Program, information processing method, and information processing device.

Abstract

The aim is to improve the likelihood of obtaining valid answers using AI-generated responses. [Solution] The present invention is a program that causes a computer to execute a reception step of receiving a prompt from a user requesting the presentation of a method for resolving a problem of an information processing device; an acquisition step of acquiring information indicating the operation history of the information processing device; and a display step of displaying a display that asks the user for additional information related to the information processing device that is used in providing the solution, wherein the reception step further accepts the additional information input by the user, and the display means displays the answer generated based on the information indicating the operation history and the additional information.

Description

Technical Field

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

Background Art

[0002] There is a use case of trying to solve troubles that occur in devices and services using generative AI (Artificial Intelligence). In solving troubles, in order for generative AI to output an answer that leads to the solution of the trouble, it is necessary for the user to correctly input information such as the situation of the trouble that has occurred and the operations performed to generative AI without contradiction.

[0003] In Patent Document 1, as an example of a process for solving troubles, when a user makes an inquiry to an operator, a method of displaying the past operation history to the operator to assist in grasping the trouble situation has been proposed.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] For example, the history of operations such as when an external device is restarted is recorded on a recording medium as a log, but for example, the history of operations such as when a user cleans an external device is not recorded on the recording medium as a log. Therefore, generative AI can grasp the operations recorded as logs through communication, but cannot grasp the operations not recorded in the logs through communication.

[0006] As a result, an answer effective for solving the trouble cannot be output, and there is a possibility that the user cannot solve the trouble.

[0007] This invention was made in view of the above-mentioned problems, and aims to improve the likelihood of obtaining effective answers using generation AI. [Means for solving the problem]

[0008] The program of the present invention, First Information processing device Occurred A first reception step in which a prompt is received from the user requesting a solution to the problem, and the First Information showing the operation history of the information processing device From the server that manages information indicating the operation history of the first information processing device The acquisition process to be acquired, the received prompt, and the information indicating the acquired operation history. 、 The aforementioned First Information showing how to operate the information processing device. and The first input step involves inputting the following into the generating AI: The status of the operation of the first information processing device included in the operation method described above, Entered Before Based on the information showing the operation history and the information showing the operation method The aforementioned generating AI cannot be identified. In this case, the information generated by the generating AI is based on the input prompt, the information indicating the operation history, and the information indicating the operation method. to chase A first display step of displaying information for the user to inquire about additional information, before Note Trouble The additional information used to generate information indicating a solution, This shows the status of the aforementioned operation. A second receiving step of receiving the additional information from the user; a second input step of inputting the additional information received from the user into the generating AI; and the generating AI generating the additional information based on the input prompt, the information indicating the operation method, and the additional information. Trouble A second display step involves displaying information that shows the solution, The status of the operation included in the above operation method, Entered Before Based on the information showing the operation history and the information showing the operation method The aforementioned generating AI can be identified. In this case, the generation AI generates the following based on the input prompt, the information indicating the operation history, and the information indicating the operation method. Trouble A third display step that displays information indicating the solution A second information processing device, which is different from the first information processing device described above.It is characterized by causing a computer to execute.

Advantages of the Invention

[0009] According to the present invention, the answer accuracy of the generative AI is improved.

Brief Description of the Drawings

[0010] [Figure 1] A diagram showing an example of the system configuration and hardware configuration in Example 1 [Figure 2] A diagram showing an example of the software configuration in Example 1 [Figure 3] A diagram showing an example of the database configuration in Example 1 [Figure 4] A diagram showing an example of the agent manifest in Example 1 [Figure 5] A diagram showing an example of the Plugin app manifest in Examples 1 and 2 [Figure 6] A flowchart diagram of the input verification process in Example 1 [Figure 7] A diagram showing an example of the screen displayed during the execution of the input verification process in Example 1 [Figure 8] A diagram showing an example of the system configuration and hardware configuration in Example 2 [Figure 9] A diagram showing an example of the software configuration in Example 2 [Figure 10] A diagram showing an example of the database configuration in Example 2 [Figure 11] A diagram showing an example of the agent manifest in Example 2 [Figure 12] A diagram showing an example of the screen displayed during the execution of the input verification process in Example 2

Modes for Carrying Out the Invention

[0011] Hereinafter, modes for carrying out the present invention will be described with reference to the drawings.

[0012] (Example 1) This document describes one embodiment of a device / service support chatbot system (hereinafter referred to as the support system) using the generation AI according to the present invention. Here, as an example of the support system, a device management service and a generation AI service operating on the cloud will be described. Furthermore, as an example of a device targeted by the support system, a printer will be used as an example in this embodiment. Note that devices other than printers, such as MFPs (Multifunction Peripherals), multifunction printers, fax machines, and projectors, will also be included. In the following, the support system may also be referred to as the device / service support chatbot system.

[0013] Furthermore, the generation AI cloud 120, client computer 140, and printer 160 are examples of information processing devices and control devices, respectively. The support system according to the present invention can be used not only for inquiries about equipment but also for inquiries about services.

[0014] In this application, "generative AI" refers to a technology that automatically generates various types of content, such as text, images, music, and videos, similar to those created by humans, by utilizing deep learning and machine learning techniques.

[0015] In this embodiment, "contradictory to facts" refers to a situation where the content entered by the user differs from the operation history recorded in the log, or where the content entered by the user is incomplete when compared with the actual operation history.

[0016] Figure 1 is a block diagram illustrating an example of the system configuration and hardware configuration of the support system in this embodiment.

[0017] The support system consists of a device management cloud 100, a generation AI cloud 120, a client computer 140, and a printer 160, all connected via a network 180. The general-purpose computers that realize the device management cloud 100 and the generation AI cloud 120 are implemented, for example, by utilizing hardware resources supplied on demand through virtualization technology. The device management cloud 100 and the generation AI cloud 120 may also be servers. The client computer 140 has the configuration of a general-purpose computer, such as a personal computer. The network 180 may be either wired or wireless.

[0018] In the device management cloud 100, the CPU 101 performs control and calculations based on application programs stored in the ROM 103 or external memory 110. CPU stands for Central Processing Unit, and ROM stands for Read Only Memory. Furthermore, the CPU 101 comprehensively controls each device connected to the system bus 111. The CPU 101 also opens various registered windows based on commands indicated by a mouse cursor (not shown) on the display 109 and performs various data processing operations.

