Answering device, answering method, program, and recording medium

Abstract

Provided are an answering device and the like that make it possible to flexibly respond to the content of a question from a user and derive a suitable answer. An answering device (10) according to one embodiment of the present invention comprises a processor (11). For an inquiry from a user, the processor (11) performs determination regarding a viewpoint including a degree of objectivity and determines an answering method on the basis of the determination result relating to the viewpoint.

Description

Answering device, answering method, program, and recording medium

[0001] One embodiment of the present invention relates to an answering device, an answering method, a program, and a recording medium that can flexibly respond to the content of a user's question and derive an appropriate answer.

[0002] In recent years, the functionality of large language models (LLMs) has improved significantly, enabling them to quickly and accurately respond to user inquiries. Furthermore, these LLMs are now available for use in Internet services and specific applications, making them readily available in a variety of settings. In particular, companies offering a variety of products and services often employ LLMs in help systems that automatically respond to user inquiries and simulation systems that automatically suggest products and other features suitable for each user through conversation.

[0003] Therefore, technologies related to the above-described help systems and the like are already known. One example of such technologies is the technology described in Patent Document 1. Patent Document 1 discloses a technology related to a photo book creation system and a server device that allows editing and ordering of a photo book by speaking to a smart speaker.

[0004] This technology relates to a photobook creation system that includes a server device that stores image data received from a user terminal, and a smart speaker that is communicatively connected to the server device and outputs audio and collects user speech. The server device has an image analysis unit that analyzes the image data and detects objects or text in the image, a dialogue processing unit that understands the user's voice input through the smart speaker, generates a response to the user, and outputs the response to the user via the smart speaker, an image selection unit that selects multiple images from the images to be saved when the user's voice is an instruction to create a photobook, a comment generation unit that generates and adds comments to the selected images based on the objects or text detected by the image analysis unit, and an editing processing unit that generates photobook data using the images selected by the image selection unit and the comments, and sends a preview screen of the photobook data to the user terminal.

[0005] Furthermore, Patent Document 2 discloses a technology relating to an automatic voice response system that can respond in real time to inquiries from drivers.

[0006] This technology relates to an automatic voice response system for motorsports, which is a system that automatically provides voice responses to inquiries from drivers, and which comprises: a first storage unit that stores inquiry patterns and response patterns in association with each other; an input unit that inputs voice data from the driver that represents the content of the inquiry; a generation unit that generates an inquiry pattern by analyzing the input voice data of the driver; a first determination unit that determines whether the generated inquiry pattern is stored in the first storage unit; a selection unit that, if the generated inquiry pattern is not stored in the first storage unit, estimates and selects an inquiry pattern similar to the generated inquiry pattern from the inquiry patterns stored in the first storage unit; a registration unit that newly registers the generated inquiry pattern and a response pattern associated with the selected inquiry pattern in the first storage unit; and a first output unit that generates and outputs voice data that represents the content of a response to an inquiry from the driver based on the newly registered response pattern.

[0007] JP 2019-160186 A JP 2019-185243 A

[0008] However, even with the ever-evolving LLM, it is not always possible to output a desirable answer depending on the content of the user's question. For example, this applies to questions that contain a large amount of emotional content or questions that are composed entirely of emotional content. In such cases, the LLM cannot accurately estimate the content or intent of the question, making it difficult to provide an appropriate answer. Furthermore, with older help systems that do not use the LLM but instead use a table of question-answer pairs, it is difficult to provide any substantive response at all.

[0009] One embodiment of the present invention has been made in consideration of the above circumstances, and aims to provide an answering device, answering method, program, and recording medium that can flexibly respond to the content of a user's question and derive an appropriate answer.

[0010] The above object can be achieved by an answering device according to any one of the following items [1] to

[23] : [1] An answering device including a processor, wherein the processor determines a viewpoint, including objectivity, regarding an inquiry from a user, and determines a replying method based on the determination result regarding the viewpoint.

[0011] [2] The response device according to [1], wherein the processor generates a response to the query using a response method.

[0012] [3] The response device according to [1] or [2], wherein the processor determines that the more a user's preferences are reflected in the query, the lower the objectivity of the query.

[0013] [4] The response device according to any one of [1] to [3], wherein the processor includes the ambiguity of the query as a viewpoint, and determines the response method based on the determination result of at least one of the ambiguity and the objectivity.

[0014] [5] The response device according to [4], wherein the processor determines the ambiguity, and if the ambiguity is higher than a first threshold, determines a response method that reduces the ambiguity.

[0015] [6] The response device described in [5], wherein the processor determines the objectivity when the ambiguity is equal to or less than a first threshold, and when the objectivity is higher than a second threshold, determines the response method to be a unique response.

[0016] [7] The response device described in [5], wherein the processor determines the objectivity when the ambiguity is equal to or less than a first threshold, and determines a response method using multiple responses when the objectivity is equal to or less than a second threshold.

[0017] [8] A response device described in any of [1] to [7], in which a storage device stores information regarding multiple response methods in advance, and a processor determines which of the multiple response methods to use based on the judgment results regarding the received query and perspective.

[0018] [9] A response device according to any one of [1] to [8], wherein the processor sequentially answers questions to the user about a plurality of items necessary for creating a product.

[0019]

[10] A response device described in any of [1] to [9], wherein a storage device stores information on items necessary for creating a product, and a processor reads the information on the items necessary for creating the product based on the viewpoint determination result, identifies a first item among the necessary items that is not satisfied by the query, and generates a question for the user based on the first item as at least part of the response to the query.

[0020]

[11] The answering device according to [9] or

[10] , wherein the processor changes the user interface of the answer depending on the content of the user's input.

[0021]

[12] A response device as described in

[10] or

[11] , in which a storage device stores information on user interfaces corresponding to required items, and a processor selects a user interface corresponding to a question generated based on the first item based on the user interface information and the first item, and outputs the selected user interface to an output device.

[0022]

[13] The answering device according to any one of [8] to

[12] , wherein the processor generates a user interface based on the determined answering method.

[0023]

[14] The answering device according to any one of [1] to

[13] , wherein the processor determines the presentation format of the answering method depending on the objectivity.

[0024]

[15] The response device according to any one of [1] to

[14] , wherein the processor includes a text response in at least a portion of the response to a query having an objectivity higher than a threshold.

[0025]

[16] The response device according to any one of [1] to

[15] , wherein the processor includes an image response in at least a portion of the responses to a query whose objectivity is lower than a threshold.

