Computer systems and programs

Abstract

To control how NPCs communicate using a generation unit. [Solution] When a communication event occurs during content provision in which the target character 7 communicates with the other character 8, the server system 1100 sets background information 728 based on at least one of the following: the status of the content, the status of the target character 7, the status of the other character 8, and the relationship between the target character 7 and the other character 8. Then, based on the communication transmission information and the background information 728, it generates comprehensive generation instruction information 730, which is given to the server-side generation AI 20 to generate expression information that causes the target character 7 to express communication.

Description

Technical Field

[0001] The present invention relates to a computer system and the like.

Background Art

[0002] Communication with NPCs (Non Player Characters) that appear in video games is an important factor affecting the charm of the game world. For example, various technologies have been devised regarding speech, which is a representative example of communication.

[0003] Patent Document 1 describes a technique in which a plurality of lines and a plurality of display conditions are associated and stored in advance, and even if the user does not input a line, a line that conforms to the display conditions is automatically selected according to the game situation at that time and displayed within the game screen.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0008] In this context, "communication" refers to the use of expressions (such as speech, gestures, musical performance, singing, hand signals, etc.) by a target character to convey information to another character. In the case of a conversation, one utterance from the target character to the other character constitutes one instance of communication.

[0009] The generation unit, exemplified by generative AI, is trained to generate data in the format specified in the generation instruction information, based on the information contained within that instruction information. According to the first invention, the computer system can provide the generation unit with generation instruction information that includes at least one of the following: the status of the content, the status of the target character, the status of the opponent character, and the relationship between the target character and the opponent character. The generation unit generates representation information corresponding to these, so the computer system can cause the NPC to perform communication according to at least one of the following: the status of the content, the status of the target character, the status of the opponent character, and the relationship between the target character and the opponent character. In other words, the control of communication between NPCs appearing in a video game can be realized using the generation unit.

[0010] The second invention is a computer system in which, in the above-described computer system, the communication information setting means sets the communication information when the conditions for the occurrence of the communication based on the progress status, which is the status of the content and / or the status of the target character (for example, the current content status condition 557c and the current target character status condition 557d in Figure 6) are met, and the background information setting means sets the background information when the conditions for the occurrence are met (for example, YES in step S12 → step S24 in Figure 30).

[0011] According to the second invention, the computer system can set communication information and background information when the progress status, which is the status of the content and / or the status of the target character, satisfies the occurrence requirements.

[0012] The third invention is a computer system further comprising the above-described computer system, wherein preconditions for satisfying the occurrence requirements are defined (for example, pre-content status conditions 557a and pre-target character status conditions 557b in Figure 6), and satisfaction monitoring means for monitoring whether the occurrence requirements are satisfied when the preconditions are met (for example, satisfaction monitoring unit 210 in Figure 28 and step S12 in Figure 30).

[0013] According to the third invention, for example, a computer system uses conditions concerning a past situation from the perspective of the present as preconditions, and monitors whether the conditions for occurrence are met when these preconditions are satisfied. In situations where the preconditions are not met, it is not necessary to monitor whether the conditions for occurrence are met.

[0014] The fourth invention is a computer system in which, in the above-described computer system, the representation information acquisition control means controls the amount of information of the generation instruction information in a variable manner based on the billing information of the user playing the content (for example, the information amount setting function 536 in Figure 36).

[0015] According to the fourth invention, the computer system can variably control the amount of information in the generation instruction information based on the billing information.

[0016] The fifth invention is a computer system in which the background information includes at least the history of the content's status (for example, the pre-content status condition 557a in Figure 6, which corresponds to the range calculated by the status history range setting function 580 in Figure 7) and / or the history of the target character's status (for example, the pre-target character status condition 557b in Figure 6, which corresponds to the range calculated by the status history range setting function 580 in Figure 7).

[0017] According to the fifth invention, the computer system becomes capable of setting background information, including the history of the content's situation and the history of the target character's situation.

[0018] The sixth invention is a computer system in which, in the above-described computer system, the background information setting means determines the scope of the circumstances to be included in the background information (for example, step S22 in Figure 30).

[0019] According to the sixth invention, the computer system becomes capable of determining the scope of the contextual history to be included in the background information.

[0020] The seventh invention is a computer system in the above computer system, wherein candidates for control items of the communication control include the lines of the target character, the speech voice of the target character, the motion of the target character, and the effects during communication.

[0021] According to the seventh invention, the computer system can perform communication control using any one of the lines of the target character, the speech voice of the target character, the motion of the target character, and the effects during communication.

[0022] The eighth invention is a computer system in the above computer system, wherein the communication control execution means executes the communication control by including at least the lines of the target character among the control item candidates as control items.

[0023] According to the eighth invention, the computer system can realize at least communication in which the target character makes a statement.

[0024] The ninth invention is a computer system in the above computer system, wherein the expression information acquisition control means generates the generation instruction information including item designation information (for example, item designation information 734 in FIG. 8) for designating the control item candidates to be the control items, and gives it to the generation unit, thereby acquiring the expression information for the control items, and the communication control execution means executes the communication control by controlling the control items based on the expression information.

[0025] According to the ninth invention, the computer system can designate the expression information to be generated by the generation unit with the control item designation information. <00OO097> The tenth invention is a computer system in which, in the above-described computer system, the representation information acquisition control means variably determines the item designation information according to the number of target characters (for example, modified example 4).

[0027] According to the tenth invention, the computer system can change the representation information generated by the generation unit according to the number of target characters.

[0028] The eleventh invention is a computer system in which the generation unit comprises a control item-specific generation unit for each candidate control item (for example, the control item-specific generation AI23 (23a, 23b, ...) in Figure 9, the control item-specific generation unit 232 in Figure 28), and a control unit that causes the control item-specific generation unit to generate representation information related to the control item based on the generation instruction information (for example, the control unit AI21 in Figure 9, the control unit 234 in Figure 28).

[0029] According to the eleventh invention, the computer system can generate representation information using a generation unit having a control item generation unit and a control unit.

[0030] The twelfth invention is a computer system in which, in the above-described computer system, the control unit variably determines the generation order (for example, the generation order control item list 740 in Figure 29), and causes the control item generation unit to generate representation information related to the control item according to the said generation order.

[0031] According to the twelfth invention, a computer system can change the order in which it generates representation information.

[0032] The 13th invention further comprises, in the above-mentioned computer system, a control time length determination means (for example, a control time length determination unit 220 in Figure 28) for determining the control time length, which is the length of the communication control, The aforementioned expression information acquisition control means is a computer system that includes the control time length in the generation instruction information and provides it to the generation unit (for example, the performance policy description 735 in Figure 8).

