Interaction support apparatus, interaction support method, and interaction support program

Abstract

To facilitate smooth interaction between a generative AI and each of a plurality of users even when the users have different attributes.SOLUTION: An interaction support apparatus 16 acquires attribute data of a target user. The interaction support apparatus 16 acquires an interaction condition corresponding to the attribute data of the target user, from an interaction condition database 18 that stores interaction conditions set in advance for each attribute data of users. The interaction support apparatus 16 modifies interaction data output from a generative AI in accordance with the interaction condition of the target user. The interaction support apparatus 16 outputs the modified interaction data.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a dialogue support device, a dialogue support method, and a dialogue support program. [Background technology]

[0002] BACKGROUND ART A mentoring system is known that generates information for input to an artificial intelligence (AI) suitable for mentoring without user intervention and provides mentoring to a user (see, for example, Patent Document 1).

[0003] Also, an information processing device is known that analyzes image data and easily generates a layout for a test using a terminal device (see, for example, Patent Document 2). The new layout generation AI used in this information processing device generates a layout of content that combines elements according to the performance and size of the display of the output terminal device that displays the layout (see, for example, paragraph

[0034] of Patent Document 2).

[0004] Also known is a patent claim creation support device that reduces the variation in quality when creating patent claims or reduces the time required to create claims (see, for example, Patent Document 3). This creation support device graphically displays claims included in claims of past patent applications on a display, accepts changes to the displayed graphic, and generates new claims in text format based on the accepted changes. The creation server corresponding to this creation support device is equipped with a claim generation AI engine that generates and updates claims by having AI learn information from past patent applications. The creation server processes input information with this claim generation AI engine, automatically generating patent claims corresponding to the input keywords (see, for example, paragraph

[0017] ). [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 7416390 [Patent Document 2] Japanese Patent Publication No. 2023-87805 [Patent Document 3] Japanese Patent Application Publication No. 2023-112150 Summary of the Invention [Problem to be solved by the invention]

[0006] Incidentally, when considering a situation in which a human user and a generation AI are conversing, it may be preferable to change the content of the information output by the generation AI depending on the user's attributes. For example, the definition of event A in user U1's country of origin X may differ from the definition of event A in user U2's country of origin Y. In this case, if the generation AI outputs information based on the definition of event A in country Y when conversing with user U1, user U1, who is from country X, may feel uncomfortable, which could result in a problem in which the conversation between the user and the generation AI is not smooth.

[0007] The techniques disclosed in the above Patent Documents 1 to 3 generate some information using a generation AI, but are not techniques that take into account the attributes of the user.

[0008] The present disclosure has been made in consideration of the above points, and aims to facilitate smooth dialogue between users and a generation AI when the generation AI interacts with each of multiple users, even if the attributes of each user are different. [Means for solving the problem]

[0009] In order to achieve the above-mentioned object, the dialogue support device of the present disclosure is a dialogue support device that supports a dialogue between a target user and a generation AI (Artificial Intelligence), and includes: a first acquisition unit that acquires attribute data of the target user; a second acquisition unit that acquires dialogue conditions corresponding to the attribute data of the target user from a database in which dialogue conditions previously set for each user's attribute data are stored; a modification unit that modifies the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; and an output unit that outputs the modified dialogue data.

[0010] In addition, the dialogue support method disclosed herein is a dialogue support method executed by a computer constituting a dialogue support device that supports dialogue between a target user and a generation AI (Artificial Intelligence), and includes acquiring attribute data of the target user, acquiring dialogue conditions corresponding to the attribute data of the target user from a database in which dialogue conditions previously set for each user's attribute data are stored, modifying dialogue data output from the generation AI in accordance with the dialogue conditions of the target user, and outputting the modified dialogue data.

