Information processing system, information processing method and program

Abstract

To provide a technology for improving the quality of dialogue with users. [Solution] According to one aspect of the present invention, an information processing system is provided, which has at least one processor, and the processor is configured to execute the following steps by reading a program: a progress control step controls the progress of an interview with a first answering subject; a first presentation step causes the first answering subject to present a first question in the interview, the first question being generated based on progress conditions preset by at least a management subject; an acquisition step acquires first evaluation information for an evaluation target, the evaluation target being an object to be evaluated by the first answering subject in the interview, the first evaluation information including information describing the results of an evaluation of the evaluation target by a reference subject different from the first answering subject; and a second presentation step causes the first answering subject to present a second question generated based on at least the first evaluation information in the interview.

Description

[Technical Field]

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

[0002] Patent Document 1 discloses a technology for returning appropriate answers to user questions to the user. In Patent Document 1, a content table for interactive processing is created by having a computer execute a function for registering a plurality of contents (59) including question information (58E) that defines questions expected from the user and answer information (58F) that defines answers to the questions, a function for registering classification information (58B-D) that defines classifications to which the questions correspond, and a function for registering transition destination classification information (58J-L) that defines classifications to which the next expected question from the user should be narrowed down. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-024765 Summary of the Invention [Problem to be solved by the invention]

[0004] However, there is room for improvement in dialogue systems to improve the quality of dialogue with users.

[0005] In view of the above circumstances, the present invention provides a technique for improving the quality of dialogue with a user. [Means for solving the problem]

[0006] According to one aspect of the present invention, there is provided an information processing system comprising at least one processor, the processor being configured to execute the following steps by reading a program: a progress control step controlling the progress of an interview with a first answering subject; a first presentation step having the first answering subject present a first question in the interview, the first question being generated based on progress conditions preset by at least a management subject; an acquisition step acquiring first evaluation information for an evaluation target, the evaluation target being an object to be evaluated by the first answering subject in the interview, the first evaluation information including information describing the results of an evaluation of the evaluation target by a reference subject different from the first answering subject; and a second presentation step having the first answering subject present a second question in the interview, the second question being generated based on at least the first evaluation information.

[0007] According to this aspect, a technique for improving the quality of interaction with a user is provided. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a configuration diagram illustrating an information processing system 1 according to an embodiment of the present invention. [Figure 2] 2 is a block diagram showing functions realized by a control unit 23 and the like in the information processing device 2. FIG. [Figure 3] 3 is a flowchart showing an outline of information processing according to the present embodiment. [Figure 4] FIG. 2 is an activity diagram showing the flow of information processing according to the present embodiment. [Figure 5] FIG. 5 is a diagram showing an example of a setting screen 500. [Figure 6] FIG. 6 is a diagram showing an example of an interview screen 600. [Figure 7] FIG. 7 is a diagram showing another example of the interview screen 700. [Figure 8] FIG. 2 is a diagram showing an example of an evaluation information table T. DETAILED DESCRIPTION OF THE INVENTION

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings. Various features shown in the following embodiments can be combined with each other.

[0010] Incidentally, the program for realizing the software appearing in one embodiment may be provided as a non-transitory computer-readable medium, or may be provided so that it can be downloaded from an external server, or may be provided so that the program is started on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] Furthermore, various information processing according to an embodiment may realize input and output corresponding to the input. Here, the form of information referenced in such information processing (hereinafter referred to as reference information) is not limited as long as an output is obtained as a result of the input. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a decision formula such as a regression formula constructed using a statistical method), a trained model that has previously learned the correlation between input and output, or a generative AI such as a large-scale language model (these models include parameters that establish the correlation between input and output) or a visual language model that can output a desired result in response to a prompt.

[0012] In one embodiment, a "unit" may include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In one embodiment, various information is handled, and this information is represented, for example, by physical values ​​of signal values ​​representing voltage and current, high and low signal values ​​as a binary bit set consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculations can be performed on a circuit in the broad sense.

[0013] Furthermore, a circuit in the broad sense is a circuit realized by at least an appropriate combination of a circuit, circuitry, processor, memory, etc. The processor may be a general-purpose processor or a dedicated circuit. That is, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0014] 1. Hardware Configuration In this section, the hardware configuration of this embodiment will be described.

[0015] FIG. 1 is a configuration diagram showing an information processing system 1 according to this embodiment. The information processing system 1 includes a user terminal 3 and an information processing device 2, which are connected via a network 4. In the information processing system 1, an interview is provided for an answering user. The interview proceeds by the information processing system 1 presenting questions to the answering user and obtaining answers to the questions from the answering user. The user terminals 3 include a first user terminal 3A associated with the administrative user and a second user terminal 3B associated with the answering user. The first user terminal 3A and the second user terminal 3B have substantially the same configuration, and will be referred to as user terminal 3 when there is no need to distinguish between them.

[0016] In this embodiment, the administrative user is an example of a management entity involved in generating questions for the answering user. The answering user is an example of an answering entity that answers the generated questions. The administrative user may be a person who manages the information processing system 1, or may be a person who wishes to obtain information from the answering user. Note that, when a question is generated by inputting predetermined information into reference information of a generation AI or the like, the reference information may be the administrative entity. Also, when an answer is generated by inputting a generated question into reference information of a generation AI or the like, the reference information may be the answering entity. In particular, the administrative entity and the answering entity may be AI agents that autonomously investigate and collect necessary information based on the purpose of the interview. In the following explanation, the case where the administrative entity is the administrative user and the answering entity is the answering user will be mainly illustrated.

[0017] The purpose of the interview is arbitrary, for example, marketing research on a product (provision of a product or service). Information desired to be obtained from the responding user is, for example, information about the product or product category, information about the provider (corporation or individual) who provides the product, and information about the user who uses the product. The interview may also be an interview or the like aimed at obtaining information about the respondent's own abilities, preferences, etc.

[0018] Information about products, etc. includes any information related to products, etc., and may include, for example, the actual use of the product, etc. (frequency of use, situations in which it is used, reasons for use, use in combination with other products, reasons for continued use, satisfactions and dissatisfaction with use, etc.), information about purchasing behavior (place of purchase, purchasing motivation, purchase frequency, purchase price, products compared and considered and reasons for comparison, etc.), evaluation of the appealing elements (impressions of the appealing elements emphasized in advertisements or packages, degree of empathy when someone says, "this product seems to be good," degree of match between the appealing elements and one's own values, etc.), relative evaluation with competing products, emotional background and value background (psychological background as to why one chooses that product, etc., self-image or image of life that one feels one can obtain through the product, expressions that cause an adverse reaction), information about price sensitivity, information sources that influenced one, purchasing decision-making process, latent needs, proposed opinions, etc.

[0019] The information about the provider may include any information about the provider, such as whether or not the provider is known and how they became aware of the provider, the level of trust and liking for the provider, the degree of empathy for the provider's values, differences in image compared to other providers, requests for improvement for the provider, impressions of the provider seen in advertisements, etc.

[0020] Therefore, the Admin User may include corporations and individuals that provide the product, etc., corporations and individuals that provide competing products, etc., brand operators, sales agents, licensors and licensees, and any other corporations and individuals involved in the manufacture, sale, advertising, etc. of the product, etc. The Admin User may also be an individual belonging to any of these organizations. The Answering User is the consumer of the product, etc., and in the case of BtoC products, this corresponds to a general consumer. Furthermore, in the case of BtoB products, this corresponds to a corporation or individual that is a consumer. Furthermore, the Answering User may be the entity that provides the product, etc., or the corporation or individual in charge of distributing the product, etc.

[0021] In one embodiment, the information processing system 1 is comprised of one or more devices or components. For example, if the information processing system 1 is comprised only of an information processing device 2, the information processing device 2 can be the information processing system 1. If the information processing system 1 is comprised only of a user terminal 3, the user terminal 3 can be the information processing system 1. If the information processing system 1 includes the information processing device 2 and the user terminal 3, the combination of the information processing device 2 and the user terminal 3 can be the information processing system 1. More specifically, the information processing system 1 may include one or more elements selected from a group including the information processing device 2 and the user terminal 3. The elements not selected may not be included in the information processing system 1, but may be electrically connected to the selected elements as external elements. These components will be described below.

[0022] <Information processing device 2> 1, the information processing device 2 includes a communication unit 21, a storage unit 22, and a control unit 23, and these components are electrically connected via a communication bus 20 inside the information processing device 2. Each component will be further described.

