Question answering system, question answering method, and question answering program

Abstract

This system provides a question-answering system that reduces the burden on respondents in preparing answers to questions from speakers. [Solution] The question answering system of the present invention comprises: a question acquisition means for acquiring question information input to a first terminal; an answer acquisition means for acquiring a plurality of answer candidate pieces of information generated by a generation AI based on the question information; an output means for outputting the acquired plurality of answer candidate pieces of information to a second terminal; and a selection acceptance means for receiving the selection of at least one answer candidate piece of information from the plurality of answer candidate pieces of information from the second terminal.

Description

Technical Field

[0001] The present invention relates to a question-and-answer system, a question-and-answer method, and a question-and-answer program.

Background Art

[0002] As the background art of this technical field, there is Japanese Patent Application Laid-Open No. 2002-197196 (Patent Document 1). In this publication, it is described that "when a question is issued from a user terminal, the administrator terminal assigns one of the expert terminals to this question and allows it to be viewed from the expert terminal. When an answer is created on the expert terminal, the administrator terminal views and modifies this answer and allows the answer to be viewed on the user terminal that issued the question." (See the abstract).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the above Patent Document 1, there was a concern that with the increase in users, the burden on experts creating answers to questions from users would increase in creating answers. Therefore, an object of the present invention is to provide a question-and-answer system capable of reducing the burden of creating answers to questions from users.

Means for Solving the Problems

[0005] In order to solve the above problems, for example, the configuration described in the claims is adopted.

Effects of the Invention

[0006] According to the present invention, it is possible to reduce the burden on answerers in creating answers to questions from counselors. Other issues, configurations, and effects not mentioned above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0007] [Figure 1] Figure 1 shows an example of an overall configuration diagram of a question answering system. [Figure 2] Figure 2 shows an example of the hardware configuration of the management server. [Figure 3] Figure 3 shows an example of the various types of information stored in the auxiliary storage device of the management server. [Figure 4] Figure 4 shows an example of the hardware configuration of the answer generation server. [Figure 5] Figure 5 shows an example of the hardware configuration of a user terminal. [Figure 6] Figure 6 shows an example of a question answering flow. [Figure 7] Figure 7 shows an example of the flow for obtaining candidate answer sentences. [Figure 8] Figure 8 shows an example of the display flow of reference materials. [Figure 9] Figure 9 shows an example of a medical appointment scheduling and medical evaluation flow chart. [Figure 10] Figure 10 shows an example of the judgment flow for inspection information. [Figure 11] Figure 11 shows an example of the first screen of a user terminal. [Figure 12] Figure 12 shows an example of the second screen of a user terminal. [Figure 13] Figure 13 shows an example of the third screen of a user terminal. [Figure 14] Figure 14 shows an example of the fourth screen of a user terminal. [Figure 15] Figure 15 shows an example of the fifth screen of a user terminal. [Modes for carrying out the invention]

[0008] (1) Outline of the Embodiment Figure 1 shows an example of the configuration of a question answering system 100 (hereinafter simply referred to as system 100). As shown in Figure 1, system 100 includes a management server 200 and an answer generation server 400. The management server 200 and the answer generation server 400 are connected to a knowledge base 300 and multiple user terminals 500 via a network. The network can be wired or wireless, and each terminal can send and receive information via the network.

[0009] Each server and terminal in System 100 may be a mobile device such as a smartphone, tablet, mobile phone, or personal digital assistant (PDA), or a wearable device such as glasses, a wristwatch, or clothing. Alternatively, it may be a stationary or portable computer, or a server located in the cloud or on a network. Functionally, it may also be a VR (Virtual Reality) terminal, an AR (Augmented Reality) terminal, or an MR (Mixed Reality) terminal. Alternatively, it may be a combination of multiple such terminals. For example, a combination of one smartphone and one wearable device can logically function as a single terminal. Other types of information processing terminals may also be used.

[0010] Each server and terminal in System 100 is equipped with a processor that runs an operating system, applications, and programs; main memory such as RAM (Random Access Memory); auxiliary storage such as IC cards, hard disk drives, SSDs (Solid State Drives), and flash memory; a communication control unit such as a network card, wireless communication module, or mobile communication module; input devices such as a touch panel, keyboard, mouse, voice input, and camera; and output devices such as monitors and displays. The output devices may also be devices or terminals that transmit information for output to external monitors, displays, printers, or other equipment.

[0011] The main memory stores various programs and applications (also referred to as modules or processing units). By the processor executing these programs and applications, each functional element of the entire system 100 is realized. Note that these modules (processing units) may be implemented in hardware, such as by integration. Also, each module may be an independent program or application, or may be implemented in the form of sub-programs, functions, etc. within one integrated program or application.

[0012] In this specification, each module is described as the subject (subject) that performs processing. In reality, it is the processor that processes various programs and applications (modules) that executes the processing. The auxiliary storage device stores various databases (DBs). A "database" is a functional element (storage unit) that stores a data set so as to be able to handle any data operations (such as extraction, addition, deletion, overwriting, etc.) from the processor or an external computer. The implementation method of the database is not limited. For example, it may be a database management system, a spreadsheet software, or a text file such as XML or JSON.

[0013] The system 100 has a question-answering function that generates a plurality of candidate answer sentences for a question input by a consulter, allows the answerer to select an appropriate candidate answer sentence, make necessary corrections, and output them to the consulter terminal 500A. Also, the system 100 provides a knowledge base 300 for the answerer to view reference materials necessary for editing the candidate answer sentences. Further, the system 100 makes a predetermined determination based on the input of the results of a user's simple inspection. The details of the configuration and processing of such a system 100 will be described in order below.

[0014] (2) Configuration of the management server 200 Figure 2 is a diagram showing an example of the hardware configuration of the management server 200. The management server 200 is composed of, for example, servers arranged on the cloud. The management server 200 mainly realizes the following functions. • A function to categorize the entered questions. • A function to assign respondents to questions. • A function that outputs the generated candidate answer information for a question to the respondent terminal 500B. • A function to output the candidate answer information edited by the respondent to the consultant's terminal 500A. • A function that allows respondents to view reference materials of their choice. • Function to output the judgment result generated based on the input test result.

[0015] Here, the consultant terminal 500A (first terminal) refers to the terminal used by the consultant among the user terminals 500. The respondent terminal 500B (second terminal) refers to the terminal used by the respondent among the user terminals 500. For example, if system 100 is applied to a consultation service regarding dental treatment, the consultant is assumed to be a user who has some concerns about the health of their teeth.

[0016] Respondents may include, for example, the following: • A dentist who is affiliated with a specific dental clinic and is engaged in providing dental care. • Dental hygienists who are affiliated with a specific dental clinic and engage in dental treatment assistance and other related activities under the guidance of a dentist. • A dentist who previously worked in dental care but is currently not practicing dentistry and is providing home-visit services. • Dental hygienists who previously worked in dental practice assisting with clinical procedures but are currently not engaged in clinical practice and are working from home.

[0017] As shown in Figure 2, the main memory 201 stores programs and applications such as the question acquisition module 211, classification module 212, allocation module 213, prompt generation module 214, answer acquisition module 215, selection acceptance module 216, editing module 217, link output module 218, judgment module 219, output module 220, reservation module 241, and evaluation module 242. The processor 203 executes these programs and applications to realize each functional element of the management server 200.

[0018] (Question Acquisition Module 211) The question acquisition module 211 acquires question information entered into the client terminal 500A. For example, the question acquisition module 211 acquires question information entered as text data. The question information may also be entered as audio data or images. The question acquisition module 211 is an example of a question acquisition means in the present invention.

[0019] (Classification Module 212) The classification module 212 classifies the acquired question information into predetermined categories. For example, the classification module 212 classifies the text data, which is the question information, according to the content of the question using natural language processing. Specifically, it performs morphological analysis on the text data, segments the question sentence into morphemes, and then performs vector transformation on each morpheme. Then, based on the similarity of the vector values ​​of each morpheme, it classifies the content of the question sentence into a predetermined category. The classification module 212 is an example of a classification means in the present invention.