[0019] Here, the display 109 functions as a display unit. It may also function as an operation unit. Furthermore, each process may be executed not only by the mouse cursor, but also by input via the keyboard 108.

[0020] RAM102 functions as the main memory and work area of ​​the CPU101. RAM stands for Random Access Memory. ROM103 is a read-only memory that functions as a storage area for basic I / O programs, etc. I / O stands for Input / Output. The ROM103 or external memory 110 stores the operating system program (OS), which is the control program for the CPU101. Furthermore, the ROM103 or external memory 110 stores files and other various data used during processing based on the above application programs.

[0021] The network interface 104 connects to the network 180 and performs network communication. This allows the CPU 101 to communicate with external devices via the network interface 104. The keyboard interface 105 controls input from the keyboard 108 and a pointing device (not shown). The display interface 106 controls the display on the display 109. The external memory interface 107 controls access to external memory 110, such as a hard disk drive (HDD) or solid-state drive (SSD). External memory 110 stores boot programs, various applications, user files, editing files, etc.

[0022] The device management cloud 100 operates with the CPU 101 executing the basic I / O program and OS written to ROM 103 or external memory 110. The basic I / O program is written to ROM 103, and the OS is written to either ROM 103 or external memory 110. When the computer is powered on, the initial program load function in the basic I / O program writes the OS from ROM 103 or external memory 110 to RAM 102, and the OS starts running.

[0023] The system bus 111 connects each device, enabling each part to communicate with one another. Hardware resources such as the CPU 101, ROM 103, and external memory 110 that constitute the device management cloud 100 are supplied on demand using virtualization technology. Because these hardware resources are supplied on demand using virtualization technology, the device management cloud 100 is configured as a virtual server on a cloud computing environment.

[0024] The hardware configuration of the generation AI cloud 120 and client computer 140 is the same as that of the equipment management cloud 100, so the explanation is omitted.

[0025] In printer 160, the network interface 161 connects to network 180 for network communication. The USB (Universal Serial Bus) interface 162 connects directly to client computer 140 for communication. Printer 160 supports both network connection via network interface 161 and USB connection via USB interface 162.

[0026] The CPU 163 outputs an image signal as output information to the printer 169 via the printer I / F 168 connected to the system bus 172, based on a control program. The control program is stored in ROM 165 or external memory 171. The CPU 163 is configured to communicate with a computer via network I / F 161 or USB I / F 162. Furthermore, the CPU 163 executes processing based on application programs stored in ROM 165 or external memory 171.

[0027] RAM164 functions as the main memory and work area of ​​the CPU163, and is configured to allow memory capacity expansion by optional RAM connected to an expansion port (not shown). RAM164 is used for output information expansion area, environment data storage area, NVRAM (Non-Volatile Random Access Memory), etc. ROM165 is a read-only memory that functions as a storage area for basic I / O programs, etc.

[0028] External memory 171 corresponds to a hard disk drive (HDD), solid state drive (SSD), IC card, etc., and stores various data such as boot programs, various applications, user files, and editing files. This ROM 165 or external memory 171 stores the control program and application programs of the CPU 163, font data used when generating the above output information, and information used on the printer 160.

[0029] The operation unit I / F 166 manages the interface with the operation unit 167 and outputs image data to be displayed to the operation unit 167. It also receives information input by the user via the operation unit 167. The operation unit 167 corresponds to an operation panel or touch panel equipped with switches and LED indicators for operation, and transmits information input by the user to the operation unit I / F 166. The operation unit 167 may include a keyboard or a display. In this embodiment, the operation unit 167 has a display unit such as a display or touch panel, and various screens and information are displayed on it.

[0030] The printer interface 168 outputs an image signal (an example of output information) to the printer 169 (printer engine). The external memory interface (memory controller) 170 controls access to external memory 171 such as a hard disk drive (HDD), solid-state drive (SSD), or IC card. Furthermore, the aforementioned external memory is not limited to one; at least one or more may be provided, and the system may be configured to allow connection of multiple external memories. In addition, the printer 160 may have an NVRAM (not shown) and be configured to store printer mode setting information from the operation unit 167. The system bus 172 connects each device, thereby enabling each part to communicate with one another.

[0031] In this system, any number of device management clouds 100, generation AI clouds, client computers 140, and printers 160 can be connected, and multiple units of each may be connected.

[0032] Figure 2 is a block diagram illustrating an example of the software configuration of the support system in this embodiment.

[0033] First, the software configuration of the device management cloud 100 is shown. Each software function is implemented through the control of the CPU 101 of the device management cloud 100.

[0034] In the device management cloud 100, the device management application 202 and each module are stored and managed, for example, in external memory 110. However, the storage location is not limited to external memory 110; other storage media may be used, or the software may be stored outside the device management cloud 100 and accessed when needed.

[0035] These are program modules that are loaded into RAM 102 and executed at runtime by the OS or modules that utilize it. Furthermore, the device management application 202 can be added to external memory 110 HDDs or SSDs supplied on demand via virtualization technology in a cloud computing environment.

[0036] The network module 200 communicates with the generated AI cloud 120 and printer 160 via a network using any communication protocol.

[0037] The web server service module 201 provides a service that responds with an HTTP response when it receives an HTTP request. The web server service module 201 may also request the device management application 202 to generate the HTTP response. HTTP stands for HyperText Transfer Protocol, and is a type of communication protocol.

[0038] The device management application 202 is an application that manages the printer 160 connected to the device management cloud 100 and network 180. The device management application 202 is implemented as a program that processes requests to a Web API provided, for example, by the Web server service module 201. API stands for Application Programming Interface.

[0039] As described above, the device management application 202, together with the Web server service module 201, implements a cloud service for managing the printer 160. In the device management application 202, the Web API module 203 generates HTTP responses by calling each module as needed in response to requests from the Web server service module 201.

[0040] As an example of a module called from Web API module 203, we will explain using the device management module 204. Of course, Web API module 203 may call other modules as well.

[0041] The device management module 204 obtains device information and logs from the printer 160, which is connected to the device management cloud 100 via the network module 200 and the network 180. Any communication protocol can be used to obtain device information and logs from the printer 160. An example of a communication protocol used by the device management module 204 is HTTPS (Hypertext Transfer Protocol Secure).

[0042] The device management module 204 stores device information obtained from each module of the printer 160 in the tables of the database server service module 205, which will be described later. It also retrieves (obtains) device information from each table as needed.