[0033] According to the 13th invention, a computer system becomes capable of determining the control time length for communication control.

[0034] The fourteenth invention is a computer system further comprising a control time length determination means for determining the length of the communication control, wherein the expression information acquisition control means provides the control time length to the control unit including the control time length in the generation instruction information, the control unit determines candidate control items to be the control items based on the control time length, and causes the control item generation unit corresponding to the candidate control item to generate expression information relating to the control item.

[0035] According to the 14th invention, the computer system can cause the generation unit to generate representation information of control items according to the control time length of the communication.

[0036] The 15th invention is a computer system in which, in the above-described computer system, the control unit causes the control item generation unit to regenerate the expression information related to the control item, based on the expression information related to the control item generated by the control item generation unit, in order to modify the expression information.

[0037] According to the 15th invention, the computer system becomes capable of regenerating the representation information generated by the control item generation unit.

[0038] The sixteenth invention is a computer system in which the generation unit comprises a generation unit on the server system side that provides the content (for example, the server-side generation AI20 and server-side generation unit 230 in Figure 1) and a generation unit on the user terminal side of the content (for example, the terminal-side generation AI30 and user terminal-side generation unit 230t in Figure 1).

[0039] According to the 16th invention, the computer system can generate representation information using both the generation unit on the server system side and the generation unit on the terminal side.

[0040] The 17th invention is a computer system in which, in the above-described computer system, the generation unit on the server system side has a control unit (for example, control unit 234 in Figure 28) that determines, based on the generation instruction information, whether to generate the representation information in the generation unit on the server system side or in the generation unit on the user terminal side.

[0041] According to the 17th invention, the computer system can decide whether to generate the representation information in the generation unit on the server system side or in the generation unit on the user terminal side.

[0042] The 18th invention is a program for causing a computer system to perform communication control when a target character communicates with a given opponent character in content featuring multiple characters, and is a program for causing the computer system to function as follows: transmission information setting means for setting transmission information for the communication; background information setting means for setting background information based on at least one of the following: the state of the content, the state of the target character, the state of the opponent character, and the relationship between the target character and the opponent character; expression information acquisition control means for generating generation instruction information based on the transmission information and the background information, and providing the generation instruction information to a given generation unit to perform control for acquiring expression information when the target character communicates; and communication control execution means for executing the communication control based on the expression information.

[0043] According to the 18th invention, it is possible to realize a program that can enable a computer system to perform the same functions as the first invention. [Brief explanation of the drawing]

[0044] [Figure 1] A system configuration diagram showing an example of a content delivery system. [Figure 2] A diagram illustrating an example of content. [Figure 3] A diagram illustrating the overview of communication control related to communication events. [Figure 4] A diagram illustrating a series of communications. [Figure 5] A diagram showing an example of the data structure of character initial settings data. [Figure 6] A diagram showing an example of the data structure of communication event definition data. [Figure 7] A diagram illustrating the situation history range setting function. [Figure 8] A diagram showing an example of a comprehensive generation instruction information written in natural language. [Figure 9] A diagram illustrating an example of server-side AI configuration and the generation of representational information for communication control. [Figure 10] A diagram to explain the machine learning of the integrated AI. [Figure 11] A diagram illustrating an example configuration when a unified AI is composed of multiple AIs, and explaining the machine learning characteristics of each. [Figure 12] A diagram illustrating the machine learning process used in scenario generation AI. [Figure 13] A diagram showing an example of control item-specific generation instruction information for scenario generation, written in natural language. [Figure 14] A diagram illustrating the generation of representational information from scenario text data. [Figure 15] A diagram illustrating machine learning in speech generation AI. [Figure 16] A diagram showing an example of generation instruction information for each control item used in speech generation, described in natural language. [Figure 17] A diagram showing an example of the data structure of speech expression information. [Figure 18] A diagram illustrating machine learning in motion generation AI. [Figure 19] A diagram showing an example of motion generation instruction information for each control item, written in natural language. [Figure 20] A diagram showing an example of the data structure for motion representation information. [Figure 21] A diagram illustrating machine learning in VFX generation AI. [Figure 22] A diagram showing an example of generation instruction information for VFX generation, categorized by control item, written in natural language. [Figure 23] A diagram showing an example of the data structure of VFX (Visual Effects) representation information. [Figure 24] A diagram illustrating the machine learning process used in AI for generating sound effects (SE). [Figure 25] A diagram showing an example of control item-specific generation instruction information for SE generation, written in natural language. [Figure 26] A diagram showing an example of the data structure of SE representation information. [Figure 27] A diagram showing examples of programs and data stored by a server system. [Figure 28] A diagram showing an example of the functional configuration of the server processing unit. [Figure 29] A diagram showing an example of the data structure of communication generation and management data. [Figure 30] A flowchart illustrating the processing flow related to content delivery executed by a server system. [Figure 31] Flowchart continuing from Figure 30. [Figure 32] A flowchart illustrating the pass / fail determination process. [Figure 33] A diagram showing an example of prerequisite evaluation instruction information written in natural language. [Figure 34] A diagram showing an example of communication information evaluation instruction information written in natural language. [Figure 35] A diagram showing an example of impression evaluation instruction information written in natural language. [Figure 36] A diagram illustrating a modified example. [Modes for carrying out the invention]

[0045] Examples of embodiments of the present invention will be described below, but it goes without saying that the embodiments to which the present invention can be applied are not limited to the following embodiments.

[0046] Figure 1 is a system configuration diagram showing an example of the configuration of a content provision system according to this embodiment. The content provision system 1000 is a computer system that provides content offering virtual experiences in a virtual space, such as video games, virtual activities, and shopping, to multiple registered users. In other words, the content provision system 1000 makes it possible to provide virtual experiences to users.

[0047] The content provision system 1000 is a computer system that includes a server system 1100 and user terminals 1500 for each user, all connected via a network 9 for data communication.

[0048] Network 9 refers to a communication path capable of data transmission. In other words, Network 9 includes not only LANs (Local Area Networks) using dedicated lines (dedicated cables) or Ethernet (registered trademark) for direct connections, but also telephone networks, cable networks, and the Internet.

[0049] The server system 1100 is a computer system that performs various processes such as managing and controlling registered user information and various controls related to content provision (for example, controlling the progress of a game), and has a database 1140.