[0020] For example, if System 100 is applied to a consultation service related to dental treatment, the question categories may include the following: • Consultation regarding tooth loss • Consultation regarding chipped teeth • Consultation regarding cracked teeth • Consultation regarding teeth grinding • Consultation regarding sensitive teeth • Consultations regarding children's tooth eruption and replacement, etc.

[0021] (Assignment module 213) The assignment module 213 assigns respondents to the question information entered by the caller. Specifically, the assignment module receives assignment requests from the respondent terminal 500B, which has received the entered question information and classified category, requesting to be assigned as a respondent to that question, and assigns the respondent to that question. Assignment requests are valid on a first-come, first-served basis, in principle.

[0022] For example, when System 100 is applied to a consultation service regarding dental treatment, the home-visit dentists who act as respondents will have different areas of expertise based on their individual knowledge as dentists. Therefore, respondents will check the question and its category and then decide whether they are able to answer it. If they decide they can answer the question, they will operate the respondent terminal 500B to send an assignment request for that question. On the other hand, if they decide that it is difficult to answer the question based on their own knowledge, the respondent will not send an assignment request. In this case, the question will be assigned based on assignment requests from other respondents. The assignment module 213 is an example of an assignment means in this invention. In this way, the quality of the answers can be ensured by having respondents send assignment requests based on their own judgment.

[0023] (Prompt generation module 214) The prompt generation module 214 generates prompts to be input to the generation AI that generates candidate answer sentences for question information. The prompt generation module 214 generates prompts based on the question information. The prompt generation module 214 may also generate prompts based on the question information and the category of the question. The prompt generation module 214 is an example of a prompt generation means in the present invention.

[0024] (Answer acquisition module 215) The answer acquisition module 215 acquires multiple candidate answer information for the question information. The answer acquisition module 215 inputs a prompt to the generation AI and acquires multiple candidate answer sentences generated by the generation AI. The answer acquisition module 215 is an example of an answer acquisition means in the present invention.

[0025] (Selection acceptance module 216) The selection reception module 216 receives the selection of at least one answer candidate from among multiple answer candidate information from the respondent terminal 500B. This selects the answer candidate sentence to be adopted from among the multiple answer candidate information. The selection reception module 216 is an example of a selection reception means in the present invention.

[0026] (Editing Module 217) The editing module 217 edits the selected candidate answer information based on editing operations performed on the candidate answer sentences by the respondent terminal 500B. As a result, the candidate answer sentences generated by the generation AI are edited by the respondent to become appropriate for the specific questions from individual consultants. The editing module 217 is an example of an editing means in the present invention.

[0027] (Link output module 218) The link output module 218 outputs a link display tag that provides access to the knowledge base 300. The knowledge base 300 stores reference materials that respondents refer to when editing candidate answer sentences. In other words, if the respondent is a dentist who has moved away from clinical practice, they may need to verify their specialized knowledge when carefully considering their answer to a question. For this reason, in system 100, various types of knowledge used in answers are classified and stored in the knowledge base 300 according to knowledge categories, and respondents can refer to them as they wish. The link output module 218 is an example of a link output means in the present invention.

[0028] (Determination module 219) The judgment module 219 makes a judgment on the test results based on the test information of the client entered into the client terminal 500A. The judgment means further acquires comment information generated by the AI ​​based on the test information. In other words, the judgment information includes the following information. • Symptom score: The symptoms and their score estimated based on the test information of the person seeking advice. • Comment information: Information on recommended actions to improve symptoms. The determination module 219 is an example of the determination means of the present invention.

[0029] (Output module 220) The output module 220 outputs information such as the following in response to a request from the user terminal 500. The system outputs the question information and category information 223, which indicates the category to which the question information is classified, to the respondent terminal 500B. The AI ​​generates multiple candidate answer information, which is then output to the respondent's terminal 500B. • Link display tags associated with reference materials stored in knowledge base 300 are output to the respondent terminal 500B. • The reference materials stored in the knowledge base 300 are output to the respondent terminal 500B. • The candidate answer information edited by the respondent is output to the consultant's terminal 500A. • The system outputs a symptom score and comment information indicating the assessment to the client's terminal 500A. The output module 220 is an example of the output means of the present invention.

[0030] (Reservation module 241) The reservation module 241 recommends to the client terminal 500A, from which the response information has been output, information on multiple hospitals, including the hospital to which the respondent who edited the response information belongs, as potential hospitals for treatment. Furthermore, the reservation module 241 makes a medical appointment for the hospital selected on the user terminal 500A. The reservation module 421 is an example of the reservation means of the present invention.

[0031] (Evaluation Module 242) The evaluation module 242 acquires review information about the medical treatment received at the clinic, which is entered into the consultation terminal 500A by the user who received treatment at the clinic. Furthermore, the evaluation module 242 performs an aggregated evaluation of the acquired review information. The evaluation module 242 is an example of the evaluation means of the present invention.

[0032] As shown in Figure 2, the auxiliary storage device 202 stores information such as consultant information 221, respondent information 222, category information 223, question text information 224, candidate answer text information 225, answer text information 226, link information 227, and judgment information 228. Other information may also be stored in the auxiliary storage device 202. The auxiliary storage device 202 is an example of a storage means in the present invention.

[0033] (3) Various information stored in the management server 200 Figure 3 shows an example of the various types of information stored in the auxiliary storage device 202 of the management server 200. Note that this figure only illustrates one example of the data structure for the various types of information; substantially equivalent information may be represented using other data structures. The contents of each of these types of information will be explained in order below.

[0034] <Consultant Information 221> First, let's explain the consultant information 221. Consultant information 221 mainly consists of information about consultants who, as users of system 100, use system 100 to ask various questions. In the illustrated example, the consultant table is stored as consultant information 221.

[0035] (Consultant Table) The client table associates a unique identifier, the client ID, with information such as name, address, device information, contact details, date of birth, medical history, password, and registration date and time. The name field stores the name information of the person seeking advice, which corresponds to the consultation ID. Alternatively, a pseudonym or handle name may be stored instead of the name. The address field stores the address information of the person seeking advice, which corresponds to the person seeking advice ID.

[0036] The terminal information stores information about the user who uses the service, corresponding to the user ID. The contact information section stores the contact details of the person seeking advice, corresponding to the person seeking advice's ID. The date of birth field stores the date of birth of the person seeking advice, which corresponds to the consultation ID.

[0037] The medical history section stores information about the client's medical history, corresponding to the client's client ID. Medical history refers to records of illnesses and other conditions the client has suffered from in the past, and includes information such as the following: • Names of illnesses / symptoms previously diagnosed (e.g., malocclusion, temporomandibular joint disorder) • Medications currently being taken, and any past drug side effects • Allergy information • Experience of traffic accidents • Whether or not you have given birth Other matters concerning your current health condition. Medical history is entered, for example, based on the client's self-reporting. The My Number system stores the My Number of the person seeking advice, which corresponds to the consultant's ID.

[0038] The password field stores the password that the user, corresponding to the user ID, is required to enter when logging into system 100. The hash value may be managed using a hash value. At the time of registration, a series of records corresponding to the consultant ID are stored in the consultant table, along with the date and time information registered. The client table may also include other columns.

[0039] <Respondent information 222> Next, we will explain the respondent information 222. The respondent information 222 mainly stores information about respondents who, as users of system 100, answer questions from the consultant. In the example shown, the respondent table is stored as the respondent information 222.

[0040] (Respondent Table) The respondent table associates a unique identifier, the respondent ID, with information such as name, address, device information, contact details, job title, assigned category, affiliated hospital, number of responses, password, and registration date and time. The name field stores the respondent's name information, which corresponds to the respondent ID. Alternatively, a pseudonym or handle name may be stored instead of the actual name. The address field stores the address information of the respondent, corresponding to the respondent ID.

[0041] The terminal information includes information about the respondent terminal 500B used by the respondent, corresponding to the respondent ID. The contact information section stores the contact details of the respondent, corresponding to the respondent's ID.