[0043] The database server service module 205 manages data and stores and retrieves data in response to requests from other modules. The database server service module 205 may reside on a different device from the device management cloud 100, or it may be a database service on a cloud computing environment, as long as it is accessible from the device management application 202.

[0044] Figure 3 shows an example of the table configuration within the database server service module 205. Figure 3(a) shows the device management table 300, Figure 3(b) shows the job log management table 301, Figure 3(c) shows the operation log management table 302, Figure 3(d) shows the sensor log management table 303, and Figure 3(e) shows the troubleshooting work rule management table 304. Note that the table configuration in Figure 3 is merely an example, and a different table configuration may be used.

[0045] The device management table 300 is a table stored in the external memory 171 of the device management cloud 100, and is a table that manages device information related to the printer 160 managed by the device management application 202.

[0046] The information managed in the device management table 300 is as follows: for example, Device ID, Device Name, Model Name, IP Address, Serial Number, and Last Updated. Here, the Device ID is an identifier that uniquely identifies printer 160, and the Last Updated date and time indicates the last update date and time when the record was updated with information obtained from printer 160.

[0047] The job log management table 301 is a table stored in the external memory 171 of the device management cloud 100, and is a table that stores job logs acquired by the device management module 204 from the printer 160.

[0048] The information managed in the job log management table 301 includes, for example, the device identifier (Device ID), job log identifier (Job Log ID), and job type (Job Type). Other information managed includes the job execution start time (Start Time), job execution end time (End Time), job execution user name (User Name), job execution result (Result), and job execution result error code (Error Code).

[0049] Here, a "job" refers to a processing task, such as printing, that a user can perform on printer 160. The Job Log ID is an identifier that uniquely identifies the log of a job. The Job Execution Result Error Code is a code used to uniquely identify the cause of an error in a job.

[0050] The operation log management table 302 is a table stored in the external memory 171 of the device management cloud 100, and is a table that stores operation logs acquired by the device management module 204 from the printer 160.

[0051] The information managed in the operation log management table 302 is as follows: for example, Device ID, Operational Log ID, Operation Type, Date and Time of Operation, and User Name of the user who performed the operation.

[0052] Here, the Operational Log ID is an identifier that uniquely identifies the operation log. The Operation Type is the type of operation that can be performed on printer 160, such as a restart.

[0053] The sensor log management table 303 is a table stored in the external memory 171 of the device management cloud 100, and is a table that stores sensor logs acquired by the device management module 204 from the printer 160.

[0054] The information managed in the sensor log management table 303 is as follows: for example, Device ID, Sensor Log ID, Sensor Type, Sensor Information, and Log Acquisition Date and Time.

[0055] Here, the Sensor Log ID is an identifier that uniquely identifies the sensor log. The Sensor Type is the type of sensor information that can be obtained from the printer 160, such as Wi-Fi signal strength.

[0056] The troubleshooting work rule management table 304 is a table stored in the external memory 171 of the equipment management cloud 100, and is a table that manages the recommended work rules for troubleshooting the printer 160.

[0057] The information managed in the troubleshooting rule management table 304 includes, for example, the rule identifier (Rule ID), troubleshooting type (Troubleshooting Type), and rule (Rule).

[0058] Here, the Rule ID is an identifier that uniquely identifies a work rule. The Troubleshooting Type is the type of troubleshooting required for printer 160, such as resolving Wi-Fi connection problems. The Rule is the actual recommended work rule for troubleshooting, such as restarting printer 160 and the Wi-Fi router, checking the Wi-Fi connection status, and reinstalling the printer driver when resolving Wi-Fi connection problems.

[0059] Next, returning to Figure 2, an example of the software configuration of the generation AI cloud 120 is shown. Each software function is implemented under the control of the CPU 101 of the generation AI cloud 120. In addition, each application module that makes up the generation AI cloud 120 is stored and managed in external memory 110.

[0060] These are program modules that are loaded into RAM 102 and executed at runtime by the OS or modules that utilize it. Furthermore, each application constituting the generated AI cloud 120 can be added to external memory 110 (HDD or SSD) supplied on demand via virtualization technology on the cloud computing environment.

[0061] The network module 220 communicates with the device management cloud 100 and the client computer 140 using any communication protocol. The web server service module 221 provides a service that responds with an HTTP response when it receives an HTTP request from the generation AI client application 242 on the client computer 140. The web server service module 221 may also request the generation AI application 223 to generate an HTTP response.

[0062] Agent 222 is an artificial intelligence system application that generates responses to user input such as text, and is configured to be customizable according to purpose and application. In this embodiment, Agent 222 consists of the generation AI application 223 (described later), the AI ​​orchestrator 227, the user data layer 228, the AI ​​base model 229, and an agent manifest including an agent description.

[0063] Figure 4 shows an example of the implementation of the agent manifest for agent 222 in this embodiment. The agent manifest states that agent 222 is an agent whose purpose is to point out inconsistencies between the user's declaration and the facts based on the printer 160 logs and troubleshooting work rules, and to prompt the user to confirm and correct. In this embodiment, the action taken in response to user input is determined based on the contents of this agent manifest through the process described later in Figure 6.

[0064] The Generative AI Application 223 is an artificial intelligence system application that manages the UX (User Experience) by generating responses to user input such as text. The Generative AI Application 223 is an element that makes up Agent 222 and works in conjunction with the AI ​​Orchestrator 227, AI Platform Model 229, AI Infrastructure 230, etc., which will be described later. The Generative AI Application 223 is implemented as a program that executes processing in response to requests to the Web API provided by the Web Server Service Module 221, for example. As described above, the Generative AI Application 223, together with the Web Server Service Module 221, realizes a cloud service for Generative AI.

[0065] In the generation AI application 223, the front-end application 224 receives input from the user and generates and returns a response in cooperation with the AI ​​orchestrator 227, AI base model 229, AI infrastructure 230, etc. In the front-end application 224, the Web API module 225 generates an HTTP response by calling each module as needed in response to a request from the Web server service module 221.

[0066] The Generative AI Application 223 can be extended, enhanced, and customized by adding Generative AI knowledge, skills, and experiences. One way to extend the functionality of the Generative AI Application 223 is through a plug-in mechanism that extends its skills by interacting with external web services using natural language.

[0067] Plugin application 226 is an application that adds specific functions to generation AI application 223 through the plug-in mechanism of generation AI application 223. In this embodiment, Plugin application 226 implements printer management functions for printer 160 by calling the web service of device management application 202 on device management cloud 100. Plugin application 226 then adds printer management functions for printer 160 to generation AI application 223 through the plug-in mechanism of generation AI application 223.

