Information processing program, information processing method, information processing system, and information processing terminal

Abstract

To provide an information processing program for efficiently answering queries.SOLUTION: There is provided a program that is executed by a computer which has a control unit, the program comprising: a step in which the control unit displays a first image that includes an inquiry input box for receiving input of an inquiry sentence with respect to a machine learning model; a step in which the control unit inputs, into the machine learning model, the inquiry sentence input into the inquiry input box; and a step in which the control unit displays a second image, which includes a query group and an answer input box for receiving input of an answer to the query group that are generated on the basis of output from the machine learning model.SELECTED DRAWING: Figure 6

Description

[Technical field]

[0001] The present disclosure relates to an information processing program, an information processing method, an information processing system, and an information processing terminal. [Background technology]

[0002] Businesses that provide products and services have established contact points such as help desks and call centers to respond to inquiries from customers. At the contact points, for example, an operator (a person who responds) responds to inquiries received via e-mail or a form on a website later, or an operator receives inquiries by phone and responds immediately. Patent Document 1 discloses a technology that uses a chatbot to improve the efficiency of call center operations in response to inquiries from customers. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-28783 A Summary of the Invention [Problem to be solved by the invention]

[0004] Inquiries from users via e-mail, forms, or telephone often lack enough information to provide an appropriate response. For this reason, operators have had to ask additional questions to obtain information that complements the content of the inquiry, and ask again to clarify the intention and details of the inquiry. Chat services also complement the content interactively using a rule-based (scenario-based) mechanism that predefines patterns of questions to ask again according to the content. However, chat services have standardized patterns of questions to ask again, which makes the dialogue flow unnatural and takes time to create rules. In addition, it can be difficult to respond to inquiries that deviate from the patterns.

[0005] Therefore, the present disclosure has been made to solve the above problem, and has an object to provide an information processing program that can efficiently respond to inquiries. [Means for solving the problem]

[0006] In order to achieve the above-mentioned object, the information processing program of the present disclosure is a program executed by a computer having a control unit, and includes the steps of: the control unit displaying a first image including a query input field for inputting a query statement for the machine learning model; the control unit inputting the query statement inputted in the query input field to the machine learning model; and the control unit displaying a second image generated based on the output of the machine learning model, the second image including a set of questions and an answer input field for inputting answers to the set of questions.

[0007] In addition, in order to achieve the above-mentioned object, an information processing method according to the present disclosure is a method executed in an information processing system, and includes the steps of: displaying a first image including a query input field for inputting a query statement for a machine learning model; inputting the query statement inputted in the query input field into the machine learning model; and displaying a second image generated based on the output of the machine learning model, the second image including a set of questions and an answer input field for inputting answers to the set of questions.

[0008] In addition, in order to achieve the above-mentioned objective, the information processing system of the present disclosure is a system including a computer having a control unit, wherein the control unit displays a first image including a query input field for inputting a query statement for the machine learning model, the control unit inputs the query statement inputted in the query input field to the machine learning model, and the control unit displays a second image including a set of questions generated based on the output of the machine learning model and an answer input field for inputting answers to the set of questions.

[0009] In addition, in order to achieve the above-mentioned object, the information processing terminal of the present disclosure is a terminal having a control unit, which displays a first image including a query input field for inputting a query statement for a machine learning model, which acquires the query statement to be inputted to the machine learning model, and which displays a second image including a set of questions and an answer input field for inputting answers to the set of questions, the second image being generated based on the output of the machine learning model to which the query statement has been inputted. Effect of the Invention

[0010] According to the present disclosure, it is possible to provide an information processing program that can efficiently answer inquiries. [Brief description of the drawings]

[0011] [Figure 1] 1 is an overall view of an information processing system 1. FIG. [Diagram 2] FIG. 2 is a diagram illustrating a hardware configuration of a server device 100. [Diagram 3] 2 is a block diagram showing a functional configuration of the server device 100. FIG. [Figure 4] FIG. 2 is a block diagram showing the functional configuration of a terminal device 200. [Diagram 5] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 6] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 7] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 8] 2 is a flow diagram showing an example of information processing in the information processing system 1. FIG. [Figure 9] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 10] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 11] FIG. 2 is a diagram showing an example of a display on the terminal device 200. [Figure 12] FIG. 2 is an overall view of an information processing system 2. [Figure 13]FIG. 2 is an overall view of an information processing system 3. [Figure 14] FIG. 13 is a diagram showing an example of a display on an operator terminal device 1400. [Figure 15] FIG. 13 is a diagram showing an example of a display on an operator terminal device 1400. [Figure 16] FIG. 13 is a diagram showing an example of a display on an operator terminal device 1400. [Figure 17] FIG. 13 is a diagram showing an example of a display on an operator terminal device 1400. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In all drawings describing the embodiments, common components are given the same reference numerals, and repeated explanations will be omitted. Note that the following embodiments do not unduly limit the contents of the present disclosure described in the claims. In addition, not all components shown in the embodiments are necessarily essential components of the present disclosure.

[0013] <Summary of the Invention> The information processing system according to the present invention is a system that enables efficient response to inquiries in call center operations and the like that respond to inquiries from users such as customers. In the present invention, an inquiry received from a user via an inquiry form or the like is input to a machine learning model, a set of questions for obtaining information necessary to properly respond to the inquiry is output, and the set of questions and an answer input field for inputting answers to the set of questions are generated and presented to the user. By asking additional questions for information that is lacking in order to respond to a user inquiry before responding to the inquiry, the burden on an operator or the like responding to the inquiry until the final response can be reduced. <Embodiment 1> (Configuration of Information Processing System 1)

[0014] Fig. 1 is an overall view of an information processing system 1. The overall view of the information processing system according to this embodiment will be described with reference to Fig. 1.

[0015] The information processing system 1 according to the present embodiment includes a server device 100 and a terminal device 200. The server device 100 and the terminal device 200 are communicatively connected via a network NW. The network NW is configured from a WAN (Wide Area Network), a LAN (Local Area Network), and the like. Note that, in FIG. 1, for the sake of simplicity, only one terminal device 200 is illustrated, but the information processing system 1 may include multiple terminal devices 200.

[0016] The server device 100 is a device used by a business entity that handles inquiries such as call center operations, a platform provider that provides the business entity with a system related to the call center operations, etc. The server device 100 is realized by a general-purpose computer such as a workstation, a server computer, a stationary PC (Personal Computer), a laptop PC, etc. The server device 100 may also be logically realized by cloud computing.

[0017] In response to a request from the terminal device 200, the server device 100 performs control so as to display an image for making an inquiry on the terminal device 200. For example, in response to an access request from the terminal device 200, the server device 100 may perform control so as to display an image such as an inquiry form by providing a Web page.

[0018] The server device 100 also has a machine learning model. The machine learning model may include a system capable of learning a large amount of data and generating new sentences and content, and may be, for example, a large language model (LLM). Note that the machine learning model may be one provided by a device or system other than the server device 100.