[0042] The job title field stores the job title of the respondent, corresponding to the respondent ID. Examples of job titles include dentist and dental hygienist. The assigned category stores the category assigned to the respondent corresponding to the respondent ID. For example, in the case of a dental hygienist, the assigned category would be "Dental and Oral Health Support," and they can answer general questions such as the prevention of cavities and periodontal disease, and dental checkups. Furthermore, even if a dental hygienist is able to answer a category, if the answer to the question requires mentioning medical procedures, the dental hygienist responding can operate the respondent terminal 500B to pass the answer to a respondent with a medical license.

[0043] On the other hand, dentists can answer questions about all aspects of dental care, but if they have a particular area of ​​specialization, they can specify that area. Examples of areas of specialization include the following: • General dentistry (cavity treatment, periodontal disease treatment) • Orthodontics (dental straightening, bite correction) Pediatric Dentistry • Oral and maxillofacial surgery (tooth extraction, treatment of jaw and facial diseases / trauma) • Implant treatment In System 100, the classification of assigned categories and the classification of question categories may be exactly the same as the classification of knowledge categories and question categories described later, or there may be categories that do not coincide with each other.

[0044] The "Affiliated Clinic" field stores the identifier of the clinic to which the respondent belongs, corresponding to the respondent ID. If the respondent (a dentist or dental hygienist) is not currently affiliated with a specific dental clinic, the "Affiliated Clinic" field will be left blank.

[0045] The response count stores the number of times the respondent corresponding to the respondent ID has responded in the past. The response count may be cumulative, or it may be information that indicates the actual frequency of responses, such as the number of responses within the most recent specified period.

[0046] The password contains the password that the respondent corresponding to the respondent ID is required to enter when logging into system 100. The hash may be managed using a hash value. At the time of registration, a series of records corresponding to the respondent ID are stored in the respondent table, along with the date and time information registered. The respondent table may also include other columns.

[0047] <Category Information 223> Next, we will explain category information 223. Category information 223 mainly consists of information about the category of the question (question category). In the example shown, the category table is stored as category information 223.

[0048] (Category Table) The category table holds information about the categories of the question information. The category table associates a unique identifier, the category ID, with the category name, category content, display tag string, respondent qualifications, registration date and time, and other information.

[0049] The category name contains the name of the category, which corresponds to the category ID. The category content stores the information for the category corresponding to the category ID. The display tag string stores the tag string that will be displayed on user terminal 500 when displaying the category corresponding to the category ID.

[0050] The respondent qualifications section stores the qualifications (job titles) of respondents who can answer questions for the category corresponding to the category ID. In other words, questions are classified into those that can be answered by dental hygienists and those that can only be answered by dentists, depending on the content of the question. Therefore, the respondents who can answer the question are set according to the category of the question.

[0051] The registration date and time information is stored in the category table, containing a series of records corresponding to the category ID. Note that the category table may also have other columns.

[0052] <Question Information 224> Next, we will explain the question text information 224. The question text information 224 mainly consists of text data about the questions entered by the person seeking advice. In the example shown, the question text table is stored as the question text information 224.

[0053] (Question Table) The question table holds the text data of questions entered by the user seeking advice. The question table associates the user ID, question content, question category, assigned respondent, registration date and time, etc., with each question ID, which is a unique identifier.

[0054] The Consultant ID stores the identifier of the consultant who entered the question text, which corresponds to the Question Text ID. The question content contains text data that represents the content of the question, corresponding to the question ID.

[0055] The question category stores the category into which the question text corresponding to the question text ID is classified. The assigned respondent will have an identifier stored that corresponds to the question text ID.

[0056] The registration date and time will store the date and time information of the registration in the question table, which corresponds to a series of records with question IDs. The question table may also have other columns.

[0057] <Answer candidate sentence information 225> Next, we will explain the candidate answer information 225. The candidate answer information 225 consists of information on candidate answer sentences generated by the generation AI in response to the question. In the example shown, the candidate answer table is stored as the candidate answer information 225.

[0058] (Table of possible answers) The answer candidate sentence table holds information about the answer candidate sentences generated by the generation AI. In the answer candidate sentence table, each answer candidate sentence ID, which is a unique identifier, is associated with the question ID, answer candidate number, answer candidate sentence content, registration date and time, etc.

[0059] The question ID stores the identifier of the question corresponding to the candidate answer that corresponds to the candidate answer ID. The answer candidate number stores the output number of the answer candidate sentence, which corresponds to the answer candidate sentence ID. In this embodiment, the generating AI generates three candidate answer sentences in response to a single question sentence. In this case, candidate answer sentences No. 1 to No. 3 will be generated for a single question sentence. Note that the generating AI can arbitrarily select two or more candidate answer sentences.

[0060] The "Suggested Answer" field contains text data representing the content of the suggested answer corresponding to the suggested answer ID. The registration date and time will store the date and time information of the registration of a series of records corresponding to the candidate answer ID in the candidate answer table. The table of suggested answer sentences may also include other columns.

[0061] <Answer text information 226> Next, we will explain the response information 226. The response information 226 consists of information about the response text, which is the content edited by the respondent for each candidate response text. In the example shown, the response text table is stored as the response information 226.

[0062] (Answer Table) The response table holds information about responses edited by respondents. Each response ID, a unique identifier, is associated with the question ID, candidate response ID, editing information, registration date and time, and other relevant details.

[0063] The question ID stores the identifier of the question corresponding to the answer that corresponds to the answer ID. The answer candidate sentence ID stores the identifier of the answer candidate sentence corresponding to the answer sentence with the answer sentence ID.

[0064] The response content contains text data that represents the content of the response, corresponding to the response ID. At the time of registration, a series of records corresponding to the response ID are stored in the response table, along with the date and time information registered. The response table may also include other columns.

[0065] <Link Information 227> Next, we will explain the link information 227. The link information 227 mainly consists of information about links that access the knowledge base 300. In the example shown, the link table is stored as the link information 227.

[0066] (Link Table) The link table holds information about links for each reference material in Knowledge Base 300. The link table associates a unique identifier, the link ID, with the category ID, link display tag, reference material path information, registration date and time, and other relevant details.

[0067] The Knowledge Category ID stores the identifier of the knowledge category of the reference material corresponding to the link, which is equivalent to the Link ID. A knowledge category is a classification set according to the content described in the reference material. In System 100, the knowledge category classification assigned to the reference material and the question category classification to which the question is classified may be exactly the same, or there may be categories that do not match each other. The link display tag contains a display tag indicating the knowledge category of the reference material corresponding to the link with the link ID.

[0068] The path information for reference materials stores path information indicating the address within knowledge base 300 that allows access to the reference material corresponding to the link with the link ID. The registration date and time stores the date and time information of the registration in the link table, which corresponds to a series of records that are linked to the link ID. The linked table may also have other columns.

[0069] <Judgment information 228> Next, we will explain the judgment information 228. The judgment information 228 consists of information determined based on the input inspection information. In the illustrated example, the judgment result table is stored as the judgment information 228.

[0070] (Judgment Result Table) The judgment result table holds information determined based on the entered test information. In the judgment result table, the unique identifier, the judgment ID, is associated with the consultant ID, entered test information, symptom score, comment information, registration date and time, etc.

[0071] The Consultant ID stores the identifier of the consultant who is the subject of the assessment, which corresponds to the Assessment ID. The input inspection information stores the inspection information that is the subject of the judgment, corresponding to the judgment ID. Details of the inspection information will be described later.

[0072] The symptom score stores the symptom score obtained through the judgment process corresponding to the judgment ID. Details of the symptom score will be described later. The comment information stores the comment information obtained through the judgment process corresponding to the judgment ID. Details of the comment information will be described later. The registration date and time will store the date and time information of the series of records corresponding to the judgment ID that were registered in the judgment result table. The judgment result table may also include other columns.

[0073] <Clinic information 229> Next, we will explain the clinic information 229. The clinic information 229 consists of information about the dental clinic to which the respondent dentist or dental hygienist belongs. In the example shown, the clinic table is stored as the clinic information 229.

[0074] (Clinic table) The clinic table holds information about dental clinics to which dentists or dental hygienists belong. Each clinic in the table is associated with a unique identifier, the clinic ID, along with its name, address, phone number, email address, website URL, registration date, and other relevant information.

[0075] The name field stores the name of the clinic, which corresponds to the clinic ID. The location field stores the clinic's address information, which corresponds to the clinic ID. The phone number will contain the phone number of the clinic, which will be equivalent to the clinic ID. The email address will contain the clinic's email address, which corresponds to the clinic ID.