[0026]

[17] The response device according to any one of [1] to

[16] , wherein the processor accepts a query including an image as the query.

[0027]

[18] The response device according to

[17] , wherein the processor accepts a query including a plurality of images as a query.

[0028]

[19] The response device according to

[18] , wherein the processor accepts a query requesting the creation of a composite image using at least some of the multiple images, and further accepts instructions for the composite image in a subsequent query.

[0029]

[20] The response device described in

[19] , wherein when the processor generates a composite image in response to a query or changes the composite image in response to an instruction, the processor generates text indicating the meaning of the generation or change and outputs the text to the output device.

[0030]

[21] The response device according to

[19] or

[20] , wherein the processor creates a photo book as a composite image in response to the inquiry.

[0031]

[22] The response device according to claim

[21] , wherein the processor accepts a query for a part of the photo book as the query.

[0032]

[23] The response device according to

[22] , wherein the processor generates a plurality of photobooks as a response if the objectivity of the query is lower than a threshold.

[0033] The above object can also be achieved by the reply method described in

[24] below:

[24] A method for replying to an inquiry, the reply method including the steps of: determining, by a processor, a viewpoint including objectivity with respect to the inquiry from the user; and determining, by the processor, a reply method based on the result of the determination regarding the viewpoint.

[0034] The above object can also be achieved by the program described in

[25] below:

[25] A program for causing a computer to execute each step included in the reply method described in

[24] .

[0035] The above object can also be achieved by the recording medium described in

[26] below:

[26] A computer-readable recording medium having recorded thereon a program for causing a computer to execute each step included in the reply method described in

[24] .

[0036] According to one embodiment of the present invention, there are provided an answering device, an answering method, a program, and a recording medium that can flexibly respond to the content of a user's question and derive an appropriate answer.

[0037] 1 is a diagram illustrating an example of a system configuration including an answering device according to the present embodiment; FIG. 2 is a diagram illustrating an example of a hardware configuration of an answering device according to the present embodiment; FIG. 3 is a diagram illustrating functional units of an answering device according to the present embodiment; FIG. 4 is a diagram illustrating an example of a text information table according to the present embodiment; FIG. 5 is a diagram illustrating an example of an image information table according to the present embodiment; FIG. 6 is a diagram illustrating an example of determination results of objectivity and ambiguity according to the present embodiment; FIG. 7 is a diagram illustrating an example of a product information table according to the present embodiment; FIG. 8 is a diagram illustrating an example of a flow information table according to the present embodiment; FIG. 9 is a diagram illustrating an example of a screen transition according to the present embodiment; FIG. 10 is a diagram illustrating an example of a generated content DB according to the present embodiment; FIG. 11 is a diagram illustrating an example of a flow of an answering method according to the present embodiment; FIG. 12 is a diagram illustrating an example of a screen according to the present embodiment; FIG. 13 is a diagram illustrating an example of a screen according to the present embodiment; FIG. 14 is a diagram illustrating an example of a screen according to the present embodiment; FIG. 15 is a diagram illustrating an example of a screen according to the present embodiment;

[0038] Specific embodiments of the present invention will be described below. For ease of explanation, the following description may be given in terms of a GUI (Graphic User Interface). Furthermore, since the basic data processing technologies (communication / transmission technologies, data acquisition technologies, data recording technologies, data processing / analysis technologies, machine learning technologies, image processing technologies, visualization technologies, etc.) required to realize the present invention are well-known technologies, a description thereof will be omitted.

[0039] In addition, in this specification, the concept of "device" includes not only a single device that performs a specific function, but also a combination of multiple devices that exist independently and in a distributed manner but cooperate (link) to perform a specific function.

[0040] Furthermore, in this specification, the term "user" refers to a user of the response device of the present invention, specifically, a person who receives appropriate creation support using the functions of the response device of the present invention as a customer of a content creation service such as a photo book or photo print. Of course, this includes not only customers of such content creation services, but also users of help systems for various products, services, etc.

[0041] In addition, in this specification, the term "person" refers to an entity that performs a specific action, and includes individuals, groups, corporations such as companies, and organizations, as well as computers and devices that constitute artificial intelligence (AI). Artificial intelligence (AI) realizes intelligent functions such as inference, prediction, and judgment using hardware and software resources. The algorithm of the artificial intelligence is arbitrary, and examples include expert systems, case-based reasoning (CBR), convolutional neural networks (CNN), deep neural networks (DNN), Bayesian networks, and subsumption architectures.

[0042] <<One Embodiment of the Present Invention>> [Configuration of the Response System] In one embodiment of the present invention (hereinafter referred to as this embodiment), a situation will be illustrated in which a response system 5 is configured by a response device 10, a user terminal 30, and a server computer 40, all of which are connected to a network 1 shown in Fig. 1. Among these, the user terminal 30 is an information processing device that allows a user to access the response device 10 in order to check or input information required to use a creation service for content that the user wishes to create, such as a photo book. Note that the following description will be given taking the photo book as an example of content in this embodiment.

[0043] In addition, the user terminal 30 accepts user input of images, audio, etc. (hereinafter referred to as materials) that will be used to create content by the response device 10 or the server computer 40, or acquires them by photographing or recording them, and uploads them to the response device 10.

[0044] In this specification, an "image" is defined as an image composed of multiple pixels and represented by the gradation values ​​of each pixel, and includes, for example, an image in which the user or their friends are the subject. Digital image data (hereinafter referred to as "image data") that defines an image at a set resolution is generated by compressing data in which the gradation values ​​for each pixel are recorded using a predetermined compression method. Examples of types of image data include lossy compressed image data such as JPEG (Joint Photographic Experts Group) format, and lossless compressed image data such as GIF (Graphics Interchange Format) or PNG (Portable Network Graphics) format.

[0045] Furthermore, the user may read the materials and input text information such as inquiries and creation instructions on the UI (User Interface) of the response device 10 without using the user terminal 30. In this case, the UI of the response device 10 reads the materials such as photo prints owned by the user, the screen of the user terminal 30 (displaying still images or videos), and audio being played on the user terminal 30 by photographing or collecting sound. Alternatively, the UI reads data of such materials from a recording medium presented by the user. In this case, the user may directly input the text information using the UI such as a keyboard, mouse, or touch panel.