[0019] The terminal device 200 is an information processing terminal used by a user, and is, for example, a PC (Personal Computer). However, the terminal device 200 is not limited to a PC, and may be any device, such as a smartphone, a tablet, or XR (Cross Reality) (AR: Augmented Reality, MR: Mixed Reality, VR: Virtual Reality), as long as it can display information from the server device 100 and input a query or the like.

[0020] The terminal device 200 displays an image of an inquiry form or the like received from the server device 100. The terminal device 200 also transmits an inquiry or the like input by a user to the server device 100. For example, the terminal device 200 may access the server device 100 using a web browser for viewing web pages, an API (Application Programming Interface), an application, or the like to receive data such as elements and specifications constituting a web page, and display the web page. An image of the inquiry form drawn on the web page and a confirmation of input contents are presented to the user of the terminal device 200.

[0021] Note that an "image" such as an inquiry form displayed on the terminal device 200 is a user interface (UI (User Interface)) for a user to interact with a system, an application, or the like. The UI may include, for example, visual elements such as text boxes and buttons that a user directly operates, a layout that is information on how these visual elements are arranged, text for providing information to the user, and graphics that display the visual elements. The UI may further include feedback information that indicates how the system responds when the user takes an action, and interaction information that indicates actions and methods for the user to interact with the interface. In this specification, an "image" displayed on the terminal device 200 includes at least graphics that are part of the UI.

[0022] (Hardware configuration of server device 100 and terminal device 200) Fig. 2 is a diagram showing a hardware configuration of the server device 100. The hardware configuration of the server device 100 according to this embodiment will be described with reference to Fig. 2.

[0023] The server device 100 includes a processor 11, a memory 12, a storage 13, a communication IF 14, and an input / output IF 15.

[0024] The processor 11 is hardware for executing an instruction set described in a program, and is composed of an arithmetic unit, a register, peripheral circuits, etc. The processor 11 functions as a "control unit" and controls the functions of the server device 100 by executing processes based on a program stored in the memory 12 described later.

[0025] The memory 12 is for temporarily storing programs and data to be processed by the programs, etc. The memory 12 is realized by a volatile memory such as a DRAM (Dynamic Random Access Memory). The programs stored in the memory 12 are programs for executing the information processing method according to the present invention, etc.

[0026] The storage 13 is a storage device for saving data, and is realized by, for example, a hard disc drive (HDD), a flash storage (a solid state drive (SSD), a universal serial bus (USB) flash drive, etc.), etc.

[0027] The communication IF 14 is an interface for transmitting and receiving signals so that the server device 100 can communicate with an external device, for example, a wireless communication device for wireless communication, a USB connector for serial communication, or the like.

[0028] The input / output IF 15 is an interface between an input device for receiving input from a user and an output device for presenting information to the user. The input device is realized by, for example, a keyboard, a mouse, a touch panel, a button, a microphone, etc. The output device is realized by, for example, a display, a printer, a speaker, etc.

[0029] Although the hardware configuration of the server device 100 has been described above, the hardware configuration of the terminal device 200 may be the same as the hardware configuration shown in Fig. 2. The operation of each component of the terminal device 200 is realized by the processor 11 according to the program stored in the memory 12, similar to the above-mentioned server device 100.

[0030] (Functional configuration of server device 100) Fig. 3 is a block diagram showing a functional configuration of the server device 100. The functional configuration of the server device 100 according to the first embodiment will be described with reference to Fig. 3.

[0031] The server device 100 includes a communication unit 110, a storage unit 120, an input unit 130, an output unit 140, a control unit 150, and a large-scale language model 160.

[0032] The communication unit 110 performs processing for the server device 100 to communicate with other devices. The other devices may be PCs connected via a network NW, or terminals such as smartphones and tablets. The other devices may also be input / output devices, such as flash memory and HDD, that input and output data.

[0033] The storage unit 120 is configured with, for example, a volatile memory such as a RAM, a HDD, a flash storage, etc., and stores data used by the server device 100 and computer programs used for various processes. The computer programs may be installed from a predetermined server or the like, or may be installed from a computer-readable portable recording medium using a known setup program or the like. The predetermined server may be constructed, for example, by an external computer that can be accessed remotely. The portable recording medium is, for example, a CD-ROM (Compact Disc Read Only Memory), a DVD-ROM (Digital Versatile Disc Read Only Memory), a USB flash memory, etc.

[0034] The input unit 130 is an interface for inputting data and commands to the server device 100 .

[0035] The output unit 140 is an interface that outputs information.

[0036] The control unit 150 is a processor such as a CPU (Central Processing Unit) that controls each function of the server device 100 and operates based on a program previously stored in the storage unit 120. The control unit 150 receives sentences such as inquiries and answers input by the user from the terminal device 200 via the communication unit 110, and sends them to the large-scale language model 160. The control unit 150 also outputs information for generating an image to be displayed on the terminal device 200 based on the output from the large-scale language model 160, etc.

[0037] The control unit 150 transmits information for generating an image to the terminal device 200 via the communication unit 110, and controls the terminal device 200 to display the image. The image displayed by the terminal device 200 is, for example, a first image including a query input field for inputting a query for the machine learning model (large-scale language model 160), or a second image including a group of questions and an answer input field for inputting answers to the group of questions, generated based on the output of the machine learning model (large-scale language model 160). The image displayed by the terminal device 200 will be described later.

[0038] The large-scale language model 160 is a natural language processing model that has learned general-purpose language patterns from a large amount of data through pre-learning. The large-scale language model 160 may be in a zero-shot state (a state in which no additional learning has been performed), or may be fine-tuned to a task corresponding to a problem that the user wants to solve. The problem that the user wants to solve may be, for example, a response to a product failure, repair, or loss of belongings. It may also be a response to a procedure in an organization such as a company. However, the above-mentioned problem that the user wants to solve is only an example, and is not limited to these. By fine-tuning the large-scale language model to a task corresponding to a problem that the user wants to solve, the user's intention can be more easily interpreted even if the query is freely written by the user.

[0039] The large-scale language model 160 receives text such as an inquiry or answer input by a user from the control unit 150. The large-scale language model 160 determines the problem that the user wants to solve from the input text, and checks whether the input text lacks information necessary to solve the problem. If the large-scale language model 160 determines that the necessary information is lacking, it outputs a set of questions. The set of questions is a question for inputting information related to the inquiry. The set of questions may be one or more. The large-scale language model 160 may not be provided in the server device 100, but may be provided in a device (not shown) connected to the server device 100 via a network or directly. The text input to the large-scale language model 160 may be a response to a set of questions that is further generated by the response to the set of questions output as described above.

[0040] (Functional configuration of terminal device 200) Fig. 4 is a block diagram showing a functional configuration of the terminal device 200. The functional configuration of the terminal device 200 according to the first embodiment will be described with reference to Fig. 4.

[0041] The terminal device 200 includes a communication unit 210, a storage unit 220, an input unit 230, an output unit 240, and a control unit 250.

[0042] The communication unit 210 performs processing for the terminal device 200 to communicate with other devices. The function of the communication unit 210 is similar to that of the communication unit 110 of the server device 100, so a duplicated description will be omitted.