[0011] In addition, the dialogue support program disclosed herein is a dialogue support program to be executed by a computer constituting a dialogue support device that supports dialogue between a target user and a generation AI (Artificial Intelligence), and is a dialogue support program that acquires attribute data of the target user, acquires dialogue conditions corresponding to the attribute data of the target user from a database in which dialogue conditions previously set for each user's attribute data are stored, modifies the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user, and outputs the modified dialogue data. [Effects of the Invention]

[0012] According to the dialogue support device, dialogue support method, and dialogue support program disclosed herein, when a generation AI dialogues with each of multiple users, even if the attributes of each user are different, dialogue between the user and the generation AI can be facilitated. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram illustrating a dialogue support system according to an embodiment. [Figure 2] 1 is a block diagram illustrating a schematic configuration of a dialogue support system according to an embodiment. [Figure 3] 1 is a block diagram showing a hardware configuration of a dialogue support device according to an embodiment. [Figure 4] FIG. 2 is a diagram for explaining various databases according to the embodiment. [Figure 5] 10 is a flowchart showing the flow of a dialogue assistance process in the embodiment. [Figure 6] 10 is a flowchart showing the flow of a dialogue condition database construction process in the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] An example of an embodiment of the present disclosure will be described below with reference to the drawings. In this embodiment, a dialogue support system including a dialogue support device according to the present disclosure will be described as an example. Note that the same reference numerals are used in the drawings to designate identical or equivalent components and parts. Furthermore, the dimensions and proportions of the drawings are exaggerated for the sake of explanation and may differ from the actual proportions.

[0015] FIG. 1 is a diagram for explaining an embodiment. FIG. 1 shows a user U1, a user U2, and a generation AI4. The scene shown in FIG. 1 is an example of a scene in which a conversation about double-cooked pork is taking place between users U1, U2, and generation AI4. For example, assume that user U1 is from country X (e.g., Japan), and that in country X, double-cooked pork is presumed to contain cabbage. Also, assume that user U2 is from country Y (e.g., China), and that in country Y, double-cooked pork is presumed to contain garlic leaves rather than cabbage. Also, assume that users U1 and U2 are both chefs.

[0016] In this case, for example, assume that a dialogue C1 has taken place between user U1 and generation AI4, as shown in Figure 1. As shown in Figure 1, user U1 asks generation AI4 to "write out a recipe for double-cooked pork," to which generation AI4 responds, "Stir-fry the pork, remove it, and add the cabbage..." Then, user U1 responds, "Pork is OK, but it's important to remove it once," and further requests generation AI4 to "Add eggplant and green peppers in addition to the cabbage."

[0017] On the other hand, in dialogue C2 shown in Figure 1, user U2 asks generation AI4 to "write out the recipe for Twice Cooked Pork," to which generation AI4 responds, "Stir-fry the pork, remove it, and then add the cabbage..." User U2 then responds, "Pork is OK, but it's important to remove it once." However, in dialogue C2 shown in Figure 1, user U2 points out, "It's not cabbage, it's garlic leaves."

[0018] As shown in dialogue C2 in Figure 1, in user U2's home country Y, the definition of "twice-cooked pork" is that it is a dish containing garlic leaves, whereas the definition given by generation AI4 is that it is a dish containing cabbage. For this reason, the dialogue between user U2 and generation AI4 is not smooth.

[0019] On the other hand, as shown in dialogue C1 in Figure 1, in user U1's country of origin X, double-cooked pork is defined as a dish that contains cabbage, so user U1 and generation AI4 can have a smooth dialogue.

[0020] Because the definition of Twice Cooked Pork differs from country to country, if the generation AI4 were to converse with multiple users based on a single definition of Twice Cooked Pork, it would be unable to converse smoothly with the users. On the other hand, users U1 and U2, who share the attribute of chef, share a common understanding that Twice Cooked Pork contains pork and that the meat is removed once.

[0021] Therefore, when the generation AI interacts with each of multiple users, if the attributes of each user are different, a problem may arise in which the interaction between the user and the generation AI cannot be carried out smoothly.

[0022] Therefore, the dialogue support device of the embodiment changes the dialogue data output from the generation AI in accordance with the attributes of the user, thereby smoothing the dialogue between the user and the generation AI. This will be specifically described below.

