the form in English, saying, "I need help with my order." The server sends this to translation software, which translates it into Japanese as, "I need help with my order."

[2002] 5. Confirmation by the operator

[2003] The operator will review the translated inquiry and enter the appropriate response in Japanese.

[2004] The server sends this response back to the translation software, which translates it into English.

[2005] The server sends the translated response to the user's device.

[2006] What happens: The operator sees the inquiry "I need help with my order" and types "We have it in stock. We will process your order" in Japanese. The server sends this response to translation software, which translates it into English as "The item is in stock. We will process your order." The server then sends this translated response to the user's device.

[2007] 6. Leveraging logs

[2008] The server keeps a log of all queries and responses.

[2009] The server analyzes these logs and generates insights for improving the site.

[2010] Specific operation: The server saves logs such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock." These logs are periodically analyzed to generate insights, such as "There are many inquiries about a particular product," which can be used to improve the site.

[2011] Prompt Sentence Examples

[2012] User asks: "Do you have this item in stock?"

[2013] Prompt for generative AI model: "A user is asking, 'Do you have this item in stock?' Please generate an appropriate answer."

[2014] The flow of the identification process in the first embodiment will be described with reference to FIG.

[2015] Step 1: Receiving user inquiries

[2016] Input: A user uses a device to type a query into a website chatbot.

[2017] What happens: A user types into the chatbot, "Do you have this item in stock?"

[2018] Data processing: The device sends this input to the server.

[2019] Output: The server receives the query and logs it in the database.

[2020] Specific operation: The server records "User ID: 1234, Enquiry: Is this item in stock?"

[2021] Step 2: Generate answers with an AI chatbot

[2022] Input: The query received by the server.

[2023] What happens: The server sends the query "Do you have this item in stock?" to the generative AI model.

[2024] Data computation: Generative AI models generate appropriate answers.

[2025] Output: The generative AI model generates the answer "In stock" and sends it back to the server.

[2026] Specific behavior: The server sends this response to the user's device, and the user sees the message "In stock" on the chatbot's screen.

[2027] Step 3: Processing the purchase request

[2028] Input: The user enters a purchase request.

[2029] What happens: The user types, "I want to buy this product."

[2030] Data processing: The device sends this request to the server.

[2031] Output: The server receives the request, checks inventory, and processes the order.

[2032] Specific operation: The server checks the inventory, and if the item is in stock, it notifies the user through the generative AI model that "your order has been accepted."

[2033] Step 4: Escalate the case

[2034] Input: The query the bot cannot resolve.

[2035] What happens: The bot informs the user, "We can't solve this problem. Please submit the form in English."

[2036] Data processing: User submits form in English.

[2037] Output: The server sends the content to translation software, which translates it into Japanese.

[2038] What happens: The server translates "I need help with my order" to "I need help with my order."

[2039] Step 5: Confirmation by operator

[2040] Input: The translated query.

[2041] Specific operation: The operator confirms the inquiry "I need help with my order" and types in Japanese "We have the item in stock. We will take your order."

[2042] Data calculation: The server sends this response back to the translation software, which translates it into English.

[2043] Output: Sends the translated answer to the user's device.

[2044] Specific behavior: The server translates the message into English as "The item is in stock. We will process your order" and sends it to the user.

[2045] Step 6: Leverage the logs

[2046] Input: A log of all queries and responses.

[2047] Specific operation: The server saves a log such as "User ID: 1234, Inquiry: Is this product in stock?, Answer: Yes, it is in stock."

[2048] Data calculations: Our servers analyze these logs and generate insights for improving the site.

[2049] Output: Insights for improving your site.

[2050] Specific operation: The server generates insights such as "there are many inquiries about a particular product" and uses this insight to improve the site.

[2051] (Application example 1)

[2052] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[2053] Conventional online purchasing support systems sometimes failed to provide prompt and appropriate answers to user questions, and operator resources were often insufficient, especially when multilingual support was required. Furthermore, there was a lack of a way to effectively utilize user inquiry logs to improve the site. This resulted in a poor user experience and reduced willingness to purchase.

