Answer generation system, answer control device and program

Abstract

Providing answers that are optimized for each user. [Solution] The answer generation system comprises a user terminal used by a user, an answer control device that generates personal classification information for the user based on the user's date of birth and multiple divination methods, and generates prompts based on questions from the user and the user's personal classification information generated on the user terminal, and an answer generation device that generates answers to questions based on the prompts.

Description

[Technical Field]

[0001] The present invention relates to an answer generation system, an answer control device, and a program. [Background technology]

[0002] In recent years, AI (Artificial Intelligence) has been used in a variety of fields. AI includes predicting future information through pattern recognition and data analysis, as well as generative AI, which generates new information based on existing information.

[0003] For example, Patent Document 1 discloses an AI answering system that generates answers to inquiries as a system using a generation AI. Specifically, Patent Document 1 describes that the AI ​​answering system has a learning unit that has undergone machine learning using inquiries, general information, answers, and satisfaction indices related to the answers, and an answer generating unit that generates answers to be sent to multiple mobile devices in response to inquiries from the multiple mobile devices based on the learning results of the learning unit. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 7393725 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the answers generated in Patent Document 1 are generalized, and it may not be possible to provide an appropriate answer for each user.

[0006] Therefore, an object of the present invention is to provide an answer that is optimized for each user. [Means for solving the problem]

[0007] An answer generation system according to a representative embodiment of the present invention comprises a user terminal used by a user, an answer control device that generates personal classification information for the user based on the user's date of birth and multiple divination methods, and generates prompts based on questions from the user and the user's personal classification information generated on the user terminal, and an answer generation device that generates answers to questions based on the prompts. [Effects of the Invention]

[0008] According to the present invention, it is possible to provide answers that are optimized for each user. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of an overview of a response generation system according to a first embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating a configuration of a response control device. [Figure 3] FIG. 1 is a diagram illustrating a configuration of a response generation device. [Figure 4] FIG. 10 is a diagram illustrating an example of the configuration of a user terminal. [Figure 5] FIG. 2 is a flow diagram illustrating a response generation method according to the first embodiment of the present invention. [Figure 6] FIG. 10 is a flow diagram relating to account registration. [Figure 7] FIG. 10 is a flow diagram relating to the registration of a target person for whom compatibility diagnosis with a user is to be performed. [Figure 8] FIG. 10 is a diagram illustrating an example of a response in Example 3. DETAILED DESCRIPTION OF THE INVENTION

[0010] (Embodiment 1) Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all drawings for explaining the embodiments, the same components are generally designated by the same reference numerals, and repeated description thereof will be omitted as appropriate.

[0011] <Answer Generation System Overview> The answer generation system according to this embodiment is a system that uses a generation AI to generate answers to questions sent from a user terminal. The answer generation system generates the user's personal classification information (also referred to as "user classification information") using multiple divination techniques based on user information including the user's date of birth. In addition, a prompt to be input to the generation AI is generated based on the user's question and the user's personal classification information sent from the user terminal. Then, the generation AI generates an answer to the question based on the prompt. This makes it possible to provide an answer that is optimized based on the user's characteristics.

[0012] The answer generation system according to this embodiment can be applied to fields such as human resource recruitment support (career aptitude test, interview design, turnover rate prediction, etc.), education and career support (study style suggestions, optimization of teaching methods, etc.), love consultation and matchmaking (compatibility test, advice based on emotional tendencies (personality tendencies)), etc. However, the answer generation system according to this embodiment can also be applied to fields other than these.

[0013] 1 is a diagram illustrating an example of an overview of an answer generation system according to Embodiment 1 of the present invention. As shown in FIG. 1, the answer generation system 1 includes an answer control device 10, an answer generation device 20, and a user terminal 30.

[0014] The components of the answer generation system 1 are connected via a network NET, and information (data) is sent and received between the components. The network NET is a WAN (Wide Area Network) such as the Internet, or a LAN (Local Area Network).