[0076] The website URL contains information about the URL of the clinic's website, which corresponds to the clinic ID. The registration date and time will store the date and time information registered in the clinic table, which corresponds to a series of records equivalent to the clinic ID. The clinic table may also include other columns.

[0077] <Medical Information 230> Next, we will explain the medical information 230. The medical information 230 consists of information regarding appointments and treatment history at the dental clinic for the patient seeking dental care. In the example shown, the medical information 230 stores an appointment table and a treatment history table.

[0078] (Appointment Schedule Table) The appointment scheduling table holds information about appointments made by the patient to a dental clinic. Each appointment scheduling table has a unique identifier, the appointment ID, associated with the clinic ID, dentist ID, patient ID, question ID, appointment date and time, registration date and time, and other relevant information.

[0079] The hospital ID stores the identifier of the hospital related to the medical appointment, which is equivalent to the reservation ID. The Dentist ID stores the identification information of the dentist scheduled to handle the appointment, which corresponds to the Appointment ID. The dentist's identification information may be the Respondent ID in the Respondent Table, or it may be other identification information. When making an appointment after a question from a user and a response from a respondent using System 100, the respondent can be selected as the attending physician.

[0080] The patient ID stores the identification information of the patient (consultant) who made a medical appointment, which corresponds to the appointment ID. The patient's identification information may be the consultant ID in the consultant table, or it may be other identification information.

[0081] The Question ID stores the identifier of the question if the medical appointment corresponding to the Reservation ID was made through a question from a user using System 100 and an answer from a respondent.

[0082] The reservation date and time will store the appointment time slot corresponding to the reservation ID. At the time of registration, a series of records corresponding to the reservation ID are stored in the appointment table, along with the date and time information registered. The appointment scheduling table may also include other columns.

[0083] (Medical History Table) The treatment history table holds information about the treatment history of a patient at a dental clinic. The treatment history table associates a unique identifier, the history ID, with the appointment ID, patient ID, dentist ID, status, treatment details, registration date and time, and other relevant information.

[0084] The reservation ID stores an identifier for the medical appointment corresponding to the medical history, which is equivalent to the history ID. The patient ID stores identification information for patients who have received medical treatment, corresponding to the history ID. This patient identification information may be the consultant ID in the consultant table, or it may be other identification information.

[0085] The Dentist ID stores the identification information of the dentist who actually handled the treatment history, which corresponds to the History ID. The dentist's identification information may be the respondent ID in the respondent table, or other identification information. If a different dentist handled the treatment than the one scheduled for the appointment, the information of that different dentist will be stored in the appointment table.

[0086] The status field stores the status of the medical history, which corresponds to the history ID. The medical history status includes, for example, the following information: • Appointment scheduled • Consulted • Change of reservation • Cancel reservation • No-show

[0087] The medical treatment details store the medical treatment information corresponding to the history ID in the medical history. At the time of registration, a series of records corresponding to the history ID are stored in the medical history table, along with the date and time information registered. The medical history table may also include other columns.

[0088] <Review Information 231> Next, we will explain review information 231. Review information 231 consists of review information in which the consultant and patient evaluated the respondent or dental clinic. In the example shown, the response review table and the treatment review table are stored as review information 231.

[0089] (Answer Review Table) The response review table holds information on the evaluations of respondents entered by the user. The evaluation of a respondent includes an evaluation score and review comments. The response review table associates a unique identifier, the response review ID, with the response text ID, evaluation score, review comments, respondent's comments, fraudulent review flag, registration date and time, etc.

[0090] The response ID stores the identifier of the response being reviewed, which corresponds to the response review ID. The evaluation score includes the evaluation score that the person who received the answer entered as part of the review content corresponding to the answer review ID. The review comments contain specific comments from the user regarding the answer they received, corresponding to the answer review ID. Review comments may include both positive and negative content.

[0091] The respondent's comments will store comments entered by the respondent for the review corresponding to the response review ID. For example, it is expected that positive reviews will receive comments encouraging continued dental care, while negative reviews will receive comments including apologies.

[0092] The fraudulent review flag stores an indicator that the review in question is fraudulent. The fraudulent review flag is entered when a review corresponding to the response review ID is determined to be fraudulent, for example, through a report from the respondent or through maintenance by the system 100 administrator. Examples of fraudulent reviews include the following: • Reviews containing false information that could damage the credibility of the respondent. • A review based on an extremely one-sided argument and severely lacking in objectivity. • Reviews that go beyond legitimate criticism and defame the respondent in question. • Reviews that contain immoral, obscene, discriminatory, or extremist language that may offend others. • Reviews containing content completely unrelated to the answer. • Other reviews that are not beneficial to other users in the selection of respondents

[0093] The registration date and time will store the date and time information for a series of records corresponding to the response review ID that were registered in the response review table. The response review table may also include other columns.

[0094] (Clinical Review Table) The treatment review table holds information on evaluations of dental clinics entered by the patient. These evaluations include an evaluation score and review comments. The treatment review table associates a unique identifier, the treatment review ID, with the history ID, evaluation score, review comments, clinic comments, fraudulent review flag, registration date and time, and other relevant information.

[0095] The History ID stores an identifier for the medical history subject to review, which corresponds to the Medical Review ID. The evaluation score includes the evaluation score entered by the patient for the medical treatment they received, which corresponds to the content of the review, or the medical review ID. The review comments contain specific comments entered by the patient about the medical treatment they received, corresponding to the medical review ID. Review comments may include both positive and negative content.

[0096] The comments section of the clinic will store the responses entered by the dental clinic for reviews corresponding to the treatment review ID. For example, it is expected that comments encouraging continued dental care for positive reviews will be entered, and comments including apologies will be entered for negative reviews.

[0097] The fraudulent review flag stores an indicator that the review in question is fraudulent. The fraudulent review flag is entered when a review corresponding to a medical review ID is determined to be fraudulent, for example, through a report from the clinic or through maintenance by the administrator of system 100. Examples of fraudulent reviews include the following: • Reviews containing false information that could damage the reputation of the clinic in question. • A review based on an extremely one-sided argument and severely lacking in objectivity. • Reviews that go beyond legitimate criticism and slander the clinic or its staff. • Reviews that contain immoral, obscene, discriminatory, or extremist language that may offend others. • Reviews containing content completely unrelated to the medical history. • Other reviews that are not helpful to other users when selecting a dental clinic.

[0098] The registration date and time will store the date and time information for a series of records corresponding to the medical review ID that were registered in the medical review table. The clinical review table may also include other columns.

[0099] (4) Configuration of the response generation server 400 Next, the configuration of the response generation server 400 will be described. Figure 4 shows an example of the hardware configuration of the response generation server 400. The response generation server 400 consists of one or more servers located on the cloud, for example.

[0100] As shown in Figure 4, the main memory 401 of the answer generation server 400 stores programs and applications such as the generation AI module 411. The processor 403 executes these programs and applications to realize each functional element of the answer generation server 400.

[0101] Among the functional elements to be implemented, the generation AI module 411 is a deep learning model that has been pre-trained on a large dataset. In other words, this module is an LLM (i.e., a Large-Scale Language Model). This module outputs a response to an input prompt (in other words, an instruction). This module has a deep learning model pre-trained using a large dataset, and therefore can be used without training data or additional training. Alternatively, additional training may be performed using the content of the knowledge base 300 utilized in system 100.

[0102] (5) Configuration of user terminal 500 Next, the configuration of the user terminal 500 will be described. Figure 5 shows an example of the hardware configuration of the user terminal 500. The user terminal 500 is a terminal device such as a smartphone, tablet, notebook PC, or desktop PC.

[0103] As shown in Figure 5, the main memory 501 of the user terminal 500 stores programs and applications such as the question answering module 511, the server cooperation module 512, and the display module 513. The processor 503 executes these programs and applications to realize each functional element of the user terminal 500.

[0104] The question answering module 511 works in cooperation with the server integration module 512 to output an answer to a question entered by the user. The server integration module 512 works in conjunction with other servers, such as the answer generation server 400, to obtain answers to questions. The display module 513 implements a function to display the response text in response to user input. The specific display methods will be described later.