[2054] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[2055] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in English when the bot cannot resolve the issue, an operator confirming the response in Japanese translated by the AI, and the AI ​​translating the response into English and sending it; a means for making the above flow multilingual; a means for users to input questions about products as an application installed on a smartphone and the AI ​​chatbot providing an immediate response; a means for prompting users to submit a form in English for questions that cannot be resolved and translating the form into Japanese by the AI; and a means for an operator confirming the response and the AI ​​translating the response back into English and sending it to the user. This enables quick and appropriate responses to users' questions, facilitates multilingual support, compensates for a lack of operator resources, and improves the user experience.

[2056] "User" refers to the consumer or user making an online purchase.

[2057] "Questions" refer to questions or concerns users have about products or services.

[2058] "Online purchasing" refers to the act of purchasing goods or services over the Internet.

[2059] An "artificial intelligence chatbot" is a program that automatically provides answers to users' questions.

[2060] "Logs" refer to records of user inquiries and interactions with chatbots.

[2061] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[2062] "Bot unresolved" refers to a situation where the chatbot is unable to provide an appropriate answer to the user's question.

[2063] "Form submission in English" refers to a form that allows users to enter questions or requests in English.

[2064] "Operator" refers to a human representative who responds to user inquiries.

[2065] "Artificial intelligence translation" refers to the process of using artificial intelligence to translate text into different languages.

[2066] "Multilingual" refers to providing services in multiple languages.

[2067] "Applications installed on a smartphone" refers to software programs that run on a smartphone.

[2068] "Providing immediate answers" refers to providing a quick response to a user's question.

[2069] "Encouraging form submission in English" means instructing users to enter their questions in English.

[2070] "Translating into Japanese" refers to converting English text into Japanese.

[2071] "Translate back into English and send" means translating the Japanese response into English and sending it to the user.

[2072] A system for implementing this invention includes an AI chatbot that answers users' questions, supports online purchases, and uses the results to improve the website. The system also includes a means for accepting form submissions in English when the bot cannot resolve the issue, for an operator to confirm and respond in Japanese using AI translation, and for the AI ​​to retranslate the response into English and send it back. The system also includes a means for making the above flow multilingual.

[2073] Hardware and software used

[2074] Hardware:

[2075] Smartphone

[2076] software:

[2077] Python

[2078] OpenAI API

[2079] Googletrans library

[2080] Data processing and calculation

[2081] Accepting user input:

[2082] Users input questions about products through a smartphone application, which processes the input as text data within the application.

[2083] Answered by an AI chatbot:

[2084] The server uses the OpenAI API to generate answers to the user's questions, which are then returned to the user as text data.

[2085] Translation features:

[2086] The server uses the GoogleTrans library to translate between English and Japanese. When a user submits a form in English, the text is translated into Japanese and sent to the operator. The operator's response is also translated from Japanese to English and sent to the user.

[2087] Operator Intervention:

[2088] Any questions that the chatbot cannot solve are forwarded to an operator, who types the answer in Japanese, which is then translated back into English and sent to the user.

[2089] Specific examples

[2090] Example 1:

[2091] A user asks, "What size is this item?" The AI ​​chatbot immediately replies, "This item is a size medium."

[2092] Example 2:

[2093] If a user asks, "Can this product be shipped internationally?" and the chatbot cannot resolve the question, it prompts the user to submit a form in English. The user types, "Can this product be shipped internationally?", and the AI ​​translates it into Japanese as, "Can this product be shipped internationally?" The operator replies, "Yes, it can be shipped internationally," and the AI ​​translates it into English as, "Yes, it can be shipped internationally," and sends it to the user.

[2094] Prompt Sentence Examples

[2095] User: What size is this item?

[2096] AI Chatbot: This product is size M.

[2097] User: Can this item be shipped internationally?

[2098] AI Chatbot: Unable to resolve. Please submit the form in English.

[2099] User: Can this product be shipped internationally?

[2100] AI Translation: Can this item be shipped internationally?

[2101] Operator: Yes, you can.

[2102] AI Translation: Yes, it can be shipped internationally.

[2103] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2104] Step 1:

[2105] A user inputs a question about a product through a smartphone application. The input text data is processed within the application and sent to the server.

[2106] Input: User question text

[2107] Output: Text data sent to the server

[2108] Step 2:

[2109] The server uses the OpenAI API to generate answers to user questions, and the generated answers are stored as text data on the server.