[0015] The answer control device 10 is a terminal used by the administrator of the answer generation system 1. The answer control device 10 is an information processing terminal equipped with a communication function, such as a personal computer. The answer control device 10 performs various processes related to generating answers to questions from users. Specifically, the answer control device 10 performs processes such as accepting questions from users, generating prompts to cause the answer generation device 20 to generate answers to the user's questions, receiving answers generated by the answer generation device 20 and providing them to the user, and saving user information.

[0016] The answer generation device 20 is a device that generates answers to questions from users. The answer generation device 20 is an information processing terminal with a communication function, such as a personal computer. The answer generation device 20 is equipped with a generation AI, and generates answers to user questions by inputting prompts into an answer generation model of the generation AI.

[0017] The user terminal 30 is a terminal used by a user who uses the answer generation system. The user terminal 30 is an information processing terminal equipped with a communication function, such as a smartphone or a personal computer. The user uses the user terminal 30 to send a question to the answer control device 10 and receives an answer generated by the answer generation device 20 for the question.

[0018] <<Answer control device>> 1, the answer control device 10 includes functional blocks such as a question processing unit 110, a personal classification information generation unit 120, a prompt generation unit 130, and an answer processing unit 140. These functional blocks may be configured using software, or may be configured using a combination of hardware and software.

[0019] The question processing unit 110 performs processes such as receiving and saving questions from users sent from the user terminal 30. The individual classification information generation unit 120 generates individual classification information by classifying the characteristics of individuals (including users and people whose compatibility with the user is being diagnosed) using divination techniques based on the birth date of the users, etc. The individual classification information may include information such as the individual's emotional tendencies, thought patterns, luck, values, behavior patterns, risk tolerance, etc.

[0020] The prompt generation unit 130 generates a prompt based on a question from the user and the user's personal classification information. The prompt generation unit 130 may, for example, tag and / or digitize each piece of information included in the personal classification information and generate a prompt by incorporating the tagged and / or digitized information. In this case, the prompt may be configured, for example, by combining a template with each piece of tagged and / or digitized information. Each piece of information included in the personal classification information may include, for example, an individual's emotional tendencies, thought patterns, luck, values, behavioral patterns, risk tolerance, etc. Furthermore, when performing compatibility diagnosis with a user, the prompt generation unit 130 generates a prompt based on the question from the user, the user's personal classification information, and the personal classification information of the person whose compatibility diagnosis with the user is to be performed.

[0021] A prompt is natural language-based input data for controlling the output (answer) of the generation AI (answer generation device 20). Specifically, a prompt can be said to be a specification for specifying to the generation AI "under what conditions and context you want the answer to be given."

[0022] The prompt is input to the answer generation model 220 included in the answer generation device 20, and instructs the generation of an answer to the question from the user. The answer processing unit 140 performs processes such as receiving answers to the user's questions generated in the answer generation device 20, transmitting the answers to the user terminal 30, and saving the answers.

[0023] Fig. 2 is a diagram illustrating the configuration of a response control device 10. As shown in Fig. 2, the response control device 10 includes a processor 11, a main memory 13, a storage 15, and a communication interface 17. The response control device 10 may also include a monitor 19 as necessary.

[0024] The communication interface 17 is a communication device that transmits and receives various information between the answer generation device 20 and the user terminal 30 via the network NET.

[0025] The storage 15 includes various storage areas, such as a program storage area 15a, an account information storage area 15b, a fortune-telling information storage area 15c, a personal classification information storage area 15d, a question storage area 15e, a prompt storage area 15f, and an answer storage area 15g.

[0026] The program storage area 15a stores various programs such as a basic program such as an OS (Operating System) that operates the answer control device 10, and a program that causes the processor 11 to realize the functional blocks of the answer generation system 1. The program storage area 15a may also store various programs such as an API (Application Programming Interface) that links with the user terminal 30. The program storage area 15a also stores parameters for each program. In this way, the storage 15 functions as a program recording medium.