[0023] <Communications Department 21> The communication unit 21 is preferably a wired communication means such as USB, IEEE1394, Thunderbolt (registered trademark), wired LAN network communication, etc., or a wireless communication means such as wireless LAN network communication, mobile communication such as 3G / LTE / 5G, Bluetooth (registered trademark), etc. The information processing device 2 transmits and receives various information to and from external devices via the communication unit 21 and the network.

[0024] <Storage section 22> The storage unit 22 stores various pieces of information defined above. The storage unit 22 is a storage device such as a solid state drive (SSD) and stores various pieces of information such as data and programs related to the information processing device 2 executed by the control unit 23. Alternatively, the storage unit 22 is a memory such as a random access memory (RAM) and stores temporarily required information (arguments, arrays, etc.) related to program calculations. Note that the various pieces of information may be configured to be stored in a distributed manner across multiple external devices using blockchain technology or the like.

[0025] <Control unit 23> The control unit 23 is a processor that processes and controls the overall operations related to the information processing device 2. The control unit 23 is configured to include, for example, a central processing unit (CPU) and a graphics processing unit (GPU). The control unit 23 realizes various functions related to the information processing device 2 by reading out predetermined programs stored in the storage unit 22.

[0026] For example, the control unit 23 reads various pieces of information stored in a storage area that is at least a part of the memory unit 22, and writes the read information into a memory that is at least a part of the memory unit 22, thereby executing various information processes based on the acquired information. That is, information processing by software stored in the memory unit 22 is specifically realized by the control unit 23, which is an example of hardware, and can be executed as each functional unit included in the control unit 23. The control unit 23 is not limited to being single, and the configuration may include multiple control units 23 for each function.

[0027] The information processing device 2 is a computer such as a server device or a workstation, and may be in an on-premise form or a cloud form. As a cloud-based information processing device 2, the above-mentioned functions and processes may be provided in the form of, for example, SaaS (Software as a Service) or cloud computing.

[0028] <User terminal 3> 1, each user terminal 3 includes a communication unit 31, a memory unit 32, a control unit 33, an output interface (hereinafter, I / F) 34, and an input I / F 35, and these components are electrically connected via a communication bus 30 inside the user terminal 3. The explanations of the communication unit 31, the memory unit 32, and the control unit 33 are the same as the explanations of each unit in the information processing device 2, so redundant explanations will not be repeated.

[0029] <Output I / F34> The output I / F 34 is configured to output various information so that the user can recognize it through his or her five senses. The output I / F 34 is configured to output various stimuli generated by the control unit 33 so that the user can perceive them, and for example, the various stimuli are output as images, sounds, tactile stimuli, smells, tastes, etc. The output I / F 34 is configured to output the various stimuli individually or in combination. The output I / F 34 may be included in the housing of the user terminal 3 or may be externally attached.

[0030] The output I / F 34 is, for example, any display device capable of displaying images, such as a CRT display, a liquid crystal display, an organic EL display, a plasma display, or a projector. The image displayed on the output I / F 34 may be generated based on image information stored in the storage unit 32 or image information generated outside the information processing device 2 or the like and received via the communication unit 31. The image displayed on the output I / F 34 includes a graphical user interface (GUI) that can be operated by the user.

[0031] The output I / F 34 is, for example, an audio output device such as a speaker, headphones, bone conduction earphones, hearing aids, or cochlear implants that can output audio. The output I / F 34 is, for example, a tactile output device that provides contact stimuli such as vibrations or temperature changes, or non-contact stimuli such as air spray. The output I / F 34 is, for example, an olfactory output device that provides a change in smell by spraying aroma or the like into the air. The output I / F 34 is, for example, a taste output device that induces a change in taste in the user by electrical stimuli or the like.

[0032] The input I / F 35 is configured to accept operation inputs from the user. The operation inputs are transferred as command signals to the communication unit 31 via the communication bus 30, and the control unit 33 executes predetermined control and calculations. The input I / F 35 may be included in the housing of the user terminal 3 or may be attached externally. For example, the input I / F 35 may be configured as a touch panel integrated with the output I / F 34, and accept operation inputs such as tapping and swiping based on coordinate information of a touch operation performed by the user. The input I / F 35 may be a mouse, a QWERTY keyboard, a touchpad, a switch button, a game controller, or the like.

[0033] The input I / F 35 may also accept user operation inputs by employing a voice recognition device that acquires voice, a gesture detection device that detects the posture or movement of an object, a gaze detection device that detects the user's gaze direction, a biosignal detection device that detects various biosignals of the user, a position information detection device that acquires position information such as latitude and longitude information, etc. The input I / F 35 may also include a camera that can acquire images (still images, videos, etc.). The camera may be provided in the user terminal 3, or may be configured to acquire images captured by a separately provided camera.

[0034] 2. Functional configuration In this section, the functional configuration of this embodiment will be described. As described above, the information processing system 1 includes at least one processor, and the processor is configured to include the following units by reading a program. In other words, the program causes a computer to function as each unit of the information processing system 1.

[0035] 2 is a block diagram showing functions realized by the control unit 23 and the like in the information processing device 2. The information processing device 2, which is an example of the information processing system 1, includes a progress control unit 231, a presentation unit 232, an acquisition unit 234, an identification unit 235, a recording unit 236, and an artificial intelligence unit 237.

[0036] The progress control unit 231 is configured to be able to execute a progress control step. In the progress control step, the progress control unit 231 advances the interview with the answering user.

[0037] The presenting unit 232 is configured to be able to execute a presenting step. In the presenting step, the presenting unit 232 controls processing for causing the output I / F 34 of the user terminal 3 to present various types of information. In particular, the presenting unit 232 presents various types of information via the first user terminal 3A so that the administrative user can recognize it. Furthermore, the presenting unit 232 presents various types of information via the second user terminal 3B so that the answering user can recognize it. For example, as the first presenting step, the presenting unit 232 presents a first question in an interview to the answering user. Furthermore, as the second presenting step, the presenting unit 232 presents a second question in an interview to the answering user.

[0038] The presentation unit 232 may draw various screens to be displayed on the display device of the user terminal 3 and transmit them to the user terminal 3, or may transmit display information including visual information such as images including still images or moving images, icons, and messages to the user terminal 3 so that various screens are drawn on the user terminal 3. The control unit 33 of the user terminal 3 draws the screen to be displayed on the display device based on the received display information and information such as parameters defining the various visual information stored in the storage unit 32. That is, at least some of the functions of the presentation unit 232 may be executed by the control unit 33 of the user terminal 3.

[0039] Furthermore, the presentation unit 232 may present and notify various types of information so that the user can perceive it via the five senses. For example, the user terminal 3 may output various types of information via the output I / F 34 so that the user can perceive it via the five senses. The output I / F 34 outputs various types of stimuli so that the user can perceive them based on instructions from the presentation unit 232. For example, the various types of stimuli may be output as sounds, tactile stimuli, smells, tastes, etc. The output I / F 34 may be configured to output the various types of stimuli alone or in combination, and these may be output together with images displayed on a display device. The output I / F 34 may be included in the housing of the user terminal 3 or may be externally attached.

[0040] The acquisition unit 234 is configured to be able to execute an acquisition step. In the acquisition step, the acquisition unit 234 executes processing to acquire various pieces of information from the memory unit 22 or an external device. The acquisition unit 234 is configured to be able to acquire various pieces of information by reading out various pieces of information stored in a storage area that is at least a part of the memory unit 22 and writing the read out information into a working area that is at least a part of the memory unit 22.

[0041] The identification unit 235 is configured to be able to execute an identification step. In the identification step, the identification unit 235 executes processing for identifying various pieces of information used in information processing, based on the information acquired by the acquisition unit 234. The identification unit 235 may input predetermined information into the reference information, thereby outputting various pieces of information used in information processing.

[0042] The recording unit 236 is configured to be able to execute a recording step. In the recording step, the recording unit 236 records various information in a predetermined storage area. The predetermined storage area may be at least one of the storage unit 22, the storage unit 32, and a storage area provided in a predetermined external device.

[0043] The artificial intelligence unit 237 is configured to receive input from each functional unit and return the instructed output. The artificial intelligence unit 237 is configured to be able to access the artificial intelligence module. The artificial intelligence unit 237 receives a predetermined input and instructs the artificial intelligence module to return the instructed output. The artificial intelligence module used by the artificial intelligence unit 237 may be stored in at least one of the memory unit 22 and the memory unit 32, or may be stored in an external artificial intelligence server different from the information processing device 2 and the user terminal 3. The artificial intelligence unit 237 inputs predetermined input information to the artificial intelligence module and causes the artificial intelligence module to output predetermined output information. The artificial intelligence module used by the information processing device 2 may be a common one, or may be prepared individually for each functional unit.