[0043] The storage unit 220 is configured, for example, by a volatile memory such as a RAM, a flash memory, a HDD, etc. The storage unit 220 stores data used by the terminal device 200 and computer programs used for various processes. The functions of the storage unit 220 are similar to those of the storage unit 120 of the server device 100, so duplicated explanations will be omitted.

[0044] The input unit 230 is an interface that accepts user input to the terminal device 200. The input unit 230 is, for example, but is not limited to, a keyboard, a touch panel, or a microphone that detects voice input. The input unit 230 accepts inquiries and answers entered by the user.

[0045] The output unit 240 is an interface that outputs information to notify the user of the terminal device 200. The output unit 240 is, for example, a display or a speaker that outputs audio, but is not limited to these. The output unit 240 displays, for example, a WEB page (image) generated based on data received from the server device 100 or the like on a display.

[0046] The control unit 250 is a processor such as a CPU (Central Processing Unit) that controls each function of the terminal device 200 and operates based on a program previously stored in the storage unit 220. The control unit 250 receives data necessary for generating an image from the server device 100 via the communication unit 210. Then, the control unit 250 generates an image based on the received data and causes the output unit 240 to display the image.

[0047] (Example of display on terminal device 200) Fig. 5 is a diagram showing an example of a display on terminal device 200. With reference to Fig. 5, an inquiry form will be described as an example of a display on output unit 240 of terminal device 200.

[0048] Window 500 shows a web page displayed on a web browser started in terminal device 200, and has image 510. Image 510 is a form for inputting information necessary for responding to an inquiry, and is a first image including an inquiry input field for inputting an inquiry for a machine learning model. Image 510 includes inquiry input fields 512, 514, 516 and icon 518. The contents and number of items to be input, required items, optional items, and the like, and the layout can be freely set by a business entity that handles inquiries such as a call center.

[0049] The inquiry input field 512 is an input field for inputting the name of the user making the inquiry. The inquiry input field 514 is an input field for inputting the email address of the user making the inquiry. For example, this is used when sending a reply to a received inquiry by email. In the example of FIG. 5, "Takkyubo Taro" is input in the inquiry input field 512, and "example@xxxxx.com" is input in the inquiry input field 514.

[0050] The inquiry input field 516 is an input field for inputting an inquiry to the machine learning model. The user can freely write the content of the inquiry in the inquiry input field 516. In the example of FIG. 5, "The camera I bought at the store didn't work properly, so please fix it or refund my money." is input in the inquiry input field 516.

[0051] Icon 518 is a button for confirming whether the contents entered in inquiry input fields 512, 514, 516 have been accepted as the contents to be sent to the contact point (for example, a contact point such as a call center or a help desk). When icon 518 is pressed by the user, an image for confirming the entered contents may be displayed on output unit 240. For example, the contents of image 510 in window 500 may be rewritten to display an image for confirming the contents, or a Web page may be switched and displayed. Also, a new window may be generated and displayed.

[0052] In response to pressing of the icon 518, the query sentence entered in the query input field 516 may be transmitted to the server device 100 and input to the large-scale language model 160 (see FIG. 3). Alternatively, when the user enters an input in the query input field 516, the content of the input may be simultaneously transmitted in real time to the server device 100. The query sentence transmitted in real time from the terminal device 200 to the server device 100 may be input to the large-scale language model 160, and based on the output, a set of questions and an answer input field may be automatically displayed on the output unit 240 of the terminal device 200.

[0053] Fig. 6 is a diagram showing an example of a display on terminal device 200. With reference to Fig. 6, an image for confirming the contents entered in an inquiry form (see Fig. 5) will be described as an example of a display on output unit 240 of terminal device 200.

[0054] The window 600 shows a WEB page displayed on the WEB browser started up on the terminal device 200, and has an image 610. The image 610 shows the contents (name, email address, inquiry content) entered in the inquiry input field of the inquiry form (see FIG. 5), a group of questions for inputting information related to the inquiry, and an answer input field for the group of questions. In other words, the image 610 is a second image that includes a group of questions generated based on the output of a machine learning model and an answer input field for inputting answers to the group of questions. The image 610 includes a question field 612, an answer input field 614, and icons 616 and 618.

[0055] The question field 612 includes a group of questions output by the large-scale language model 160 (see FIG. 3) from the input inquiry. In the example of FIG. 6, the questions include "the manufacturer and model number of the camera" and "the store and date of purchase". The inquiry is "the camera I bought at the store did not work properly, so please repair it or refund me." The large-scale language model 160 has determined that the information required to properly answer the inquiry is insufficient, and has generated a question to prompt the user to input information related to the "camera I bought at the store" included in the inquiry. Although two questions are displayed in the question field 612, the number of questions generated by the large-scale language model 160 is not limited to two, and may be one, or three or more. The number of questions to be generated may be preset by the inquiry support desk. However, if the large-scale language model 160 determines that the information required to properly answer the inquiry is sufficient, no question is generated.

[0056] The answer input field 614 is an input field for inputting answers to the questions written in the question field 612. The user can freely enter answers to the questions in the answer input field 614. Note that, although the answers are "optional" in the example of FIG. 6, they may be "required" depending on the content of the output question so that the operator can provide an appropriate answer. The size of the answer input field 614 may be changed depending on the number and content of questions included in the question field 612.

[0057] Furthermore, the position where the questions in the question field 612 are arranged is not limited to the position shown in Fig. 6. The questions only need to be presented in a manner that the user can recognize, and may be written inside the answer input field 614, for example. Furthermore, in the example of Fig. 6, the answer input field 614 is surrounded by a frame, but the answer input field 614 does not necessarily need to be surrounded by a frame. It only needs to be presented in a manner that allows the user to recognize the area where the answer is to be input.

[0058] Icon 616 is a button for returning to the image that was displayed before transitioning to image 610 currently being displayed on terminal device 200. When icon 616 is pressed by the user, the image displayed on terminal device 200 returns to the image of the inquiry form (see FIG. 5).

[0059] Icon 618 is a button for sending to the contact the "Name", "Email address", "Inquiry content" and the content entered in the answer input field 614 that have been confirmed as the content entered by the user. When the user presses icon 618, the display transitions to an image notifying completion of receipt (FIG. 7 described later).

[0060] Fig. 7 is a diagram showing an example of a display on terminal device 200. With reference to Fig. 7, an image notifying completion of receipt of an inquiry will be described as an example of a display on output unit 240 of terminal device 200.

[0061] The window 700 shows a WEB page displayed in the WEB browser launched in the terminal device 200, and has an image 710. The image 710 is an image showing the contents sent to the contact point (the contents accepted by the contact point). More specifically, the image 710 shows the name of the inquiring user "Tokutei Taro", the email address "example@xxxxx.com", the inquiry contents "The camera I bought at the store didn't work properly, so please repair it or refund it", which are input in the inquiry input field 512, 514, and 516 (see FIG. 5), and the answer "The camera is a C52X0 made by XX company. I bought it at △△ store on XX month XX day, 2023", which is the answer input in the answer input field 614 (see FIG. 6). It is preferable that the contents accepted by the contact point are displayed in an easy-to-read format with good visibility in order to make it easy to understand the contents.