[0027] The account information storage area 15b stores account information of users registered in the answer generation system 1. The user's account information may include, for example, a user ID, login information (such as a password), login history, and various information input by the user as user information, such as the user's name, contact information (such as an email address or phone number), date of birth, time of birth, place of birth, and gender. The account information storage area 15b also stores, as target person information, information such as the name and date of birth of the target of the compatibility diagnosis described above.

[0028] The divination information storage area 15c stores information (divination information) related to the divination methods used to generate the individual classification information. In this embodiment, for example, the individual's date of birth and six types of divination methods are used to generate the individual classification information. The six types of divination methods are, for example, four pillars of destiny, nine star kigaku, Shukuyo astrology, Western astrology, animal fortune telling, and Yin Yang and the five elements.

[0029] The personal classification information storage area 15d stores personal classification information (user classification information) of each user generated based on the user information and fortune-telling technique in the answer control device 10. Each piece of personal classification information is stored in association with the corresponding user. Each piece of personal classification information may include information such as the emotional tendencies, thought patterns, fortune, values, behavioral patterns, and risk tolerance of the corresponding individual.

[0030] Furthermore, the personal classification information storage area 15d stores classification information of a person who is the target of a compatibility diagnosis with a user (also referred to as "target person classification information"). The target person classification information is also stored in the personal classification information storage area 15d in association with the corresponding user.

[0031] The question storage area 15e stores questions from each user received from the user terminal 30. Each question is stored in association with the corresponding user.

[0032] The prompt storage area 15f stores a prompt generated for each question generated by the answer control device 10. Each prompt is stored in association with either the corresponding question or the corresponding user.

[0033] The answer storage area 15g stores answers received from the answer generation device 20. Each answer is stored in association with at least one of the corresponding question, the corresponding prompt, and the corresponding user.

[0034] The main memory 13 temporarily stores various information such as programs and parameters read from the program storage area 15a, calculation results in the processor 11, questions from the user terminal 30, personal classification information, prompts, and answers to questions. The main memory 13 also temporarily stores various information transmitted and received via the communication interface 17.

[0035] The processor 11 reads and executes programs stored in the main memory 13, thereby configuring the functional blocks that drive the components of the answer control device 10, the functional blocks that realize the answer generation system 1 (FIG. 1), etc., with software, and executes various processes related to the answer generation system 1. Note that some of the components and functions of the answer control device 10 may be configured using cloud services, or may be configured by combining hardware and software.

[0036] <<Answer generation device>> 1, the answer generation device 20 includes functional blocks such as a prompt processing unit 210 and an answer generation model 220. These functional blocks may be configured using software, or may be configured using a combination of hardware and software.

[0037] The prompt processing unit 210 performs processes such as receiving and storing prompts sent from the answer control device 10 .

[0038] When a prompt is input, the answer generation model 220 generates an answer based on the prompt as text data. The answer generation model 220 is a generation AI equipped with a text generation model such as a large language model (LLM). Note that the answer generation model 220 may be configured as a voice generation model that generates an answer based on the prompt as voice data, or may be configured as a combination of a text generation model and a voice generation model.

[0039] The answer generation model 220 adjusts the wording of the answer according to the personal classification information of the user who asked the question. For example, the answer generation model 220 generates an answer using logical and sociable language for a user classified as a "partial wealth type" according to Four Pillars of Destiny, generates an answer using language that emphasizes order, procedure, and a sense of security for a user classified as a "direct official type," and generates an answer using language that indicates originality, freedom, and the future for a user classified as a "scarred official type." The answer generation model 220 can also adjust the language of the answer according to types other than these in Four Pillars of Destiny and types classified by divination methods other than Four Pillars of Destiny.

[0040] The answer processing unit 230 performs processes such as sending answers to questions from users that are generated in the answer generation device 20 and saving the answers.

[0041] Fig. 3 is a diagram illustrating the configuration of an answer generation device. As shown in Fig. 3, the answer generation device 20 includes a processor 21, a main memory 23, a storage 25, and a communication interface 27. The answer generation device 20 may also include a monitor 29 as necessary.