[0044] The AI ​​module used by the AI ​​unit 237 includes reference information configured to return a predetermined output in response to an input. The reference information is configured to identify the correspondence between input information and output information, and may include, for example, at least information regarding the correlation between the two. The reference information may include a table, a function, a simple algorithm, etc.

[0045] The artificial intelligence module is an AI (Artificial Intelligence) equipped with a learning model such as a language model, such as a Transformer including a Generative Pretrained Transformer (GPT) and a Bidirectional Encoder Representations from Transformers (BERT), or a Recurrent Neural Network (RNN), and may be a generative AI or an AI agent.

[0046] In the reference information, a correlation between input information and output information can be constructed by statistically analyzing data recording input information and corresponding output information. The reference information may be a trained model, a language model, or the like that has been trained to be able to output output information in response to input information. In these, parameters calculated, tuned, or the like through training constitute the correlation of the reference information. Furthermore, when the reference information is configured to probabilistically calculate a predetermined output for an input, the probability constitutes the correlation between the input and the output. In other words, the correlation need only be configured to generate an output having a predetermined correspondence with the input, and does not have to be a correlation in the strict sense. In the following description, redundant explanations regarding the method of constructing reference information will not be repeated.

[0047] A trained model can be configured to perform a specific task by a learning method such as supervised learning, unsupervised learning, or self-supervised learning. In supervised learning, machine learning is performed using training data (training data). The training data includes pairs of input data and output data (correct answer data) for learning.

[0048] A language model is an example of a learning model based on a machine learning algorithm. Specific examples of machine learning algorithms include nearest neighbor algorithms, naive Bayes algorithms, decision trees, support vector machines, and deep learning using neural networks. The artificial intelligence module can apply the above algorithms as appropriate. Furthermore, a language model may not only be one trained for a specific task, but also a general-purpose model that can be used for a wide range of tasks (for example, large language models (LLMs) trained on massive amounts of data).

[0049] Generative AI performs various tasks using general-purpose models, including language models. Specifically, generative AI is configured to process a wide range of tasks, such as understanding sentence patterns and context, answering questions, and generating text, images, and audio, in response to given input (prompts), and return the specified output. Prompts may be provided in any format, including text, images, and audio. Tasks may also be processed interactively, alternating between receiving input and generating and outputting information.

[0050] General-purpose models for generative AI are not limited to language models. General-purpose models may include, for example, image generation models (e.g., diffusion models, generative adversarial networks), speech models, video generation models, and multimodal models. Image generation models generate new images from portions of input text or images. Speech models realize speech synthesis (text-to-speech conversion) and speech conversion (e.g., speaker conversion). Video generation models generate continuous image and audio information to generate video content. Multimodal models enable cross-modal inference and generation by processing and integrating different types of data, such as text, images, audio, and sensor data, in the same model. In this way, generative AI does not rely on a single modality but can perform diverse and advanced tasks by integrating multiple modalities.

[0051] AI agents perform various tasks using artificial intelligence modules, including generative AI. Specifically, when a specific instruction is input, an AI agent is configured to break down the processing required to accomplish the specified task into subtasks, actions, etc., and to collect and analyze necessary data, generate and execute programs, etc. An AI agent may autonomously learn to achieve its goals by autonomously creating and executing plans and evaluating the results of its execution. Instructions given to an AI agent consist of abstract content such as goals and objectives and do not necessarily include specific actions.

[0052] The artificial intelligence modules may be stored in multiple artificial intelligence servers. The artificial intelligence servers may provide services using the learning models, or may execute various tasks using the learning models. The artificial intelligence servers may be configured to be accessible to the generation AI, accept prompt inputs such as text, images, and voice, and generate and respond to the prompts.

[0053] 3. Information Processing This section describes each step of the information processing method executed by the information processing system 1. The information processing method is executed by a processor included in at least one of the user terminal 3 and the information processing device 2 reading out a program.

[0054] 3.1 Overview of Information Processing FIG. 3 is a flowchart showing an outline of information processing of this embodiment. As shown in FIG. 3, the progress control unit 231 controls the progress of the interview with the first respondent. The presentation unit 232 (first presentation unit) causes the first respondent to present a first question in the interview as a first presentation step. The first question is generated based on at least a progress condition preset by the management entity. The acquisition unit 234 acquires first evaluation information for the evaluation target. The evaluation target is an object to be evaluated by the first respondent in the interview. The first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference entity different from the first respondent. The presentation unit 232 (second presentation unit) causes the first respondent to present a second question generated based on at least the first evaluation information in the interview as a second presentation step. According to this aspect, the effectiveness of the interview can be improved.

[0055] 3.2 Details of information processing 4 is an activity diagram showing the flow of information processing in this embodiment. As shown in FIG. 4, first, the first user terminal 3A accepts settings regarding the interview progress conditions made by the administrative user. In the information processing device 2, when the acquisition unit 234 acquires information regarding the progress conditions (action A101), the progress control unit 231 generates an interview room based on the set progress conditions (action A102).

[0056] In this embodiment, the progress control unit 231 individually conducts interviews based on the same progress conditions with multiple answering users, including the first to third answering users. That is, the progress control unit 231 generates multiple interview rooms associated with the same progress conditions. Then, a request to participate in an interview associated with the interview room from each answering user is accepted at the second user terminal 3B. When the information processing device 2 receives the participation request, the progress control unit 231 associates the answering user with the interview room and starts the interview. Note that the progress control unit 231 may conduct interviews with multiple answering users in parallel so that at least a portion of the time periods overlap, or may conduct the interviews so that the time periods do not overlap.

[0057] In other words, the progress control unit 231 controls the progress of the interview with the first respondent. Also, the progress control unit 231 progresses the interview with the second respondent, which is different from the first respondent, separately from the interview with the first respondent. Furthermore, the progress control unit 231 progresses the interview with the third respondent, which is different from the first and second respondent, separately from the interviews with the first and second respondent.

[0058] As will be described later, the presenting unit 232 presents a question (second question) to the answering user based on answers, etc. (rating information, which will be described later) from other answering users. In this embodiment, a question is presented to the first answering user based on answers, etc. from one or more second answering users. Also, a question is presented to one or more third answering users based on answers, etc. from the first answering user. In the following description, when the first to third answering users are not to be distinguished, they may be collectively referred to as answering users.

[0059] The presenting unit 232 presents to the answering user a question text related to the question (first question) generated based on the progress conditions. The question text is generated by inputting the progress conditions into the generation AI. The next question (first question) may also be generated based on the content of the answering user's answer to the question. In this case, the question text related to the next question is generated by inputting information indicating the progress conditions and the content of the answer into the generation AI. This provides the answering user with a highly interactive interview, improving the quality of the dialogue.

[0060] The progress condition includes at least objective information indicating information desired to be acquired from the first respondent. The objective information includes at least the information desired to be acquired from the respondent user, such as the information described above, such as "to clarify what kind of people are refraining from purchasing rice due to the price increase of rice, and why." That is, it is preferable that the objective information be described so as to identify at least one of the respondent user's subject (what kind of person), target (rice), behavior (whether to refrain from purchasing rice), and background (the price increase of rice). Note that the information desired to be acquired from the respondent user does not need to be specified in the objective information. In this example, the "subject" is described as one piece of information desired to be acquired from the respondent user, but a specific persona is not specified. Therefore, the specific persona of the respondent user also constitutes part of the desired information.

[0061] The question (first question) may be generated without being input to the generation AI. For example, the progress conditions may specify the question sentence and the order in which multiple questions are presented, and the presentation unit 232 may present the question sentence specified in the progress conditions to the answering user as is.

[0062] Furthermore, it is preferable that the progress conditions include at least one piece of question information. The question information is information describing the content of a question (first question) to be presented to the answering user. The question information may include multiple questions, and for each question, the content of the question, how to proceed with the question, and supplementary information related to the question are set. The content of the question may include the content of a common question and a follow-up question. In the interview, a common question is presented to all answering users, and follow-up questions are presented as necessary depending on the answers of the answering users.

[0063] The progress conditions may include information indicating how the interview will proceed, such as the interview flow, question format, response format, and in-depth perspective.