[0050] The server system 1100 has a control board 1150 mounted on the main unit 1101. The control board 1150 is equipped with various microprocessors such as a CPU (Central Processing Unit) 1151, a GPU (Graphics Processing Unit), and a DSP (Digital Signal Processor), various IC memories 1152 such as VRAM, RAM, and ROM, and a communication device 1153. Some or all of the functions mounted on the control board 1150 may be implemented using an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a SoC (System on a Chip).

[0051] Although Figure 1 depicts the server system 1100 as a single server device, it may also be implemented using multiple devices. For example, the server system 1100 may be configured with multiple servers, each responsible for a specific function, connected to each other via an internal bus or network 9 for data communication.

[0052] The server system 1100 has a server-side generated AI 20. The server-side generating AI 20 is implemented using a machine learning-based AI model on hardware employing a multi-core architecture (e.g., a group of GPUs and memory, a group of AI chips, etc.). The hardware for the server-side generating AI 20 is not limited to being exclusive to the server system 1100. The server-side generating AI 20 may be implemented as a single multimodal generating AI model, or it may be configured as an AI group having one or more generating AIs for each type of data to be generated (e.g., text, speech, motion, etc.). In other words, the server-side generating AI 20 is the generation unit 230 on the server system side.

[0053] User terminal 1500 serves as a Man-Machine Interface (MMIF) for user 2 to play content as a player. For example, if the content to be played is a game, user terminal 1500 functions as a gameplay terminal for user 2, who is the player. Although only one user terminal 1500 is depicted in Figure 1, in actual operation, it is common for multiple user terminals 1500 to communicate and connect to the server system 1100 simultaneously.

[0054] The user terminal 1500 is a computer system that can connect to the network 9, such as a personal computer, smartphone, wearable computer, portable game console, home game console, or tablet computer.

[0055] The user terminal 1500 is a computer comprising an operation input device, an image display device, a communication device, and a control board 1550 for performing calculations. Examples of operation input devices include a touch panel 1506, a keyboard, a game controller, and a mouse. Examples of image display devices include a touch panel 1506, a head-mounted display, and a glasses-type display.

[0056] The control board 1550 is equipped with a CPU 1551, various microprocessors such as a GPU and DSP, various IC memories 1552 such as VRAM, RAM, and ROM, and a communication module 1553 that connects to the network 9. These elements mounted on the control board 1550 are electrically connected via bus circuits and the like, enabling data reading and writing, and signal transmission and reception. Part or all of the control board 1550 may be an ASIC, FPGA, or SoC.

[0057] The control board 1550 stores programs and various data necessary to realize the functions of the user terminal 1500 in the IC memory 1552. The user terminal 1500 realizes the functions of a play terminal for playing content by executing a predetermined application program.

[0058] The user terminal 1500 has a terminal-side generated AI 30. The terminal-side generation AI 30 is implemented using a machine learning-based AI model on hardware employing a multi-core architecture (e.g., a group of GPUs and memory, a group of AI chips, etc.). The hardware of the terminal-side generation AI 30 is not limited to being exclusive to the user terminal 1500. The terminal-side generation AI 30 may be implemented as a single multimodal generation AI model, or it may be configured as an AI group having one or more generation AIs for each type of data to be generated (e.g., text, voice, motion, etc.). In other words, the terminal-side generation AI 30 is the user terminal-side generation unit 230t. In this embodiment, the terminal-side generation AI 30 is assumed to be a standalone generation AI that generates scenario text including character dialogue text.

[0059] Figure 2 is a diagram illustrating an example of content. The content provision system 1000 provides a video game as content, in which a player character 4 and NPCs 5 (5a, 5b, ...) appear in a game space constructed in a virtual three-dimensional space. The game genre can be set as appropriate, but for example, as shown in the example of game screen W2, it may be an action RPG (Role Playing Game) in which the story unfolds as the player character 4 and NPCs 5 fight together as allies against enemy NPCs 6.

[0060] Figure 3 is a diagram illustrating the overview of communication control related to communication events, which are in-game events. During gameplay, when a communication event occurs, a series of communications take place between multiple NPCs (NPCs 5) within that event. In the example shown in Figure 3, as illustrated in the example of event screen W4, NPCs 5a and 5b encounter and fight an enemy NPC 6d after becoming separated from the player character 4. Before the battle begins, NPCs 5a and 5b communicate with each other to confirm their willingness to fight.

[0061] Figure 4 is a diagram illustrating a series of communications. A series of communications consists of a sequence of N individual communications (the parts enclosed by dotted lines in Figure 4) that are executed according to a given execution order (where N is an integer greater than or equal to 1). The number of individual communications that make up a series of communications is determined by the status of the content at that time and the status of the communicating NPC5 (details below).

[0062] Focusing on individual communication, there is a target character 7 that transmits the information and a counterpart character 8 that receives the information. In individual communication, there is only one target character 7, but the counterpart character 8 may be one or multiple depending on the game's progress.

[0063] Information to be communicated is the content or information you want to convey to the other party through communication. Information is conveyed through various forms of communication by the target character 7. These forms of communication include, for example, displaying dialogue text, emitting spoken audio of the dialogue text, and the target character 7's motion (including gestures, hand signals, etc.). Other forms of communication include the display of visual effects (VFX) and the emission of sound effects (SE; BGM). The choice of which form of communication to use depends on the content and the NPC's situation at the time (details below).

[0064] The server system 1100, for each individual communication, has the server-side AI 20 generate and acquire expression information for the target character 7, which is NPC 5, to express the communication information. Then, it uses the acquired expression information to cause the target character 7 to express the communication. This is called communication control for the target character 7.

[0065] Returning to Figure 3, the server system 1100 instructs the server-side generation AI 20 and the terminal-side generation AI 30 to generate representation information to show that NPC 5a and NPC 5b are communicating. To this end, the server system 1100 generates comprehensive generation instruction information 730 to be given to the server-side generation AI 20, and in doing so, includes (or attaches) background information 728 to the comprehensive generation instruction information 730.

[0066] Background information 728 is information about the factors that the server-side AI 20 uses to decide how to express the content to be conveyed in communication, and it is information about various background aspects of the communication. Background information 728 is set based on at least one of the following: the content status, the status of target character 7, the status of opponent character 8, the personality of target character 7, the personality of opponent character 8, the relationship between target character 7 and opponent character 8, and information about the communication event that occurred.

[0067] In this embodiment, the status of the content corresponds to the progress of the game. The content status is indicated by the content status log 706, which the server system 1100 continuously records at predetermined intervals or after each game operation, starting from the moment the content is provided. The content status log 706 may also be replay data that allows the game to be reproduced.