[0042] The communication interface 27 is a communication device that transmits and receives various information between the response control device 10 and the user terminal 30 via the network NET.

[0043] The storage 25 includes various storage areas, such as a program storage area 25a, an answer generation model storage area 25b, a prompt storage area 25c, and an answer storage area 25d.

[0044] The program storage area 25a stores various programs such as a basic program such as an OS (Operating System) that operates the answer generation device 20, and a program that causes the processor 21 to realize the functional blocks of the answer generation system 1. The program storage area 25a may also store various programs such as an API (Application Programming Interface) that links with the user terminal 30. The program storage area 25a also stores parameters for each program. In this way, the storage 25 functions as a program recording medium.

[0045] The answer generation model storage area 25b stores models that constitute the answer generation model 220 in the processor 21. The prompt storage area 25c stores prompts received from the answer control device 10. The answer storage area 25d stores answers generated by the answer generation device 20. Each answer is stored in association with either the corresponding prompt or the corresponding user.

[0046] The main memory 23 temporarily stores various information such as programs and parameters read from the program storage area 25a, calculation results in the processor 21, information related to prompts, information related to answer generation, etc. The main memory 13 also temporarily stores various information transmitted and received via the communication interface 27.

[0047] The processor 21 reads and executes programs stored in the main memory 23, thereby configuring, with software, functional blocks that drive the components of the answer generation device 20, functional blocks that realize the answer generation system 1 (FIG. 1), etc., and executes various processes related to the answer generation system 1. Note that some of the components and functions of the answer generation device 20 may be configured using a cloud service, or may be configured by combining hardware and software.

[0048] <<User terminal>> Fig. 4 is a diagram illustrating the configuration of a user terminal. As shown in Fig. 4, the user terminal 30 includes a processor 31, a main memory 33, a storage 35, a communication interface 37, a camera 38, a microphone 39, and a display unit 36. The communication interface 37 is a communication device that communicates with the answer control device 10 and the answer generation device 20 via a network NET.

[0049] The storage 35 includes various storage areas such as a program storage area 35a, an account information storage area 35b, a question storage area 35c, and an answer storage area 35d.

[0050] The program storage area 35a stores basic programs such as an OS that runs the user terminal 30, and applications that cause the processor 31 to implement the functional blocks of the answer generation system 1. The program storage area 35a may also store various programs such as an API (Application Programming Interface) that links with the answer control device 10. The program storage area 35a also stores parameters corresponding to each program. In this way, the storage 35 functions as a program recording medium.

[0051] The account information storage area 35b stores account information of the user who uses this user terminal 30. The account information may include various information such as a user ID, login information, login history, the user's name, contact information (e.g., email address, phone number, etc.), and date of birth.

[0052] The question storage area 35c stores questions from users generated on the user terminal 30. The question storage area 35c may store questions generated on the user terminal 30. This allows new questions to be created using past questions stored in the question storage area 35c.

[0053] The answer storage area 35d stores answers received from the answer control device 10. Each answer is stored in association with the corresponding question.

[0054] The main memory 33 stores various programs read from the program storage area 35a and corresponding parameters, etc. The main memory 33 also temporarily stores information to be transmitted and received via the communication interface 37, etc.

[0055] The processor 31 reads and executes programs and parameters stored in the main memory 33, thereby configuring, in software, functional blocks that drive the components of the user terminal 30, functional blocks that realize the answer generation system 1, etc. Note that some of the functional blocks may be configured in hardware, or may be configured in combination with software and hardware.

[0056] The camera 38 and microphone 39 can be used, for example, when sending a question using video and audio.

[0057] The display unit 36 ​​displays a user interface for operating the terminal, screens related to the answer generation system 1 (for example, an account registration screen, a login screen, a question input screen, and an answer display screen), etc. The display unit 36 ​​may also have a touch input function, which allows the user to perform desired input operations by touching the display screen. The input device may also be provided separately from the display unit 36.