[0064] An interview flow is information that defines the relationships between common questions. For example, the interview flow defines the order in which common questions are presented to the respondent user, and the appropriate follow-up questions are presented before moving on to the next common question, so that the interview goal is achieved. The interview flow may also define the timing and conditions for presenting a second question (described later) to the respondent user.

[0065] The question format is information that defines the format of the question, including rules for paragraph division and font settings (such as making important parts bold). The response format is information that defines the format of the response when responding to the answer of the answering user, including expressing gratitude for the answer. In the following, when there is no particular distinction between question and answer, the term "question" may be used to refer to them collectively. Furthermore, a question may be composed of a response to the previous answer and a question related to the next question.

[0066] Probing perspectives are information that stipulate the rules by which the generation AI autonomously generates probing questions. When probing perspectives are input into the generation AI, questions are output to explore the reasons, specific examples, background, etc. for important information contained in the answers, points of doubt, and points that should be further explored. Exception handling is information for proceeding with the interview exceptionally regardless of the interview flow.

[0067] FIG. 5 is a diagram showing an example of a setting screen 500. The setting screen 500 accepts input of interview progress conditions by an administrative user. The setting screen 500 is displayed on the output I / F 34 of the first user terminal 3A before an interview room is generated. The setting screen 500 includes a goal input area 510, a question input area 520, a question input area 530, and a button 540. Predetermined information is input into the goal input area 510, the question input area 520, and the question input area 530 in response to operations by the administrative user via the input I / F 35. By operating the button 540 after each item has been input, the input information is sent to the information processing device 2 as a progress condition. The button is an example of a GUI that is displayed on a display device and accepts operational input by a user.

[0068] The purpose input area 510 is an area for accepting input of purpose information. The question input areas 520 and 530 are areas for accepting input of question information for each question number. The question input areas 520 and 530 may accept input of common questions, supplementary information, and follow-up questions (specific question sentences or perspectives for follow-up). Questions (first questions) are classified by question number designated by the administrative user, and the question groups included in each section include common questions and follow-up questions. The common questions are the questions presented to the answering user first in each question group. The follow-up questions are additional questions presented after the common questions and in relation to the common questions. The question group included in the question input area 520 constitutes the first section, and the question group included in the question input area 530 constitutes the second section. The questions included in the first and second sections can be distinguished as constituting independent topics.

[0069] The setting screen 500 may further include areas for accepting the interview flow, question format, response format, number of questions, perspective of in-depth investigation, exception handling, etc., and these may be set for each question input area. Furthermore, each item may be set commonly for all categories, or may be set for each category in each question input area.

[0070] Returning to FIG. 4, the progress control unit 231 controls the progress of the interview with the first respondent. First, a request from the answering user to participate in an interview associated with the interview room is accepted at the second user terminal 3B (action A103). When the information processing device 2 receives the participation request, the progress control unit 231 associates the answering user with the interview room and starts the interview. Note that, in response to accepting the participation request from the answering user, the progress control unit 231 may generate an interview room associated with the answering user and start the interview. The participation request may be generated, for example, by the answering user inputting a predetermined URL (Uniform Resource Locator) via the input I / F 35.

[0071] In an interview conducted in each interview room, the presentation unit 232 presents a question (first question) to each answering user. The question is presented to the output I / F 34 of the second user terminal 3B. The second user terminal 3B accepts an answer to the question presented by the answering user and transmits the accepted answer to the information processing device 2.

[0072] That is, in an interview with the first answering user, the presenting unit 232 causes the first answering user to present a first question (action A105), and the acquiring unit 234 acquires the first answering user's answer to the first question (action A106). Also, in an interview with the second answering user, the presenting unit 232 causes the second answering user to present a first question (action A103), and the acquiring unit 234 acquires the second answering user's answer to the first question (action A104). Furthermore, in an interview with the third answering user, the presenting unit 232 causes the third answering user to present a first question (action A107), and the acquiring unit 234 acquires the third answering user's answer to the first question (action A108).

[0073] The artificial intelligence unit 237 inputs the progress conditions into the generation AI, thereby causing it to output a question sentence to be presented to the answering user. The artificial intelligence unit 237 first inputs objective information included in the progress conditions into the generation AI, thereby causing the generation AI to store the objective of the interview. The artificial intelligence unit 237 also inputs the setting items common to each question described above into the generation AI, thereby causing the generation AI to store the basic procedure for the interview. Next, the artificial intelligence unit 237 inputs question information related to the first question as a prompt into the generation AI, thereby causing the generation AI to output a question sentence related to the first question. The presentation unit 232 transmits the output question sentence to the second user terminal 3B, causing it to be presented in a recognizable manner to the answering user. Note that the objective information and the setting items common to each question may be input to the generation AI together with the question information. In other words, the presentation unit 232 (first presentation unit) presents a first question to the first answering entity in the interview as a first presentation step. The first question is generated based on at least the progress conditions preset by the management entity. The first question is generated by inputting at least a part of the progress conditions into the generation AI.

[0074] Furthermore, the artificial intelligence unit 237 may input the progress conditions and the content of the answer acquired from the answering user to the generation AI, thereby causing it to output a question sentence. For example, the artificial intelligence unit 237 may input the content of the answer to the previous question along with question information to the generation AI, thereby causing it to output a question sentence regarding the next question. Furthermore, the artificial intelligence unit 237 may input only the content of the answer to the generation AI, thereby causing it to output a question sentence regarding the next question, taking into account the progress conditions and answer history input up to that point. In other words, the presentation unit 232 (first presentation unit) causes the generation AI to present to the first answering subject a next first question, which is generated based on the progress conditions and the content of the answer to the first question by the first answering subject, as a first presentation step.

[0075] The presenting unit 232 may present the question specified in the question information included in the progress conditions to the answering user as is. That is, the first question may be generated without being input to the generation AI.

[0076] FIG. 6 is a diagram showing an example of an interview screen 600. The interview screen 600 is output to, for example, the output I / F 34 of the second user terminal 3B associated with the first answering user. The interview screen 600 includes an opening sentence 610, an answer sentence 611, question sentences 620, 621, and 622, an answer input field 630, and a button 640. The question sentences 620, 621, and 622 are question sentences of a question (first question) presented by the presenting unit 232. The opening sentence 610 and the answer sentence 611 indicate the content input by the first answering user. The content of the answer by the first answering user is input into the answer input field 630 by the first answering user operating the input I / F 35. Then, by operating the button 640, the content of the answer is transmitted to the information processing device 2 and output to the interview screen 600.

[0077] On second user terminal 3B, an interview room in which the first answering user can participate is output prior to interview screen 600. Then, interview screen 600 relating to the selected interview room is output to output I / F 34. Note that interview screen 600 may also be output by inputting the aforementioned predetermined URL via input I / F 35. Start statement 610 is a command statement instructing the first answering user to start the interview. When start statement 610 is input, artificial intelligence unit 237 inputs the progress conditions and start statement 610 to the generation AI, thereby starting the interview.

[0078] When the interview begins, the presenting unit 232 causes a question sentence 620 regarding the first question, question 1 (first question), to be presented on the interview screen 600. Question 1 includes content inquiring about the family composition of the first answering user. When the acquiring unit 234 acquires an answer sentence 611 as an answer to the question sentence 620, the artificial intelligence unit 237 causes the generating AI to input the answer sentence 611, thereby outputting a question sentence 621 regarding the next question (first question). The presenting unit 232 causes the output question sentence 621 to be presented on the second user terminal 3B.

[0079] The question 621 includes the phrase "Thank you for your hard work raising your children" as a response. The response is generated based on the content of the answer and the progress conditions, since the answer 611 includes the phrase "My children attend elementary school and kindergarten."

[0080] Furthermore, question 621 includes the phrase, "Could you also tell me the gender of your child?" This question is a follow-up question (first question) related to question 1 included in question 620. For example, the progress condition may include supplemental information such as, "Ask follow-up questions until the ages and genders of the family members are determined," as question information related to question 1. In this case, question 621 is generated in response to the fact that "the gender of the child is unknown" in answer 611. The content of the follow-up question in question 621 may be set as question information, or may be autonomously generated by a generation AI based on the supplemental information.

[0081] Question 622 is a question related to question 2 (first question) and includes the phrase, "How many kilograms of rice does your whole family eat in a month?". An answer to question 622 is entered in answer input field 630. Question 2 is a question included in a group of questions that constitutes a different category from question 1, and is generated when the purpose of question 1 is achieved. In other words, question 622 can be generated when the answer to question 1 reveals "the ages and genders of the first answerer's family members."