[0068] A single content status log 706 stores various data describing the game's progress at the time of recording. For example, content status log 706 includes the play time at the time of recording (e.g., elapsed time since the start of play), the game world date and time, the game progress, the game stage ID and event ID being played, and data on game world background objects. Of course, other data may also be included as appropriate.

[0069] Character Status Log 708 is a log related to the content status, but it is a character-specific log prepared for each character appearing (Player Character 4, NPC 5, Enemy NPC 6), and stores various information indicating the latest status of the character at the time of recording. The status of target character 7 and the status of opponent character 8 will be shown in the character status log 708 corresponding to those characters.

[0070] A single character status log 708 may include, for example, the character ID, the time of play at the time of recording, the character's position coordinates, the character's level, and ability parameter values ​​(e.g., HP, MP, skill points, etc.). Other information such as status ailments, equipment information, and the name of the group the character belongs to may also be included as appropriate.

[0071] The personality of target character 7, the personality of opponent character 8, and the relationship between target character 7 and opponent character 8 are all described in the personality information 530 of the character initial setup data 520, which is prepared for each NPC 5.

[0072] Figure 5 shows an example of the data structure of character initial setting data 520. The character initial setup data 520 includes the character ID 522, character model data 524, personality information 530, and character relationship information 535. Of course, other data may be included as appropriate.

[0073] Personality information 530 includes basic setting information 531, personality description text 532, speech pattern description text 533, and gesture characteristic description text 534. Of course, other data may be included as appropriate.

[0074] Basic settings information 531 stores information such as the character's name, age, and race. Character description text 532 is a natural language description that explains the character's personality. Speech style description text 533 is a natural language description of the character's speech style. The gesture characteristic description text 534 is a natural language description that explains the characteristics of the character's gestures, body language, and mannerisms.

[0075] Character relationship information 535 is various information that indicates the relationship between the character in question and the related character (player character 4, other NPC 5). For example, one character relationship information 535 stores the related character ID and the relationship description text in association. The relationship description text is a natural language explanation that describes the relationship between the character in question and the related character (what the character thinks).

[0076] Information about the communication events that occur is predefined by communication event definition data 550, which is prepared for each event, as shown in Figure 6.

[0077] A single communication event definition data 550 includes an event ID 552, a target character ID 554, and event occurrence requirements 556. It also includes background information 560, expression form information 562, communication information 564, and performance policy information 566. Of course, other data may be included as appropriate.

[0078] The starting character ID 554 indicates which NPC5 will be the target character 7 in the first individual communication.

[0079] The event occurrence requirement 556 specifies the conditions that must be met for the communication event to occur. This requirement is described by combining one or more subconditions 557 with AND or OR. Examples of subconditions 557 may include pre-content status condition 557a, pre-target character status condition 557b, current content status condition 557c, current target character status condition 557d, current opponent character status condition 557e, character relationship condition 557f, etc.

[0080] Pre-content status condition 557a and pre-target character status condition 557b indicate the conditions that must be met regarding the status of past content and the status of past target characters, respectively. For example, pre-content status condition 557a could be "battling in a dungeon" or "the event in the dungeon where the boss character appears has not yet occurred." For example, pre-target character status condition 557b could be "was inside a dungeon" or "target character 7 used a specified number of healing items." In other words, each sub-condition 557 is set as a condition that must be met regarding the circumstances leading up to the current situation.

[0081] The extent to which the historical context needs to be considered to determine whether sub-condition 557 is met is determined each time using a predetermined historical context range setting function 580, as shown in Figure 7.

[0082] The situation history range setting function 580 calculates a log reference range specification value (for example, a past range of play time, the number of logs to go back in time, etc.) that references the content status log 706 and the character status log 708 (see Figure 3), using parameter values ​​that describe the game situation as variables. The parameter values ​​that describe the game situation here may be, for example, content progress (game progress, number of game stages cleared, etc.), the character level of target character 7, the number of opponent characters 8, the intimacy level between target character 7 and opponent character 8, etc.

[0083] Returning to Figure 6, the current content status condition 557c, the current target character status condition 557d, and the current opponent character status condition 557e indicate the conditions that must be met for the current content status, the status of target character 7, and the status of opponent character 8, respectively. For example, content status condition 557c could be "an event has occurred in a village under attack by enemies" or "target character 7 and opponent character 8 are alone together." Current target character status condition 557d could be "reached within a predetermined distance from the village under attack by enemies." Opponent character status condition 557e could be "HP and MP are both below half."

[0084] Character relationship condition 557f specifies the conditions that must be met regarding the relationship between target character 7 and opponent character 8.

[0085] Furthermore, these sub-conditions 557 may effectively be treated as "not specified" or "unlimited" in some cases.

[0086] The background situation information 560 is text data that describes the sequence of events, the situation in which target character 7 and opponent character 8 are located, etc., in natural language.

[0087] The expression form information 562 is text data that describes in natural language which character communicates in what form during the communication event. The expression form information 562 also describes the correspondence between the character type and the communication information (first communication information and second communication information in the example in Figure 6).

[0088] The communication information 564 is text data that describes the content that the initiating character 7 wants to convey to the other character 8 in a communication event, with the first communication information being the content that the other character 8 wants to convey to the initiating character 7, and the second communication information being the content that the other character 8 wants to convey to the initiating character 7. The communication information 564 may not be specific text data, but may also contain predetermined data that means "not set". In this case, the conversation in the communication event corresponds to ad-libs generated by the server-side AI 20.

[0089] The production policy information 566 is information that indicates how the communication event should be directed and is written in natural language. For example, the production policy information 566 includes the control time length that can be used for a series of communications in the communication event, the number of times individual communications will be executed, and instructions regarding VFX and sound effects.

[0090] Figure 8 shows an example of a comprehensive generation instruction information 730 written in natural language. The comprehensive generation instruction information 730 includes a description of the purpose of generation 731, a description of the premise situation 732, a description of the expression form 733, a description of the production policy 735, reference information 736, and setting communication information 737. Of course, other descriptions may also be included as appropriate. The background information 728 consists of the description of the premise situation 732 to the reference information 736.

[0091] The generation purpose description 731 is a common template. The premise description 732 is a copy of the premise information 560 (see Figure 6) for the communication event that will occur. The expression form description 733 is a copy of the expression form information 562 (see Figure 6) of the communication event that occurs. The expression form description 733 also serves as item specification information 734, which specifies candidate control items that should be used as control items for the target character 7 in individual communication.

[0092] The production policy description 735 is a copy of the production policy information 566 (see Figure 6) for the communication events that will occur.