[0023] <Dialogue support system 10 according to the embodiment> FIG. 2 is a block diagram showing a schematic configuration of a dialogue support system 10 according to an embodiment. As shown in the figure, the dialogue support system 10 includes a user terminal 12-1 operated by a human user U1, a user terminal 12-2 operated by a human user U2, a server 14 on which a generation AI4 operates, and a dialogue support device 16. The user terminals 12-1 and 12-2, the server 14 on which the generation AI4 operates, and the dialogue support device 16 are communicatively connected via a network 30 such as the Internet. Note that while FIG. 2 illustrates two users U1 and U2 and two user terminals 12-1 and 12-2, more users and user terminals may actually exist.

[0024] User terminal 12-1 is operated by user U1. User U1 interacts with generated AI4 by operating user terminal 12-1. User terminal 12-2 is operated by user U2. User U2 interacts with generated AI4 by operating user terminal 12-2. User terminal 12-1 and user terminal 12-2 are realized by a computer such as that shown in FIG. 3, which will be described later. In the following, when user terminal 12-1 and user terminal 12-2 are not distinguished from each other, they will simply be referred to as user terminal 12. In the following, when user U1 and user U2 are not distinguished from each other, they will simply be referred to as user U.

[0025] The server 14 operates the generation AI 4. This allows a dialogue to take place between the user U operating the user terminal 12 and the generation AI 4. Specifically, the generation AI 4 is stored in a memory unit (not shown) of the server 14, and the generation AI 4 operates in accordance with control processing by a control unit (not shown) of the server 14. The generation AI 4 is an AI realized by known machine learning technology and known artificial intelligence technology, for example, an AI based on a large-scale language model. The server 14 is realized by a computer such as that shown in FIG. 3, which will be described later.

[0026] Note that when a dialogue takes place between the user U and the generated AI 4, the dialogue is carried out via the dialogue support device 16. As described above, the dialogue support device 16 supports the dialogue between the user U and the generated AI 4.

[0027] FIG. 3 is a block diagram showing the hardware configuration of a dialogue support device 16 according to an embodiment. As shown in FIG. 3, the dialogue support device 16 includes a CPU (Central Processing Unit) 42, a memory 44, a storage device 46, an input / output I / F (Interface) 48, a storage medium reading device 50, and a communication I / F 52. Each component is communicably connected to one another via a bus 54. Note that, in this embodiment, a case is described in which a generated AI provided via a network 30 from a server 14, which is an external generated AI system different from the dialogue support device 16, is used, but this is not limited to this. For example, the dialogue support device 16 may store a generated AI in its own storage device 46 and use the generated AI.

[0028] The storage device 46 stores a dialogue support program for executing each process described below. The CPU 42 is a central processing unit that executes various programs and controls each component. That is, the CPU 42 reads the program from the storage device 46 and executes the program using the memory 44 as a work area. The CPU 42 controls each component and performs various arithmetic processes in accordance with the program stored in the storage device 46.

[0029] The memory 44 is made up of RAM (Random Access Memory) and serves as a working area to temporarily store programs and data. The storage device 46 is made up of ROM (Read Only Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), etc., and stores various programs including the operating system and various data.

[0030] The input / output I / F 48 is an interface for inputting data from an external device and outputting data to an external device. Also, input devices for various inputs, such as a keyboard or a mouse, and output devices for various information outputs, such as a display or a printer, may be connected. A touch panel display may be used as the output device, allowing it to function as an input device.

[0031] The storage medium reader 50 reads data stored in various storage media such as CD (Compact Disc)-ROM, DVD (Digital Versatile Disc)-ROM, Blu-ray Disc, and USB (Universal Serial Bus) memory, and writes data to the storage media.

[0032] The communication I / F 52 is an interface for communicating with other devices, and uses standards such as Ethernet (registered trademark), FDDI, and Wi-Fi (registered trademark).

[0033] Next, the functional configuration of the dialogue support device 16 will be described. As shown in Fig. 2, the dialogue support device 16 functionally includes a first acquisition unit 20, a second acquisition unit 22, a change unit 24, an output unit 26, and a construction unit 28. A predetermined storage area of ​​the dialogue support device 16 also includes a user database 17, a dialogue condition database 18, and a dialogue history database 19. Each functional configuration is realized by the CPU 42 reading out each program stored in the storage device 46, expanding it into the memory 44, and executing it.