[0082] 4, the explanation will be continued. In this embodiment, the presenting unit 232 presents to the answering user a question (second question) that is generated based on the content of the answer to the first question by another answering user. This aspect can improve the effectiveness of the interview.

[0083] In the following description, information used to generate a second question to be presented to each answering user may be referred to as first evaluation information. Furthermore, information of the content of an answer by each answering user that is used to generate a second question to be presented to other answering users may be referred to as second evaluation information. That is, in an interview with the first answering user, the content of the answer by the second answering user is an example of first evaluation information. Furthermore, the content of the answer by the first answering user is an example of second evaluation information. Note that, in an interview with the third answering user, the content of the answers by the first answering user and the second answering user are an example of first evaluation information. Therefore, the second question to be presented to the third answering user may be generated based on at least one of the answers by the first answering user and the second answering user. Note that, when there is no particular distinction between the first evaluation information and the second evaluation information, they may be collectively referred to as evaluation information.

[0084] The content of the answer used as the evaluation information may be in any format and may include text, audio, images, videos, sensor data, etc., or may be a combination of multiple formats among these. For example, the evaluation information from the answering user may be text or audio entered as an answer by the answering user in the interview, images or videos captured by a camera equipped in the user terminal 3B, or a screenshot of the screen displayed on the user terminal 3B.

[0085] In other words, the presenting unit 232 (second presenting unit) presents, as a second presenting step, a second question generated based on at least the first evaluation information to the first answering subject in the interview.

[0086] First, the acquiring unit 234 acquires an answer from the second answering user in an interview in which the second answering user participates. The answer from the second answering user is an answer to a first question generated based on the progress conditions. Furthermore, the identifying unit 235 identifies, from the answer from the second answering subject, first evaluation information to be used for generating a second question to be presented to the first answering subject.

[0087] The identification unit 235 identifies, from among the answers by the second answering user, information describing the results of evaluating the evaluation target in the interview as first evaluation information. The recording unit 236 records the identified first evaluation information in a predetermined storage area (action A109).

[0088] The identification unit 235 may identify the first evaluation information based on the progress conditions. For example, the identification unit 235 identifies information to be obtained from the respondent user in the interview as the evaluation target. In the above example, the evaluation targets include, for example, the object (rice), the content of the behavior toward the object (whether to refrain from purchasing), and the background (price increase of rice). The first evaluation information includes information describing the results of the second respondent user's evaluation of at least any of these. For example, evaluation information may include whether or not the user likes rice, the preferred brand of rice, the frequency of eating rice, the amount of rice consumed, changes in these, and opinions regarding the price increase of rice. Furthermore, impressions of the rice provider, opinions regarding the rice brand and the introduction of the rice provider (e.g., catchphrase), etc. may also be included in the first evaluation information. In other words, in addition to the product, etc., which is the direct evaluation target, related products, etc., and the provider, etc., who provides them, may be evaluation targets evaluated by the evaluation information. Furthermore, the evaluation information may include any information evaluated as being relevant to the purpose of the interview. The range of the relevance may be defined in the progress conditions.

[0089] In this way, the acquisition unit 234 acquires first evaluation information (such as an answer by the second answering user) for the evaluation target. The first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference subject different from the first answering subject. The evaluation target is an object that is evaluated by the first answering subject in an interview. Typically, the reference subject is the second answering subject. The first evaluation information may be acquired after the interview with the second answering subject has ended, or may be acquired in real time in parallel with the progress of the interview.

[0090] Furthermore, the identification unit 235 may input the acquired answer of the second answering user to the generation AI to identify the first evaluation information. The artificial intelligence unit 237 inputs the acquired answer of the second answering user and the progress conditions to the generation AI to output the first evaluation information that is relevant to the purpose of the interview. Note that the acquisition unit 234 may acquire information indicating the answer of the second answering user selected by the admin user to be used as the first evaluation information. The identification unit 235 may acquire this information and identify the information indicating the answer of the second answering user to be used as the first evaluation information based on the selection by the admin user.

[0091] Furthermore, when a plurality of pieces of first evaluation information are identified (for example, when a plurality of pieces of first evaluation information are identified from answers by a plurality of second answering users), the identification unit 235 may identify the first evaluation information to be used for generating the second question from among them. The artificial intelligence unit 237 inputs the plurality of pieces of first evaluation information to the generation AI, thereby outputting the first evaluation information to be used for generating the second question.

[0092] Furthermore, the identification unit 235 may identify the first evaluation information based on a history of answers by the first answering user to previous questions (first questions). For example, if the attributes (age, gender, family structure, etc.) of the first answering user can be identified from the answer history, the identification unit 235 may identify first evaluation information by a second answering user with similar attributes. Furthermore, the identification unit 235 may identify first evaluation information that is closely related to a question (first question) in which the first answering user is particularly interested. A question of interest is, for example, a question that has received an answer containing specific content or content that is highly relevant to the purpose of the interview. The artificial intelligence unit 237 may input multiple pieces of first evaluation information and the answer history of the first answering user to a generation AI, thereby outputting first evaluation information used to generate a second question.

[0093] The artificial intelligence unit 237 may input at least a part of the progress conditions (for example, objective information) to the generation AI together with at least one of the answer of the second answering user, the answer of the first answering user, and the first evaluation information. On the other hand, if the progress conditions are stored in the generation AI, the artificial intelligence unit 237 may not need to input at least a part of the progress conditions to the generation AI again.

[0094] In other words, the first evaluation information used to generate the second question is identified from the plurality of pieces of first evaluation information acquired by inputting at least one of at least a part of the progress conditions and the answer to the first question by the first answering subject into the generation AI. According to this aspect, the answers of other answering subjects to be utilized can be effectively identified.

[0095] Then, the presenting unit 232 generates a question (second question) based on the identified first evaluation information and progress conditions, and causes the second user terminal 3B of the first answering user to present the question (second question) (action A105). For example, the artificial intelligence unit 237 inputs the first evaluation information into the generation AI, thereby outputting a question to be presented to the answering user. The second question can be generated based on the input first evaluation information, the progress conditions stored in the generation AI, and the content of the answer by the first answering user.

[0096] The second question may include the content of the answer by the second answering user as is. In this case, the content of the answer by the second answering user may be presented on the second user terminal 3B of the first answering user without being input to the generation AI. On the other hand, the content of the answer by the second answering user may be edited as appropriate. For example, the content of the answer may be abstracted and integrated with the content of answers by other second answering users, and part of the content of the answer may be deleted or changed. For example, if the answer contains information that could identify the personal information of the second answering user, that part may be deleted or changed. Furthermore, only the expression, such as the writing style, may be changed.

[0097] 7 is a diagram showing another example of an interview screen 700. The interview screen 700 is output to the output I / F 34 of the second user terminal 3B of the first answering user. The interview screen 700 shows an example of a second question generated based on at least the first evaluation information by the second answering user. The interview screen 700 shows a state after the dialogue on the interview screen 600 has continued.

[0098] Interview screen 700 includes answer sentences 710 and 711, question sentences 720 to 722, an answer input field 730, and a button 740. Interview screen 700 shows an interview in the same interview room as interview screen 600 shown in Fig. 6, and shows a dialogue continuing from question sentence 622 in interview screen 600. Therefore, question sentence 622 is common to question sentence 720.

[0099] Question sentence 720 is a first question generated by the generation AI based on the progress conditions. Answer sentence 710 is an answer from the first answering user to question sentence 720. Question sentence 721 is a first question generated by the generation AI based on the progress conditions and answer sentence 710. Answer sentence 711 is an answer from the first answering user to question sentence 721. Question sentence 722 is a second question generated by the generation AI based on the first evaluation information. The content of the answer by the first answering user is input into answer input field 730 by the first answering user operating input I / F 35. Then, by operating button 740, the content of the answer is sent to the information processing device 2 and output to the interview screen 700.

[0100] The artificial intelligence unit 237 inputs the content of the answer by the second answering user to the generation AI, causing it to output question sentence 722 as the second question. The artificial intelligence unit 237 may input the content of the answer by the second answering user to the generation AI, causing it to output the first evaluation information, and input the outputted first evaluation information to the generation AI, causing it to output question sentence 722. In this way, question sentence 722 can be generated based on the first evaluation information, the progress conditions, and the content of the answer by the first answering user.