[0062] 5 to 7, examples of display on the terminal device 200 and transitions of images displayed have been described above. In the example of FIG. 6, the answer input field 614 is generated as an input field different from the query input fields 512, 514, and 516. However, the answer input field 614 for inputting an answer to a group of questions does not have to be an input field different from the query input field 516 and the like. For example, a group of questions generated simultaneously with input of a query may be presented to the user in the query input field 516 (see FIG. 5), so that the user is prompted to input the answer in the query input field 516. Also, instead of presenting a group of questions, a sample query may be presented to encourage the user to rewrite the query.

[0063] Also, window 500 ("first screen") in which image 510 ("first image") is displayed and window 600 ("second screen") in which image 610 ("second image") are displayed are different windows (screens), but the display may be switched by tabs within the same window. Also, within the same window, the display may transition from the first image to the second image, for example, as when switching between web pages. Also, like an SPA (Single Page Application), content (images) may be dynamically rewritten within a single web page, and information may be provided without transitioning web pages.

[0064] Furthermore, image 510 ("first image") and image 610 ("second image") may be in different display areas within the same window. For example, the images may be arranged in left and right, or top and bottom display areas within the window. Furthermore, the images may be arranged in different display layers, such as by arranging image 610 above image 510 so that at least a portion of image 510 overlaps with image 610 within the window, or may be displayed modally.

[0065] (Information processing flow in information processing system 1) Fig. 8 is a flow diagram showing an example of information processing in the information processing system 1. The flow of processing related to an inquiry in the information processing system 1 will be described with reference to Fig. 8. Note that the flow of processing is an example, and is not limited to the order shown in Fig. 8. In addition, in the following, the screen displayed by the terminal device 200 may be an image displayed on at least a part of the display area of ​​the output unit 240 such as a display, or an image displayed entirely. It may also be an image displayed in a window of a web browser or the like.

[0066] 5 to the server device 100 in response to the display request, and the server device 100 transmits information required to generate the inquiry screen to the terminal device 200 in response to the display request. The control unit 250 of the terminal device 200 displays the inquiry screen on the output unit 240 based on the received information. Step S101 is an example of a process in which the control unit 150 of the server device 100 displays a first image including a query input field for inputting a query for the machine learning model.

[0067] In step S102, the control unit 250 of the terminal device 200 accepts an inquiry text entered by the user in an inquiry input field of the inquiry screen displayed on the output unit 240. The control unit 250 transmits the entered inquiry text to the server device 100, for example, in response to a transmission instruction from the user. Note that the control unit 250 may automatically transmit the inquiry text to the server device 100 at the same time as the user inputs it, without waiting for a transmission instruction from the user.

[0068] In step S103, the control unit 150 of the server device 100 inputs the query sentence received in step S102 and input into the query input field into a machine learning model. In this embodiment, the machine learning model is the large-scale language model 160.

[0069] In step S104, the large-scale language model 160 of the server device 100 determines the problem that the user wants to solve from the input query sentence, and checks whether the information required to solve the problem is insufficient. If the large-scale language model 160 determines that the input query sentence lacks the information required to solve the problem (Y in step S104), the process proceeds to step S105. On the other hand, if it determines that the necessary information is not insufficient (N in step S104), the process proceeds to step S108.

[0070] In step S105, the large-scale language model 160 outputs a set of questions. The set of questions is questions for inputting information related to the inquiry sentence. The control unit 150 of the server device 100 outputs information required to generate an image including the set of questions and an answer input field for inputting answers to the set of questions, based on the output of the large-scale language model 160. The image including the answer input field is, for example, a confirmation screen for input contents as shown in FIG. 6.

[0071] In step S106, the control unit 150 of the server device 100 transmits information required to generate a confirmation screen (see FIG. 6) to the terminal device 200, and the control unit 250 of the terminal device 200 displays the confirmation screen on the output unit 240 based on the received information. Step S106 is an example of a process in which the control unit 150 of the server device 100 causes the terminal device 200 to display a second image, which is generated based on the output of the machine learning model and includes a group of questions and answer input fields for inputting answers to the group of questions.

[0072] In step S107, the control unit 250 of the terminal device 200 accepts an answer entered by the user in the answer entry field of the confirmation screen displayed on the output unit 240.

[0073] On the other hand, in step S104, if the large-scale language model 160 of the server device 100 determines that necessary information for the input query sentence is sufficient (N in step S104), then in step S108, the control unit 150 of the server device 100 outputs information necessary for generating a confirmation screen for confirming the content input in the query input field of the inquiry screen, and transmits the information to the terminal device 200. The control unit 250 of the terminal device 200 displays the confirmation screen on the output unit 240 based on the received information. Note that the confirmation screen displayed in step S108 is, for example, a screen obtained by deleting the question field 612 and the answer input field 614 from the image 610 in FIG. 6. Also, after step S104, the process may proceed to step S109 without performing the process of displaying the confirmation screen as shown in step S108.

[0074] In step S109, the control unit 250 of the terminal device 200 transmits the contents entered in the inquiry input field and the answer input field to the server device 100 in response to an operation to transmit the input contents from the user to the inquiry destination. The control unit 150 of the server device 100 outputs information required to generate an acceptance completion screen notifying the contents received from the terminal device 200 as the accepted contents, and transmits the information to the terminal device 200. The control unit 250 of the terminal device 200 displays the acceptance completion screen on the output unit 240 based on the received information. The control unit 150 of the server device 100 outputs information required to generate an acceptance completion screen notifying the completion of the acceptance of the inquiry, for example, as shown in FIG. 7.

[0075] This is the end of the process related to the inquiry in the information processing system 1. The operator responding to the inquiry can confirm the inquiry content including the answers to the questions generated by the large-scale language model 160.

[0076] (Explanation of effect) As described above, in the information processing system according to this embodiment, an inquiry received from a user via an inquiry form or the like is input to a machine learning model, a set of questions for obtaining information necessary to properly answer the inquiry is output, and a set of questions and an answer input field for inputting answers to the set of questions are generated and presented to the user. By automatically asking additional information that is lacking in order to answer a user inquiry before answering the user inquiry, an operator or the like who handles the inquiry does not need to ask the user again, and the burden on the operator until the final answer is given can be reduced.

[0077] Furthermore, in this embodiment, a large-scale language model is used to generate questions to obtain the information necessary to properly respond to a query, so that the content of the question can be expressed more appropriately than in a rule-based system in which patterns of reflective questions according to the situation are predefined.

[0078] In addition, in this embodiment, an answer input field different from the inquiry input field is automatically generated. Unlike a time-dependent dialogue format such as a chatbot, an answer input field different from the inquiry input field is generated, so that the user can instantly understand and input additional information required to obtain an appropriate answer.