[0068] The Plugin application 226 consists of an application manifest, which includes a description of the application in natural language. Figure 5 shows an example implementation of the application manifest for Plugin application 226 in this embodiment. The application manifest states that Plugin application 226 is an application that has the skill to provide logs and work rules for printer 160. The application manifest also contains commands to call the linked web service and parameters to be used when calling the web service.

[0069] As shown in Figure 6, the process described later selects a plugin with the skills suitable for generating responses to user input based on the contents of this application manifest, and the command is executed.

[0070] Furthermore, the troubleshooting steps for resolving Wi-Fi issues in this embodiment include restarting the printer, restarting the Wi-Fi router, checking the Wi-Fi connection status, and reinstalling the printer driver.

[0071] Let's return to the explanation in Figure 2. The AI ​​orchestrator 227 is an element that makes up agent 222, and it operates in the background from natural language input by the user to natural language output (response), controlling business logic such as selecting and executing plugins suitable for response generation.

[0072] The user data layer 228 provides a means for storing and accessing user data. User data is information associated with the user's login account, such as the user's email address, calendar information, the team the user belongs to, colleagues, and accessible files. The AI ​​foundation model 229 consists of generative AI models such as LLMs (Large Language Models). The AI ​​infrastructure 230 configures and manages the cloud and GPUs (Graphics Processing Units) that constitute the infrastructure of the generative AI cloud.

[0073] Next, an example of the software configuration of client computer 140 is shown. Each software program's function is implemented under the control of the client computer 140's CPU 101. Furthermore, each module constituting client computer 140 is stored and managed in external memory 110. These are program modules that are loaded into RAM 102 and executed at runtime by the OS or modules that utilize them.

[0074] The network module 240 communicates with the generated AI cloud 120 and printer 160 via a network using any communication protocol. The printer driver 241 generates a print job and sends the print job to the printer 160 via the network module 240.

[0075] The printer driver 241 receives the print job execution results from the printer 160 via the network module 240, and the received results are displayed on the display 109 of the client computer 140.

[0076] The generation AI client application 242 sends an HTTP request message to the generation AI cloud 120 via the network module 240 and receives an HTTP response message from the generation AI cloud 120. The HTTP response message received by the generation AI client application 242 is displayed on the display 109 of the client computer 140. Access from the client computer 140 to the generation AI cloud 120 is performed through the generation AI client application 242.

[0077] Next, an example of the software configuration of printer 160 is shown. Each software function of printer 160 is implemented under the control of the printer 160's CPU 163. Furthermore, in printer 160, various modules are stored and managed in ROM 165 or external memory 171, and are loaded into RAM 164 and executed at runtime.

[0078] The I / F module 260 communicates with the device management cloud 100 and client computers 140 using any communication protocol via the USB module 261 or the network module 262. The USB (Universal Serial Bus) module 261 connects directly to the client computer 140 for communication. The network module 262 connects to the network 180 for network communication.

[0079] The printer 160 supports both USB connection via the USB module 261 and network connection via the network module 262, and the connection method can be switched in the device settings.

[0080] The print module 263 receives print jobs sent from the printer driver 241 of the client computer 140 via the I / F module 260 and executes the print jobs. The print module 263 also creates a log of the print job execution results and saves it to the device management module 264.

[0081] The device management module 264 manages the device information of the printer 160. The device management module 264 receives requests for device information acquisition from the device management module 204 of the device management cloud 100 via the network module 262 and returns the device information. The device management module 264 also receives requests for log acquisition from the device management module 204 of the device management cloud 100 via the network module 262 and returns the job execution result log. The UI module 265 draws the UI displayed on the operation unit 167 of the printer 160 and receives user input values ​​entered by the user through UI operations on the operation unit 167.

[0082] Next, using Figure 6, we will explain an example of the process by which the generating AI cloud 120 verifies the facts in response to the user-inputted prompt for a description of the trouble situation regarding the printer 160, points out inconsistencies with the facts, and prompts the user to confirm and correct the explanation. In addition, using Figure 7, we will show an example of the screen display of the generating AI client application 242 that is shown to the user. The screen in Figure 7 is an example of a screen generated by the generating AI client application 242 controlled by the CPU 101 of the client computer 140, and is displayed on the display 109.

[0083] This screen may also be displayed on the control panel 167 of the printer 160. In this case, the response generated by the language model through the following process will be displayed on the control panel 167 of the printer 160. This process is triggered when the user enters a prompt on the generated AI client application 242 describing the trouble situation with the printer 160.

[0084] Furthermore, each software program 220-230 of the generation AI cloud 120 is executed and operated by the CPU 101 of the generation AI cloud 120. In addition, each software program 200-205 of the device management cloud 100 is executed and operated by the CPU 101 of the device management cloud 100.

[0085] In S600, the front-end application 224 of the generative AI application 223, which runs on the generative AI cloud 120, receives a prompt from the generative AI client application 242, which runs on the client computer 140. Here, a prompt refers to an instruction entered by the user using natural language on the generative AI client application 242. In other words, the front-end application 224 receives a natural language prompt from the user.

[0086] This embodiment describes troubleshooting Wi-Fi connection issues related to printer 160. Therefore, this embodiment explains an example where the user enters a prompt containing a string describing the problem with printer 160. Note that the input prompt describing the problem with printer 160 is not limited to Wi-Fi connection issues as in this embodiment, but may contain other information.

[0087] The prompt entered by the user can be a string of text describing the situation of the problem that has occurred, a string of text asking a question such as "How do I do ~?", or a string of text giving an instruction such as "Tell me the solution to ~". In any context, the user's act of entering a prompt to ask for a response from the generating AI can be said to be a request for the presentation of information.

[0088] In the example screen display in Figure 7, the logged-in username 700 is displayed, and the prompt content entered by the user is displayed in the prompt area 701. The user can input a string (prompt) into an input field (not shown) using, for example, a keyboard, and instruct the generated AI to input (send) the prompt by selecting the send button (object) (not shown).

[0089] In S601, the front-end application 224 checks the prompt entered in S600 from the perspectives of fairness, reliability, safety, privacy, security, inclusivity, transparency, and accountability. If any of the prompt checks fail, the conversation is terminated and this process is ended. In this case, a warning message such as "The check failed, so the conversation will be terminated" or "The prompt is inappropriate" may be displayed. If the prompt check by the front-end application 224 is successful, the process proceeds to S602.