[0046] The response device 10 stores materials transmitted from the user terminal 30 as inquiries and responses at any time in an image information table 212 (described later in FIGS. 3 and 5) to prepare for future interactions with the user and photo book creation. The response device 10 also stores text information transmitted from the user terminal 30 as inquiries and responses at any time in a text information table 211 (described later in FIGS. 3 and 4) to prepare for future interactions with the user and photo book creation.

[0047] The answering device 10 is an information processing device that determines the objectivity and ambiguity of the material and text information using a determination AI 221 (see FIG. 3) included in the answering device 10 or a function provided by the server computer 40, and determines a method of answering to an inquiry based on the determination result. The answering device 10 is also an information processing device that generates an answer using the above-mentioned answering method using an LLM 230 (see FIG. 3) or a generation AI 240 (see FIG. 3) included in the answering device 10 or a function provided by the server computer 40, and responds to the user terminal 30 with the answer. Furthermore, the answering device 10 is an information processing device that creates a photo book, which is content, using the generation AI 240 included in the answering device 10 or a function provided by the server computer 40, and responds to the user terminal 30 with the photo book. Each of these processes is executed autonomously by the answering device 10 according to the answering method of the present invention, or is executed in response to a user instruction received via a predetermined UI.

[0048] Furthermore, as described above, the server computer 40 is a server device that provides the answering device 10 with functions similar to the determination AI 221, the LLM 230, and the generation AI 240 via the network 1. Of course, if the answering device 10 is configured not to require the provision of the functions of the determination AI 221, the LLM 230, and the generation AI 240, the server computer 40 does not need to be included in the answering system 5.

[0049] In addition to the form in which the response device 10 is connected to an external device such as a user terminal 30 or a server computer 40 via a network 1, the response device 10 may be connected directly to the internal bus wiring of the response device 10 and the interface of the external device.

[0050] [Configuration Example of Answering Device] Next, a configuration example of the answering device 10 according to the present embodiment will be described with reference to FIGS. 2 and 3 . Specifically, the answering device 10 is configured by a server device, a desktop personal computer (PC), a notebook PC, a tablet terminal, a smartphone, or the like. Note that the answering device 10 is not limited to a computer owned by an operator of the content creation service or accessible via the network 1. For example, the answering device 10 may be configured by a terminal that is owned and operated by an operator other than the operator, such as a terminal installed in a store or a station, but that can be used by a user when visiting the store or station. Note that the following description will be given taking as an example a case where the answering device 10 is configured by a computer owned by an operator of the creation service, specifically a server device.

[0051] As shown in FIG. 2, the computer constituting the response device 10 includes a processor 11, an auxiliary storage device 12, a main storage device 13, an input device 14, an output device 15, and a communication device 16.

[0052] The processor 11 is composed of, for example, a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), an MCU (Micro Controller Unit), a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), a TPU (Tensor Processing Unit), or an ASIC (Application Specific Integrated Circuit).

[0053] The auxiliary storage device 12 may be composed of, for example, a flash memory, a hard disc drive (HDD), a solid state drive (SSD), a flexible disc (FD), a magneto-optical disc (MO disc), a compact disc (CD), a digital versatile disc (DVD), a secure digital card (SD card), or a universal serial bus memory (USB memory).

[0054] Such auxiliary storage device 12 may be built into the computer main body constituting the response device 10, or may be attached to the computer main body in an external format. Alternatively, the auxiliary storage device 12 may be configured by a NAS (Network Attached Storage) or the like. Furthermore, the auxiliary storage device 12 may be an external device, such as an online storage or a database server, that can communicate with one of the computers constituting the response device 10 via a communication network.

[0055] The auxiliary storage device 12 also stores a program 121, such as an operating system (OS) and an application for executing a content generation process. When the program 121 is read and executed by the processor 11, the computer constituting the response device 10 performs the functions of the reception unit 21, determination unit 22, response unit 23, generation unit 24, and output unit 25 shown in Fig. 3, and specifically, generates a response using a response method based on the objectivity and ambiguity of the material and text information provided by the user, and responds to the user terminal 30, thereby executing a series of processes associated with creating a photo book.

[0056] The main storage device 13 is configured by semiconductor memories such as a ROM (Read Only Memory) and a RAM (Random Access Memory). The processor 11 loads the program 121 onto the main storage device 13 and executes it there.

[0057] The input device 14 is a device that accepts user input operations and is configured, for example, by a keyboard, a mouse, a touch panel, a camera unit, etc. The input device 14 may also include a photographing device implemented in a digital camera, a microphone for collecting sound, etc. The output device 15 is configured, for example, by a display, a speaker, a printer, etc.

[0058] The communication device 16 may be configured, for example, by a network interface card or a communication interface board, etc. The computer configuring the reply device 10 can communicate with other devices connected to the network 1, such as the Internet and a mobile communication line, i.e., the user terminal 30 and the server computer 40, via the communication device 16.

[0059] 3, the answering device 10 has a receiving unit 21, a determining unit 22, an answering unit 23, a generating unit 24, and an output unit 25. These functional units are realized by the processor 11 of the answering device 10 executing the program 121 and working in cooperation with other hardware devices of the answering device 10. In addition, the determining AI 221 in the determining unit 22, the LLM 230 in the answering unit 23, and the generating AI 240 in the generating unit 24 may be called from the server computer 40 and executed.

[0060] [Functional Units in Content Generation Device] Next, the processes executed by each functional unit, namely, the reception unit 21, the determination unit 22, the response unit 23, the generation unit 24, and the output unit 25, provided in the processor 11 (i.e., the CPU) of the response device 10 will be described.

[0061] (Reception Unit) The processing of the reception unit 21 includes processing of receiving various images and text information, which are material for the user's questions or answers, from the input device 14 or from the user terminal 30 via the communication device 16, and storing these in, for example, the auxiliary storage device 12. The storage processing in the reception unit 21 also includes processing of storing the received text information in the text information table 211 and storing the material images in the image information table 212.

[0062] 4 shows an example of the configuration of the text information table 211 of this embodiment. The text information table 211 is a collection of records including values ​​for user ID, case, date and time, input, and output. Among these, the user ID is identification information of the user who input the text information from the user terminal 30, and may also be identification information of the user terminal 30 (the same applies below). Furthermore, if the response device 10 is operated to be used by a single user, this user ID is not required. The date and time is the date and time when the text information was input.