[2110] Input: User question text

[2111] Data processing: Generate answers to questions using OpenAI APIs

[2112] Output: Generated answer text

[2113] Step 3:

[2114] The server sends the generated answer to the user's smartphone application, through which the user can confirm the answer.

[2115] Input: Generated answer text

[2116] Output: Answer text displayed in the user's smartphone application

[2117] Step 4:

[2118] If the user is not satisfied with the chatbot's answer, a message prompting them to submit a form in English is displayed. The user can then enter their question in English and submit it to the server.

[2119] Input: User's question text in English

[2120] Output: English question text sent to the server

[2121] Step 5:

[2122] The server uses the GoogleTrans library to translate the user's English question into Japanese, and the translated text is sent to the operator.

[2123] Input: User's question text in English

[2124] Data processing: Translated from English to Japanese using the Googletrans library

[2125] Output: Question text in Japanese sent to the operator

[2126] Step 6:

[2127] The operator inputs the answer in Japanese and sends it to the server, which receives the answer.

[2128] Input: Operator's response text in Japanese

[2129] Output: Answer text in Japanese sent to the server

[2130] Step 7:

[2131] The server uses the GoogleTrans library to translate the operator's Japanese response into English, and the translated text is sent to the user's smartphone application.

[2132] Input: Operator's Japanese response text

[2133] Data processing: Translate from Japanese to English using the Googletrans library

[2134] Output: Answer text in English that will be displayed on the user's smartphone application

[2135] Step 8:

[2136] The user can check the English response from the operator through the smartphone application.

[2137] Input: Answer text in English

[2138] Output: Answer text in English that will be displayed on the user's smartphone application

[2139] Example 2

[2140] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[2141] Conventional AI chatbot systems only support certain languages, making it difficult to support multiple languages. Furthermore, the process for generating appropriate responses to user inquiries is complicated, making it difficult to respond when there are insufficient operator resources. This leads to a poor user experience and insufficient log collection for site improvement.

[2142] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[2143] In this invention, the server includes a means for receiving user input data and determining the language, a means for changing the AI ​​chatbot's settings based on the determined language, and a means for generating an appropriate response using a generative AI model based on the set language. This enables multilingual support and provides prompt and appropriate responses to user inquiries. Furthermore, even if there are insufficient operator resources, the system automatically responds, enabling log collection for improving the user experience and site improvement.

[2144] An "artificial intelligence chatbot" is a program that automatically generates responses to users' questions and assists them in online purchases.

[2145] "Artificial intelligence translation" is a technology that automatically translates text between different languages.

[2146] A "generative AI model" is an artificial intelligence algorithm that generates appropriate responses based on user input.

[2147] "Language determination" is the process of automatically identifying the language of text entered by a user.

[2148] "Reconfiguration" is the process of adjusting the system's operating modes and parameters based on the determined language.

[2149] "Response generation" is the process of generating an appropriate reply to a user's input.

[2150] "Multilingual" refers to the system's ability to respond to user inquiries in several different languages.

[2151] An "operator" is a person whose role is to respond to user inquiries.

[2152] "Resource allocation" refers to preparing the personnel and equipment necessary for the system to operate normally.

[2153] "User experience" is a general term for the satisfaction and ease of use that users feel when using a system.

[2154] "Site Improvement" is the process of improving the functionality and design of a website based on user usage data and feedback.

[2155] MODE FOR CARRYING OUT THE INVENTION

[2156] This invention relates to a multilingual AI chatbot system. A specific embodiment of this system is described below.

[2157] System configuration

[2158] The system consists of three main components: the user, the terminal, and the server. The user makes a query through the terminal, and the terminal sends the data to the server. The server processes the received data and generates an appropriate response to send to the terminal.

[2159] Hardware and software used

[2160] Hardware: Terminals include devices such as PCs, smartphones, and tablets. Servers are high-performance computer systems, including cloud servers.

[2161] Software: The server uses a generative AI model (e.g., GPT-4) to generate the response, and translation software such as Google Cloud Translation API to determine the language.

[2162] Program processing

[2163] When the server receives a user inquiry, it first determines the language. It uses the Google Cloud Translation API to determine the language. Based on the determined language, the server changes the AI ​​chatbot's settings. For example, if French is detected, the settings are changed to support French.