[0034] The user database 17 stores the user ID and attribute data of each of a plurality of users U in association with each other. For example, the user database 17 stores table data such as that shown in FIG. 4(A).

[0035] The dialogue condition database 18 stores dialogue conditions that are set in advance for each attribute data of the user U. For example, the dialogue condition database 18 stores table data such as that shown in FIG. 4(B). As will be described later, the dialogue condition database 18 is a database constructed based on dialogue history data between each of the multiple users U and the generated AI 4. The dialogue condition database 18 may also be constructed in advance based on some prior information.

[0036] The dialogue history database 19 stores dialogue history data between each of a plurality of users U and the generated AI 4. For example, the dialogue history database 19 stores table data such as that shown in FIG. 4(C).

[0037] (Dialogue support processing) When the user U and the generated AI 4 have a dialogue, each unit of the dialogue support device 16 executes the following process.

[0038] The first acquisition unit 20 acquires attribute data of the target user who is the conversation partner of the generation AI 4. For example, when issuing a user ID to a user using the dialogue support device 16, the user's attribute data is acquired by having the user enter their own attribute data. Alternatively, for example, a system administrator other than the user may register the user's attribute data in the user database 17 of the dialogue support device 16. For example, the system administrator may acquire the user's attribute data from an existing employee database or the like by linking the dialogue support device 16 with the employee database or the like. In this case, the user himself / herself does not need to enter the attribute data. Alternatively, the first acquisition unit 20 may acquire the user's attribute data from the dialogue between the user and the generation AI using known natural language processing technology or known artificial intelligence technology. Examples of the attribute data include country of origin, occupation, gender, age, and place of residence. This allows the user database 17 shown in FIG. 4(A) to be constructed, making it possible to acquire attribute data from the user ID. For this purpose, the first acquisition unit 20 refers to the user database 17 and acquires attribute data of the target user based on the user ID input when the target user uses the dialogue support device 16.

[0039] The second acquisition unit 22 acquires, from the dialogue condition database 18, dialogue conditions corresponding to the attribute data of the target user acquired by the first acquisition unit 20. For example, if the attribute data of the target user indicates origin in Japan, occupation as chef, and gender as male, dialogue condition G1 shown in FIG. 4(B) is acquired.

[0040] The modification unit 24 modifies the dialogue data output from the generation AI 4 in accordance with the dialogue conditions of the target user acquired by the second acquisition unit 22. The modification unit 24 modifies the dialogue data by using a known natural language processing technique or a known artificial intelligence technique.

[0041] For example, the modification unit 24 modifies the dialogue data already output from the generation AI 4 based on a rule corresponding to the dialogue condition. Specifically, for example, if the dialogue condition is "The vegetable to put in the Twice Cooked Pork is garlic leaves," the modification unit 24 replaces "cabbage" with "garlic leaves" in the dialogue data already output from the generation AI 4.

[0042] Alternatively, for example, the modification unit 24 modifies the dialogue data output from the generation AI 4 by inputting a prompt to the generation AI that modifies the dialogue data output from the generation AI 4 and that corresponds to the dialogue conditions. Specifically, for example, if the dialogue condition is "The vegetable to be put in Twice Cooked Pork is garlic leaves," the modification unit 24 inputs a prompt to the generation AI 4 such as "The dialogue partner's country of origin is Country Y, so the vegetable to be put in Twice Cooked Pork is garlic leaves. Please create an answer based on this assumption." to the generation AI 4, thereby changing "cabbage" in the dialogue data to "garlic leaves." Note that the modification unit 24 may input a prompt to the generation AI 4 after recognizing data already output from the generation AI 4, thereby generating appropriate dialogue data.

[0043] The output unit 26 outputs the changed dialogue data to the user terminal 12 of the target user. The output unit 26 also stores dialogue history data between the target user and the generated AI 4 in the dialogue history database 19.

[0044] (Dialogue condition database 18 construction process) Next, when constructing the dialogue condition database 18, each unit of the dialogue support device 16 executes the following process. As mentioned above, the dialogue condition database 18 may be constructed in advance based on some prior information.