[0090] In S602, the AI ​​orchestrator 227 adjusts the prompts based on the operation method of agent 222 defined in the agent manifest. In this embodiment, based on the contents of the agent manifest shown in Figure 4, the prompts are adjusted to point out inconsistencies with the facts based on the printer 160 log and troubleshooting work rules in response to the user's input of a trouble description.

[0091] In S603, the AI ​​orchestrator 227 acquires context. Here, context refers to information that indicates the background, circumstances, etc., related to the user's conversation. Examples of context acquired here include information about the user's past conversations and activities with the generative AI, which are used to provide personalized responses and suggestions based on the user's needs and preferences.

[0092] In S604, the AI ​​orchestrator 227 updates the context based on user data obtained from the user data layer 228. The user data obtained here includes information about the relationship between the user, the user's activities, and the data of the organization to which the user belongs, such as information about the user's past emails, chats, documents handled, and meetings attended. Updating refers to processing such as adding new information to the obtained context, organizing it, or both.

[0093] The AI ​​orchestrator 227 adjusts the prompts based on the updated context and inputs them into the LLM of the AI ​​base model 229. Here, prompt adjustment involves, for example, removing forbidden words or editing the context to make it easier for the LLM to interpret, and this adjustment is not a mandatory process.

[0094] When the AI ​​orchestrator 227 receives a response (output) from the LLM, it proceeds to process S605 to obtain additional data.

[0095] At S605, the AI ​​orchestrator 227 requests the Plugin application 226 to retrieve Plugin information.

[0096] At S640, the Plugin application 226 receives a request from the AI ​​orchestrator 227 to retrieve Plugin information. At S641, the Plugin application 226 returns the Plugin information, including the application manifest, to the AI ​​orchestrator 227.

[0097] In S606, the AI ​​orchestrator 227 receives the Plugin information returned from the Plugin application 226. This allows the AI ​​orchestrator 227 to obtain the Plugin information.

[0098] In S607, the AI ​​orchestrator 227 determines whether to call Plugin application 226 to log printer 160 based on the description of the application command contained in the application manifest of Plugin application 226. The AI ​​orchestrator 227 may also make this determination by querying, for example, the LLM of the AI ​​base model 229 to find the most suitable Plugin application 226 for the answer based on the description of the application command contained in the application manifest. The description of the application command contained in the application manifest here refers to Plugin application 226.

[0099] If the AI ​​orchestrator 227 determines that it will call the Plugin application 226, the process proceeds to S608; otherwise, it proceeds to S612.

[0100] In this embodiment, it is assumed that a Plugin application 226, which has management functions for the printer 160, is called in order to obtain logs from the printer 160 and point out any inconsistencies with the facts regarding the input prompt.

[0101] The AI ​​orchestrator 227 may perform other controls to ensure that the Plugin application 226, which has management functions for the printer 160, is reliably called. For example, the AI ​​orchestrator 227 may always call the Plugin application 226. Alternatively, for example, if the AI ​​orchestrator 227 can communicate with the printer 160 via a network, it may be configured to directly obtain information stored in the printer 160 without calling various applications.

[0102] In S608, the AI ​​orchestrator 227 generates a new prompt that includes the prompt entered by the user, the updated context, and information from the Plugin application 226, and inputs it into the LLM of the AI ​​base model 229. Based on the response from the LLM to the input information, the AI ​​orchestrator 227 obtains the function and parameters for calling the web service declared in the Plugin application 226.

[0103] In S609, the AI ​​orchestrator 227 uses the function and parameters obtained in S608 to call the web service declared in the Plugin application 226. In the example screen display in Figure 7, the prompt area 701 displays in natural language an explanation to the user and a request for permission to call the web service declared in the Plugin application 226 in S609, as a response from the generated AI.

[0104] Furthermore, the user can respond to this by selecting from three options: "Always Allow," "Allow Once," or "Cancel." In this embodiment, we will continue the explanation assuming the user has selected either "Always Allow" or "Allow Once." Note that the process may terminate if the user selects "Cancel."

[0105] In S660, the device management application 202 of the device management cloud 100 receives a log acquisition request via a web service call. In S661, the device management application 202 acquires the logs of the printer 160 stored in the database server service module 205 via the device management module 204. In this embodiment, the device management application 202 identifies the target printer 160 from the device management table 300 and acquires the respective logs from the job log management table 301, the operation log management table 302, and the sensor log management table 303.

[0106] Furthermore, the method for acquiring logs from the printer 160 does not necessarily have to be via the cloud. For example, the user may upload logs acquired from the printer 160 to the generated AI client application 242 when prompted. Also, the logs to be acquired are not limited to those stored within the printer 160. For example, logs from devices such as PCs, servers, and smartphones that the printer 160 communicates with may also be acquired, and the device management application 202 may acquire these as well.

[0107] In S662, the device management application 202 returns (sends) the printer 160 log acquired in S661 to the AI ​​orchestrator 227. In S610, the AI ​​orchestrator 227 receives the log returned from the device management cloud 100. In S611, the AI ​​orchestrator 227 integrates the log received in S610 into the context.

[0108] In S612, the AI ​​orchestrator 227 determines whether to call Plugin application 226 to obtain work rules based on the description of the app command contained in the app manifest of Plugin application 226. The AI ​​orchestrator 227 may also make this determination by querying, for example, the LLM of the AI ​​base model 229 to search for the most suitable Plugin application 226 in response based on the description of the app command contained in the app manifest. Here, the description of the app command contained in the app manifest refers to Plugin application 226.

[0109] If the AI ​​orchestrator 227 decides not to call the Plugin application 226, the process proceeds to S613; otherwise, it proceeds to S617.

[0110] In this embodiment, it is assumed that a Plugin application 226, which has the functionality to manage the printer 160, is called in order to acquire the work rules and check for any omissions or errors in the input prompts. The AI ​​orchestrator 227 may perform other controls to ensure that the Plugin application 226, which has the functionality to manage the printer 160, is reliably called.

[0111] In S613, the AI ​​orchestrator 227 generates a new prompt that includes the prompt entered by the user, the updated context, and information from the Plugin application 226, and sends (inputs) it to the LLM of the AI ​​base model 229. Based on the response from the LLM, the AI ​​orchestrator 227 obtains the function and parameters for calling the web service declared in the Plugin application 226.