[0105] (6) Processing of answers to questions Next, we will explain the question answering process performed by the management server 200. Figure 6 shows an example of the question answering flow 600. As shown in Figure 6, the client terminal 500A accepts the input of the question (step S610). The client terminal 500A accepts the question entered by the client. The question is sent to the management server 200.

[0106] Next, the management server 200 retrieves the question text (step S620). Specifically, the question retrieval module 211 of the management server 200 retrieves the question text sent from the client terminal 500A and records it as a new record in the question text table.

[0107] Next, the management server 200 classifies the question text (step S621). Specifically, the classification module 212 of the management server 200 analyzes the text data of the question text using natural language processing and classifies it into predetermined categories. The question category information obtained through the classification process is added to the question text table.

[0108] Next, the management server 200 outputs the question text, question category, and information regarding the consultant's address to the respondent terminal 500B (step S622). Specifically, the output module 220 of the management server 200 obtains the text data of the question text, the category information of the question text, and the address information of the consultant who asked the question by referring to the respective data tables, and sends them to the respondent terminal 500B.

[0109] Here, address information may refer to all information included in address information, but from the perspective of protecting personal information, it is preferable that it be only a part of the address information. Specifically, for example, if the address is "A Prefecture B City D3-chome 1-ban 1-go", then "A Prefecture B City D3-chome", which is the place name, may be considered as address information for the person seeking advice. Alternatively, "A Prefecture B City", which is part of the place name, may be considered as address information, or the entire address, including the street number, "A Prefecture B City D3-chome 1-ban", may be considered as address information.

[0110] Next, the respondent terminal 500B displays the question and category tag (step S630). Specifically, the display module 513 of the respondent terminal 500B displays the question and category tag sent from the management server 200 on the display, which is the output device 505, and presents it to the respondent. The respondent checks the presented question and category tag and determines whether or not it is a question they can answer. In other words, in step S630, the respondent candidate is asked whether or not they can answer the question.

[0111] Furthermore, at this time, the display module 513 of the respondent terminal 500B displays information about the consultant's address along with the question. Specifically, some of the consultant's address information will be displayed on the respondent terminal 500B along with the question and question category. Therefore, potential respondents can confirm which region the consultant lives in.

[0112] Therefore, potential respondents can check the presented question and the address of the person seeking advice, and prioritize answering questions from people who are close to their own dental clinic. In other words, answering questions from people in the same area as the dental clinic is an activity that encourages patients to seek treatment at their clinic. Furthermore, it is convenient for the person seeking advice to receive an answer from a dentist at a dental clinic near their home, as they will not have to explain their health concerns from scratch again during subsequent treatment.

[0113] In other words, System 100 displays the address of the person who asked the question along with the question itself, and the dentist who answers can select a question to answer based on the person's address, thereby guiding new patients to their dental clinic. Therefore, by using System 100, prospective patients can communicate with dental clinics near their homes before their appointment and confirm treatment plans in advance, allowing them to approach treatment with a sense of satisfaction. In addition, dentists who answer the questions can acquire prospective patients through the system. As a result, System 100 can promote the advancement of local dental care.

[0114] Next, the respondent terminal 500B sends an assignment request (step S631). Specifically, the server communication module 512 of the respondent terminal 500B sends an assignment request to the management server 200 in response to the respondent's request operation for an assignment request, in which the respondent has expressed a desire to take on the role. On the other hand, if the respondent does not wish to take on the role, this process is not executed, and the series of processes ends.

[0115] Next, the management server 200 assigns the question (step S623). Specifically, the assignment module 213 of the management server 200, in response to receiving the assignment request, assigns the respondent to the request and the person in charge of that question. As a result, a new record for the assigned respondent is recorded in the question table.

[0116] Next, the management server 200 obtains multiple candidate answer sentences (step S624). Specifically, the answer acquisition module 215 performs a procedure to obtain candidate answer sentences for the question sentence based on the process of assigning respondents to the question sentence, that is, in response to the execution of the assignment process. This process will be described in detail with reference to Figure 7.

[0117] Figure 7 shows an example of the flow for acquiring candidate answer sentences shown in Figure 6. As shown in Figure 7, in the flow for acquiring candidate answer sentences, the answer acquisition module of the management server 200 acquires the question sentence (step S710).

[0118] Next, the management server 200 generates a prompt (step S712). Specifically, the prompt generation module 214 of the management server 200 generates a prompt based on the information in the question. In this case, the prompt generation module 214 may also generate a prompt based on the category of the question. The generated prompt is sent to the answer generation server 400.

[0119] Next, the response generation server 400 receives a prompt (step S721). The response generation server 400 then inputs the received prompt to the generation AI module 411 and generates candidate response sentences (step S722). In this embodiment, the generation AI module 411 generates three candidate response sentences. Subsequently, the candidate response sentences are output to the management server 200 (step S723).

[0120] Next, the management server 200 receives the candidate answer sentences (step S713). Specifically, the answer acquisition module 215 of the management server 200 receives multiple candidate answer sentences sent from the answer generation server 400. As a result, a new record is recorded in the candidate answer sentence table. This completes the process of step S625 shown in Figure 6.

[0121] Next, as shown in Figure 6, the management server 200 outputs multiple candidate answer sentences to the respondent terminal 500B (step S625). Specifically, the output module 220 of the management server 200 sends the acquired candidate answer sentences to the respondent terminal 500B as candidate answer sentences for the question.

[0122] Next, the respondent terminal 500B displays multiple candidate answer sentences (step S632). Specifically, the display module 513 of the respondent terminal 500B displays the multiple candidate answer sentences sent from the management server 200 side by side on the display of the output device 505. The manner in which the candidate answer sentences are displayed on the respondent terminal 500B at this time will be described later.

[0123] Next, the respondent terminal 500B accepts the selection of the most suitable answer candidate sentence (step S633). Specifically, the respondent checks the multiple answer candidate sentences displayed on the respondent terminal 500B's display and selects the one that is appropriate as an answer to the question. At this time, the respondent selects the one that is most convenient, taking into account that they will edit it later. In other words, even if there is a slight error in the content of the answer candidate sentence, the respondent may select the answer candidate sentence that requires the least effort to generate the answer sentence after editing. The selection acceptance module 216 of the management server 200 accepts the information of the selected answer candidate sentence. In this way, the system 100 allows the respondent to obtain multiple answer candidate sentences, thereby reducing the burden of creating answers to questions from users.

[0124] Next, the respondent terminal 500B accepts editing operations for the selected candidate answer sentence (step S634). Specifically, the respondent terminal 500B accepts the editing operations entered by the respondent and outputs them to the management server 200.

[0125] Next, the management server 200 receives the edited content sent from the administrator terminal (step S627). Specifically, the respondent terminal 500B edits the candidate response text according to the edited content and records a new record in the response text table. After that, the output module 220 of the management server 200 outputs the response text to the consultant terminal 500A (step S628).

[0126] Next, the user terminal 500A displays the answer (step S611). Specifically, the display module 513 of the user terminal 500A displays the answer sent from the management server 200 on the display, which is the output device 505. This allows the user to confirm the answer to the question. After confirming the answer, the user enters an evaluation score and review comments about the content of the answer. The entered information is sent from the user terminal 500A to the management server 200 and recorded as a new record in the answer review table. With this, the question-answer flow 600 is completed.

[0127] (7) Regarding the display processing of reference materials Next, we will explain the process of displaying reference materials performed by the management server 200. Figure 8 shows an example of the reference material display flow 800. As shown in Figure 8, the link output module 218 of the management server 200 outputs a link display tag (step S810). The link display tag contains a string that expresses the title and content of the reference material in a short sentence.

[0128] In this system 100, the link output module 218 may output a link display tag for the corresponding category item in the knowledge base 300 based on the question category in which the question information is classified. In this case, link information 227 of reference materials included in the knowledge category corresponding to the question category in which the question is classified can be recommended to the respondent terminal 500B. This makes it possible to smoothly provide the respondent with the link information 227 of the reference materials that the respondent needs.