[0093] Reference information 736 includes the following information (or address information): (1) to (3). (1) A copy 736a of the personality information 530 (see Figure 5) of the initiating character (NPC5 with character ID 554, the initiating character of the communication event that occurs) and the other character 8. (2) A copy 734b of the character relationship information 535 from the perspective of the initiating character to the opposing character 8, and the character relationship information 535 from the perspective of the opposing character 8 to the initiating character (see Figure 5). (3) A copy of the log 734dc of the content status log 706 and the character status log 708 (see Figure 3) within the status history range.

[0094] The configuration information 737 is a copy of the communication information 564 (see Figure 6) of the communication event that occurred, and indicates the communication information set for that event.

[0095] Figure 9 is a diagram illustrating an example configuration of the server-side generated AI20 and the generation of representation information for communication control. The server-side generated AI 20 includes a general AI 21 and control item-specific generated AIs 23 (23a, 746b, ...) prepared for each control item.

[0096] Control items are control items in communication control, and they represent the types of forms of communication expression. Examples of control items include dialogue text (including narration), spoken voice of a character reading the dialogue text, character motion, visual effects (VFX), and auditory effects (SE, BGM).

[0097] Therefore, the server-side generation AI 20 includes a scenario generation AI 23a that generates a scenario for a communication event including dialogue text, a speech voice generation AI 23b that generates spoken audio, and a motion generation AI 23c that generates motion. It also includes a VFX generation AI 23d that generates VFX, and a SE generation AI 23e that generates sound effects and background music.

[0098] The overall AI 21 determines, in the order of execution of individual communications, which control item-specific generation AI 23 will be responsible for generating the control items related to each individual communication, and in what order. It then generates control item-specific generation instruction information for the assigned generation AI and provides it to each assigned generation AI, and obtains control item-specific expression information 766 (756a, 756b, ...) based on the generated data. The overall AI 21 can be said to be in charge of supervision or production, directing the various control item-specific generation AI 23 to generate the expression information 766.

[0099] The generated representation information 766 is associated with the individual communication execution order 762 and the target character ID 764 for the individual communication, and is stored as a single communication dataset 760. Then, the communication dataset 760 is applied to the NPC5 indicated by the target character ID 764 in the communication control of the individual communication execution order 762.

[0100] The overall AI 21 sets the assigned generation AI from among the control item-specific generation AIs of the server-side generation AI 20 and the control item-specific generation AIs of the terminal-side generation AI 30. In this embodiment, the terminal-side generation AI 30 consists only of the scenario generation AI 33a. Therefore, in this embodiment, the overall AI 21 sets the scenario generation AI responsible for scenarios including dialogue text to either the server-side scenario generation AI 23a or the terminal-side scenario generation AI 33a. Specifically, the former may be designated as the responsible generation AI if the amount of information in the communication information included in the comprehensive generation instruction information 730 reaches a predetermined threshold, i.e., if it is likely to be a long line of dialogue; otherwise, the latter may be designated as the responsible generation AI. Of course, the terminal-side generation AI 30 may also be configured to have multiple generation AIs for different control items, similar to the server-side generation AI 20.

[0101] Figure 10 is a diagram illustrating the machine learning capabilities of the integrated AI21. The integrated AI 21 receives the comprehensive generation instruction information 730 prepared for training, a list of available generation AIs for each control item, load information for each generation AI for each control item, and annotations as input. It is pre-trained to output a list of generation control items in the order of generation for each individual communication, and generation instruction information for each generation item for each individual communication, and it supports generation instructions in natural language.

[0102] As mentioned above, the comprehensive generation instruction information 730 includes various types of information used for communication control, such as the control time length, content status, character status, communication information, and production policy.

[0103] The list of generation order control items for each individual communication shows the control items for communication control in the order they are generated, corresponding to the execution order of the N individual communications that make up the series of communications performed by the communication event that occurred.

[0104] Candidate control items for individual communication include dialogue text (including narration), spoken voice of the character reading the dialogue text, character motion, visual effects (VFX), and auditory effects (SE, BGM). Which of these control items is applied is determined by the overall AI 21 based on various information included in the comprehensive generation instruction information 730, such as the control time length, content status, character status, communication information settings, and performance policy settings.

[0105] For example, if, as the first individual communication, target character 7 mutters without moving, the dialogue text and spoken audio are selected as control items. If the control items include dialogue text, the overall AI 21 prioritizes the generation of the dialogue text as the first priority. The second priority and subsequent priorities may be assigned in order of the control items that are expected to take the longest to generate.

[0106] Figure 11 is a diagram illustrating an example configuration of the overall AI 21 when it is composed of multiple AIs, and the machine learning processes of each. The overall AI 21 may include a control item setting AI 21a, a generation 담당 setting AI 21b, and a generation instruction information generation AI 21c.

[0107] The control item setting AI 21a is pre-trained to take the comprehensive generation instruction information 730 and annotations prepared for learning as input and output a list of generation order control items for each individual communication, and it supports generation instructions in natural language. Based on the description of the control time length included in the performance policy description 735 of the comprehensive generation instruction information 730, the control item setting AI 21a determines the control items for each individual communication so that the series of communications is completed within the given time limit.

[0108] The generation-responsible AI 21b is pre-trained to take as input the individual communication-specific generation order control item list output from the control item setting AI 21a prepared for training, the available control item-specific generation AI list, load information for each control item-specific generation AI, and annotations, and to output a list of generation AIs responsible for generating each input control item (control item-specific generation AI list). Furthermore, the generation-responsible AI 21b can handle generation instructions in natural language.

[0109] The AI ​​21c, which generates generation instructions, is pre-trained to take the comprehensive generation instructions prepared for training, the list of AIs responsible for each control item output by the generation assignment setting AI 21b, and annotations as input, and to output generation instructions for each control item corresponding to the input generation AI. Furthermore, the AI ​​21c supports generation instructions in natural language.

[0110] Figure 12 is a diagram illustrating the machine learning process of the scenario generation AI 23a. The scenario generation AI23a is implemented, for example, as a Large Language Model (LLM) of the Generative Pretrained Transformer (GPT) type.

[0111] The scenario generation AI23a is pre-trained to output scenario text data by inputting generation instruction information for each control item prepared for scenario generation, comprehensive generation instruction information 730, and annotations, and supports generation instructions in natural language.

[0112] In generating a scenario, the scenario generation AI 23a determines the content of individual communications so that a series of communications are completed within the control time length included in the production policy description 735 of the overall generation instruction information 730, and outputs scenario text data that includes these communications.