[0058] <Divination> As already mentioned, divination techniques are used to generate the personal classification information. Here, the divination techniques used in this embodiment will be explained. Note that, although six types of divination techniques are exemplified here, the divination techniques used to generate the personal classification information are not limited to these.

[0059] <<Four Pillars of Destiny>> Four Pillars of Destiny is a traditional divination method originating in China that predicts an individual's fortune, personality, career aptitude, etc. based on the date and time of birth. The "Four Pillars" refer to the four pillars of the year, month, day, and hour, and fate is predicted by combining each of the "Four Pillars" with the "Heavenly Stems" and "Earthly Branches."

[0060] The Heavenly Stems are the ten heavenly stems: Jia, Yi, Bing, Ding, Wu, Ji, Geng, Xin, Ren, and Gui. The Earthly Branches are the twelve earthly branches: Rat, Ox, Tiger, Rabbit, Dragon, Snake, Horse, Sheep, Monkey, Rooster, Dog, and Pig. The five elements (wood, fire, earth, metal, and water) are added to the Heavenly Stems.

[0061] Four Pillars of Destiny has a statistical aspect and is more systematic than other fortune-telling methods. In addition, Four Pillars of Destiny can take into account the time of birth, so the fortune-telling results will differ even for people with the same birthday if the time of birth is different.

[0062] <<Nine Star Ki Studies>> Kyusei Kigaku is a divination method that uses the nine stars of the year of birth to determine personality and fortune, and utilizes compass direction to improve one's luck. The nine stars are "Water Star No. 1 by White," "Earth Star No. 2 by Black," "Wood Star No. 3 by Green," "Wood Star No. 4 by Green," "Earth Star No. 5 by Yellow," "Vegetable Star No. 6 by White," "Vegetable Star No. 7 by Red," "Earth Star No. 8 by White," and "Fire Star No. 9 by Purple." Each person belongs to one of these nine stars. The nine stars are also classified into the five elements (wood, fire, earth, metal, and water) of energy.

[0063] << Shukuyo Astrology >> Shukuyo astrology is a divination method originating in India that predicts a person's personality, fortune, and compatibility based on the 27 constellations known as the "Shukuyo" star constellations. It is deeply connected to Buddhism and Onmyodo, and is particularly effective at diagnosing compatibility in human relationships. By utilizing it in business, it can help with team formation and partner selection, taking into account compatibility with others.

[0064] <<Western Astrology>> Western astrology is a form of divination based on the idea that the movement of celestial bodies influences a person's personality and destiny. Western astrology predicts personality, fortune, suitable career, and compatibility based on the position of the celestial bodies (horoscope) at the moment of birth. Western astrology can take into account the date, time, and place of birth.

[0065] <<Animal Fortune Telling>> "Animal Fortune Telling" is a uniquely Japanese form of fortune telling based on the Four Pillars of Destiny. Based on the person's date of birth, the person is classified into one of 12 animal characters, and their personality and compatibility are diagnosed. The 12 animal characters are wolf, rooster, monkey, cheetah, leopard, lion, tiger, raccoon, koala, elephant, sheep, and Pegasus. Each animal character is further subdivided into a total of 60 personality types.

[0066] <<Yin-Yang Five Elements>> "Yin-Yang and the Five Elements" is a system based on ancient Chinese philosophy that explains the principles of the natural world and human society. It is a combination of the "Yin-Yang ideology" and the "Five Elements ideology," and is used in a variety of fields, including traditional Eastern medicine, astrology, feng shui, calendars, martial arts, and political thought.

[0067] Yin and Yang is a concept that explains all things in terms of two opposing yet complementary forces: "yin" and "yang." "Yin" is a force that embodies concepts such as dark, cold, static, feminine, and introverted (e.g., the moon, night, and water). On the other hand, "yang" is a force that embodies concepts such as bright, warm, dynamic, masculine, and extroverted (e.g., the sun, day, and fire). "Yin" and "yang" are not just in opposition to each other; they also influence each other, maintaining a balance that is said to be what makes up the world.