[0045] The construction unit 28 constructs the dialogue condition database 18 based on the dialogue history data stored in the dialogue history database 19. The dialogue condition database 18 is constructed based on different types of attribute data included in the attribute data of each of the multiple users U and the dialogue history data between each of the multiple users U and the generation AI 4. As described above, the attribute data of a given user U includes multiple attributes. For example, the attribute data of a given user U includes country of origin, occupation, gender, age, and place of residence. Therefore, the construction unit 28 constructs the dialogue condition database 18 using this multiple attribute data.

[0046] Specifically, when the first attribute data included in the attribute data of user U1 of the multiple users is the same as the first attribute data included in the attribute data of user U2 of the multiple users, the construction unit 28 extracts common events between the interaction history data between user U1 and generated AI4 and the interaction history data between user U2 and generated AI4. User U1 is an example of a first user, and user U2 is an example of a second user.

[0047] For example, as shown in FIG. 1, between user U1 and user U2, both chefs, the fact that "pork is added to twice-cooked pork" and the fact that "the meat is removed once when twice-cooked pork is cooked" are events commonly recognized as information related to twice-cooked pork. In this case, since the first attribute data "chef" included in the attribute data of user U1 and the first attribute data "chef" included in the attribute data of user U2 are the same, construction unit 28 extracts the common events "pork is added to twice-cooked pork" and "the meat is removed once when twice-cooked pork is cooked" from the dialogue history data between user U1 and generation AI4 and the dialogue history data between user U2 and generation AI4. Construction unit 28 extracts the common events from the dialogue history data by using known natural language processing or known artificial intelligence technology.

[0048] Then, the construction unit 28 associates the first attribute data "chef" with the event data "pork is added to double-cooked pork" and "the meat is removed once when double-cooked pork is served," which represent common events, and stores them in the dialogue condition database.

[0049] On the other hand, the construction unit 28 considers a case where the second attribute data "Country of origin: Japan" included in the attribute data of user U1 is different from the second attribute data "Country of origin: China" included in the attribute data of user U2. In this case, the construction unit 28 associates the event "Put cabbage in twice-cooked pork" included in the dialogue history data between user U1 and generation AI4 with the second attribute data of user U1 "Country of origin: Japan" and stores them in the dialogue condition database 18. On the other hand, the construction unit 28 associates the event "Put garlic leaves in twice-cooked pork" included in the dialogue history data between user U2 and generation AI4 with the second attribute data of user U2 "Country of origin: China" and stores them in the dialogue condition database 18.

[0050] This creates a dialogue condition database 18 in which attribute data and dialogue conditions are associated with each other.

[0051] Next, the operation of the dialogue support system 10 according to this embodiment will be described.

[0052] First, the dialogue support process will be described. When a dialogue is started between the user U and the generated AI 4, the CPU 42 of the dialogue support device 16 reads out the dialogue support program from the storage device 46, loads it into the memory 44, and executes it. As a result, the CPU 42 functions as each functional component of the dialogue support device 16, and the dialogue support process shown in FIG. 5 is executed.

[0053] In step S100, the first acquisition unit 20 acquires attribute data of the target user who is the conversation partner of the generated AI 4. Specifically, the first acquisition unit 20 acquires attribute data associated with the user ID of the target user by referring to the user database 17.

[0054] In step S102, the second acquisition unit 22 acquires, from the dialogue condition database 18, dialogue conditions corresponding to the attribute data of the target user acquired in step S100.

[0055] In step S104, while a dialogue is taking place between user U and generated AI4, the modification unit 24 modifies the dialogue data output from generated AI4 in accordance with the dialogue conditions of the target user acquired in step S102.

[0056] In step S106, the output unit 26 outputs the dialogue data changed in step S104 to the user terminal 12 of the target user.

[0057] Next, the dialogue condition database construction process will be described. During a dialogue between each of the multiple users U and the generation AI 4, multiple dialogue history data are stored in the dialogue history database 19. When the dialogue support device 16 receives a request signal for construction of the dialogue condition database 18, the CPU 42 of the dialogue support device 16 reads out a dialogue condition database construction process program from the storage device 46, loads it into the memory 44, and executes it. As a result, the CPU 42 functions as each functional component of the dialogue support device 16, and the dialogue condition database construction process shown in FIG. 6 is executed.