[0063] The input is a description of an inquiry about information desired by the user for generating a photo book or a response to a question from the response device 10, i.e., text, linked to the ID of the text. This text information is a stored sentence or group of words accepted by the UI when the user enters information on the input device 14 or the user terminal 30. On the other hand, the output is a description of a question or response, i.e., text, generated by the response device 10 in response to the inquiry or response from the user, linked to the ID of the text. However, this output value is generated and stored when the response device 10 responds to the user terminal 30, and is not stored before the inquiry from the user terminal 30 or before the response to the question (for convenience of explanation, in FIG. 4, values ​​are stored as if each output value has already been generated and responded to).

[0064] FIG. 5 shows an example of the configuration of the image information table 212 in this embodiment. The image information table 212 is a collection of records including values ​​for user ID, case, date and time, input image, and output image. Among these, the user ID is identification information of the user who input the image as material, and may be identification information of the user terminal 30. Furthermore, if the response device 10 is operated to be used by a single user, this user ID is not required. The date and time is the date and time when the image as material was input. The input image is file information of the image as material received from the user terminal 30 as described above. The output image is file information of the image generated by the response device 10 and sent to the user terminal 30 as a response to an inquiry from a user.

[0065] The method of receiving text information and materials (images) in the receiving unit 21 is not particularly limited, but includes acquiring the images by reading, using a camera unit, a photo print of an image taken by a photographing unit or the like of the user terminal 30 or a playback video. The receiving unit 21 may also acquire the images by downloading the image data, which is the material, from an external device or a web server or the like via the network 1.

[0066] (Determination Unit) The processing of the determination unit 22 includes processing for applying the user inquiry obtained by the processing of the reception unit 21, i.e., text information and materials, to, for example, the determination AI 221 to determine the objectivity and ambiguity of the text information and materials. The determination AI 221 has a model that has undergone deep learning using, for example, a set of a sentence and the objectivity and ambiguity of the sentence (e.g., appropriate values ​​determined by a knowledgeable person) as training data, and outputs values ​​of objectivity and ambiguity by inputting text information into the model. Note that, for materials such as images, the determination AI 221 does not determine the objectivity or ambiguity, and determines that both the objectivity and ambiguity are below the standard.

[0067] In the processing of the determination unit 22, the objectivity and ambiguity values ​​determined by the above determination are temporarily stored in, for example, the main storage device 13 (see FIG. 6) and are referred to during processing by the response unit 23. In the example shown in FIG. 6, the ambiguity and objectivity are determined for each input (text). Both the ambiguity and the objectivity are defined as events that take a value range of minimum: 0 to maximum: 100, but this is merely an example, and a granularity level such as minimum: C to maximum: A, or a value range such as OK or NG may also be specified.

[0068] (Answering Unit) The processing of the answering unit 23 includes determining a response method for the user's inquiry based on the ambiguity and objectivity of the inquiry determined by the determining unit 22. The processing of the answering unit 23 also includes generating a response based on the response method using the LLM 230 or the generation AI 240 and sending the response to the user terminal 30. When generating this response, product specification information indicated in the product information table 231, information on each procedure associated with product ordering indicated in the flow information table 232, etc. may be used as appropriate. Furthermore, a user interface may be generated by the generation AI 240, for example, based on the determined response method. In this case, the generation AI 240 is assumed to have a model that has learned the correct user interface pattern for each response method (described below) through deep learning.

[0069] The answering unit 23 may adopt the following patterns as an algorithm for determining the answering method. For example, if the ambiguity is higher than a first threshold (e.g., 50), a pattern may be adopted in which a replying method that reduces the ambiguity is determined. This replying method that reduces the ambiguity may, for example, present examples of events that may correspond to keywords contained in the query and ask the user whether the events match the purpose or intent of the query, or present examples of multiple events and ask the user which of them matches the purpose or intent of the query.

[0070] Another pattern is to determine the objectivity when the ambiguity is equal to or less than a first threshold, and to decide on a response method that provides a unique answer when the objectivity is higher than a second threshold (e.g., 50). This response method, when the content of the inquiry is clear and the objectivity is sufficient, presents the correctness or incorrectness of the phenomenon suggested by the inquiry (e.g., whether or not the understanding is correct), or specific values ​​(e.g., product specifications such as price, number of pages, and size), or asks the user about these.

[0071] In addition, when the ambiguity is equal to or less than a first threshold, the objectivity is determined, and when the objectivity is equal to or less than a second threshold, the answering method is determined so that the user selects an appropriate answer from multiple answers. These multiple answers may be stored in advance by the answering unit 23 or generated on an ad hoc basis based on the product information table 231 or materials. When generating answers based on information in the product information table 231, the LLM 230 generates options based on information (e.g., the product information table 231) about the product to be created (e.g., a photo book) and the specifications of the product, such as multiple sizes, number of pages, and theme. When generating answers based on materials, the generation AI 240 generates multiple correction patterns and product examples that match the perspectives (e.g., the subject's atmosphere, such as cute or bright, or the subject's behavior, such as travel or work) indicated by the materials or text information.

[0072] Another pattern is to determine an answer method that includes at least a part of the answer in text for an inquiry with an objectivity higher than a threshold. Because the objectivity of the inquiry is high, requesting an answer in text is likely to lead to obtaining a unique answer. On the other hand, there is also a pattern in which an answer method that includes at least a part of the answer in image for an inquiry with an objectivity lower than a threshold. In the case of an inquiry with low objectivity, it is easier to confirm the intent of the inquiry more efficiently by presenting, for example, specific candidate images or product examples (photo books) generated by estimating the user's intent, rather than presenting an answer to the user in text.

[0073] Based on the flow indicated by the flow information table 232, the reply unit 23 generates questions about required items, i.e., items necessary for product creation, which are included in the UI of each step constituting the flow and are indicated by the product information table 231, in the LLM 230, and provides the answers to the user with these questions. In this case, the reply unit 23 compares the information included in the inquiry with the required items in the product information table 231, and identifies the required items that were not satisfied by the inquiry. Then, the reply unit 23 generates questions to be asked of the user about the items in the LLM 230 as at least part of the answer to the inquiry.

[0074] FIG. 7 shows an example of the configuration of a product information table 231 in this embodiment. This product information table 231 is a table that specifies the specifications of a product that is the subject of a user inquiry. The product information table 231 is a collection of records that link the product's name, size, recommended pixel count, price, and values ​​of required items with an ID that identifies the product as a key. Of these, the size describes specifications related to the size of, for example, a photo book or postcard. The recommended pixel count specifies the number of pixels of an image required depending on the size. The price is a value that specifies the unit cost of creating the target product. The required items are values ​​that specify items for which information is required when creating the target product.