[0113] Furthermore, the scenario generation AI 23a has been reinforced through prior training so that, upon input of training scenario text data, it outputs a classification of the series of communications depicted in the given scenario. In other words, the scenario generation AI 23a can estimate the personalities of the characters involved, based on the preconditions under which the series of communications depicted in the scenario take place, and output the classification result.

[0114] Furthermore, the scenario generation AI 23a has been reinforced through prior training so that, when given a text to be evaluated and a reference text, it can output an evaluation of whether the reference text can be considered to contain the text to be evaluated. In other words, the scenario generation AI 23a can compare and evaluate the semantic content of the text to be evaluated (e.g., scenario text data) with the semantic content of the reference document (e.g., communication information 564 in the communication event that occurred; see Figure 6).

[0115] Furthermore, the scenario generation AI 23a has been reinforced through prior training so that, when given a text to be evaluated and a reference text, it can output an evaluation of whether the impression given to the reader of the text to be evaluated matches the impression specified by the reference text. In other words, the scenario generation AI 23a can compare the impression evaluation of the content of the text to be evaluated (e.g., scenario text data) with the impression evaluation of the content of the reference text (e.g., communication policy information 568 in the communication event that occurred; see Figure 6).

[0116] Figure 13 shows an example of control item-specific generation instruction information 742a for scenario generation, which is provided to the scenario generation AI 23a written in natural language. For the content enclosed in <> in the text of the control item-specific generation instruction information 742a, the scenario generation AI 23a refers to the overall generation instruction information 730 (see Figure 8), which is provided together with the control item-specific generation instruction information 742a for scenario generation.

[0117] Figure 14 is a diagram illustrating the generation of representation information 766 from scenario text data. The scenario generation AI 23a generates scenario text data 750, detects and separates character-specific lines in the scenario individually (for example, as a single unit of quotation marks in the scenario), and generates individual line data 752. In addition, the scenario generation AI 23a detects and separates character-specific action descriptions (silent gestures, silent body language) from the stage directions of the scenario individually, and generates action description data 754 without dialogue.

[0118] The individual dialogue data 752 stores the dialogue character ID indicating the target character 7 that speaks the dialogue, the order in which the individual communication is performed, and the dialogue text data in association with each other. The individual dialogue data 752 becomes the expression information 766a of the dialogue text.

[0119] The action description data 754 without dialogue includes a dialogue character ID indicating the target character 7 performing the action without dialogue, the order in which individual communications are executed, and the action description text.

[0120] Motion expression information 766c is generated based on motion description data 754 without dialogue. The motion description text is included in the control item-specific generation instruction information 742 and provided to the motion generation AI 23c, which then determines what type of motion the motion generation AI 23c will generate.

[0121] Figure 15 is a diagram illustrating the machine learning process of the speech generation AI 23b. The speech generation AI23b is pre-trained to take in speech generation instruction information for each control item prepared for training, as well as annotations, and output speech representation information 766b, and supports generation instructions in natural language.

[0122] Figure 16 shows an example of generation instruction information 742b for speech generation, categorized by control item, which is provided to the speech generation AI 23b, which is written in natural language. The control item-specific generation instruction information 742b for generating spoken speech includes prerequisites, the dialogue text to be converted into speech, the production policy, and information about the personality and tone of voice of the target character 7 to speak.

[0123] The prerequisites and production guidelines are copied from the prerequisite situation description 732 and production guideline description 735 (see Figure 8) of the comprehensive generation instruction information 730. The dialogue text to be converted to speech is copied from the expression information 766a of the dialogue text. Regarding the personality and speech patterns of target character 7, information about the personality and speech patterns of target character 7 is copied from reference information 736 (see Figure 8) of the comprehensive generation instruction information 730.

[0124] Figure 17 shows an example of the data structure of the speech expression information 766b. The speech expression information 766b includes the target character ID, the order in which individual communication is performed, and the generated speech data. The speech data is audio data in which the speech generation AI 23b pronounces the input dialogue text in the input context, with the input character's personality and in the input tone.

[0125] Figure 18 is a diagram illustrating the machine learning process of motion generation AI23c. The motion generation AI23c is pre-trained to take control item-specific generation instruction information and annotations prepared for training as input, and to output motion data representation information 766c, and supports generation instructions in natural language.

[0126] Figure 19 shows an example of motion generation instruction information 742c for each control item used for motion generation, which is provided to the motion generation AI 23c written in natural language. The motion generation instruction information 742c for each control item includes information on the dialogue text or action description without dialogue to match the motion, and the status of the target character 7. It also includes information on the personality of the target character 7 that will perform the motion, the gesture characteristics of the target character 7 that will perform the motion, and the direction policy.

[0127] For dialogue text, the expression information 766a of the dialogue text is copied (see Figure 14). For action descriptions without dialogue, the action description data 754 without dialogue is copied (see Figure 14). Regarding the status of target character 7, the information concerning the status history of target character 7 is copied from reference information 736 of the comprehensive generation instruction information 730 (see Figure 8). The personality and mannerisms of target character 7 are derived from the personality information of the originating target character in the reference information 736 of the comprehensive generation instruction information 730. Regarding the production policy, the production policy description 735 of the comprehensive generation instruction information 730 (see Figure 8) is copied.

[0128] Figure 20 shows an example of the data structure of motion representation information 766c. Motion representation information 766c includes the target character ID, the order in which individual communications are performed, and the generated motion data. The motion data is a motion used when the target character 7 speaks dialogue text or a motion that performs an action description without dialogue, and is a motion that is in line with the direction policy and the personality and gesture characteristics of the target character 7.

[0129] Figure 21 is a diagram illustrating the machine learning process of the VFX generation AI23d. The VFX generation AI 23d is pre-trained to take VFX generation control item-specific generation instruction information and annotations prepared for training as input and output VFX expression information 766d, and it supports generation instructions in natural language.

[0130] Figure 22 shows an example of VFX generation instruction information 742d, categorized by control item, provided to the VFX generation AI 23d, which is written in natural language. The VFX generation instruction information 742d, which is categorized by control item, includes information on dialogue text or action descriptions without dialogue, and the direction policy. For dialogue text, the expression information 766a of the dialogue text is copied (see Figure 14). For action descriptions without dialogue, the action description data 754 without dialogue is copied (see Figure 14). For the direction policy, the direction policy description 735 of the comprehensive generation instruction information 730 (see Figure 8) is copied.

[0131] Figure 23 shows an example of the data structure of VFX representation information 766d. The VFX representation information 766d includes the target character ID, the order in which individual communication is performed, and the generated VFX data. The VFX data is data for displaying when the target character 7 speaks dialogue text or performs actions without dialogue, and is data for executing visual effects in accordance with the directorial policy.