[0129] Next, the respondent terminal 500B displays a link display tag (step S820). Specifically, the display module 513 of the respondent terminal 500B displays the link display tag along with the candidate answer sentence. This allows the respondent editing the candidate answer sentence to view the reference materials necessary for the editing process. The second screen example S2 and the third screen example S3 of the respondent terminal 500B at this time will be described later.

[0130] Next, the respondent terminal 500B accepts the selection of link display tags (step S821). Specifically, the respondent selects link display tags for reference materials that they need to check when editing their candidate answer. The information of the selected link display tags is sent to the management server 200.

[0131] Next, the management server 200 outputs a link display tag for the relevant reference material (step S812). Specifically, the link output module 218 of the management server 200 refers to the link table for the link display tag sent from the respondent terminal 500B, identifies the path information of the associated reference material, and returns the path information to the respondent terminal 500B.

[0132] Next, the respondent terminal 500B displays reference materials (step S822). Specifically, the display module 513 of the respondent terminal 500B accesses the path information sent from the management server 200 and displays the corresponding reference materials on the display. The respondent edits the candidate answer sentences while reviewing the reference materials. This completes the reference material display flow 800.

[0133] (8) Processing of medical appointments and medical evaluations Next, we will explain the processing of medical appointment scheduling and medical evaluation performed by the management server 200. Figure 9 shows an example of the medical appointment scheduling and medical evaluation flow 900. In system 100, after the respondent answers the questions from the consultant as shown in Figure 6, the consultant can, for example, schedule a medical appointment as a patient at the dental clinic to which the respondent belongs and receive treatment. In addition, the patient who has received treatment can evaluate the treatment by entering their own evaluation as a review. We will now explain these medical appointment scheduling and medical evaluation processes in order.

[0134] In step S611 shown in Figure 6, the consultant, having confirmed the respondent's answer, enters information indicating their desire to make a medical appointment into the consultant terminal 500A, as shown in Figure 9 (step S910). Specifically, for example, if the consultant, having learned from the respondent's answer that their concerns about their dental health require urgent treatment, they make a medical appointment as a patient at the clinic where the respondent works. Here, in system 100, the reservation module 241 of the management server 200 refers to the patient's address and the clinic's location to recommend several candidate clinics, including the clinic where the respondent works. Candidate clinics include, for example, clinics located within a predetermined range from the patient's address. The patient can select a clinic from among the multiple candidate clinics to receive treatment. Furthermore, if the clinic has multiple dentists, the patient can also designate the dentist who answered their questions as their assigned dentist.

[0135] On the other hand, if the responding dentist is not affiliated with a clinic, the patient cannot receive treatment from that dentist. In this case, the patient may choose any clinic from the recommended list of clinics and make an appointment. Furthermore, if the patient feels uneasy about the respondent's answer, or if the clinic has poor reviews from other users, the patient may choose a different clinic from the one the respondent is affiliated with and make an appointment. Information regarding appointment preferences entered into the consultation terminal 500A is sent to the management server 200.

[0136] Next, the management server 200 accepts the reservation (step S920). Specifically, the reservation module 241 of the management server 200 obtains information regarding the reservation request sent from the consultation terminal 500A. Then, the reservation module 241 executes the reservation processing (step S921). Specifically, the reservation module 241 records the obtained reservation information as a new record in the medical appointment table in the medical information 230. At this time, the question ID from the patient (consultant) is associated with it.

[0137] Next, the output module 220 of the management server 200 outputs the reservation information to the user terminal 500A (step S922). The user terminal 500A then displays the reservation information sent from the management server 200 and presents it to the user (step S911). At this time, the user's question, the answer from the respondent, or links to navigate to other screens containing this information may be displayed along with the reservation information.

[0138] In this way, System 100 allows for the scheduling of appointments for consultations as needed, following a question-and-answer process in which respondents answer questions from individuals concerned about their health. Patients can then designate the dentist who answered their questions as their attending physician. Furthermore, the appointment information is linked to the questions and answers that motivated the visit. As a result, in System 100, patients can avoid the need to explain their dental health concerns again by being examined by the dentist who answered their questions, and can receive treatment with a sense of satisfaction. On the other hand, dentists can guide patients to their clinic through the action of answering questions, and can facilitate communication with patients by addressing their concerns during the consultation.

[0139] Next, after receiving medical treatment as a patient, the user enters a review into the user terminal 500A (step S912). Specifically, the user enters an evaluation score and review comments regarding the response of the clinic and dentist they received treatment from. The entered information is then sent from the user terminal 500A to the management server 200.

[0140] Next, the management server 200 accepts the review (step S923). Specifically, the evaluation module 242 retrieves the evaluation score and review comments sent from the client terminal 500A and records them as new records in the medical review table.

[0141] Next, the management server 200 shares the reviews (step S924). Specifically, the output module 220 of the management server 200 displays the evaluation score and review comments for the clinic being reviewed in response to display requests from other user terminals 500. In other words, prospective patients can check the evaluations of the clinic and dentist to which the respondent belongs from other users when making their appointment, and consider whether to receive treatment at the respondent's clinic or look for another clinic. From the dentist's perspective, their treatment is evaluated retrospectively, which can motivate them to provide appropriate treatment.

[0142] Furthermore, the evaluation module 242 can, for example, respond to a request for aggregated evaluation from an administrator by referring to the review information 231 and aggregating the results of reviews from multiple patients for a particular hospital. Based on the aggregated review results, the following actions may be taken, for example: • Clinics with a certain level of high ratings are given priority in being recommended as potential treatment locations. • Clinics with a rating below a certain level will not be recommended as potential treatment locations. • Providing guidance to clinics with a rating below a certain level regarding improvements to customer service during consultations. • As part of handling complaints, System 100 administrators will conduct interviews with patients. With the above steps completed, the process of scheduling appointments and evaluating medical treatments is finished.

[0143] (9) Regarding the processing of inspection information Next, we will explain the process of determining the test information performed by the management server 200. Figure 10 shows an example of the test information determination flow 1000. The system 100 has the function of calculating a score for the estimated disease based on various test information and responding with recommended actions for improvement as comments. Examples of the test information that is input include the following: • Results of the initial assessment using a home-based enzyme activity measurement kit for periodontal disease-causing bacteria • Gum color information obtained through gum chewing checks Note that the inspection information that can be determined by system 100 is not limited to these examples. The process for determining such inspection information will be explained in order.

[0144] As shown in Figure 10, the user terminal 500A first accepts the input of test information (step S1010). Specifically, the user terminal 500A launches dedicated application software and displays an input form for test information. The user enters the necessary information into the input form according to the format. For example, if the test information is the result of a primary determination using an enzyme activity measurement kit, the user enters the result of that primary determination. Also, if the test information is the color information of gum obtained from a gum chewing check, the user enters the color or an image of the gum after chewing.

[0145] Next, the management server 200 acquires the test information (step S1020). Specifically, the judgment module 219 of the management server 200 acquires the input test information and records a new record in the judgment information. The judgment module 219 then makes a judgment on the test information (step S1021). At this time, the judgment module 219 calculates a score related to the estimated symptoms by comparing the input test information with the pre-set judgment criteria. The score may be calculated by, for example, the following method. This method estimates symptom scores by inputting test information into a machine learning model that has learned the relationship between test information and symptom scores. This method uses a mathematical model that calculates symptom scores by inputting the values ​​of the acquired test information into variables, with each item of the test information being used as a variable.

[0146] Furthermore, the judgment module 219 generates prompts to be input to the generating AI based on the inspection information, and inputs them to the response generation server 400 to obtain comment information. The comment information includes, for example, the following information. • Information regarding the explanation of symptom scores • Information on recommended actions to improve the symptom score. • Information regarding suggestions for detailed examinations at medical institutions. • Information to warn about potential health conditions that may arise as symptoms progress. • Information indicating whether symptoms have improved or worsened compared to the previous examination. The judgment module 219 records the symptom score and comment information as new records in the judgment result table.

[0147] Next, the management server 200 outputs the judgment information to the client terminal 500A (step S1022). Specifically, the output module 220 outputs the symptom score and comment information included in the judgment information to the client terminal 500A. The client terminal 500A then displays the judgment information (step S1012). Specifically, the display module 513 of the client terminal 500A displays the symptom score and comment information transmitted from the management server 200 on the display, which is the output device 505. The display method of the client terminal 500A at this time will be described later. With the above steps completed, the inspection information determination process by system 100 is finished.