[0112] In S614, the AI ​​orchestrator 227 uses the function and parameters obtained in S613 to call the web service declared in the Plugin application 226. In the example screen display in Figure 7, the prompt area 701 displays in natural language an explanation to the user about calling the web service declared in the Plugin application 226 in S614 and requests permission to do so, as a response from the generated AI.

[0113] Furthermore, the user can respond to this by selecting from three options: "Always Allow," "Allow Once," or "Cancel." In this embodiment, we will continue the explanation assuming the user has selected either "Always Allow" or "Allow Once." Note that the process may terminate if the user selects "Cancel."

[0114] Note that while Figure 7 shows an example where the explanation and permission acquisition in S614 and S609 are displayed simultaneously, the system is not limited to this. In other words, the system may be configured to explain and confirm to the user separately whether to allow log searching and to search for work rules in each step.

[0115] In S663, the device management application 202 of the device management cloud 100 receives a request to retrieve work rules via a web service call. In S664, the device management application 202 retrieves work rules via the device management module 204, which contain strings indicating how to troubleshoot printer 160 stored in the database server service module 205. In this embodiment, the device management application 202 retrieves work rules from the troubleshooting work rule management table 304.

[0116] In S665, the device management application 202 returns (sends) the troubleshooting work rules for printer 160, which were acquired in S664, to the AI ​​orchestrator 227. In S615, the AI ​​orchestrator 227 receives the work rules returned from the device management cloud 100.

[0117] In S616, the AI ​​orchestrator 227 integrates the work rules received in S615 into the context.

[0118] In S617, the AI ​​orchestrator 227 determines whether it is appropriate to generate a response (answer). If the AI ​​orchestrator 227 determines that it is appropriate to generate a response, it proceeds to S618; otherwise, it returns to S604. In other words, if the AI ​​orchestrator 227 determines that it is not appropriate to generate a response, it repeats the necessary processing from S604 to S616, continuously performing inference by LLM until it determines in S617 that it is appropriate to generate a response.

[0119] In S618, the AI ​​orchestrator 227 sends (inputs) all the information collected in the above process to the LLM of the AI ​​base model 229 and generates a response. However, it is not mandatory for all the information collected in the above process to be input into the LLM; if there is any information that is not necessary for generating the response, that information does not need to be input into the LLM.

[0120] AI Orchestrator 227 checks the generated responses from the perspectives of fairness, reliability, safety, privacy, security, inclusivity, transparency, and accountability.

[0121] In S619, the AI ​​orchestrator 227 returns a response to the front-end application 224. The front-end application 224 returns a response to the generating AI client application 242 running on the client computer 140. As a result, the generating AI client application 242 displays the response (answer).

[0122] The response is a troubleshooting method generated in response to the nature of the problem, and since its content varies widely, a diagrammatic explanation is omitted. For example, the response (answer) may include general troubleshooting methods for the problem, such as "Perform operation X," "Press button X," or "Check the connection status of X." The problem will be resolved when the user performs the operation according to the displayed troubleshooting method.

[0123] In the example screen display in Figure 7, the prompt area 701 displays in natural language the response from the generating AI, stating that there was a record of successful printing up to three days ago, and that the printer 160 has been restarted and the Wi-Fi connection status has been checked, as confirmed by the acquired logs. On the other hand, the generating AI application 223 cannot confirm from the user input prompts and logs whether the Wi-Fi router restart and printer driver reinstallation, which are included in the acquired work rules, have been performed.

[0124] Therefore, Figure 7 shows an example where, in S617, the AI ​​orchestrator 227 determines that it is not appropriate to generate a response, returns to S604, and attempts to obtain additional information by querying the user on the prompt area 701.

[0125] Here, the generated AI application 223 displays a message asking the user whether a specific operation has been performed. The text "Have you tried restarting the Wi-Fi router?" in Figure 7 corresponds to this. Note that the displayed content is not limited to this; for example, a message asking about the operation performed, such as "What operation did you perform?", may also be displayed.

[0126] When these inquiries are displayed, the user provides additional information. In Figure 7, this corresponds to the strings "Yes, but it didn't work" and "No, I haven't tried." When the inquiries are displayed, the user enters these strings as prompts. The additional information related to the information processing device is received when the user enters these prompts into an input field (not shown).

[0127] Based on its interaction with the user in prompt area 701, the AI ​​orchestrator 227 confirms that the Wi-Fi router has been restarted but the printer driver has not been reinstalled. Upon confirming this, the AI ​​orchestrator 227 determines in S617 that it is possible to generate a response.

[0128] Furthermore, Figure 7 displays the results of fact-checking performed on the user input prompt in the prompt area 701, using the response generated by the AI ​​orchestrator 227 in S618, in natural language. Here, it is shown that there was a record of successful printing up to three days ago, that the printer 160 has been restarted and the Wi-Fi connection status has been checked, and that the Wi-Fi router restart included in the work rules has also been performed. On the other hand, it is shown that the printer driver reinstallation included in the work rules has not been performed.

[0129] Furthermore, the user's prompt included a statement that they had tapped the casing but the problem was not resolved. However, LLM determined that this information did not conform to the work rules and did not contribute to resolving the problem, and therefore it was excluded from the fact-finding results.

[0130] In the example shown in Figure 7, only tasks that have been verified are displayed on the prompt area 701. However, tasks that have been verified and those that cannot be verified may be displayed separately. Furthermore, tasks that do not conform to the work rules may be indicated as not contributing to troubleshooting and therefore unnecessary. If there is a conflict between the factual verification from the log and the user's report, the system may be able to set which to prioritize and switch whether or not to perform the factual verification.

[0131] Finally, at S620, the front-end application 224 determines whether the conversation is ending. If it determines that the conversation will continue based on user input or other factors (Yes at S620), it returns to S600. If it determines that the conversation is ending, it terminates the process.

[0132] Through the above process, in response to a prompt requesting a solution to a problem, the generating AI cloud 120 queries for additional information based on the log showing the device's operation history and the work rules for resolving the problem, prompting the user to confirm and correct their explanation. Furthermore, if there is a contradiction between the operation history revealed by the log and the user's response, the AI ​​cloud can confirm this with the user.

[0133] As a result, the generating AI can identify the cause of a problem from sufficient information and provide solutions. In other words, the likelihood of obtaining an effective answer from the generating AI increases.