[0068] The five elements are the five basic elements that make up all things: wood, fire, earth, metal, and water. Their mutual relationships—mutual generation and mutual destruction—are described as follows: Wood represents the quality of growth and expansion, the season of spring, the direction of the east, and the color of blue / green. Fire represents the quality of heat and rising, the season of summer, the direction of the south, and the color of red. Earth represents the quality of stability and center, the season of the earthly life, the direction of the center, and the color of yellow. Metal represents the quality of contraction and hardness, the season of autumn, the direction of the west, and the color of white. Water represents the quality of cooling and falling, the season of winter, the direction of the north, and the color of black. The mutual relationship of mutual generation indicates a mutually beneficial relationship, such as when wood gives birth to fire. "Mutual destruction" indicates a relationship of mutual inhibition and opposition, such as "water extinguishes fire."

[0069] <How to use the answer generation system> Next, a method for generating an answer to a question from a user using the answer generation system 1 will be described. Fig. 5 is a flow diagram illustrating the answer generation method according to the first embodiment of the present invention. Fig. 5 shows the steps from creating a question in the user terminal 30 to displaying the answer on the user terminal 30.

[0070] First, in step S10, the user logs in or registers an account. If account registration has already been completed, the user logs in to the answer generation system 1 by entering a user ID and login information on the login screen displayed on the user terminal 30.

[0071] On the other hand, a user who has not registered an account will register an account. Fig. 6 is a flow diagram related to account registration. The user operates the user terminal 30 to input user information into a predetermined account registration screen and transmits it to the response control device 10 (step S110). The user information may include various information such as the user's name, contact information (e.g., email address, telephone number, etc.), date of birth, time of birth, place of birth, gender, etc.

[0072] Then, the response control device 10 determines whether there are any errors in the input contents (step S120). For example, required input items may be set, and whether there are any errors in the input contents may be determined based on whether the input into the required input items has been completed.

[0073] If there are no errors in the input details (step S120, Yes), a user ID and login information (password, etc.) are issued and an account for this user is registered (step S130). In addition, this information and user information are saved as account information.

[0074] On the other hand, if it is determined in step S120 that the input contents are incomplete (step S120, No), the user is notified that the input contents are incomplete (step S140), and the process returns to step S110 to correct the input contents. Note that the response control device 10 may forcibly terminate the account information registration process if the account information registration cannot be completed within a predetermined number of repetitions or within a predetermined time from the start of input of the account information.

[0075] The answer control device 10 generates personal classification information for a user who has registered an account (step S150). The personal classification information generation unit 120 generates personal classification information (user classification information) for the user based on the user's date of birth and multiple divination techniques (for example, the six divination techniques mentioned above). The generated personal classification information is saved in association with the user's account information. Note that the generation of personal classification information may be performed when a question from the user is received or when a prompt is generated.

[0076] In step S20, a target person for whom compatibility diagnosis with the user is to be performed is registered. Note that registration of a target person for whom compatibility diagnosis with the user is to be performed can be performed at any time. FIG. 7 is a flow diagram related to the registration of a target person for whom compatibility diagnosis with the user is to be performed. The user operates the user terminal 30 to input target person information into a predetermined target person registration screen and transmits it to the answer control device 10 (step S210). The target person information may include various information such as the target person's name, date of birth, place of birth, gender, etc. Note that a target person registration field may be provided on the account registration screen.

[0077] Then, the response control device 10 determines whether there are any errors in the input contents (step S220). For example, required input items may be set, and whether there are any errors in the input contents may be determined based on whether the input into the required input items has been completed.

[0078] If there are no errors in the input contents (step S220, Yes), the target person information of this target person is linked to the corresponding user and registered and saved (step S230). In this way, the target person is registered.

[0079] On the other hand, if it is determined in step S220 that the input contents are incomplete (step S220, No), the user is notified that the input contents are incomplete (step S240), and the process returns to step S210 to correct the input contents. Note that the answer control device 10 may forcibly terminate the registration process of the target person information if the registration of the target person information cannot be completed within a predetermined number of repetitions or within a predetermined time from the start of input of the target person information.