[0148] (10) Screen example Next, we will describe an example screen on user terminal 500 in system 100.

[0149] (10-1) First screen example S1 Figure 11 shows an example of the first screen S1 of the user terminal 500. This figure shows an example of the screen of the respondent terminal 500B where multiple answer candidate sentences are displayed in step S632 shown in Figure 6. As shown in Figure 11, in the first example screen S1, the display field C1 first shows that this question has been assigned. The content of the question is displayed in display field C2, and the question category to which the question is classified is displayed in display field C3. The numbers of the candidate answer sentences are displayed in display field C4, and by selecting a number, the corresponding candidate answer sentence can be displayed. The selected candidate answer sentence is displayed in display field C5. The respondent can review multiple candidate answer sentences and select the answer sentence they wish to use. Display field C5 can be moved downwards by scrolling the screen upwards. The respondent can also perform any editing on the candidate answer sentences displayed in text data format in display field C5. Alternatively, the respondent is not required to perform any editing.

[0150] (10-2) Second screen example S2 Figure 12 shows an example of the second screen S2 of the user terminal 500. This screen is located below the first example screen S1. The second example screen is displayed by scrolling upwards from the first example screen S1.

[0151] As shown in Figure 2, in the second example screen S2, below the display field C5 for suggested answer sentences, which are a continuation from the first example screen, there is a display field C6 for links to recommended reference materials that are recommended to be reviewed when answering. The link display tags show the same categories as the question categories shown in Figure 11. Therefore, by reviewing the recommended reference materials, it becomes possible to efficiently edit the suggested answer sentences.

[0152] Furthermore, below the display field C6 for the recommended reference materials link display tag, there is a dropdown menu C7 for other reference materials. The display method when the dropdown menu is selected is explained in Figure 13.

[0153] Additionally, below the display area C7 for other reference materials, a screen transition button C8 is displayed. When the respondent presses the screen transition button C8, the editing of the candidate answer is considered complete, and the respondent is redirected to the final check screen before submitting the answer.

[0154] (10-3) Example of the third screen S3 Figure 13 shows an example of the third screen S3 of the user terminal 500. This figure shows an example of the screen of the respondent terminal 500B where the link display tag is displayed in step S820 shown in Figure 8. As shown in Figure 13, in the third example screen S3, clicking the "Other Reference Materials" section C7 displays a list of knowledge categories for the reference materials stored in the knowledge base 300 in the "Details" section C9. All knowledge categories are displayed in the "Details" section C9, and all can be selected by scrolling the screen to navigate between screens. Therefore, the respondent is not limited to the same category as the automatically selected question category, but can check reference materials for all knowledge categories they wish to review.

[0155] (10-4) Fourth screen example S4 Figure 14 shows an example of the fourth screen S4 of the user terminal 500. This figure is the first example of the consultant terminal 500A where the judgment information is displayed in step S1012 shown in Figure 10. This screen example shows the screen when the results of the primary judgment using the enzyme activity measurement kit for periodontal disease causative bacteria are entered. As shown in Figure 14, in the fourth screen example S4, the assessment information D1 displays the case score, explanation, and recommended actions as comments. The person seeking advice can review this information to check their risk of developing periodontal disease, and if necessary, modify their lifestyle or undergo a thorough examination at a medical institution, enabling prompt medical intervention for the person seeking advice.

[0156] (10-5) Example of the 5th screen S5 Figure 15 shows an example of the fifth screen S5 of the user terminal 500. This figure is a second example of the consultant terminal 500A where the judgment information is displayed in step S1012 shown in Figure 10. This screen example shows the screen when the results of the gum chewing check are entered. As shown in Figure 15, in the fifth example screen S5, the case score and explanation are displayed in the judgment information D2. By reviewing this information and objectively understanding their own chewing condition, the user can, for example, increase their awareness of the importance of chewing in their daily meals.

[0157] (11) Variant Next, a modified example of system 100 will be described. In the embodiment described above, the respondent selected one answer from several candidate sentences, but this is not limited to this configuration. For example, the respondent may select multiple answer candidates to integrate those sentences and then perform any necessary editing.

[0158] Furthermore, the link output module 218 may output a link display tag for the corresponding category item from the knowledge base 300 based on the generated candidate answer information. In other words, the system 100 may further categorize the candidate answer information output from the answer generation server 400, and based on the resulting categorization, select the knowledge category of the reference material to recommend to the respondent.

[0159] Furthermore, although the above embodiment shows a configuration in which the display module 513 of the respondent terminal 500B displays the address of the person who asked the question along with the question text on the respondent terminal 500B, this is not limited to this configuration. In other words, the system 100 may select respondent candidates to be asked whether they can answer the question based on the address of the person who asked the question. Specifically, the classification module 212 of the management server 200 may refer to the address of the person who asked the question for the question it received, compare it with the address of the hospital to which the respondent belongs, and display the question only to respondents who belong to a hospital within a predetermined range that they can visit. Such selection of question texts based on address may also be performed by comparing the address of the person who asked the question with the address of the respondent.

[0160] Furthermore, system 100 may select potential answerers to ask whether they can answer the question based on information about the consultant stored in the consultant table, or other information. Examples of such selection processes for potential answerers to ask whether they can answer the question include the following: • Based on the patient's medical history, a process is used to select potential respondents who belong to a clinic specializing in that particular condition. • The process involves estimating the age group of the person seeking advice based on their date of birth, and then selecting potential respondents who belong to clinics specializing in specific age groups, such as pediatric dentistry or implants. • Refer to review information 231 to select potential respondents from highly-rated clinics.

[0161] Furthermore, in the system 100, when recommending potential medical clinics, the system may also select potential medical clinics to recommend to the consultant based on information about the consultant stored in the consultant table, or other information.

[0162] Furthermore, while the above embodiment shows an example of applying System 100 to various consultation services related to dental care, it is not limited to this. System 100 may also be applied to consultation services related to medical care other than dentistry, or to consultation services in various fields requiring a certain level of expertise, such as legal consultations, various administrative procedures, and real estate transactions.

[0163] Furthermore, System 100 may also integrate with My Number Portal (Registration Act), an online personal information inquiry system operated by the government, via API to obtain user health information linked to My Number. Examples of health information linked to My Number include the following: • Information about the user's medical history (including dental care) • Past medical records (including dental care) • Past prescription information • Past health checkup information • Information regarding lifestyle habits, etc., that the individual has previously reported during medical examinations or health checkups.

[0164] System 100 allows users to view their health information linked to their My Number (Social Security Number). System 100 may also use the user's health information linked to their My Number to select potential respondents to answer questions. Furthermore, System 100 may use the user's health information linked to their My Number to select potential clinics to recommend to patients. Specifically, for a patient with a history of diabetes who is considering tooth extraction, a dentist with expertise in tooth extraction for diabetic patients may be selected as a respondent, and the clinic where that dentist works may be selected as a potential clinic and recommended to the patient.

[0165] Furthermore, System 100 allows users to designate a primary care physician. In other words, System 100 allows users to register their favorite respondents from among those who have answered their questions in the past. Specifically, when a user asks a new question, they can specify a respondent they have already registered as a favorite respondent. In this case, the question will, in principle, be assigned to one of the favorite respondents (if there are multiple favorite respondents). Respondents can also check the number of people who have registered them as a favorite respondent, allowing them to quantitatively understand the level of trust they have earned from users and showcase their past consultation experience. The number of registrations is also shared with other users and used in selecting respondents.

[0166] Furthermore, System 100 allows users to set a primary care physician. Specifically, System 100 allows users to register their favorite clinics or dentists from among those they have previously received treatment from as their favorite physicians. In this case, when a patient makes a new appointment, they can receive recommendations for potential clinics limited to those registered as their favorite physicians. Clinics or dentists can also check the number of patients who have registered their clinic as their favorite physician, allowing them to quantitatively understand the level of trust patients have in them. The number of registrations is also shared with other users and used to help them select physicians.