[0132] Figure 24 is a diagram illustrating the machine learning process of SE generation AI23e. The SE generation AI23e is pre-trained to take control item-specific generation instruction information and annotations prepared for training as input and output SE representation information 766e, and supports generation instructions in natural language.

[0133] Figure 25 shows an example of control item-specific generation instruction information 742e for SE generation, which is provided to the SE generation AI23e written in natural language. The control item-specific generation instruction information 742e for SE generation includes information on dialogue text or action descriptions without dialogue, and the direction policy. For dialogue text, the expression information 766a of the dialogue text is copied. For action descriptions without dialogue, the action description data 754 without dialogue is copied. For the direction policy, the direction policy description 735 of the comprehensive generation instruction information 730 is copied.

[0134] Figure 26 shows an example of the data structure of SE representation information 766e. SE expression information 766e includes the target character ID, the order in which individual communication is performed, and the generated SE data and BGM data. The SE data and BGM data are sound data emitted when the target character 7 speaks dialogue text or performs actions without dialogue, and are data for performing auditory effects in accordance with the production policy.

[0135] Figure 27 shows an example of programs and data stored by the server system 1100. The server system 1100 stores the server program 501, the distribution client program 503, the trained AI model 508, user information 600 for each user 2, and play data 700 in the IC memory 1152. It also stores content initial setup data 510 in the database 1140. Of course, other data may be stored as appropriate.

[0136] The server system 1100 performs the server program 501 on the CPU 1151, thereby realizing the function of a server processing unit 200s, as shown in Figure 28. The server program 501 may include a generation AI program 502 for realizing the functions of a server-side generation AI 20 and a control item-specific generation AI 23. Alternatively, the generation AI program 502 may be stored separately.

[0137] The server processing unit 200s performs various controls related to content provision. Specifically, the server processing unit 200s includes a user information management unit 202, a billing control unit 204, a content provision control unit 206, a sufficiency monitoring unit 210, a transmission information setting unit 212, a background information setting unit 214, and an expression information acquisition control unit 216. It also includes a communication control execution unit 218, a control time length determination unit 220, a generation unit 230, an available generation unit monitoring unit 236, a processing load monitoring unit 238, and a timing unit 290.

[0138] The user information management unit 202 executes the prescribed player registration procedure (user registration procedure) and controls the registration and management of user information 600 for each user 2.

[0139] The billing control unit 204 processes payments for content. For example, it processes payments for playing content and for purchasing paid items usable in the content. Once the billing control unit 204 has processed a payment, it causes the user information management unit 202 to update the billing information contained in the user information 600.

[0140] The content provision control unit 206 controls the progress of the content and provides it to the user terminal 1500. In this embodiment, the content is a video game, so the control will be for the progress of gameplay.

[0141] The satisfaction monitoring unit 210 has defined preconditions for satisfying the event occurrence requirements 556 (see Figure 6) for each communication event, and monitors whether the event occurrence requirements are satisfied when the preconditions are met. Specifically, if the event occurrence requirements 556 include preconditions for content status 557a and preconditions for target character status 557b, the satisfaction monitoring unit 210 monitors whether the event occurrence requirements 556 are satisfied, including whether these conditions are met during content provision.

[0142] The communication information setting unit 212 sets the communication information for individual communications. Specifically, the communication information setting unit 212 refers to the communication information 564 of the communication event definition data 550 that satisfies the event occurrence requirement 556 (see Figure 6).

[0143] The background information setting unit 214 sets background information 728 based on at least one of the following: the status of the content, the status of the target character 7, the status of the opponent character 8, and the relationship between the target character 7 and the opponent character 8 (see Figure 3).

[0144] The expression information acquisition control unit 216 generates generation instruction information based on the transmission information 564 and background information 728, and provides this generation instruction information to a given generation AI (Artificial Intelligence) to perform control to acquire expression information 766 when the target character 7 engages in individual communication.

[0145] The communication control execution unit 218 performs communication control based on the representation information 766.

[0146] The control time length determination unit 220 determines the control time length, which is the length of the communication control. Specifically, the control time length determination unit 220 refers to the time length specified in the performance policy information 566 of the communication event definition data 550 (see Figure 6).

[0147] The generation unit 230 performs various generation processes related to the acquisition of representation information 766. The server-side generation AI 20 is an example of this (see Figure 9). The generation unit 230 includes a control item-specific generation unit 232 for each candidate control item, and a general unit 234 that causes the control item-specific generation unit 232 to generate representation information 766 based on the overall generation instruction information 730.

[0148] The available generation unit monitoring unit 236 monitors the available control item generation unit 232.

[0149] The processing load monitoring unit 238 monitors the processing load status for each control item generation unit 232.

[0150] The timing unit 290 uses the system clock to determine the current date and time 900, as well as various times related to the provision of content.

[0151] Returning to Figure 27, the distribution client program 503 is downloaded and executed by the user terminal 1500, thereby enabling the user terminal 1500 to function as a game client for various controls and as a human-machine interface for the content provision system 1000.

[0152] A pre-trained AI model 508 is prepared for each AI that makes up the server-side generation A20 (overall AI 21, control item-specific generation AI 23).

[0153] The content initial setup data 510 stores various initial setup data necessary for the execution of the content. For example, the content initial setup data 510 includes character initial setup data 520 (see Figure 5), communication event definition data 550 (see Figure 6), and situation history range setting function 580 (see Figure 7).

[0154] User information 600 is prepared for each user 2 who has completed the prescribed registration procedure, and stores various information associated with that user. One user information 600 includes, for example, a user account, content save data, and billing information (for example, billing history associating billing date and time with billing amount, billing history statistics such as cumulative billing amount, number of billings, and billing frequency).

[0155] Play data 700 stores various information related to content provision and is updated by the content provision control unit 206. The play data 700 includes game world background object control data 702 for each object that constitutes the background of the game world, such as the terrain of the game world, and character control data 704 for each character, such as player character 4 and NPC 5. It also includes content status log 706 and character status log 708 (see Figure 3), generated AI monitoring data 710, and communication generation management data 720.

[0156] The generated AI monitoring data 710 is prepared separately for each generated AI, the server-side generated AI 20 and the terminal-side generated AI 30, and stores a availability flag for the AI ​​in question, as well as processing load status information (e.g., CPU usage, memory usage, etc.).

[0157] The communication generation and management data 720 stores various data related to the generation and acquisition of expression information 766 and other data related to a series of communications performed in a communication event.