[0134] (Example 2) In Example 1, when a user of printer 160 enters a prompt describing the trouble situation to request a solution to the problem, the generating AI cloud 120 requests additional information and prompts the user to confirm and correct the explanation.

[0135] This technology is not limited to the examples above; for example, it can also be applied when a service technician for an MFP (Multifunction Peripheral) uses the Generation AI Cloud 120 to create a work record after troubleshooting the MFP. In this embodiment, we show an example in which the Generation AI Cloud 120 points out inconsistencies with the facts in response to prompts describing the work entered by the MFP service technician, prompting the service technician to confirm and correct the explanation. The following explanation will focus on the differences from Embodiment 1.

[0136] Figure 8 is a block diagram illustrating an example of the system configuration and hardware configuration of the support system in this embodiment.

[0137] The support system consists of a device management cloud 100, a generation AI cloud 120, a client computer 140, and an MFP 800, all connected via network 180.

[0138] In the MFP800, the scanner I / F 801 receives an image signal (an example of input information) from the scanner 802 (scanner engine). The rest of the hardware configuration is the same as that of the printer 160 in Figure 1 of Example 1, so a description is omitted. The system configuration and hardware configuration other than the MFP800 are the same as those in Figure 1 of Example 1, so a description is omitted.

[0139] Figure 9 is a block diagram illustrating an example of the software configuration of the support system in this embodiment.

[0140] In the MFP800, the scan transmission module 900 receives scan commands from the user via the UI module 265 and generates and executes scan jobs and scan data transmission jobs. Communication protocols such as email or SMB (Server Message Block) are used for transmitting scan data. A log of the scan job and transmission job execution results is also created and saved in the device management module 264.

[0141] The FAX module 901 receives FAX jobs sent from FAX devices or MFPs (not shown) via the network module 262. These received FAX jobs are then either printed via the print module 263 or forwarded to other FAX devices or MFPs via the network module 262. In addition, it receives FAX transmission instructions from users via the UI module 265 and generates and executes FAX transmission jobs. It also creates logs of FAX reception and transmission job execution results and saves them in the device management module 264.

[0142] The software configuration of the MFP800 is the same as that of printer 160 in Figure 2 of Example 1, so the explanation is omitted. The software configuration of devices other than the MFP800 is the same as that of printer 160 in Figure 2 of Example 1, so the explanation is omitted.

[0143] Figure 10 shows an example of the table configuration within the database server service module 205 in this embodiment. The table configuration in Figure 10 is the same as that in Figure 3 of Embodiment 1, so no explanation is given, but the data is stored in each table according to this embodiment. Note that the table configuration in Figure 10 is merely an example, and a different table configuration may be used.

[0144] The device management table 300 stores device information related to the MFP800, and the job log management table 301 stores jobs such as copying that can be executed on the MFP800. The operation log management table 302 stores operation logs such as drum replacement and automatic gradation correction performed on the MFP800. The sensor log management table 303 stores sensor logs such as drum life warnings that can be obtained on the MFP800. The troubleshooting work rule management table 304 stores rules that are recommended for troubleshooting the MFP800, such as checking image quality with a copy test, checking drum life, replacing the drum unit, and cleaning the ADF and casing.

[0145] Figure 11 shows an example of the agent manifest implementation for agent 222 in this embodiment. This manifest states that agent 222 is an agent for supporting work records, which points out inconsistencies between the service technician's declaration and the facts based on the MFP800 logs and work rules that show how to resolve problems, and prompts verification and correction.

[0146] Figure 12 shows an example of the screen display of the Generative AI Client Application 242 shown to a service technician. The screen in Figure 12 is displayed on the Generative AI Client Application 242 running on the client computer 140. In this embodiment, it is assumed that a service technician will use this application once to record the repair after completing equipment repair, for example, by performing the actions described in Embodiment 1 at a customer site.

[0147] Furthermore, the process by which the generating AI cloud 120 verifies the facts in response to prompts describing troubleshooting work related to the MFP800, confirms additional information, and points out inconsistencies with the facts is the same as in Figure 6 of Example 1, so the explanation is omitted.

[0148] In S600, the front-end application 224 of the generation AI application 223, which runs on the generation AI cloud 120, receives a prompt from the generation AI client application 242, which runs on the client computer 140. In this embodiment, an example is given in which a prompt is received requesting the creation of a work record after a service technician has completed the replacement of the drum unit of the MFP800.

[0149] Note that the input prompt requesting the creation of a work record is not limited to the example in Figure 12. In the example screen display in Figure 12, the logged-in username 700 is displayed, and the prompt area 701 displays the prompt content entered by the service technician.

[0150] In S609, the AI ​​orchestrator 227 uses the function and parameters obtained in S608 to call the web service declared in the Plugin application 226. In the example screen display in Figure 12, the prompt area 701 displays in natural language an explanation to the service technician about calling the web service declared in the Plugin application 226 in S609 and the request for permission.

[0151] In S614, the AI ​​orchestrator 227 uses the function and parameters obtained in S608 to call the web service declared in the Plugin application 226. In the example screen display in Figure 12, the prompt area 701 displays in natural language the explanation to the service technician and the request for permission to call the web service declared in the Plugin application 226 in S614, as a response from the generated AI.

[0152] Note that while Figure 12 shows an example where the explanation and permission acquisition in S614 and S609 are displayed simultaneously, the system is not limited to this. In other words, the system may be configured to explain and confirm to the user separately whether to allow log searching and to search for work rules in each step.

[0153] In S619, the AI ​​orchestrator 227 returns a response to the front-end application 224. The front-end application 224 returns a response to the generating AI client application 242 running on the client computer 140. As a result, the generating AI client application 242 sees the response.

[0154] In the example screen display in Figure 12, the prompt area 701 displays in natural language the response from the generating AI, stating that a drum life warning has been detected, and that the acquired logs confirm that a copy test, drum replacement, and automatic gradation correction have been performed. On the other hand, regarding the cleaning of the ADF and enclosure, the generating AI application 223 cannot confirm the status of the cleaning from the service technician's input prompts and logs. Therefore, in S617, the AI ​​orchestrator 227 determines that it is not appropriate to generate a response, and returns to S604 to attempt to obtain additional information by querying the service technician on the prompt area 701, as shown in Figure 12.

[0155] Based on the interaction with the user in the prompt area 701, the AI ​​orchestrator 227 can confirm that the ADF and chassis have also been cleaned. Upon confirming this, S617 determines that the AI ​​orchestrator 227 is ready to generate a response.