[0079] In addition, in this embodiment, a first image including a query input field for inputting a query for the machine learning model is displayed on a first screen, and a second image including a set of questions generated based on the output of the machine learning model and an answer input field for inputting answers to the set of questions is displayed on a second screen different from the first screen. In addition, in this embodiment, the first image and the second image are displayed in different display areas. This allows the input contents and the generated set of questions to be displayed with good viewability, so that the user can easily understand the input contents and the set of questions.

[0080] <Variation 1> In the first embodiment, as shown in Fig. 6, one answer input field for inputting an answer is generated for a group of questions generated based on the output of a machine learning model. However, the number of answer input fields generated is not limited to one. An answer input field may be generated for each question in the group of questions.

[0081] Fig. 9 is a diagram showing an example of a display on terminal device 200. An image for confirming the contents entered in the inquiry form (see Fig. 5) will be described with reference to Fig. 9 as an example of a display on output unit 240 of terminal device 200. Note that common components are given the same reference numerals, and repeated description will be omitted.

[0082] The window 900 shows a Web page displayed in a Web browser launched in the terminal device 200, and has an image 910. The image 910 includes questions 912 and 916, answer input fields 914 and 918, and icons 616 and 618.

[0083] Questions 912 and 916 are questions output by large-scale language model 160 (see FIG. 3). Answer input field 914 is an answer input field for question 912, and answer input field 918 is an answer input field for question 916. In other words, the answer input fields are displayed corresponding to each question in the question set output by large-scale language model 160.

[0084] The user can input answers to the questions in the respective answer input fields. For example, "XX Company, C52X0" is input in answer input field 914 as an answer to question 912. Furthermore, nothing is input in answer input field 918 to question 916. When the user presses icon 618, the display transitions to an image notifying completion of reception (FIG. 10 described later).

[0085] Fig. 10 is a diagram showing an example of a display on terminal device 200. With reference to Fig. 10, an image notifying completion of receipt of an inquiry will be described as an example of a display on output unit 240 of terminal device 200.

[0086] The window 1000 shows a Web page displayed in a Web browser started up in the terminal device 200, and has an image 1010. The image 1010 shows the content transmitted to the contact point (the content accepted by the contact point).

[0087] Items 1012 and 1016 respectively correspond to questions 912 and 916 in Fig. 9. Answers 1014 and 1018 respectively correspond to the contents entered in answer input fields 914 and 918 in Fig. 9. Answer 1018 displays "(not entered)," indicating that no input from the user has been accepted in answer input field 918.

[0088] (Explanation of effect) By displaying answer input fields for each question in the question group, it is possible to prevent users from missing answers. Also, the operator can easily understand which question the answer is for. Furthermore, for inquiries that can be assumed to be for the purpose of harassment, by generating answer input fields for each question, it is possible to discourage the user from sending an inquiry and reduce the burden on the operator in response to such inquiries.

[0089] <Variation 2> In the server device 100 of the first embodiment, the received inquiry content and the actual answer by the operator to the inquiry content may be stored as data. By utilizing the data stored together with in-house documents for the inquiry content and information on the Internet, the answer accuracy of the large-scale language model can be improved. In addition, as described in the first modification, by generating an answer input field for each question, the correspondence relationship between the answers input to the questions becomes clear, and the answer accuracy of the large-scale language model can be improved.

[0090] Furthermore, when there is insufficient information to properly respond to a user's inquiry, the server device 100 may ask questions via an automatic reply email and have the user input them, rather than displaying a set of questions and an answer input field on the input confirmation screen and having the user input them as in embodiment 1. This makes it possible to respond without causing the user stress due to waiting, even if a set of questions is not output within a predetermined time after the inquiry is input to the large-scale language model.

[0091] The server device 100 may also use a large-scale language model to edit or adjust the text that the user freely writes in the inquiry input field or the answer input field into an appropriate form. For example, it may correct typos and grammatical corrections, or visually edit the format, such as by changing line spacing, character spacing, or dividing paragraphs. It may also adjust the structure and order of the entire text so that the content has a consistent flow, or change the wording to correct the expression so that the text is clearer and easier to understand. This reduces the burden on the user to write freely. It may also be possible to ease the psychological anxiety of the operator who checks the content of the inquiry by correcting certain abusive language, slander, and other emotional expressions.

[0092] The server device 100 may input the content received from the user into the large-scale language model and output the content of the reply such as an e-mail made by the operator. For example, when there is sufficient information to properly reply to the user's inquiry, the large-scale language model outputs a sentence showing the corresponding content. For example, when there is insufficient information to properly reply to an additional question posed to the user, the large-scale language model outputs a sentence asking about the missing content. The operator can create a reply to the inquiry based on the output sentence, which reduces the burden on the operator.

[0093] <Variation 3> Fig. 11 is a diagram showing an example of a display on the terminal device 200. With reference to Fig. 11, an inquiry text in an inquiry form (see Fig. 5) is automatically sent to the server device 100, and a group of questions and answer input fields are generated, as an example of a display on the output unit 240 of the terminal device 200. Note that the same reference numerals are used to denote common components, and repeated explanations will be omitted.

[0094] The window 1100 shows a Web page displayed in a Web browser launched in the terminal device 200, and has an image 1110. The image 1110 includes inquiry input fields 512, 514, and 516, a message 1121, questions 1122 and 1124, answer input fields 1123 and 1125, and an icon 518.

[0095] In the example of Figure 11, in an inquiry form (see Figure 5), a query entered by a user in the query input field 516 is automatically sent to the server device 100, and based on the information output by the large-scale language model 160, a message 1121, questions 1122 and 1124, and answer input fields 1123 and 1125 are displayed.

[0096] Question 1122 and question 1124 are questions output by large-scale language model 160 (see FIG. 3). Answer input field 1123 is an answer input field for question 1122, and answer input field 1125 is an answer input field for question 1124. In other words, the answer input fields are displayed corresponding to each question in the question group output by large-scale language model 160.

[0097] The user can input answers to the questions in the respective answer input fields. When the user presses icon 518, an image may be displayed to confirm the input contents, or an image may be displayed to notify completion of reception.

[0098] When a user inputs an inquiry, the inquiry is automatically input to the large-scale language model 160 of the server device 100, and the output questions and answer input fields are instantly displayed on the inquiry form. Therefore, the user can instantly understand what is missing in the input contents and can enrich the inquiry contents before sending the entire contents of the inquiry form to the server device 100, thereby improving the efficiency of inquiry work.

[0099] <Embodiment 2> The information processing system 2 according to the second embodiment is a system used by an operator who responds to verbal inquiries from users over the telephone, at a counter, or the like.

[0100] (Configuration of information processing system 2) Fig. 12 is an overall view of the information processing system 2. The overall view of the information processing system according to this embodiment will be described with reference to Fig. 12. Note that common components are given the same reference numerals, and repeated description will be omitted.

[0101] The information processing system 2 according to the present embodiment includes a server device 100, a user terminal device 300, and an operator terminal device 400. The server device 100, the user terminal device 300, and the operator terminal device 400 are communicably connected via a network NW. In FIG. 12, only one user terminal device 300 and one operator terminal device 400 are illustrated for the sake of simplicity, but there may be a plurality of each. The user terminal device 300 and the operator terminal device 400 have the same functional configuration as the above-mentioned terminal device 200.