[0080] The response control device 10 generates personal classification information for the registered target person (step S250). The personal classification information generation unit 120 generates personal classification information (target person classification information) for this target person based on the target person's date of birth and multiple divination methods (for example, the six divination methods mentioned above). The generated personal classification information is stored in association with the corresponding user. Note that the generation of the personal classification information for the target person may be performed when a question from the corresponding user is received or when a prompt is generated.

[0081] Next, in step S260, it is determined whether there are any target persons to add. If there are any target persons to add (step S260, No), the process returns to step S210 and the target persons are added. On the other hand, if there are no target persons to add (step S260, Yes), the registration of target persons ends.

[0082] Returning to the explanation of Figure 5, once the user has completed logging in to the answer generation system 1, the user operates the user terminal 30 to enter a question into the question input screen to create a question (step S30), and then sends the created question to the answer control device 10 (step S40).

[0083] In step S50, the answer control device 20 generates a prompt. The prompt generation unit 130 generates the prompt based on a question from the user and the user's personal classification information. The prompt generation unit 130 may, for example, digitize and / or tag each piece of information included in the personal classification information and generate a prompt by incorporating the digitized and / or tagged information. The information included in the personal classification information includes, for example, an individual's emotional tendencies, thought patterns, luck, values, behavioral patterns, risk tolerance, etc. Note that, when generating a prompt, the prompt may be generated using only a portion of this information. Note that, when a question requiring a compatibility diagnosis with a target person is asked, the prompt generation unit 130 generates a prompt based on the question from the user, the user's personal classification information, and the personal classification information of the corresponding target person. The answer control device 20 transmits the generated prompt to the answer generation device 20 (step S60).

[0084] In step S70, the answer generation device 20 generates an answer to the question from the user based on the prompt received from the answer control device 10. The prompt processing unit 210 inputs the prompt received from the answer control device 10 to the answer generation model 220. The answer generation model 220 recognizes the content of the question and the nature of the user from the input prompt, and generates an answer to the question from the user that is adapted to the nature of the user. The answer generation device 20 sends the generated answer to the answer control device 10 (step S80).

[0085] In step S80, the answer control device 10 transmits the answer received from the answer generation device 20 to the user terminal 30. The answer processing unit 140 transmits the answer received from the answer generation device 20 to the user terminal 30 that transmitted the corresponding question. The answer processing unit 140 may also store the answer received from the answer generation device 20 in the storage 15. The answer is stored in the storage 15 in association with at least one of the corresponding question, the corresponding prompt, and the corresponding user.

[0086] When the user terminal 30 receives the answer from the answer control device 10, it displays the received answer on the display unit 36 ​​(step S100). This allows the user to obtain the answer to the question. The user terminal 30 may also store the received answer in the storage 35. The answer is stored in association with the corresponding question.

[0087] [Example 1] Here, an example according to this embodiment will be described. For example, when a user asks a question such as "I want to change jobs. What kind of job would suit me?", the answer obtained when a conventional prompt not including personal classification information is input to the generation AI will be compared with the answer obtained when a prompt according to this embodiment including personal classification information is input to the generation AI.

[0088] When using conventional prompts, only standard, generic answers such as "The jobs that suit you are clerical work, sales, customer service, etc." are generated. In contrast, in this embodiment, by including information (personal classification information) in the prompt that the user's personality is "wealth-biased type / Scorpio type," a completely different answer is generated, suggesting specific jobs based on the user's personality, such as "You are sociable and observant, and you would be suited to an environment where you can make use of your ability to see through others' true nature. For example, a professional job that builds trust, a job that handles information, or an advisory role would be perfect for you." In this way, by using the prompts of this embodiment, detailed answers that are highly satisfying to the user are generated.