[0101] The presenting unit 232 presents the question sentence 722 as the second question in the interview room where the question sentences 720 and 721 as the first questions are presented. That is, the first presenting step of presenting the first question and the second presenting step of presenting the second question are performed in the same interview. This can improve the effectiveness of the interview conducted by the generation AI.

[0102] For example, in interview screen 700, answer 711 to question 721 does not include content that meets the conditions for progression ("Nothing in particular."), so question 722 is presented. Question 722 includes the answer of the second answering user to the question (question 3) corresponding to question 721. This allows the first answering user to respond with their own opinion, referring to the answer of the second answering user.

[0103] Furthermore, the content of question 722, "Who has a family structure similar to yours?", indicates that the attribute information of the first answering user and the second answering user is similar. Attribute information may include the answering user's address, age, gender, occupation, financial resources, family structure, nationality, language spoken, etc. The attribute information may include any information that can identify the answering user's attributes, and may also include online and offline behavioral history (such as product purchase history and browsing history of content such as web pages). The information is generated when the answers of the first answering user and the second answering user are input into the generation AI, and it is determined that the attribute information of both users is similar.

[0104] Furthermore, the content of question 722, "They said that brand rice called XX is inexpensive and has a good texture," is information corresponding to the first evaluation information input by the second answering user. The content, "What do you think about that?" is information for asking the first answering user for his / her opinion on the first evaluation information. In other words, as the second presentation step, presentation unit 232 (second presentation unit) presents a second question to the first answering subject to ask for the first answering subject's opinion on the first evaluation information. This increases the first answering subject's interest in the first evaluation information and encourages the first answering subject to answer the second question.

[0105] The structure of such question 722 may be specified in the progress conditions (e.g., question information). Furthermore, question 722 may be written with its writing style and abstraction adjusted based on the content of the answer by the first answering user. For example, if the history of answers up to that point indicates that the first answering user prefers casual conversation, question 722 in a casual style such as "What do you think about that?" may be generated.

[0106] Furthermore, question sentence 722, which is the second question, may be presented to the first answering user in real time in response to the acquisition of the first evaluation information, or may be presented at a predetermined timing. For example, question sentence 722 may be presented after an answer to a round-off first question is acquired and before the next first question is presented to the answering user. For example, if first evaluation information is acquired while one of the first questions belonging to a group of questions that are divided to form individual topics is being presented to the answering user, the second question may be presented to the answering user after all questions belonging to the group have been answered.

[0107] Specifically, assume that, on the setting screen 500 shown in FIG. 5 , a base question belonging to Question 1 and a predetermined number of follow-up questions are accepted as a question group of the first question belonging to the first section, and a base question belonging to Question 2 and a predetermined number of follow-up questions are accepted as a question group of the first question belonging to the second section. In this case, the presenting unit 232 presents a question belonging to the second section after all questions belonging to the first section have been completed, based on the set progress condition. Here, if the acquiring unit 234 acquires first evaluation information while at least one of the questions belonging to the first section is being presented to the first answering user (while all answers have not been accepted), the presenting unit 232 presents a second question based on the acquired first evaluation information before presenting the first question belonging to the second section. In other words, the presenting unit 232 (second presenting unit) executes the second presenting step in response to the progress condition, before the next first question by the presenting unit 232 (first presenting unit) is presented to the first answering subject. This allows the second question to be presented without straying from the conditions of the interview.

[0108] Furthermore, regardless of the progress condition, question sentence 722, which is the second question, may be presented to the first answering user before the next first question is presented to the first answering user if the content of the answer by the first answering user satisfies a predetermined condition. The predetermined condition may be that the answer by the first answering user to the first question does not conform to the progress condition (objective information). The predetermined condition may include, for example, that the content of the answer is unclear, that the content is little related to the question, or that a no-answer state in which no answer is entered continues for a certain period of time. Specifically, if the content of the answer is little related to the question, such as "nothing in particular" or "I don't know," the presenting unit 232 may present the second question. In this way, the second question, which has a different perspective from the first question, is presented to the first answering user, thereby increasing the first answering user's motivation to answer. In other words, the presentation unit 232 (second presentation unit) executes the second presentation step in response to the content of the answer to the first question by the first answering subject, before the next first question is presented to the first answering subject by the first presentation step.

[0109] Returning to FIG. 4, the explanation will be continued. The progress control unit 231 progresses the interview with the first answering user, and the acquisition unit 234 acquires the first answering user's answer to at least one of the first question and the second question. Furthermore, the identification unit 235 identifies, from the answers by the first answering user, second evaluation information to be used for generating a second question to be presented to the third answering user. That is, the identification unit 235 identifies, from the answers by the first answering user, information describing the results of evaluating the evaluation target in the interview as the second evaluation information. The recording unit 236 records the identified second evaluation information in a predetermined storage area (action A110). In this way, the acquisition unit 234 acquires the second evaluation information for the evaluation target. Note that the second evaluation information may be acquired after the interview with the first answering user is completed, or may be acquired in parallel with the progress of the interview.

[0110] For example, as shown in the interview screen 700 in FIG. 7, suppose that a first respondent user has entered the content entered in the answer input field 730 as an answer to a question 722. The answer includes information describing the results of an evaluation of the evaluation target (brand rice) and can be identified as second evaluation information. The answer is also an answer to the question 722 that includes the first evaluation information. In other words, the second evaluation information includes information describing the results of the first respondent's evaluation of the evaluation target based on the first evaluation information.

[0111] Here, the recording unit 236 may record the second evaluation information in association with the first evaluation information. For example, the recording unit 236 may associate the first evaluation information and the second evaluation information with information indicating a relevance to the progress conditions of the interview, and associate the two based on the information. This allows the identification unit 235 to easily identify that the two pieces of evaluation information relate to a common evaluation target. Furthermore, the identification unit 235 may easily identify evaluation information suitable for use in generating the second question.

[0112] 8 is a diagram showing an example of an evaluation information table T that stores evaluation information. The evaluation information table T stores answers identified as evaluation information from among answers by each answering user. The evaluation information table T stores answer content, answering user, attribute information, evaluation target, question information, and suitability in association with an interview ID and an answer ID.

[0113] The interview ID is information for identifying which progress condition the interview was conducted under. The answer ID is information for uniquely identifying the answer by each answering user. The answer content indicates the content entered by each answering user. The answering user indicates the answering user who entered the answer. The attribute information is information describing the characteristics of the answering user and may include any information that can identify the attributes of the answering user described above. The attribute information may include, for example, the user's address, age, gender, occupation, financial resources, family composition, nationality, language used, online and offline behavior history, etc. The evaluation target indicates the evaluation target whose evaluation is described by the answer. The question information indicates a question related to the answer among the questions set by the question information in the progress conditions. The compatibility is information that indicates the degree to which the answer content complies with the progress conditions; for example, the smaller the number, the more compatible the answer is. For example, evaluation information that meets the following criteria may be assigned a higher degree of suitability: the content is specific, the content is in line with the purpose of the interview, the content is a controversial opinion and the opinions of other respondent users should be sought. In addition, in an interview for marketing research purposes, evaluation information that includes content that stimulates purchasing motivation or content that can gain sympathy from other respondent users may be assigned a higher degree of suitability.

[0114] The information associated with the interview ID and answer ID may be associated by the control unit 23 using any method, for example, by inputting it into predetermined reference information. For example, the identification unit 235 may preliminarily extract words corresponding to each piece of information included in the progress conditions and identify words semantically similar to words included in the answer content. The recording unit 236 may record the identified words as evaluation targets. Furthermore, the identification unit 235 may identify the suitability of the answer content based on the number of identified words. Furthermore, the artificial intelligence unit 237 may input the answer content into a generation AI, which may output information associated with each answer.

[0115] The first evaluation information and the second evaluation information may be associated with each other based on at least any information associated with each answer content. For example, evaluation information by answering users with similar attributes is associated with each other based on attribute information. Answering users with similar attributes may be identified based on the answer content in an interview, or may be identified by referring to the history of answer content in multiple interviews associated with each answering user. This allows answering users with similar taste tendencies to be identified. Furthermore, evaluation information for a common evaluation target is associated with each other based on the evaluation target. Furthermore, evaluation information for a common question is associated with each other based on question information.