[0058] In step S200, the constructing unit 28 acquires a plurality of pieces of dialogue history data stored in the dialogue history database 19.

[0059] In step S 202 , the construction unit 28 constructs the dialogue condition database 18 based on a plurality of dialogue history data stored in the dialogue history database 19 .

[0060] The dialogue condition database 18 constructed by the process shown in FIG. 6 is used in subsequent dialogues between the user U and the generated AI 4.

[0061] As described above, the dialogue support device according to the embodiment acquires attribute data of a target user. The dialogue support device acquires dialogue conditions corresponding to the attribute data of the target user from a database in which dialogue conditions preset for each user's attribute data are stored. The dialogue support device modifies the dialogue data output from the generation AI according to the dialogue conditions of the target user. The dialogue support device outputs the modified dialogue data. This enables smooth dialogue between the user and the generation AI when the generation AI converses with each of multiple users, even if the attributes of each user are different. In particular, the dialogue history data can be tagged and systematized using the attribute data of the dialogue partner.

[0062] In the past, the generation AI often learned dialogue history data. However, once the generation AI learned the dialogue history data, as described above, the data output from the generation AI may be incorrect depending on the user's attributes. In contrast, the dialogue support device of the embodiment can smooth the dialogue between the user and the generation AI by changing the dialogue data output from the generation AI depending on the user's attributes.

[0063] The technology of the present disclosure is not limited to the above-described embodiment, and various modifications and applications are possible within the scope of the gist of this disclosure.

[0064] For example, in the above embodiment, the dialogue support device has the databases, but the present invention is not limited to this. For example, the databases may be located outside the dialogue support device.

[0065] Furthermore, the processes executed by the CPU after reading the software (program) in the above-described embodiments may be executed by various processors other than the CPU. Examples of such processors include programmable logic devices (PLDs) whose circuit configuration can be changed after fabrication, such as field-programmable gate arrays (FPGAs), and dedicated electrical circuits, such as application-specific integrated circuits (ASICs), which are processors with circuit configurations specifically designed to execute specific processes. Each process may be executed by one of these various processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, or a combination of a CPU and an FPGA). The hardware structure of these various processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor devices.

[0066] In the above embodiment, the programs are pre-stored (installed) in a storage device, but the present invention is not limited to this. The programs may be provided in a form stored on a storage medium such as a CD-ROM, DVD-ROM, Blu-ray disc, or USB memory. The programs may also be downloaded from an external device via a network.

[0067] (Addendum) The following additional notes are provided regarding aspects of the present disclosure.

[0068] (Appendix 1) A dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), a first acquisition unit that acquires attribute data of the target user; a second acquisition unit that acquires, from a database in which predetermined dialogue conditions are stored for each user's attribute data, dialogue conditions corresponding to the attribute data of the target user; a modification unit that modifies the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; an output unit that outputs the modified dialogue data; A dialogue support device comprising: (Appendix 2) The database is constructed based on dialogue history data between each of the plurality of users and the generation AI, 2. The dialogue support device according to claim 1. (Appendix 3) The database is constructed based on different types of attribute data included in the attribute data of each of the plurality of users and dialogue history data between each of the plurality of users and the generation AI. 3. A dialogue support device as described in appendix 2. (Appendix 4) further comprising a construction unit that constructs the database, The construction unit When first attribute data included in the attribute data of a first user among the plurality of users is the same as first attribute data included in the attribute data of a second user among the plurality of users, extracting common events between the dialogue history data between the first user and the generation AI and the dialogue history data between the second user and the generation AI, associating the first attribute data with event data representing the common events and storing them in the database; When the second attribute data included in the attribute data of the first user is different from the second attribute data included in the attribute data of the second user, the events included in the dialogue history data between the first user and the generation AI are associated with the second attribute data of the first user and stored in the database, and the events included in the dialogue history data between the second user and the generation AI are associated with the second attribute data of the second user and stored in the database. 4. A dialogue support device according to claim 3. (Appendix 5) the modification unit modifies the dialogue data output from the generation AI by inputting a prompt to the generation AI that modifies the dialogue data output from the generation AI and that corresponds to the dialogue conditions. 5. A dialogue support device according to any one of Supplementary notes 1 to 4. (Appendix 6) the modification unit modifies the dialogue data already output from the generation AI based on rules according to the dialogue conditions. 6. A dialogue support device according to any one of Supplementary notes 1 to 5. (Appendix 7) The generating AI is an AI based on a large-scale language model. 7. A dialogue support device according to any one of Supplementary notes 1 to 6. (Appendix 8) A dialogue support method executed by a computer constituting a dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), Acquire attribute data of the target user; acquiring, from a database in which predetermined dialogue conditions are stored for each user's attribute data, the dialogue conditions corresponding to the attribute data of the target user; Modifying the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; outputting the modified interaction data; Dialogue support methods. (Appendix 9) A dialogue support program to be executed by a computer constituting a dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), Acquire attribute data of the target user; acquiring, from a database in which predetermined dialogue conditions are stored for each user's attribute data, the dialogue conditions corresponding to the attribute data of the target user; Modifying the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; outputting the modified interaction data; Dialogue support program. [Explanation of symbols]