[0156] In Figure 12, the response result from the AI ​​orchestrator 227 in S618 is displayed in natural language on the prompt area 701, showing the results of fact-checking for the user input prompt. It indicates that a drum life warning was detected, a copy test, drum replacement, and automatic gradation correction have been performed, and that the ADF and enclosure cleaning included in the work rules have also been carried out.

[0157] Through the above process, the generating AI cloud 120 prompts the MFP service technician to enter troubleshooting information, inquiring about additional information and pointing out inconsistencies with the facts, prompting the service technician to confirm and correct the explanation. As a result, the possibility of inconsistencies in the troubleshooting information entered by the service technician into the generating AI can be reduced, and the likelihood of obtaining more accurate work records from the generating AI can be increased.

[0158] (Other examples) The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that implements one or more functions. [Explanation of symbols]

[0159] 100 Device Management Cloud 120 Generation AI Cloud 140 client computers 160 Printers 202 Equipment Management Application (Equipment Management Cloud) 222 Agent (Generating AI Cloud) 223 Generative AI Applications (Generative AI Cloud) 226 Plugin applications (Generating AI cloud) 242. Generate AI client applications (client computers) 800 MFP

Claims

1. A first receiving step of receiving a prompt from a user requesting a method for resolving a problem that has occurred in the first information processing device, An acquisition step of obtaining information indicating the operation history of the first information processing device from a server that manages information indicating the operation history of the first information processing device, A first input step involves inputting the received prompt, the acquired operation history information, and the operation method of the first information processing device into the generating AI. If the generating AI cannot identify the status of the operation of the first information processing device included in the operation method based on the input operation history information and the operation method information, a first display step is to display information generated by the generating AI based on the input prompt, the operation history information and the operation method information, which is information for inquiring the user for additional information; A second receiving step of receiving from the user the additional information used to generate information indicating a method for resolving the aforementioned trouble, the additional information indicating the status of the aforementioned operation, A second input step involves inputting the additional information received from the user into the generating AI, A second display step, which displays information indicating a method for resolving the trouble, generated by the generating AI based on the input prompt, the information indicating the operation method, and the additional information, A third display step in which, if the generating AI can identify the status of the operation included in the operation method based on the input operation history information and the operation method information, the generating AI displays information indicating a method for resolving the problem, which is generated based on the input prompt, the operation history information and the operation method information. A program characterized by causing a computer of a second information processing device, which is different from the first information processing device, to execute it.

2. The program according to claim 1, characterized in that the information indicating the operation method is a string of characters indicating a method for resolving the trouble.

3. The program according to claim 1, characterized in that the prompt requesting a method for resolving the trouble is a prompt that includes a string describing the situation of the trouble that occurred in the first information processing device.

4. The program according to claim 1, characterized in that the information for inquiring with the user is information indicating whether or not an operation not included in the operation history has been performed.

5. The program according to claim 1, characterized in that the information for inquiring about the user is information indicating that the operation performed has been inquired about.

6. The program according to claim 1, characterized in that the second reception step receives the additional information by the user inputting a string of natural language.

7. The program according to claim 1, characterized in that the acquisition step further acquires information indicating the operation history of a device with which the first information processing device communicates.

8. A first receiving step of receiving a prompt from a user requesting a method for resolving a problem that has occurred in the first information processing device, An acquisition step of obtaining information indicating the operation history of the first information processing device from a server that manages information indicating the operation history of the first information processing device, A first input step involves inputting the received prompt, the acquired operation history information, and the operation method of the first information processing device into the generating AI. If the generating AI cannot identify the status of the operation of the first information processing device included in the operation method based on the input operation history information and the operation method information, a first display step is performed to display information generated by the generating AI based on the input prompt, the operation history information and the operation method information, which is information for inquiring about additional information from the user. A second receiving step of receiving from the user the additional information used to generate information indicating a method for resolving the aforementioned trouble, the additional information indicating the status of the aforementioned operation, A second input step involves inputting the additional information received from the user into the generating AI, A second display step, which displays information indicating a method for resolving the trouble, generated by the generating AI based on the input prompt, the information indicating the operation method, and the additional information, If the generating AI can identify the status of the operation included in the operation method based on the input operation history information and the operation method information, a third display step is to display information indicating a method for resolving the trouble, which is generated by the generating AI based on the input prompt, the operation history information and the operation method information. An information processing method performed on a second information processing device different from the first information processing device, characterized by having the following:

9. The information processing method according to claim 8, characterized in that the information indicating the operation method is a string of characters indicating a method for resolving the trouble.

10. The information processing method according to claim 8, characterized in that the prompt requesting a method for resolving the trouble is a prompt that includes a string describing the situation of the trouble that occurred in the first information processing device.

11. The information processing method according to claim 8, characterized in that the information for which the user is to be contacted is information indicating whether or not an operation not included in the operation history has been performed.

12. The information processing method according to claim 8, characterized in that the information for which the user is to be contacted is information indicating that the user is to be contacted about the operations that have been performed.

13. The information processing method according to claim 8, characterized in that the second reception step involves the user inputting a string of natural language to receive the additional information.

14. The information processing method according to claim 8, characterized in that the acquisition step further acquires information indicating the operation history of a device with which the first information processing device communicates.

15. A first receiving means that receives prompts from the user requesting solutions to problems that have occurred in the device, An acquisition means for obtaining information indicating the operation history of the said device from a server that manages information indicating the operation history of the said device, A first input means for inputting the received prompt, the acquired operation history information, and the information indicating the operation method of the device into the generating AI, If the generating AI cannot identify the status of the operation of the device included in the operation method based on the input information indicating the operation history and the information indicating the operation method, a first display means is provided which displays information generated by the generating AI based on the input prompt, the information indicating the operation history and the information indicating the operation method, and which is information for inquiring the user for additional information. A second receiving means that receives from the user the additional information used to generate information indicating a method for resolving the aforementioned trouble, the additional information indicating the status of the aforementioned operation, A second input means for inputting the additional information received from the user into the generating AI, A second display means for displaying information indicating a method for resolving the trouble, generated by the generating AI based on the input prompt, the information indicating the operation method, and the additional information, When the generating AI can identify the status of the operation included in the operation method based on the input operation history information and the operation method information, a third display means displays information indicating a troubleshooting method generated by the generating AI based on the input prompt, the operation history information and the operation method information. An information processing device characterized by performing the following actions.

Images