[0167] Furthermore, in System 100, the questioner can select the attributes of the respondent. In other words, the questioner can narrow down the list of potential respondents by specifying one of the following items. • Respondent's position • Respondent's assigned category • The hospital to which the respondent belongs • Number of responses from respondents • Number of respondents who have favorited the item • Information included in review information 231 of the respondent's past answers

[0168] Furthermore, in System 100, respondents can select the attributes of the person seeking advice. In other words, respondents can pre-define the attributes of the person seeking advice for the question they are responsible for answering. Examples of such attributes include the following: • Past question categories of the person seeking advice • Presence or absence of fraudulent review flags in the response review table • Presence or absence of fraudulent review flags in the clinical review table • Information included in review information 231 previously entered by the user

[0169] Furthermore, the generation AI module 411 in system 100 may be retrained using past answers to questions from previous clients. This allows the generation AI module 411 to learn not only from already learned information but also from the latest dental medical knowledge compiled by the respondent, thereby improving the accuracy of its answers. Since the generation AI module 411, having been retrained on the latest dental medical knowledge, generates candidate answers, the respondent can broaden their own knowledge by taking on the role of answering.

[0170] Furthermore, in system 100, the respondent may ask questions to the generation AI module 411. That is, when editing the candidate answer sentences, the respondent may ask specific questions to the generation AI module 411 to confirm the information necessary for the answer, and then review the answers and reflect them in the editing process. In this case, for example, the questions entered by a respondent who is a dentist have a higher level of detail than the questions entered by a consultant who is not a dental professional, so more specific information can be obtained from the generation AI module 411, thereby improving the convenience for the respondent.

[0171] Furthermore, in System 100, multiple clinics may propose different treatment plans to a patient. In other words, in dental care, for example, in cosmetic dentistry such as implant treatment, patients have more options for choosing private treatment compared to general medical care. For this reason, various treatment options are set by different clinics, and the price range varies depending on the materials used in the treatment. In System 100, when recommending potential clinics when a patient makes an appointment, in addition to information about the clinic, the system can also propose treatment plans (treatment content and price) that the clinic can provide. The patient can review the multiple treatment plans recommended for their health concerns, select a clinic that proposes a treatment plan that meets their needs, and make an appointment. In other words, System 100 allows multiple clinics to compete for treatment plans for a patient. This is expected to broaden the range of medical plan choices available to patients and, along with shared review information 231, to improve the competitiveness of dental clinics.

[0172] (12) Others It should be noted that the present invention is not limited to the embodiments described above, and various modifications are included. For example, the embodiments described above are described in detail to make the present invention easier to understand, and are not necessarily limited to those having all the configurations described. Furthermore, it is possible to replace parts of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add configurations from other embodiments to the configuration of one embodiment. In addition, it is possible to add, delete, or replace parts of the configuration of each embodiment with other configurations.

[0173] Furthermore, each of the above configurations, functions, processing units, and processing means may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. Alternatively, each of the above configurations and functions may be implemented in software by having the processor interpret and execute programs that implement each function. Information such as programs, tables, and files that implement each function can be stored in memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0174] Furthermore, the control lines and information lines shown are those deemed necessary for explanatory purposes, and not all control lines and information lines are necessarily shown in the actual product. In reality, it is safe to assume that almost all components are interconnected. Furthermore, the above-described embodiments disclose at least the configuration described in the claims. [Explanation of Symbols]

[0175] 100...Question answering system, 200...Management server, 201...Main memory, 211...Question acquisition module, 212...Classification module, 213...Assignment module, 214...Prompt generation module, 215...Answer acquisition module, 216...Selection acceptance module, 217...Editing module, 218...Link output module, 219...Determination module, 220...Output module, 202...Auxiliary memory, 221...Consultant information, 222...Respondent information, 223...Category information, 224...Question text information, 225...Answer candidate text information, 226...Answer text information, 227...Link information, 228...Determination information, 400...Answer generation server, 401...Main memory, 411...Generation AI module, 500...User terminal, 600...Knowledge base

Claims

1. A question acquisition means that acquires question information entered into the first terminal, An answer acquisition means for acquiring multiple candidate answer information generated by a generation AI based on the aforementioned question information, An output means for outputting the acquired multiple candidate answer information together with the question information to a second terminal, A question answering system comprising: a selection receiving means for receiving the selection of at least one of the multiple candidate answer information from the second terminal.

2. The device further comprises editing means for editing the selected candidate answer information based on editing operations entered into the second terminal. The question answering system according to claim 1, wherein the output means outputs the answer text information obtained by editing the answer candidate information to the first terminal.

3. The system further comprises a classification means for classifying the acquired question information into predetermined categories, The question answering system according to claim 2, wherein the output means outputs the question information and category information indicating the category to which the question information is classified to the second terminal before the answer acquisition means acquires a plurality of the candidate answer information.

4. The question answering system according to claim 3, wherein the output means outputs to the second terminal information relating to the address of the person who entered the question information, along with the question information and the category information.

5. The system further includes an assignment means for receiving an assignment request from the second terminal, which outputs the aforementioned question information, the aforementioned category information, and the information relating to the address of the person seeking advice, for requesting the assignment of a respondent to the said question information. The question answering system according to claim 4, wherein the answer acquisition means acquires a plurality of answer sentences based on the assignment process to the question by the assignment means.

6. The system further comprises prompt generation means for generating prompts to be input to the generating AI based on the aforementioned question information, The question answering system according to claim 1 or 2, wherein the answer acquisition means inputs the prompt to the generating AI, thereby acquiring a plurality of the answer candidate information generated by the generating AI.

7. The system further comprises a classification means for classifying the acquired question information into predetermined categories, The question answering system according to claim 6, wherein the prompt generation means generates the prompt based on the question information and category information indicating the category in which the question information is classified.

8. The system further includes a link output means that outputs link information that allows access to a knowledge base containing reference materials referenced in the aforementioned editing operation. The question answering system according to claim 3, wherein the link output means outputs the link information relating to the corresponding category item in the knowledge base based on the category information in which the question information is classified.

9. The system further includes a link output means that outputs link information that allows access to the knowledge base used in the aforementioned editing operation. The question answering system according to claim 3, wherein the link output means outputs the link information relating to the corresponding category item in the knowledge base based on the generated candidate answer information.

10. To the first terminal from which the aforementioned response information was output, the system recommends information on multiple hospitals, including the hospital to which the respondent who edited the response information belongs, as potential hospitals for treatment. The question answering system according to claim 2, further comprising a reservation means for making a medical appointment for the hospital selected on the first terminal.

11. The system acquires review information regarding the medical treatment entered into the first terminal by a user who received medical treatment at the aforementioned hospital. The question answering system according to claim 10, further comprising an evaluation means for performing aggregate evaluation of the acquired review information.

12. The system further includes a determination means that makes a determination on the results of the test based on the user's test information entered into the first terminal, The question answering system according to claim 1, wherein the output means outputs determination information indicating the content of the determination to the first terminal.

13. The determination means acquires comment information generated by the AI ​​based on the inspection information, The question answering system according to claim 12, wherein the output means further outputs the comment information to the first terminal.

14. The judgment information includes a score of the user's symptoms based on the test information. The question-answering system according to claim 13, wherein the comment information includes information on recommended actions for improving the symptoms.

15. A method of answering that a computer processor performs, The above response method is performed by the processor, A question acquisition step that acquires question information entered into the first terminal, The answer acquisition step involves obtaining multiple candidate answer information generated by a generation AI based on the aforementioned question information, An output step of outputting the acquired multiple candidate answer information together with the question information to a second terminal, A response method comprising: a selection acceptance step of receiving the selection of at least one of the multiple candidate response information from the second terminal; and a response method.

16. An answer program that is executed by a computer processor, The aforementioned response program will provide the processor with the following: A question acquisition step that acquires question information entered into the first terminal, The answer acquisition step involves obtaining multiple candidate answer information generated by a generation AI based on the aforementioned question information, An output step of outputting the acquired multiple candidate answer information together with the question information to a second terminal, A response program that causes the second terminal to perform a selection acceptance step of accepting the selection of at least one of the multiple candidate response information.

Images