[2164] The server then uses the generative AI model to generate an appropriate response, which is translated into the user's language using translation software if further translation is required, and then sends the response to the device, which displays it to the user.

[2165] Specific examples

[2166] For example, consider a case where a user makes a query in French: "Bonjour, comment puis-je vous aider?" The device sends this input to the server. The server uses the Google Cloud Translation API to determine that the input is in French. The server then changes the AI ​​chatbot's settings to support French and uses the generative AI model to generate the response: "Bonjour! Comment puis-je vous aider today?" The generated response is sent to the device, which displays it to the user.

[2167] Prompt Sentence Examples

[2168] "Generate an appropriate response to a query in French. The user's input is 'Bonjour, comment puis-je vous aider?'"

[2169] In this way, the system can process queries in multiple languages, improving the user experience.

[2170] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2171] Step 1:

[2172] The user enters a query.

[2173] The user inputs a query into the input field of the terminal. For example, the user inputs "Bonjour, comment puis-je vous aider?" The input data is saved in text format on the terminal.

[2174] Step 2:

[2175] The terminal sends the input data to the server.

[2176] The terminal converts the data entered by the user into JSON format and sends it to the server using the HTTPS protocol. The input data is the text-formatted query, and the output is the JSON-formatted data sent to the server.

[2177] Step 3:

[2178] The server determines the language.

[2179] The server calls the Google Cloud Translation API to determine the language of the received data. The API parses the input data and returns a language code (e.g., "fr" for French). The input is the user's query, and the output is the determined language code.

[2180] Step 4:

[2181] The server changes the settings of the AI ​​chatbot.

[2182] The server changes the AI ​​chatbot's settings based on the determined language code. For example, if the language code is "fr", it changes the settings to support French. The input is the language code, and the output is the changed chatbot settings.

[2183] Step 5:

[2184] The server generates a response.

[2185] The server generates an appropriate response using a generative AI model (e.g., GPT-4) based on the set language. The prompt text is "Generate an appropriate response to a question in French. The user's input is 'Bonjour, comment puis-je vous aider?'." The input is the prompt text and the user's question, and the output is the generated response.

[2186] Step 6:

[2187] The server generates a response and sends it to the terminal.

[2188] The server converts the generated response into JSON format and sends it to the terminal using the HTTPS protocol. The input is the generated response, and the output is the JSON format data sent to the terminal.

[2189] Step 7:

[2190] The terminal displays the response to the user.

[2191] The terminal displays the response received from the server to the user, for example, "Bonjour! Comment puis-je vous aider aujourd'hui?" The input is the response data received from the server, and the output is the text displayed to the user.

[2192] (Application example 2)

[2193] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[2194] Conventional online purchasing support systems have limited support for user questions and insufficient multilingual support, making it difficult to provide adequate service to international users. Furthermore, limited operator resources can result in delayed responses, which can lead to a decline in user satisfaction. Furthermore, real-time translation and response generation are difficult, creating a need for an improved user experience.

[2195] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2196] In this invention, the server includes an AI chatbot means for answering users' questions, supporting online purchases, and using the logs to improve the site; a means for accepting form submissions in multiple languages ​​when the bot cannot resolve the issue, having an operator confirm and respond in an AI-translated language, and having the AI ​​translate and send the response; a means for making the above flow multilingual; a means for translating user inquiries in real time, generating responses using a generative AI model, and translating the responses back into the original language; and a means including an application to be installed on a smartphone. This enables quick and appropriate responses even to international users, and is expected to improve user satisfaction.

[2197] "User" means any person or legal entity making an online purchase.

[2198] "Questions" are questions or concerns users have when making an online purchase.

[2199] "Online purchasing" is the act of purchasing goods or services over the Internet.

[2200] An "artificial intelligence chatbot" is a program that automatically responds to users' questions.

[2201] "Logs" are data that record user activities and inquiries.

[2202] "Site Improvements" refers to improving the functionality and design of the Website to enhance the user experience.

[2203] "Bot Unresolved" refers to when an AI chatbot is unable to provide an appropriate answer to a user's question.

[2204] "Multilingual" refers to multiple different languages, especially languages ​​other than English.

[2205] "Form submission" means the act of a user submitting an inquiry or information through a web form.