[0075] Fig. 8 shows an example of the configuration of the flow information table 232 in this embodiment. Fig. 9 shows an example of UI, i.e., screen transitions, in a certain flow, shown in the flow information table 232. The flow information table 232 in this embodiment is a table that defines a series of procedures from product order acceptance to creation and payment. The flow information table 232 is a collection of records that use an ID that identifies a flow as a key to link the target product to which the flow applies, the flow, and each value of the UI.

[0076] Among these, the value of the target product is the ID of each product specified in the product information table 231. The flow is a list of an outline of each of the above steps in the order in which the steps are executed. The UI is a value that specifies the screen data (screens G40 to G45 in FIG. 9) that is delivered to and displayed on the user terminal 30 or the like in each step. This screen data also includes information on items to be received from the user in the corresponding step and items to be transmitted from the response device 10 to the user.

[0077] The answering unit 23 identifies the flow specified in the flow information table 232 that corresponds to the product specified by the user by using the value in the "Target Product" column, and responds to the user terminal 30 with a UI, i.e., screen data, corresponding to the step so that each step in the flow can be executed in order starting from "1." In response to this, the answering unit 23 generates an answer using the LLM 230 or the generation AI 240 based on the answering method, and distributes a response screen with this answer set in the area of ​​the corresponding item in the screen data to the user terminal 30. In conjunction with the distribution of the response screen including this answer, the answering unit 23 acquires information on at least the required items from the user terminal 30. The answer generated by the answering unit 23 in this way is stored in the "Output" column of the text information table 211 in a form corresponding to the query "Input."

[0078] (Generation Unit) The processing of the generation unit 24 includes processing for generating a photo book, which is content, by inputting inquiries and materials obtained from the user terminal 30 by the reception unit 21, and information about the product (such as the user's desired specifications) obtained from the user terminal 30 through the response (question) by the response unit 23, into the generation AI 240. Note that the above materials may also include those generated by the generation unit 24 by assigning text information and materials obtained from the user terminal 30 to the generation AI 240. Furthermore, the content (photo book data) generated by the generation unit 24 is stored in the generated content DB 241 or held in the main memory device 13.

[0079] Fig. 10 shows an example of the configuration of the generated content DB 241 of this embodiment. As shown in Fig. 10, this generated content DB 241 is a collection of records made up of a case, a user ID, a date and time, and content data. The case is a value that uniquely identifies content. The user ID is the ID of the user who issued the instruction to generate the photo book. The date and time is the date and time when the photo book was generated. The content data is a data file of the photo book.

[0080] (Output Unit) The processing of the output unit 25 includes processing for displaying the content generated by the generation unit 24, i.e., the photo book, on the output device 15 such as a display or on the user terminal 30. Note that the method for outputting content such as a photo book is not particularly limited, and includes, for example, the output unit 25 displaying the photo book data on the display or monitor of the output device 15 or the user terminal 30, printing the data, transmitting the data to another user, etc. In addition to the photo book described above, the content may include media consisting of one or more pages or cards on which images are published, such as postcards, message cards, electronic albums, bromides, etc.

[0081] [Example of a Flow of an Answering Method] Next, as an example of the operation of the answering device 10 in this embodiment, a processing flow of answering and content generation using the device will be described. The flow described below uses the answering method of the present invention. In other words, each step in the flow of the answering method described below corresponds to a component of the answering method of the present invention. Note that the flow below is merely an example, and some steps in the flow may be deleted, new steps may be added to the flow, or the execution order of two steps in the flow may be reversed, as long as it does not deviate from the spirit of this embodiment.

[0082] The steps in the content generation flow according to this embodiment are performed by the processor 11 included in the response device 10 in the order shown in Fig. 11. That is, in each process in the content generation flow, the processor 11 executes processing corresponding to each step in Fig. 11 among data processing defined in the application program for content generation.

[0083] Specifically, in the content generation flow according to this embodiment, the reception unit 21 first receives an inquiry including at least one of text information and materials from, for example, the user terminal 30 (S1), and stores the information included in the inquiry (text information and materials) in the text information table 211 and the image information table 212. This text information and materials are related to content products that the user desires or is interested in. During this process, the reception unit 21 delivers, to the user terminal 30, a screen G1 including a UI such as that shown in FIG. 12, for example.

[0084] The UI illustrated here is a series of chat lines ML in which a user's message Mc and a responsive message Ms from the reply unit 23 are repeatedly exchanged almost alternately. Text information and materials (images, etc.) are acquired from the user's message Mc. The example in Figure 12 shows how a user inquires about the price of a photo book, followed by an answer (including a question) and the user's response to that answer, are exchanged three times until the final answer (the price of the photo book) is reached.

[0085] Next, the determination unit 22 inputs the text information obtained from the inquiry by the reception unit 21 and stored in the text information table 211 to the determination AI 221, and determines the ambiguity (S2). Note that the details of the determination model in the determination AI 221 have already been described.

[0086] Ambiguity here can be said to indicate the degree to which the information being judged indicates some phenomenon, but cannot uniquely identify the phenomenon and can be interpreted in several ways. For example, "the price of a photo book" cannot be calculated unless multiple calculation bases (such as the size and number of pages of the photo book) are made clear, except in cases where the price is a fixed, uniform amount. In other words, the query "the price of a photo book" lacks the information necessary to uniquely identify it, and can be said to be information that contains ambiguity.

[0087] If the result of the above determination is that the ambiguity of the query does not meet the predetermined standard, i.e., the query is undesirably ambiguous (S3: N), the response unit 23 generates a question to reduce the ambiguity of the query (S4).

[0088] If the ambiguity of the inquiry does not satisfy the criteria, the answering unit 23 may adopt the above-mentioned chat-style UI and distribute it to the user terminal 30. With such a chat-style UI, questions and answers are not fixed, and flexible exchanges can be made depending on the answers from the user.

[0089] The question (method of answering) to reduce the ambiguity may be, for example, to provide an example of an event that may correspond to the keywords contained in the query and ask the user whether that event matches the intent or purpose of the query, or to provide an example of multiple such events and ask the user which of them matches the intent or purpose of the query.