[0116] The identification unit 235 then identifies, from the mutually associated evaluation information, evaluation information to be used in generating the second question. Preferably, the identification unit 235 identifies evaluation information that satisfies a predetermined condition in relation to the progress conditions in the interview with the third answering user. For example, the identification unit 235 may identify evaluation information to be used in generating the second question based on the degree of suitability. The identification unit 235 may identify, as the evaluation information to be used in generating the second question, the second evaluation information (answer ID: R002) by the first answering user, which is assigned a higher degree of suitability. Alternatively, the identification unit 235 may identify, as the evaluation information, an answer by an answering user having attributes similar to those of the third answering user. Alternatively, the identification unit 235 may identify, as the evaluation information, an answer whose question number is similar to that of the first question currently presented to the third answering user. In other words, the acquisition unit 234 acquires evaluation information by the first answering subject (second evaluation information) as the first evaluation information in the interview with the third answering subject. The second evaluation information satisfies a predetermined condition related to the progress conditions. This allows evaluation information from various respondent entities to be accumulated and updated to better suit the progress conditions, thereby enabling the generation of effective second questions.

[0117] Then, as a second presentation step, the presentation unit 232 (second presentation unit) causes the third respondent to be presented with a second question generated based on the progress conditions and the second evaluation information in an interview with the third respondent (action A107). For example, the artificial intelligence unit 237 inputs the second evaluation information into the generation AI, thereby outputting a question to be presented to the third respondent. The second question can be generated based on the input second evaluation information, the progress conditions stored in the generation AI, and the content of the answer by the third respondent.

[0118] The identification unit 235 may input the first evaluation information and the second evaluation information, which are associated with each other, to the generation AI to identify the evaluation information to be used in generating the second question. In this case, the artificial intelligence unit 237 may input at least one piece of information associated with each answer content, along with the answer content, to the generation AI to output the evaluation information to be used in generating the second question. This allows more comprehensively appropriate evaluation information to be identified, taking into account the timing of the interview, etc. In this case, the artificial intelligence unit 237 may input an instruction to the generation AI to make it more likely that evaluation information with a high degree of suitability will be reflected in the generation of the second question.

[0119] Returning to FIG. 4 , the progress control unit 231 progresses the interview with the third answering user, and the acquisition unit 234 acquires the answer from the third answering user to at least one of the first question and the second question. Furthermore, the identification unit 235 identifies third evaluation information to be used to generate a second question to be presented to another answering user from among the answers from the third answering user. The recording unit 236 records the identified third evaluation information in a predetermined storage area (action A111). In this way, the acquisition unit 234 acquires the third evaluation information for the evaluation target. The third evaluation information may be acquired after the interview with the third answering user is completed, or may be acquired in parallel with the progress of the interview.

[0120] When all questions have been asked, the progress control unit 231 ends the interview, and the control unit 23 notifies the first user terminal 3A of the interview results. For example, the control unit 23 notifies a list of the questions and answers in association with the information about the interview room so that the administrative user can identify them. The information about the interview room includes identification information for the interview room (interview ID, room name, etc.), the title of the interview, the progress conditions, and attribute information of the responding user.

[0121] 4. Variations The information processing system 1 according to this embodiment may adopt the following aspects.

[0122] In this embodiment, the description has been centered on the identification unit 235 and the artificial intelligence unit 237 accessing the generation AI as an artificial intelligence module, but this is not limiting. The above-described processing may be realized by the identification unit 235 and the artificial intelligence unit 237 accessing any artificial intelligence module.

[0123] In addition, in the example shown above, a first question is first presented to the answering user in an interview, and then a second question is presented before the next first question. However, this is not limiting. The presenting unit 232 may present the second question to the answering user before the first question, for example, the second question may be presented at the beginning of the interview. In this case, the acquiring unit 234 may acquire the evaluation information before presenting the first question. Furthermore, the presenting unit 232 may present a follow-up question that further inquires about the basis for the answer of the answering user to the second question, based on the answer of the answering user to the second question. In this case, the follow-up question may be generated based on the answer of the answering user to the second question and the progress conditions (particularly, the objective information). Furthermore, the generated question may be replaced with another first question that is preset in the question information.

[0124] In addition, while the evaluation information used to generate the second question is input by a referencing entity, and the referencing entity is the answering user, this is not limiting. The referencing entity may be, for example, a management entity involved in generating the questions for the interview, and the management entity may be an administrative user. In other words, the referencing entity refers to an entity other than the answering user in the interview that is involved in providing the evaluation information. In addition, while the evaluation information is the content of an answer to the first question by another answering user, it may also be the content of an answer to the second question.

[0125] For example, if the admin user is a provider of a product or the like, the evaluation information may be a statement introducing the product or the provider that has been set in advance by the admin user. Also, the second question may be a question asking the respondent user about their impression of the product or the provider, in which case the evaluation information may include information indicating the evaluation target itself. In other words, the evaluation target is not limited to items directly related to the objective set in the interview progress conditions (objective information), but also includes related items and their providers.

[0126] The evaluation information may also be generated by reference information (a generation AI or an AI agent) serving as a management entity. For example, the artificial intelligence unit 237 may input information about the evaluation target of the interview into the reference information, thereby outputting the evaluation information. The acquisition unit 234 may acquire information about the evaluation target input by the administrative user, and the artificial intelligence unit 237 may input this information into the reference information. The acquisition unit 234 may also acquire information that can identify the evaluation target input by the administrative user (e.g., the name of the evaluation target, such as a product), and the artificial intelligence unit 237 may input this information into the reference information, thereby outputting information about the evaluation target. The generation AI or AI agent serving as reference information may autonomously collect information about the evaluation target based on information that can identify the evaluation target. Then, the artificial intelligence unit 237 may input the output information about the evaluation target into the reference information. Note that the management entity that generates the change instruction and the management entity that sets the progress conditions may be different entities. For example, the progress conditions may be set by the administrative user, while the change instruction may be generated by the reference information serving as a management entity.

[0127] The evaluation information may also be the content of a question (second question) that the generating AI, acting as the reference subject, posed to another answering user. For example, a follow-up question is information generated by the generating AI based on answers from other answering users, and may be information describing an evaluation of the evaluation target.

[0128] In this embodiment, the interview is illustrated as proceeding through an interactive dialogue using text questions and answers, but is not limited to this. In the interview, questions may be presented to the respondent in the form of a questionnaire form, email exchanges, or the like. Furthermore, at least one of the questions and answers may be presented in a format that allows a person to recognize the content through their five senses, and may be presented as information including feedback such as audio, images, videos, touch, smell, and taste. For example, in response to the question, "What is your favorite brand of rice?", an answer may be obtained using an image of a rice package.

[0129] Furthermore, although the example has been given in which the progress conditions include objective information and question information, this is not limited thereto. The progress conditions do not have to include question information. For example, the first question may be generated by inputting objective information into the generation AI. The first question may be generated by inputting objective information and an answer to another question (at least one of the first question and the second question) into the generation AI. The second question may be generated by inputting objective information and a change instruction into the generation AI. The second question may be generated by inputting objective information, a change instruction, and an answer to another question (at least one of the first question and the second question) into the generation AI. In this way, the interview can proceed using questions autonomously generated based on the objective information using an AI agent.

[0130] 5.Other In the above embodiment, the progress control unit 231, the presentation unit 232, the acquisition unit 234, the identification unit 235, the recording unit 236, and the artificial intelligence unit 237 are described as functional units realized by the control unit 23 of the information processing device 2. However, at least some of these may be implemented as functional units realized by components other than the information processing device 2. Furthermore, while the user terminal 3 and the information processing device 2 store various pieces of information and control information processing, multiple external devices may be used in place of any of these. That is, various pieces of information and programs may be distributed and stored in multiple external devices using blockchain technology or the like. Furthermore, at least one of the devices included in the information processing system 1 may be installed outside the country in which the functions of the information processing system 1 are performed.

[0131] Furthermore, it may be provided in the following aspects.

[0132] (1) An information processing system comprising at least one processor, the processor being configured to execute each of the following steps by reading a program: a progress control step controlling the progress of an interview with a first respondent; a first presentation step having the first respondent present a first question in the interview, wherein the first question is generated based on at least a progress condition preset by a management entity; an acquisition step acquiring first evaluation information for an evaluation target, wherein the evaluation target is an object to be evaluated by the first respondent in the interview, and the first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference entity different from the first respondent; and a second presentation step having the first respondent present a second question generated based on at least the first evaluation information in the interview.

[0133] According to this embodiment, the effectiveness of the interview can be improved.

[0134] (2) In the information processing system described in (1) above, the progress conditions include objective information indicating at least the information desired to be obtained from the first answering entity, the first question is generated by inputting at least a portion of the progress conditions into a generation AI, and in the first presentation step, the generation AI presents to the first answering entity the next first question generated based on the progress conditions and the content of the answer to the first question by the first answering entity.