[2206] An "operator" is a human representative who responds to user inquiries.

[2207] "Artificial intelligence translation" is the process of converting text from one language to another using artificial intelligence.

[2208] A "verification response" is a formal response provided by an operator to a user's inquiry.

[2209] A "generative AI model" is an artificial intelligence model that generates appropriate responses to user inquiries.

[2210] "Real-time" refers to processing occurring immediately without delay.

[2211] A "smartphone" is a mobile phone that can connect to the Internet and install applications.

[2212] An "application" is a software program designed to perform a specific function.

[2213] A system for implementing this invention includes an artificial intelligence chatbot that answers user questions and assists with online purchasing. The system translates user inquiries in real time, generates responses using a generative AI model, and translates the responses back into the user's native language. The system also includes an application installed on a smartphone.

[2214] Hardware and Software Configuration

[2215] Hardware: Smartphone (iOS or Android)

[2216] Software: OpenAI API (generative AI model), GoogleTrans (translation library)

[2217] Data processing and calculation

[2218] 1. Receive user inquiries:

[2219] A user makes a query through a smartphone application. For example, the user asks in French, "Bonjour, pouvez-vous me recommander un bon livre?"

[2220] 2. Enquiry Translation:

[2221] The server uses GoogleTrans to translate the user's query into English, for example, "Bonjour, pouvez-vous me recommander un bon livre?" to "Hello, can you recommend a good book?"

[2222] 3. Generate a response:

[2223] The server uses OpenAI's generative AI model to generate responses to English queries, such as "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[2224] 4. Retranslating responses:

[2225] The server translates the generated response into the original language (French) using Google Translate. For example, "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald." is translated to "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[2226] 5. Response to the user:

[2227] The translated response is returned to the user, who receives the response in French through a smartphone application.

[2228] Specific examples

[2229] User inquiry: “Bonjour, pouvez-vous me recommander un bon livre?”

[2230] Prompt the generative AI model: "Hello, can you recommend a good book?"

[2231] Generated response: "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald."

[2232] Translated response: "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald."

[2233] In this way, the system can provide multilingual online purchasing support, enabling fast and appropriate responses for international users.

[2234] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2235] Step 1:

[2236] The user makes an inquiry through a smartphone application.

[2237] Input: User's question (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[2238] Output: User query text

[2239] Specific behavior: The user enters a question into the application's input field and presses the submit button.

[2240] Step 2:

[2241] The server uses GoogleTrans to translate the user's query into English.

[2242] Input: User's query text (e.g. "Bonjour, pouvez-vous me recommander un bon libre?")

[2243] Output: Text translated into English (e.g. "Hello, can you recommend a good book?")

[2244] What happens: The server calls the GoogleTrans API to translate the French text into English.

[2245] Step 3:

[2246] The server uses OpenAI's generative AI model to generate responses to English queries.

[2247] Input: Text translated into English (e.g. "Hello, can you recommend a good book?")

[2248] Output: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[2249] What happens: The server calls the OpenAI API, sends a prompt to the generative AI model, and generates a response.

[2250] Step 4:

[2251] The server translates the generated response back into the original language (French) using Google Translate.

[2252] Input: Generated response text (e.g., "Sure, I recommend 'The Great Gatsby' by F. Scott Fitzgerald.")

[2253] Output: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[2254] What happens: The server calls the GoogleTrans API and translates the English response text into French.

[2255] Step 5:

[2256] The server returns the translated response to the user.

[2257] Input: Response text translated into French (e.g., "Bien sur, je recommande 'Gatsby le Magnifique' de F. Scott Fitzgerald.")

[2258] Output: Response text displayed on the user's smartphone

[2259] What happens: The server sends the translated response text to the application, which displays it on the user's smartphone.

[2260] Example 3

[2261] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[2262] The previous system had issues such as delayed responses to user inquiries and difficulty in responding when operator resources were insufficient. It also lacked multilingual support, limiting support for global users. This resulted in issues such as a decline in user satisfaction and insufficient log collection for site improvement.

[2263] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[2264] In this invention, the server includes a means for the automated response device to generate a response to an inquiry using a generative AI model, a means for the automated response device to determine whether escalation is necessary based on the content and complexity of the inquiry, and a means for sending a notification to an operator when it is determined that escalation is necessary. This allows for a prompt and appropriate response to inquiries from users, enables the system to function even when operator resources are insufficient, and enables multilingual support.