[0090] The screen G1 shown in FIG. 12 illustrates a situation in which a user has sent a message Mc asking, "How much does a photo book cost?" This message Mc contains a great deal of ambiguity, and the message Ms is generated as a response, presenting various examples of photo book sizes and asking the user to choose one, as a question to reduce the ambiguity. The user responds by indicating that "A5" is the desired size. Similarly, to reduce the ambiguity, the message Ms is generated as a response, presenting various examples of the number of pages in a photo book and asking the user to choose one, as a question to reduce the ambiguity. The user responds by indicating that "16 pages" is the desired number of pages. In other words, the answering unit 23 obtains an answer to the question to reduce the ambiguity from the user terminal 30 (S5). When an answer to this question is received, the answering unit 23 returns to S2 and repeats the process until the determination in S3 is "Y."

[0091] On the other hand, if the result of the determination in S3 indicates that the ambiguity of the query satisfies the criteria (S3: Y), the determination unit 22 inputs the text information indicated by the query to the determination AI 221 and determines the objectivity (S6). Note that the details of the determination model in the determination AI 221 have already been described.

[0092] Here, objectivity can be said to indicate the degree to which the information being judged expresses some phenomenon using specific numerical values, etc., and is consistent in its understanding from person to person. In other words, information that primarily expresses emotions such as personal impressions, preferences, and tastes, and does not express specific values ​​or forms, is subject to a large range of understanding from person to person and has a low degree of objectivity. Therefore, the more or more expressions related to preferences, etc. (e.g., words, sentences, drawings) a piece of content contains, the lower its objectivity will be judged to be.

[0093] If the result of the above determination is that the objectivity of the inquiry does not meet the criteria (S7: N), the response unit 23 determines a response method in which the user selects an appropriate answer from multiple answers. These multiple answers may be stored in advance by the response unit 23 or generated on an ad hoc basis based on the product information table 231 or materials. When generating answers based on information in the product information table 231, the LLM 230 generates options based on information (e.g., the product information table 231) about the product to be created (e.g., a photo book) and the specifications of the product, such as multiple sizes, number of pages, and theme. When generating answers based on materials, the generation AI 240 generates multiple correction patterns and product examples that match the perspectives (e.g., the subject's atmosphere, such as cute or cheerful, or the subject's behavior, such as traveling or work) indicated by the materials or text information.

[0094] 13, on screen G2, the user inputs a message Mc, i.e., text, "Please correct the image to make it look cute," along with a material image Pc. Therefore, by inputting this "Please correct the image to make it look cute" to the judgment AI 221, the judgment unit 22 obtains a low degree of objectivity as a judgment result due to the presence of the emotional phrase "cute," which indicates a preference for "cute." In other words, as a result of the objectivity judgment (S6, S7), the response unit 23 determines that the inquiry has a low degree of objectivity (S7: N).

[0095] Therefore, in order to infer or supplement the intent of the inquiry with low objectivity, the answering unit 23 generates a question presenting multiple options that specifically indicate the corresponding phenomenon using the LLM 230 or the generation AI 240, and responds to the user terminal 30 with this question (S8). In the case of screen G2 in Fig. 13, the response is a message Ms indicating that three patterns "A" to "C" are presented as "correction" patterns for the "cute feeling" of the original image Pc, and an image Ps showing the proposed corrections.

[0096] In this case, the answering unit 23 instructs the LLM 230 to generate a response message confirming which of the three patterns "A" to "C" the user prefers, and obtains the message Ms, "Understood. Which of A, B, and C is your preferred correction effect?" The answering unit 23 also provides the source image Pc to the generation AI 240 as base content, and generates the image Ps by inputting a prompt, for example, "Prepare three examples of cute corrections that women prefer." The message Ms generated in this way is stored in the output column of the text information table 211. Similarly, the generated image Ps is stored in the output image column of the image information table 212.

[0097] By presenting the message Ms and image Ps, the response unit 23 obtains the user's response, "It's B," as the preferred correction effect from the user. That is, the response unit 23 obtains the user's response to the message or the like presented to confirm the correction effect (S9) and stores this in an input field in the text information table 211. Because the objectivity of such a user inquiry is insufficient, the lower part of screen G2 in FIG. 13 shows an interaction in which the message Ms and image Ps are generated and presented as described above and necessary information is obtained from the user to specify the content of the correction ("emotional feel") associated with printing the image.

[0098] On the other hand, if the determination results in S6 and S7 show that the objectivity of the inquiry satisfies the criteria (S7: Y), the reply unit 23 determines a reply method with a unique reply and sends back a reply generated by that reply method (S10). This reply method, if the content of the inquiry is clear and the objectivity is sufficient, corresponds to a reply method that presents the correctness or incorrectness of the phenomenon suggested by the inquiry (e.g., whether or not the understanding is correct), or a specific value (e.g., product specifications such as price, number of pages, and size), or asks the user about these.

[0099] 12, when the user responds with "16 pages" as the number of pages in the photo book, the response unit 23 presents the price as "1,500 yen in that case." The number of pages in the photo book, "16 pages," is an objective number that leaves no room for other interpretations, and by comparing this page count with the provisions in the "Price" column of the product information table 231, the basic fee of "1,500 yen" can be uniquely identified.

[0100] In some cases, a user's inquiry may include only material images Pc, with no text information. The example of screen G3 shown in FIG. 14 illustrates a situation in which a user wishing to create a photo book has input images as material for creating the photo book, without inputting any text. In this case, the answering device 10 starts processing from a situation in which the objectivity determination (S7) in the flow of FIG. 11 determines that the response is not objective (S7: N). In other words, the answering unit 23 executes a question (S8) presenting multiple options.

[0101] In this case, the answering unit 23 inputs the image Pc to the LLM 230 or the generation AI 240, for example, and obtains a message Ms, which is a question presenting multiple options, such as "I've generated a photobook prioritizing 'photos with friends on a trip,' what do you think?" by providing a prompt such as "Estimate the situation shown in this image and generate a message for the user to confirm whether the estimation result is correct or not." The answering unit 23 also inputs the image Pc to the generation AI 240 as base content, and obtains candidate images Ps for the photobook and responds with them by providing a prompt such as "Generate a candidate photobook made up of multiple images of friends on a trip."

[0102] After viewing the candidate images Ps, the user checks whether the photo book reflects their intentions. If the check finds no problems, the user sends an order message for the photo book from the user terminal 30 to the answering device 10. Meanwhile, the generation unit 24 of the answering device 10 receives the order message and generates a photo book based on the candidate images Ps and the element images Pc or a group of images related to the user that are separately stored in the auxiliary storage device 12 or the server computer 40, thereby completing the photo book generation procedure. On the other hand, if the user finds problems or feels that corrections or changes are necessary as a result of the check, the user inputs a message regarding their desired creation policy and responds to the answering unit 23.