[0135] According to this embodiment, the effectiveness of the interview conducted by the generation AI can be improved.

[0136] (3) In the information processing system described in (1) or (2) above, the second question is generated by inputting at least a portion of the progress conditions and the first evaluation information into a generation AI.

[0137] According to this embodiment, the effectiveness of the interview conducted by the generation AI can be improved.

[0138] (4) In the information processing system described in any one of (1) to (3) above, in the progress control step, the interview with a second respondent different from the first respondent is conducted separately from the interview with the first respondent, and the reference subject is the second respondent.

[0139] According to this embodiment, the effectiveness of the interview can be improved by utilizing the answers of other respondents.

[0140] (5) In the information processing system described in any one of (1) to (4) above, in the identification step, the first evaluation information used to generate the second question is identified, and here, the first evaluation information used to generate the second question is identified from among multiple pieces of first evaluation information obtained by inputting at least one of at least a part of the progress conditions and the answer to the first question by the first answering subject into the generation AI.

[0141] According to this aspect, it is possible to effectively identify the answers of other answering entities to be utilized.

[0142] (6) In the information processing system described in any one of (1) to (5) above, in the second presentation step, the first respondent is made to present the second question to ask for the first respondent's opinion on the first evaluation information, in the acquisition step, second evaluation information on the evaluation target is acquired, where the second evaluation information includes information describing the results of the first respondent's evaluation of the evaluation target based on the first evaluation information, and in the recording step, the second evaluation information is recorded in association with the first evaluation information.

[0143] According to this embodiment, it is possible to effectively utilize the answers of other answering subjects.

[0144] (7) In the information processing system described in (6) above, in the progress control step, the interview with a third respondent different from the first respondent is conducted separately from the interview with the first respondent, in the acquisition step, the second evaluation information is acquired as the first evaluation information in the interview with the third respondent, and in the second presentation step, the second question generated based on the progress conditions and the second evaluation information is presented to the third respondent in the interview with the third respondent, wherein the second evaluation information satisfies a predetermined condition regarding the progress conditions.

[0145] According to this embodiment, it is possible to effectively utilize the answers of other answering subjects.

[0146] (8) In the information processing system described in any one of (1) to (7) above, the second presentation step is executed in response to the progress condition before the next first question by the first presentation step is presented to the first answering subject.

[0147] According to this embodiment, the second question can be presented without departing from the conditions for proceeding with the interview.

[0148] (9) In the information processing system described in any one of (1) to (8) above, the second presentation step is executed in response to the content of the answer to the first question by the first answering entity before the next first question is presented to the first answering entity by the first presentation step.

[0149] According to this embodiment, the second question can be presented without departing from the conditions for proceeding with the interview.

[0150] (10) An information processing method, comprising the steps of the information processing system according to any one of (1) to (9) above.

[0151] (11) A program that causes a computer to execute each step of the information processing system according to any one of (1) to (9) above. Of course, this is not the case.

[0152] Finally, while various embodiments of the present invention have been described, these are presented by way of example only and are not intended to limit the scope of the invention. The novel embodiments may be embodied in various other forms, and various omissions, substitutions, and modifications may be made without departing from the spirit of the invention. Such embodiments and modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as defined in the accompanying claims. [Explanation of symbols]

[0153] 1: Information processing system 2: Information processing equipment 20: Communication bus 21: Communications Department 22: Storage section 23: Control section 231: Progress control section 232:Presentation part 234: Acquisition Department 235: Specific part 236: Recording section 237: Artificial Intelligence Department 3: User terminal 3A: First user terminal (administrative user) 3B: Second user terminal (answering user) 4: Network 500: Settings screen 600: Interview screen 700: Interview screen T: Evaluation information table

Claims

1. An information processing system, At least one processor is provided, the processor being configured to execute the following steps by reading a program: In the progress control step, the interview with the first respondent is controlled to proceed, and the interview is controlled to proceed by presenting questions to the first respondent and obtaining answers to the questions from the first respondent; In the first presentation step, a first question is presented to the first respondent in the interview, the first question is generated based on at least a progress condition preset by a management entity; In the acquisition step, first evaluation information for the evaluation target is acquired, wherein: The evaluation target is an object to be evaluated by the first respondent in the interview, The first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference entity different from the first respondent entity, In the second presentation step, a second question, which is the question generated based on at least the first evaluation information, is presented to the first respondent in the interview, an information processing system, wherein the second question is a question for asking the first respondent's opinion on the first evaluation information;

2. 2. The information processing system according to claim 1, the progress condition includes at least objective information indicating information desired to be acquired from the first respondent; The interview is controlled to proceed by presenting the first question to the respondent and, after obtaining the answer to the first question from the respondent, presenting a next first question or a second question to the respondent; the first question is generated by inputting at least a part of the progress condition into a generation AI; In the first presentation step, if the second question is not presented to the first answering entity, the generation AI presents to the first answering entity the next first question generated based on the progress conditions and the content of the answer to the first question by the first answering entity.

3. 2. The information processing system according to claim 1, An information processing system in which the second question is generated by inputting at least a portion of the progress conditions and the first evaluation information into a generation AI.

4. 2. The information processing system according to claim 1, In the progress control step, the interview with a second respondent different from the first respondent is conducted separately from the interview with the first respondent; An information processing system, wherein the reference entity is the second answering entity.

5. An information processing system, At least one processor is provided, the processor being configured to execute the following steps by reading a program: In the progress control step, the interview with the first respondent is controlled to proceed, and the interview is controlled to proceed by presenting questions to the first respondent and obtaining answers to the questions from the first respondent; In the first presentation step, a first question is presented to the first respondent in the interview, the first question is generated based on at least a progress condition preset by a management entity; In the acquisition step, first evaluation information for the evaluation target is acquired, wherein: The evaluation target is an object to be evaluated by the first respondent in the interview, The first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference entity different from the first respondent entity, In the second presentation step, a second question, which is the question generated based on at least the first evaluation information, is presented to the first respondent in the interview; In the identifying step, the first evaluation information used to generate the second question is identified, wherein: An information processing system in which the first evaluation information used to generate the second question is identified from multiple pieces of first evaluation information obtained by inputting at least one of at least a part of the progress conditions and the answer to the first question by the first answering subject into a generation AI.

6. An information processing system, At least one processor is provided, the processor being configured to execute the following steps by reading a program: In the progress control step, the interview with the first respondent is controlled to proceed, and the interview is controlled to proceed by presenting questions to the first respondent and obtaining answers to the questions from the first respondent; In the first presentation step, a first question is presented to the first respondent in the interview, the first question is generated based on at least a progress condition preset by a management entity; In the acquisition step, first evaluation information for the evaluation target is acquired, wherein: The evaluation target is an object to be evaluated by the first respondent in the interview, The first evaluation information includes information describing the results of an evaluation of the evaluation target by a reference entity different from the first respondent entity, In the second presentation step, a second question, which is the question generated based on at least the first evaluation information, is presented to the first respondent in the interview; In the second presentation step, the second question is presented to the first respondent to ask for the first respondent's opinion on the first evaluation information; In the obtaining step, second evaluation information for the evaluation target is obtained, wherein: the second evaluation information includes information describing the result of the first respondent evaluating the evaluation target based on the first evaluation information, In the recording step, the second evaluation information is recorded in association with the first evaluation information.

7. 7. The information processing system according to claim 6, In the progress control step, the interview with a third respondent different from the first respondent is conducted separately from the interview with the first respondent; In the acquiring step, the second evaluation information is acquired as the first evaluation information in the interview with the third respondent; In the second presentation step, the second question generated based on the progress condition and the second evaluation information is presented to the third respondent in the interview with the third respondent, wherein: An information processing system, wherein the second evaluation information satisfies a predetermined condition regarding the progress condition.

8. 2. The information processing system according to claim 1, An information processing system, wherein the second presentation step is executed in response to the progress condition before the next first question in the first presentation step is presented to the first answering subject.

9. 2. The information processing system according to claim 1, An information processing system in which the second presentation step is executed in response to the content of the answer to the first question by the first answering entity, before the next first question is presented to the first answering entity by the first presentation step.

10. An information processing method, comprising: An information processing method comprising the steps of the information processing system according to any one of claims 1 to 9.

11. A program, A program that causes a computer to execute each step of the information processing system according to any one of claims 1 to 9.

Images