[2265] An "automatic answering machine" is a device that automatically generates and provides responses to inquiries from users.

[2266] A "generative AI model" is a model that uses artificial intelligence to generate appropriate responses from text data.

[2267] An "inquiry" is a question or request that a user enters into the system.

[2268] "Escalation" is the process of handing over a response to an operator when it is determined that a more advanced response is required based on the content and complexity of the inquiry.

[2269] "Operator" means a person who directly responds to inquiries from users and provides necessary support.

[2270] "Notification" is information sent from the system to the operator to inform them when escalation is necessary.

[2271] "Multilingual support" refers to the ability of the system to respond to user inquiries in multiple languages.

[2272] The present invention is a system that automatically generates a response to an inquiry from a user and escalates the inquiry to an operator as necessary. A specific embodiment of this system is described below.

[2273] First, a user inputs a query into the system using a device. The device then sends the query to a server. The server then passes the received query to a generative AI model, which generates an appropriate response. For example, OpenAI's GPT-4 can be used as this generative AI model.

[2274] The server evaluates the generated response and determines whether escalation is necessary based on the nature and complexity of the inquiry. If escalation is necessary, the server notifies an operator, who then responds directly to the user's inquiry and provides the necessary support.

[2275] As a concrete example, the following prompt sentence can be input to a generative AI model:

[2276] Example prompt sentence:

[2277] User: Forgot password.

[2278] AI Chatbot: We've sent you a link to reset your password. Please check your email.

[2279] Example prompt sentence:

[2280] User: I think my account has been hacked.

[2281] AI Chatbot: This issue has been escalated to an operator. Please wait a moment.

[2282] The system is able to respond to user inquiries promptly and appropriately, and is designed to continue functioning even when operator resources are insufficient. It also supports multiple languages, enabling it to provide support to users globally.

[2283] The hardware used includes the user's device (PC, smartphone, etc.) and the server, while the software used includes a generative AI model (e.g., GPT-4) and an algorithm for evaluating the content of the query.

[2284] In this way, a system is realized in which the user, the terminal, and the server work together to process inquiries efficiently. The flow of the identification process in the third embodiment will be described with reference to FIG.

[2285] Step 1:

[2286] A user uses a terminal to enter a query into the system, which then transmits the query to the server.

[2287] Type: The user types "I forgot my password" into the chat window on the device.

[2288] Data processing: The terminal formats this text data for transmission to the server.

[2289] Output: The formatted query data is sent to the server.

[2290] Specific operation: The user enters text into the chat window on the device and presses the send button. The device sends the entered text to the server.

[2291] Step 2:

[2292] The server passes the received query to a generative AI model, such as OpenAI's GPT-4, to generate an appropriate response.

[2293] Input: The query data received by the server.

[2294] Data processing: The server converts the query data into the format required to input i...

Claims

[Claim 1] A means for analyzing the content of an inquiry entered by a user regarding a product purchase on an online site, generating a prompt sentence for an AI chatbot using a generative AI model to generate a response, and causing the AI ​​chatbot to generate a response using the generated prompt sentence; A prompt sentence that prompts an operator to determine whether escalation is necessary based on the content of the answer generated by the generative AI and the content or complexity of the inquiry input by the user, and a means for determining whether the AI ​​chatbot is unable to answer using the generative AI; If the AI ​​chatbot determines that it cannot provide an answer, it accepts the user's inquiry through a form submission that can be input in multiple languages, translates the inquiry into a specific language, and presents the translated inquiry to an operator; means for receiving a response to the inquiry from the operator, translating the response into the original language, and presenting the translated response to the user; means for estimating an emotion score indicative of the user's emotion using an emotion engine and tracking changes in the emotion score; means for notifying the operator when the emotion score falls below a certain threshold; a means for processing a purchase request from the user, when the purchase request is received, and for checking inventory and processing an order; a means for notifying the user of the completion of order acceptance through the generative AI model when the inventory check result indicates that the product is in stock; A means for storing a log of a plurality of inquiries from the user and a plurality of answers from the AI ​​chatbot and periodically analyzing the log to generate insights for improving the online site; A system including:

Images