[0103] Screen G3 in Figure 14 shows a situation in which the message "Create an album with photos of friends, not just travel photos," is input and responded to by the answering unit 23. In response to this, the answering unit 23, for example, inputs the material image Pc as base content to the generation AI 240 and assigns a prompt "Generate a photo book candidate composed of multiple images featuring friends," thereby obtaining new candidate images Ps for the photo book and responding to the user. In this case, in response to the new candidate images Ps, the answering unit 23 inputs a prompt "Generate a message notifying that the candidates have been regenerated and requesting confirmation" to the LLM 230, and obtains a message Ms "I understand. The candidates have been regenerated. Please confirm.", which is then responded to the user.

[0104] Unlike the above operation, when a user feels that corrections or changes are necessary, the response unit 23 can also respond to the case where the user responds to the response unit 23 by inputting an image Pc that suggests the desired creation policy, rather than a message about the creation policy that the user desires.

[0105] 15 shows a situation in which image Pc2, which suggests the user's intention of "I want to create an album with photos with friends, not limited to travel," is input and responded to by the answering unit 23. In response to this, the answering unit 23 can identify, for example, the concept of "photos with friends" by providing the LLM 230 with images Pc1 and Pc2 and a prompt saying, "Describe the situation common to images Pc1 and Pc2."

[0106] Therefore, the response unit 23 inputs the above images Pc1 and Pc2 as base content to the generation AI 240, and by adding, for example, a prompt such as "Generate a photo book candidate composed of multiple images of friends," obtains new candidate images Ps for the photo book and responds with this. The user can view these new candidate images Ps and reconfirm whether the photo book reflects the user's intentions. This series of processes related to the creation and confirmation of candidate images P is repeated until the user determines that there are no problems with the candidate images Ps.

[0107] After providing the unique response (S10), the response unit 23 determines whether the user's inquiry has been completed, for example, by notifying the user of a message to that effect, or by comparing the incomplete status of a procedure such as photo book generation or ordering procedure with the final step in the corresponding flow in the flow information table 232 (S11). If the response device 10 determines that the inquiry has not been completed (S11: N), the process returns to S1. On the other hand, if the response device 10 determines that the inquiry has been completed (S11: Y), the response device 10 responds to the user terminal 30 with a photo book, which is content generated by the generation unit 24 (S12), and ends this flow.

[0108] The response device 10 in this embodiment can also handle a case where the inquiry received from the user in S1 in the flow shown in Fig. 11 lacks information on items essential for generating the target product. In this case, the response unit 23 receives an inquiry about the photo book from the user (message Ms1 on screen G5 in Fig. 17) (S20), and compares the information indicated by the inquiry with the values ​​of the essential items in the record related to the photo book in the product information table 231 to determine the missing items (S21).

[0109] For example, when an inquiry about a photo book is received from a user, the reply unit 23 identifies a record with the name "Photo Book" in the product information table 231. The reply unit 23 then refers to the required field column of the identified record and obtains values ​​for, for example, "size, number of pages, and theme." Meanwhile, suppose that the message Ms1, which is an inquiry input by the user, includes values ​​for "size" and "number of pages." In this case, the reply unit 23 determines that "theme" is a missing field.

[0110] If the above determination indicates that there are no missing items (S22: N), the answering unit 23 terminates this flow. On the other hand, if the above determination indicates that there are missing items (S22: Y), the answering unit 23 generates a message Ms (see screen G5 in FIG. 17 ) as a question regarding the missing items and responds to the user (S23). In this case, the answering unit 23 generates a question by, for example, inputting a prompt "Generate a message to confirm the desired size and theme of the photo book" to the LLM 230 or the generation AI 240. The user then views this question and responds with a message Ms2 (see screen G5 in FIG. 17 ) describing the size and theme of the photo book they desire. In response, the answering unit 23 obtains the size and theme information from the message Ms2 (S24), and terminates this flow.

[0111] In the flow shown in FIG. 16 , if the determination of missing items reveals that there are missing items ( S22: Y), the response unit 23 references the record for the target product in the flow information table 232, and, for example, avoids displaying the screen "3: Size / Number of Pages Acceptance" in the "Flow" and instead displays the screen "2: Theme Selection Acceptance" ( S33 ), as shown in the flow shown in FIG. 18 . In the example of the screen transition shown in FIG. 9 , screen G40 is displayed followed by screen G41, but screen G42 is skipped (because the size and number of pages have already been acquired) and screen G43 (cover selection) is displayed in the next step. By controlling the screens in this way, unnecessary screen displays and the associated interactions with the user can be avoided, improving the efficiency of the flow.

[0112] In the above example, the content for which the response device 10 generates a response is an image or a photo book made up of the images, but the content is not limited to this. For example, the content may be video, audio, music, novels, paintings, etc. In this case, the material may be various types of video data, human conversation or singing, instrumental sound or music, or some kind of short sentence or sketch.

[0113] Although specific embodiments of the present invention have been described above, the above embodiments are merely examples given to facilitate understanding of the present invention and are not intended to limit the present invention. That is, the present invention may be modified or improved from the embodiments described below without departing from the spirit of the present invention. The present invention also includes equivalents thereof. Furthermore, embodiments of the present invention may include a combination of the above embodiments with one or more of the following modifications.

[0114] (Regarding the Computer Constituting the Answering Device) In the above embodiment, the answering device 10 of the present invention is configured by a computer indirectly operated via the network 1, such as a server device owned by a content creation service operator. However, the present invention is not limited to this. The answering device 10 of the present invention may also be configured by a computer directly available to the operator, such as a PC (Personal Computer). The server computer 40 may be, for example, a server computer for a cloud service, specifically, a server computer for an ASP (Application Service Provider), SaaS (Software as a Service), PaaS (Platform as a Service), or IaaS (Infrastructure as a Service). In this case, when necessary information is input to the user terminal 30 and the information is distributed from the answering device 10 to the server computer 40, the server computer 40 performs various processes (calculations) based on the input information, including determining ambiguity and objectivity, and generating answers, questions, and photobooks, and the calculation results are output on the user terminal 30. That is, the functions of the server computer 40 that can constitute the response device 10 of the present invention can be used on the user terminal 30 side via the response device 10. Alternatively, the response device 10 may be constituted by the above-mentioned PC or server computer, or a smartphone (a type of user terminal 30) used by an operating company or a user.