[0102] The user terminal device 300 is a terminal device used by a user who makes an inquiry to an operator. The user terminal device 300 has at least a calling function, and may be, for example, a landline phone or a mobile phone, or may be an information processing terminal that makes voice calls (such as IP phones or video phones) over the Internet.

[0103] The operator terminal device 400 is an information processing terminal used by an operator at a call center, etc. For example, any device may be used as long as it can transmit and receive information to and from the server device 100, such as a PC, a smartphone, a tablet, or an XR.

[0104] The operator terminal device 400 accepts a user inquiry as voice data from the user terminal device 300. For example, the operator terminal device 400 may convert a voice inquiry made by telephone from the user terminal device 300 into digital data. The operator terminal device 400 transmits the accepted voice data to the server device 100 in real time. Note that instead of voice data, the operator may create an inquiry sentence summarizing the main points of the user's inquiry and transmit it to the server device 100. Also, voice data in which the operator repeats the contents of the user's inquiry may be transmitted to the server device 100.

[0105] The server device 100 converts the voice data obtained from the operator terminal device 400 into text data and inputs the text data to the large-scale language model 160. At this time, the voice data may be input to the large-scale language model 160 as it is.

[0106] The large-scale language model 160 determines the problem that the user wants to solve from the input data, and checks whether the input text lacks the information necessary to solve the problem. If the large-scale language model 160 determines that the necessary information is lacking, it outputs a set of questions. At this time, the large-scale language model 160 may output a set of questions by utilizing in-house documents related to the inquiry content, information on the Internet, and the like.

[0107] The server device 100 transmits the set of questions output by the large-scale language model 160 to the operator terminal device 400. The operator terminal device 400 displays the received set of questions on a display or the like to present it to the operator. This allows the operator to ask questions to the user while checking the displayed set of questions.

[0108] The server device 100 may automatically create a response record based on the voice data received from the operator terminal device 400. For example, the server device 100 may analyze and record the call date and time, information identifying the user (such as name and contact information), the user's inquiry, the operator's response, the call duration, information identifying the corresponding operator, and the user's emotion information from the voice data of the operator and the user. By utilizing the response record, the response accuracy of the large-scale language model 160 can be improved.

[0109] (Explanation of effect) Although real-time responses are required for inquiries by phone from users, not all operators can respond immediately due to lack of knowledge, experience, etc. By having a large-scale language model determine the information necessary for the response, and if there is insufficient information, presenting the information that the user should ask the operator in real time, it is possible to reduce waiting times until a response is made and improve the efficiency of inquiry operations.

[0110] In addition, by automatically recording responses to inquiries, the workload of operators can be reduced.

[0111] <Embodiment 3> The information processing system 3 according to the third embodiment is a system used by an operator who immediately responds to non-standard inquiries (inquiries) from users via telephone, chat, etc. In this embodiment, a non-standard inquiry refers to an inquiry about a specific situation that is not typical and is difficult to classify in a standard way.

[0112] (Configuration of information processing system 3) Fig. 13 is an overall view of an information processing system 3. The overall view of the information processing system according to this embodiment will be described with reference to Fig. 13. Note that common components are given the same reference numerals, and repeated description will be omitted.

[0113] The information processing system 3 according to the present embodiment includes a server device 100, a user terminal device 300, and an operator terminal device 1400. The server device 100, the user terminal device 300, and the operator terminal device 1400 are communicably connected via a network NW. In FIG. 13, only one user terminal device 300 and one operator terminal device 1400 are illustrated for the sake of simplicity, but there may be a plurality of each. The user terminal device 300 and the operator terminal device 1400 each have a telephone function in addition to the same functions as the above-mentioned terminal device 200 (see FIG. 4). The telephone function may be realized by an analog line, a digital line, an IP line, a Hikari line, or the like.

[0114] The operator terminal device 1400 is an information processing terminal used by an operator at a call center, etc. For example, any device may be used as long as it can transmit and receive information to and from the server device 100, such as a PC, a smartphone, a tablet, or an XR.

[0115] The operator terminal device 1400 accepts non-standard inquiries from users as voice data or character string data via the user terminal device 300. For example, the inquiries may be accepted by telephone or chat. The operator terminal device 1400 also accepts responses from operators by telephone or chat, and displays a dialogue between the user and the operator relating to the inquiry on a display device that the operator can view.

[0116] Furthermore, the operator terminal device 1400 transmits user inquiry information indicating the contents of the user's inquiry in the dialogue to the server device 100. The user inquiry information may be, for example, digital data (voice data or character string data) converted based on the user's utterance received from the user terminal device 300. The user inquiry information may also be character string data input by the user. The operator terminal device 1400 may transmit the contents of the dialogue between the user and the operator to the server device 100, and the server device 100 may extract the contents of the user's inquiry from the contents of the dialogue.

[0117] The server device 100 inputs the user inquiry information received from the operator terminal device 1400 to the large-scale language model 160. As described in the first embodiment, the large-scale language model 160 determines the problem that the user wants to solve from the input user inquiry information, and determines whether or not information required to solve the problem is lacking. If the large-scale language model 160 determines that information required to solve the problem is lacking, it outputs the required information as question information.

[0118] Furthermore, the large-scale language model 160 may output solution information related to a problem that the user wants to solve. The solution information may be, for example, FAQ (Frequently Asked Questions), which is a collection of frequently asked questions and their answers.

[0119] As described above, the large-scale language model 160 may output at least one of question information and solution information in response to the input user query information. The server device 100 transmits the question information and / or solution information output by the large-scale language model 160 to the operator terminal device 1400.

[0120] The operator terminal device 1400 receives the question information and / or the solution information output by the large-scale language model 160. The operator terminal device 1400 displays at least one of the question information or the solution information on a display device that is viewed by an operator. The operator can respond to the user's inquiry by checking the question information and / or the solution information displayed on the display device.

[0121] In addition, after the dialogue between the user and the operator regarding the user's inquiry is completed, the server device 100 may create a summary of the introduction based on the dialogue. Then, based on the summary, machine learning may be performed with the user inquiry information as input data and the operator's response information, which is the content of the operator's response, as output data. This allows the server device 100 to output question information and solution information for the user inquiry information with higher accuracy.

[0122] (Example of display on operator terminal 1400)

[0123] Fig. 14 is a diagram showing an example of a display on the operator terminal device 1400. With reference to Fig. 14, an example of a display on the display device (output unit 240) of the operator terminal device 1400 will be described.

[0124] The screen 1500 is a screen of the display device of the operator terminal device 1400. The screen 1500 has a dialogue content area 1510 and a question information area 1520. Note that the screen 1500 may be displayed on the entire screen of the display device, or may be displayed on a part of the screen.