[0089] [Example 2] In Example 2, we will explain a case where the output changes depending on how the prompt is created, even for the same question. Suppose the question from the user is, "Please tell me what kind of job is suitable for me." In contrast, if a conventional prompt is used, a nondescript, average answer such as, "The jobs that are suitable for you are sales, clerical work, customer service, etc." will be generated.

[0090] In contrast, in the present embodiment, for example, "This questioner (user) is of the partial wealth type with the day stem being Gui, and has tendencies of being flexible, social, and theoretical. They have elements of Gemini and Scorpio, are strong in information, and possess both observation ability and deep insight. Please list three jobs suitable for such a personality and state the reasons for each." In this way, the content of the user classification information, instructions on the answering method for the answer generation device 20, etc. can be specifically defined by the prompt. The answer to this prompt is generated along the lines of the prompt content, such as "The occupations suitable for you are financial planner, patent-related support work, and research and analysis positions. The reason is that you are..."

[0091] [Example 3] In Example 3, the case of performing a compatibility diagnosis between the user and the target person A will be described. Assume that the question from the user is "Please rate the business compatibility between the user and Mr. A, the target person." Then, for example, a prompt is generated that includes content such as instructing to rate the classified information of the tagged user, the classified information of the tagged target person A, the compatibility scores for each divination method, and the overall compatibility score.

[0092] The answer based on this prompt is, for example, as shown in FIG. 8. FIG. 8 is a diagram illustrating the answer in Example 3. In the example of FIG. 8, for each divination method, the tagged natures of both parties and the compatibility scores between the two are rated respectively. Also, in FIG. 8, the overall compatibility score between the two is rated. Furthermore, as shown in FIG. 8, comments regarding the compatibility in the business aspect are generated. In this way, a specific answer is generated along the lines of the prompt content.

[0093] [Main effects according to the present embodiment] According to this embodiment, personal classification information for a user is generated based on the user's date of birth and multiple divination methods, a prompt is generated based on the question from the user and the personal classification information, and an answer to the question is generated based on the prompt. Furthermore, while conventional general-purpose systems could only provide standard or uniform answers, by using the user's personal classification information, it is possible to provide answers that are optimized for each user depending on the user's characteristics.

[0094] Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments. Those skilled in the art can change, add, or convert each element of the above embodiments without departing from the scope of the present invention. [Explanation of symbols]

[0095] 1...answer generation system, 10...answer control device, 20...answer generation device, 30...user terminal.

Claims

1. a user terminal used by a user; an answer control device that generates personal classification information for the user, including a type of the user classified by the divination method, based on the user's date of birth and a plurality of divination methods, and generates a prompt based on a question from the user generated by the user terminal and the personal classification information of the user; an answer generator that generates an answer to the question based on the prompt; Equipped with the answer generation device adjusts the sentence expression of the answer depending on the type of the user classified by the divination art. Answer generation system.

2. 2. The answer generation system according to claim 1, The divination method includes any of the following: Four Pillars of Destiny, Nine Star Ki, Shukuyo Astrology, Western Astrology, Animal Fortune Telling, and Yin-Yang and Five Elements. Answer generation system.

3. The answer generation system according to claim 1, When diagnosing the compatibility between the user and the target person, The answer control device generates the prompt based on a question from the user, the personal classification information of the user, and the personal classification information of the target person. Answer generation system.

4. Based on the user's date of birth and multiple divination methods, personal classification information for the user is generated, including the type of the user classified by the divination methods; a prompt is generated based on a question from the user generated on a user terminal and the personal classification information of the user; and an answer generation device is caused to generate an answer to the question based on the prompt, with the wording adjusted according to the type of the user classified by the divination methods. Answer control device.

5. A step in which the answer control device generates personal classification information of the user based on the user's date of birth and a plurality of divination methods, the personal classification information including the type of the user classified by the divination method; The answer control device generates a prompt based on a question from the user generated at a user terminal and the personal classification information of the user; an answer generation device generating an answer to the question based on the prompt, the answer having a sentence expression adjusted according to the type of the user classified by the divination art; to be executed on the processor, program.

Images