[0115] (Regarding the Processor Configuration) In this embodiment, each process is executed by a computer. The computer may execute these processes using a processor, a program, or a combination thereof. The computer may be a general-purpose computer, a computer for specific applications, a system such as a workstation, or other hardware element capable of executing a program.

[0116] The processor may be configured with one or more pieces of hardware, and the type of hardware is not limited. For example, the processor may be configured with hardware such as a central processing unit (CPU), a micro processing unit (MPU), a programmable logic device such as a field programmable gate array (FPGA), a dedicated circuit for executing specific processing such as an application specific integrated circuit (ASIC), a graphics processing unit (GPU), or a neural processing unit (NPU).

[0117] Furthermore, the processor has each unit or each means that executes various processes in this embodiment. Furthermore, the type of hardware may be a combination of different types of hardware. When multiple pieces of hardware are configured to execute one or more processes of a certain processor, the multiple pieces of hardware may exist in devices that are physically separate from each other, or may exist in the same device. Furthermore, in any of the embodiments, the order of each process performed by the processor is not limited to the order described above and may be changed as appropriate. The hardware is configured by an electric circuit (circuitry) that combines circuit elements such as semiconductor elements.

[0118] Furthermore, the present embodiment may be implemented by hardware, software, firmware, microcode, or a combination thereof. Software, firmware, and microcode may be configured by a program. A program may also be, for example, a group of program modules, each function of which may be implemented by a processor configured to perform the respective function. The program may be program code or multiple code segments stored in one or more non-transitory computer-readable media (e.g., storage media or other storages). The program may be stored in multiple non-transitory computer-readable media that reside in physically separate devices. A program code or a code segment may represent a procedure, a function, a subprogram, a routine, a subroutine, a module, a software package, a class, or any combination of instructions, data structures, or program statements. A program code or a code segment may be connected to another code segment or a hardware circuit by sending or receiving information, data, arguments, parameters, or memory contents.

[0119] REFERENCE SIGNS LIST 1 Network 5 Answering system 10 Answering device 11 Processor 12 Auxiliary storage device 121 Program 13 Main storage device 14 Input device 15 Output device 16 Communication device 21 Receiving unit 211 Text information table 212 Image information table 22 Determination unit 221 Determination AI 23 Answering unit 230 LLM 231 Product information table 232 Flow information table 24 Generating unit 240 Generating AI 241 Generated content DB 25 Output unit 30 User terminal 40 Server computer

Claims

1. A reply device having a processor, wherein the processor determines a viewpoint, including objectivity, regarding a query from a user, and determines a reply method based on the determination result regarding the viewpoint.

2. The answering device according to claim 1, wherein the processor generates an answer to the inquiry using the answering method.

3. The replying device according to claim 2, wherein the processor, in the determination, determines the objectivity to be lower the more the user's preferences are reflected in the query.

4. The response device according to claim 1, wherein the processor includes the ambiguity of the query as the viewpoint, and determines the response method based on a determination result of at least one of the ambiguity and the objectivity.

5. The answering device according to claim 4, wherein the processor determines the ambiguity, and if the ambiguity is higher than a first threshold, determines an answering method that reduces the ambiguity.

6. The answering device according to claim 5, wherein the processor determines the objectivity when the ambiguity is equal to or less than a first threshold, and determines a unique answer answering method when the objectivity is higher than a second threshold.

7. The answering device according to claim 5, wherein the processor determines the objectivity when the ambiguity is equal to or less than a first threshold, and determines a multiple answer answering method when the objectivity is equal to or less than a second threshold.

8. The response device according to claim 1, wherein a storage device stores information relating to a plurality of response methods in advance, and the processor determines which of the plurality of response methods to use based on the received inquiry and the judgment result relating to the viewpoint.

9. The reply device according to claim 1 or 2, wherein the processor sequentially answers questions to the user about a plurality of items necessary for creating a product.

10. A response device as described in claim 1 or 2, wherein a storage device stores information on items necessary for creating a product, and the processor reads the information on the items necessary for creating the product based on the determination result of the viewpoint, identifies a first item among the necessary items that is not satisfied by the inquiry, and generates a question for the user based on the first item as at least part of a response to the inquiry.

11. The answering device according to claim 9, wherein the processor changes a user interface of the answer in accordance with the content of the user's input.

12. The response device according to claim 10, wherein the storage device holds information on user interfaces corresponding to the required items, and the processor selects a user interface corresponding to a question generated based on the first item based on the information on the user interface and the first item, and outputs the selected user interface to an output device.

13. The answering device according to claim 8, wherein the processor generates a user interface based on the determined answering method.

14. The answering device according to claim 1, wherein the processor determines a presentation format of the answering method according to the degree of objectivity.

15. The answering device according to claim 1, wherein the processor includes a text answer in at least a portion of the answers to a query for which the objectivity is higher than a threshold value.

16. The response device according to claim 1, wherein the processor includes an image response in at least a portion of the responses to a query for which the objectivity is lower than a threshold value.

17. The answering device of claim 1, wherein the processor accepts a query including an image as the query.

18. The answering device of claim 17, wherein the processor accepts a query including a plurality of images as the query.

19. The response device of claim 18, wherein the processor accepts, as the query, a request to create a composite image using at least some of a plurality of images, and further accepts instructions for the composite image in a subsequent query.

20. The response device according to claim 19, wherein the processor, when generating a synthetic image in response to the query or when changing the synthetic image in response to the instruction, generates text indicating the meaning of the generation or change, and outputs the text to an output device.

21. The response device of claim 19, wherein the processor creates a photo book as the composite image in response to the query.

22. The response device according to claim 21, wherein the processor accepts, as the query, a query for a part of the photo book.

23. The response device of claim 22, wherein the processor generates a plurality of photobooks as the response if the objectivity of the query is below a threshold.

24. A method for responding to an inquiry, comprising: a step of determining, by a processor, a viewpoint, including objectivity, regarding an inquiry from a user; and a step of determining, by the processor, a response method based on the result of the determination regarding the viewpoint.

25. A program for causing a computer to execute each step included in the reply method according to claim 24.

26. A computer-readable recording medium having a program recorded thereon for causing a computer to execute each step included in the reply method described in claim 24.

Images