[0125] In the dialogue content area 1510, the dialogue between the user and the operator is displayed in chronological order. The dialogue content area 1510 includes user inquiry information 1511, 1513 indicating the contents of the user's inquiry, and operator response information 1512, 1514 indicating the contents of the operator's response. The user inquiry information may be displayed as, for example, "user remarks", in which text data obtained by transcribing a voice received from the user over the phone, or the user's chat input. The operator response information may be displayed as, for example, "operator remarks", in which text data obtained by transcribing a voice response from the operator over the phone, or the operator's chat input.

[0126] In the dialogue content area 1510, characters or display areas may be colored or indented to make it easier to distinguish between user utterances and operator utterances. Note that user utterances and operator utterances may contain interjections (e.g., "um," "ah," "hmm," etc.) used to smooth the flow of conversation, but these may be automatically deleted when or after being displayed in the dialogue content area 1510.

[0127] In the question information area 1520, question information output by the large-scale language model 160 of the server device 100 is displayed. In the example of Fig. 14, question information 1521 indicates that the user is asked about the customer number. Also, question information 1522 indicates that the user is asked about the symptoms of the malfunction.

[0128] The question information may be displayed in a manner associated with user query information, which is input data of the question output by the large-scale language model 160. In Fig. 14, for example, if the output when user query information 1511 is input to the large-scale language model 160 is question information 1521, the question information 1521 is associated with the user query information 1511 by a line segment 1531. Similarly, if the output when user query information 1513 is input to the large-scale language model 160 is question information 1521, the question information 1522 is associated with the user query information 1513 by a line segment 1532.

[0129] The manner of association is not limited to display using lines. For example, when question information 1521 is selected by clicking or the like, the corresponding user inquiry information 1511 may be highlighted, or may be color-coded or given an identifier. Conversely, when user inquiry information 1511 is selected by clicking or the like, the corresponding question information 1521 may be highlighted or the like. Also, related question information and user inquiry information may be displayed in close proximity to each other. Note that question information and user inquiry information do not necessarily have to be displayed in an associated manner.

[0130] In the blank space (lower part) of the question information 1521 and the question information 1522, information related to the question information received from the user or the operator may be input. For example, the operator may manually input the user's answer to the question, or the information may be extracted from the user's utterance to the operator's question and automatically input. In the example of FIG. 14, the entry space is provided below the question as a mode in which the information related to the question information received from the user or the operator is associated with the question information, but the display format is not limited to this. For example, the entry space may be provided in the right column, or any display format may be used as long as the association with the question information is clear. Note that since the utterance of the user or the operator remains written in the dialogue content area 1510, it is not necessarily necessary to provide an entry space for information related to the question information.

[0131] Fig. 15 is a diagram showing an example of a display on the operator terminal device 1400. With reference to Fig. 15, an example of a display on the display device (output unit 240) of the operator terminal device 1400, which is different from the example in Fig. 14, will be described.

[0132] A screen 1600 is a screen of the display device of the operator terminal device 1400. The screen 1600 has a dialogue content area 1610 and a solution information area 1620.

[0133] In the dialogue content area 1610, the dialogue between the user and the operator is displayed in chronological order, similar to the dialogue content area 1510 (see FIG. 14). In addition, in the solution information area 1620, solution information 1621 and solution information 1622 output by the large-scale language model 160 of the server device 100 are displayed.

[0134] The solution information 1621 is a FAQ related to the inquiry content indicated in the user inquiry information 1611, that is, a problem that the user wants to solve. The solution information may be displayed in a manner associated with the user inquiry information, which is input data of the solution information output by the large-scale language model 160. For example, if the output when the user inquiry information 1611 is input to the large-scale language model 160 is the solution information 1621, the solution information 1621 is associated with the user inquiry information 1611 by a line segment 1631. Similarly, if the output when the user inquiry information 1612 is input to the large-scale language model 160 is the solution information 1622, the solution information 1622 is associated with the user inquiry information 1612 by a line segment 1632.

[0135] As described in FIG. 14, the association between the solution information and the user inquiry information may be in any display format as long as the association can be understood. The solution information and the user inquiry information do not necessarily have to be displayed in an associated manner. In addition, in the example of FIG. 15, only one piece of solution information is displayed for each piece of user inquiry information, but multiple pieces of solution information may be displayed, or multiple pieces of solution information may be displayed by a predetermined operation. For example, in FIG. 15, multiple pieces of solution information may be displayed by selecting "+See more" by clicking, etc.

[0136] Fig. 16 is a diagram showing an example of a display on the operator terminal device 1400. With reference to Fig. 16, an example of a display on the display device (output unit 240) of the operator terminal device 1400, which is different from the examples of Fig. 14 and Fig. 15, will be described. As shown in Fig. 16, the question information and the solution information may be displayed in the same display area.

[0137] A screen 1700 is a screen of the display device of the operator terminal device 1400. The screen 1700 has a dialogue content area 1710 and a question information and solution information area 1720.

[0138] In the dialogue content area 1710, the dialogue between the user and the operator is displayed in chronological order.

[0139] In the question information and solution information area 1720, question information 1721 and solution information 1722 are displayed. The question information 1721 and solution information 1722 are outputs when the user inquiry information 1711 is input to the large-scale language model 160. In this case, in the example of FIG. 16, the solution information 1722 is displayed below the question information 1721. In addition, the user inquiry information 1711 and the question information 1721 are associated with each other by a line segment 1731. This allows the operator to easily understand that the user inquiry information 1711, the question information 1721, and the solution information 1722 are related information. In the example of FIG. 16, the line segment 1731 associates the user inquiry information 1711 and the question information 1721, but the user inquiry information 1711 and the solution information 1722 may be associated with each other. In addition, as described above, the association is not limited to being displayed by a line segment, and may not necessarily be displayed in an associated manner.

[0140] Fig. 17 is a diagram showing an example of a display on the operator terminal device 1400. With reference to Fig. 17, an example of a display on the display device (output unit 240) of the operator terminal device 1400, which is different from the examples in Figs. 14 to 16, will be described.

[0141] A screen 1800 is a screen of the display device of the operator terminal device 1400. The screen 1800 has a question information area 1810, a dialogue content area 1820, and a solution information area 1830. As shown in Fig. 17, the dialogue display area, the question information area, and the solution information area may be displayed in parallel.

[0142] In the question information area 1810, question information 1811 and 1812 are displayed. In the dialogue content area 1820, user inquiry information 1821 and 1823 and operator answer information 1822 and 1824 are displayed. In the solution information area, solution information 1831 and 1832 are displayed.

[0143] Question information 1811 and solution information 1831 are outputs when user query information 1821 is input into large scale language model 160, and user query information 1821 is associated with question information 1811 and solution information 1831. Similarly, question information 1812 and solution information 1832 are outputs when user query information 1823 is input into large scale language model 160, and user query information 1823 is associated with question information 1812 and solution information 1832.

[0144] In the example of Fig. 17, the question information region is arranged on the left side of the screen, the dialogue content region in the center, and the solution information region on the right side, but the arrangement is not limited to this. The arrangement of the question information region and the solution information region may be switched, or one may be arranged in the center and the dialogue content region may be arranged on the left and right. As long as the question information region, solution information region, and dialogue content region are displayed in parallel, any arrangement is acceptable.