[0069] 10 Dialogue Support System 12-1, 12-2 User terminal 14 Servers 16 Dialogue support device 17 User Database 18 Dialogue Condition Database 19 Dialogue History Database 20 First acquisition part 22 Second acquisition part 24 Changes 26 Output section 28 Construction Department

Claims

1. A dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), a first acquisition unit that acquires attribute data of the target user; a second acquisition unit that acquires, from a database in which predetermined dialogue conditions are stored for each user's attribute data, dialogue conditions corresponding to the attribute data of the target user; A modification unit that modifies the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; an output unit that outputs the modified dialogue data; A dialogue support device comprising:

2. The database is constructed based on interaction history data between each of the plurality of users and the generation AI, The dialogue support device according to claim 1 .

3. The database is constructed based on different types of attribute data included in the attribute data of each of the plurality of users and dialogue history data between each of the plurality of users and the generation AI. The dialogue support device according to claim 2 .

4. further comprising a construction unit that constructs the database, The construction unit When first attribute data included in the attribute data of a first user among a plurality of users is identical to first attribute data included in the attribute data of a second user among the plurality of users, extracting common events between the dialogue history data between the first user and the generation AI and the dialogue history data between the second user and the generation AI, associating the first attribute data with event data representing the common events and storing them in the database; When the second attribute data included in the attribute data of the first user is different from the second attribute data included in the attribute data of the second user, the events included in the dialogue history data between the first user and the generation AI are associated with the second attribute data of the first user and stored in the database, and the events included in the dialogue history data between the second user and the generation AI are associated with the second attribute data of the second user and stored in the database. The dialogue support device according to claim 3 .

5. The change unit changes the dialogue data output from the generation AI by inputting a prompt to the generation AI that changes the dialogue data output from the generation AI and that corresponds to the dialogue conditions. The dialogue support device according to claim 1 or 2.

6. The change unit changes the dialogue data already output from the generation AI based on rules according to the dialogue conditions. The dialogue support device according to claim 1 or 2.

7. The generative AI is an AI based on a large-scale language model. The dialogue support device according to claim 1 or 2.

8. A dialogue support method executed by a computer constituting a dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), Acquire attribute data of the target user; acquiring, from a database in which predetermined dialogue conditions are stored for each user's attribute data, the dialogue conditions corresponding to the attribute data of the target user; Changing the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; outputting the modified interaction data; Dialogue support methods.

9. A dialogue support program to be executed by a computer constituting a dialogue support device that supports a dialogue between a target user and a generating AI (Artificial Intelligence), Acquire attribute data of the target user; acquiring, from a database in which predetermined dialogue conditions are stored for each user's attribute data, the dialogue conditions corresponding to the attribute data of the target user; Changing the dialogue data output from the generation AI in accordance with the dialogue conditions of the target user; outputting the modified interaction data; Dialogue support program.

Images