[0145] Furthermore, the display device of the operator terminal device 1400 may be configured to display a summary generated after the dialogue is completed. In a call center, in order to improve the quality of service and avoid trouble, an operator may record the dialogue content, etc., after the user's inquiry is handled. By automatically generating a summary of the dialogue content and presenting it to the operator, the time it takes for the operator to create a record can be reduced, and the operator's response can be made more efficient. Furthermore, human error in record creation can be prevented, and the accuracy of the record can be improved.

[0146] (Explanation of effect) As described above, the information processing system according to this embodiment is a system used by an operator who responds instantly to non-standard inquiries from users, and displays a dialogue between the user and the operator on a display device viewed by the operator, and inputs user inquiry information indicating the content of the user's inquiry in the dialogue into a large-scale language model (machine learning model). Then, at least one of question information output by the machine learning model in response to the input user inquiry information or solution information for the user inquiry information is displayed on the display device. This makes it possible to compensate for the lack of knowledge, experience, etc. of the operator and improve the efficiency of inquiry work.

[0147] In this embodiment, the question information output from the large-scale language model is displayed on the display device in a form associated with the user inquiry information. Also, the dialogue between the user and the operator is displayed in chronological order. This makes it easier for the operator to grasp the flow of the dialogue, and improves the efficiency of inquiry work.

[0148] In this embodiment, the dialogue content area is displayed on the display device in parallel with at least one of the question information area where question information is displayed and the solution information area where solution information is displayed. This allows the operator to easily grasp the dialogue content and the content to be answered, thereby improving the efficiency of inquiry operations.

[0149] In addition, in this embodiment, information related to the question information received from the user or operator is displayed on the display device in a form associated with the question information, so that the operator can easily grasp the information that is lacking for the problem that the user wants to solve.

[0150] The above-described embodiment can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and modifications are included in the scope of the invention and its equivalents as described in the claims, as well as in the scope and spirit of the invention.

[0151] Furthermore, the present invention may have the following aspects as shown in the appended notes.

[0152] [Appendix 1] A program executed by a computer having a control unit in a system used by an operator to instantly respond to non-routine inquiries from users, comprising: a step of the control unit displaying a dialogue between the user and the operator on a display device viewed by the operator; The control unit inputs user inquiry information indicating the inquiry content of the user in the dialogue into a machine learning model; a step of displaying, on the display device, at least one of question information output by the machine learning model in response to the input user inquiry information or solution information for the user inquiry information, by the control unit; An information processing program comprising:

[0153] [Appendix 2] The information processing program according to [Appendix 1], wherein the output question information is displayed on the display device in a manner associated with the user inquiry information.

[0154] [Appendix 3] The control unit receives information related to the question information from the user or the operator; a step of the control unit displaying the received information on the display device in a manner associated with the question information; The information processing program according to [Appendix 2], further comprising:

[0155] [Appendix 4] The information processing program according to [Appendix 3], wherein the dialogue is displayed in chronological order in a dialogue content area of ​​the display device.

[0156] [Appendix 5] An information processing program described in [Appendix 4], wherein the dialogue content area is displayed in parallel with at least one of a question information area in which the question information is displayed and a solution information area in which the solution information is displayed.

[0157] [Appendix 6] After the dialogue between the user and the operator regarding the inquiry from the user is completed, generating a summary for the query based on the interaction; The information processing program described in [Appendix 5] further comprises a step in which the control unit performs machine learning based on the summary, with the user query information as input data and operator answer information indicating the content of the operator's answer as output data.

[0158] [Appendix 7] The information processing program according to any one of [Appendix 1] to [Appendix 6], wherein the user inquiry information is character string data converted from voice data received from the user.

[0159] [Appendix 8] The information processing program according to any one of [Appendix 1] to [Appendix 6], wherein the user inquiry information is character string data received from the user.

[0160] [Appendix 9] A method for use in a system for an operator to instantly respond to non-routine inquiries from users, comprising: displaying the dialogue between the user and the operator on a display device for viewing by the operator; inputting user query information indicating a query content of the user in the dialogue into a machine learning model; displaying, on the display device, at least one of question information output by the machine learning model in response to the input user inquiry information or solution information for the user inquiry information; An information processing method comprising:

[0161] [Appendix 10] In a system used by an operator to instantly respond to non-routine inquiries from users, a computer having a control unit is displaying the dialogue between the user and the operator on a display device for viewing by the operator; inputting user query information indicating a query content of the user in the dialogue into a machine learning model; displaying, on the display device, at least one of question information output by the machine learning model in response to the input user inquiry information or solution information for the user inquiry information; An information processing system that executes the above. [Explanation of symbols]

[0162] 1, 2, 3 information processing system, 11 processor, 12 memory, 13 storage, 14 communication IF, 15 input / output IF, 100 server device, 110, 210 communication unit, 120, 220 memory unit, 130, 230 input unit, 140, 240 output unit, 150, 250 control unit, 160 large-scale language model, 200 terminal device, 300 user terminal device, 400, 1400 operator terminal device, screen 1500, 1600, 1700, 1800.

Claims

1. A program executed by a computer having a control unit, A step of displaying a first image including a query input field for inputting a query for a machine learning model by the control unit; The control unit inputs the query sentence inputted into the query input field into the machine learning model; a step of causing the control unit to display a second image generated based on an output of the machine learning model, the second image including a group of questions and an answer input field for inputting answers to the group of questions; An information processing program comprising:

2. 2. The information processing program according to claim 1, wherein the group of questions is questions for prompting input of information related to the inquiry statement.

3. The information processing program according to claim 2 , wherein the answer input field is displayed corresponding to each question in the group of questions.

4. The information processing program according to claim 1 , wherein the control unit generates the answer input field different from the inquiry input field.

5. The information processing program according to claim 1 , wherein the control unit displays the first image on a first screen and displays the second image on a second screen different from the first screen.

6. The information processing program according to claim 1 , wherein the control unit displays the first image and the second image in different display areas.

7. 1. A method implemented in an information processing system, comprising: A step of displaying a first image including a query input field for inputting a query for the machine learning model; inputting the query sentence entered in the query input field into the machine learning model; displaying a second image, which is generated based on an output of the machine learning model, and includes a set of questions and an answer input field for inputting answers to the set of questions; An information processing method comprising:

8. An information processing system including a computer having a control unit, The control unit displays a first image including a query input field for inputting a query for a machine learning model; The control unit inputs the query sentence inputted into the query input field into the machine learning model; the control unit displays a second image, which is generated based on the output of the machine learning model, and which includes a group of questions and an answer input field for inputting answers to the group of questions; Information processing system.

9. An information processing terminal having a control unit, The control unit displays a first image including a query input field for inputting a query for a machine learning model; The control unit acquires the query sentence to be input to the machine learning model; the control unit displays a second image including a group of questions and an answer input field for inputting answers to the group of questions, the second image being generated based on an output of the machine learning model to which the query sentence is input; Information processing terminal.

Images