[0158] For example, as shown in Figure 29, the communication generation management data 720 includes an event ID 722 indicating the communication event that occurred, a target character ID 724, a partner character ID 726, and background information 728. Furthermore, the communication generation management data 720 includes overall generation instruction information 730, a list of generation order control items 740, generation instruction information for each control item 742, scenario text data 750, and individual dialogue data 752 (see Figure 14).

[0159] Furthermore, the communication generation and management data 720 includes action description data 754 without dialogue (see Figure 14) and communication datasets 760 for each individual communication (see Figure 9).

[0160] Figures 30 and 31 are flowcharts illustrating the processing flow related to content provision performed by the server system 1100.

[0161] Server system 1100 begins processing and providing content (step S10). If a communication event occurs during content delivery (YES in step S12), the server system 1100 determines the target character 7 and the other character 8 in the initial individual communication (step S14). The former is NPC5 of the target character ID 554 that initiated the event. The latter is another NPC5 located within a predetermined range from NPC5 of the target character ID 554 (see Figure 6).

[0162] Next, the server system 1100 sets the status and history range (step S22) and sets the background information 728 (step S24; see Figure 3). Next, the server system 1100 reads the transmission information 564 from the communication event definition data 550 (see Figure 6) of the event that occurred and sets the configuration transmission information 737 (see Figure 8) (step S25). Then, it creates a comprehensive generation instruction information 730 that includes the configuration transmission information 737 and background information 728 (step S26; see Figure 8).

[0163] Next, the server system 1100 provides the overall generation instruction information 730, the list of available control item-specific generation AIs, and the load information for each control item-specific generation AI to the overall AI 21 (step S28). At this time, depending on the data size of the background information 728, the background information 728 may not be included in the overall generation instruction information 730, and some or all of it may be provided as separate data from the overall generation instruction information 730.

[0164] When the overall generation instruction information 730 is provided to the overall AI 21, it outputs a list of generation order control items 740 for each individual communication that constitutes a series of communications in the event that occurred. It also assigns a responsible generation AI to each control item and outputs control item-specific generation instruction information 742 to be given to the responsible generation AI. In other words, the server system 1100 obtains the generation order control item list 740 and the control item-specific generation instruction information 742 (step S28).

[0165] Next, the server system 1100 provides the scenario generation AI 23a, which has become the responsible generation AI, with control item-specific generation instruction information 742 and overall generation instruction information 730 to generate scenario text data 750 (step S40; see Figure 14). Then, it performs a pass / fail judgment process on the generated scenario text data 750 (step S42).

[0166] Figure 32 is a flowchart illustrating the pass / fail determination process. In the pass / fail judgment process, the server system 1100 instructs the scenario generation AI 23a to estimate the preconditions under which the series of communications depicted in the generated scenario are carried out. Then, it calculates the similarity between the result of the estimated classification (hereinafter referred to as "estimated classification") and the classification of the situation described in the precondition situation information 560 (see Figure 6) of the event that occurred (step S60). Specifically, the server system 1100 creates precondition evaluation instruction information 770 written in natural language, as shown in Figure 33, and provides it to the scenario generation AI 23a.

[0167] Returning to Figure 32, if the calculated similarity of the situations meets a predetermined passing standard (YES in step S62), the server system 1100 instructs the scenario generation AI 23a to evaluate whether the communication information 564 (see Figure 6) in the event that occurred can be considered to be included in the generated scenario (step S64). Specifically, the server system 1100 creates communication information evaluation instruction information 772 written in natural language, as shown in Figure 34, and provides it to the scenario generation AI 23a.

[0168] Returning to Figure 32, if it can be considered that the communication information 564 in the event that occurred is included in the dialogue (YES in step S66), the server system 1100 has the scenario generation AI 23a estimate and classify the impression (performance effect) that the generated scenario will give to the reader. Then, it has the AI ​​23a calculate the similarity between the result of the estimation and classification and the "performance policy" indicated in the event's performance policy information 566 (see Figure 6) (step S68). Specifically, the server system 1100 creates impression evaluation instruction information 774 written in natural language, as shown in Figure 35, and provides it to the scenario generation AI 23a.

[0169] Returning to Figure 32, if the calculated similarity meets the predetermined passing criteria (YES in step S70), the server system 1100 considers that the scenario conforms to the production policy for the event in which it occurred, and determines that the scenario is acceptable (step S72).

[0170] If any of steps S62, S66, or S70 is negative, the server system 1100 determines the scenario to be unsuccessful (step S74).

[0171] Returning to Figure 31, if the scenario fails (NO in step S90), the server system 1100 discards the scenario text data 750 (step S92) and regenerates the scenario text data 750 (step S94). At this time, the randomness parameter value included in the control item-specific generation instruction information 742 for scenario generation may be slightly changed to increase the randomness before regeneration.

[0172] If the scenario is approved (YES in step S90), the server system 1100 separates individual lines of dialogue from action descriptions without dialogue from the scenario text data 750. It then creates individual line of dialogue data 752 and action description data without dialogue 754 (step S96; see Figure 14). The individual line of dialogue data 752 becomes the expression information 766a of the dialogue text.

[0173] Moving to Figure 31, the server system 1100 executes loop A in the order of execution of individual communications (steps S100 to S106). In loop A, the server system 1100 refers to the generation order control item list 740 for the individual communication to be processed, for control items other than the dialogue text in the individual communication to be processed. Then, according to that generation order, it generates and obtains the respective expression information (step S102).

[0174] Next, the server system 1100 creates a communication dataset 760 for the individual communication to be processed (step S104; see Figure 9), and terminates loop A (step S106).

[0175] Once Loop A has been executed for all individual communications, the server system 1100 applies the communication dataset 760 to the individual communication execution sequence 762. Then, it performs communication control to realize the series of communications (step S108).

[0176] If the content provision has not ended (NO in step S120), return to step S12. If the content provision is finished (YES in step S120), the server system 1100 saves the data (step S124) and terminates the series of processes.

[0177] As described above, according to this embodiment, control over NPC communication can be realized using the generation unit.

[0178] In other words, the server system 1100 automatically generates and acquires representation information 766 from the generation unit 230 based on at least one of the following: the status of the content, the status of the target character 7, the status of the opponent character 8, and the relationship between the target character 7 and the opponent character 8. This reduces the effort and time required for content creation.

[0179] [Variation] Although examples of embodiments to which the present invention is applied have been described above, the forms to which the present invention can be applied are not limited to the above forms, and components can be added, omitted, or modified as appropriate.

[0180] (Variation 1) For example, the content provision system 1000 may be implemented using a P2P (Peer to Peer) architecture with multiple user terminals 1500. In this case, programs and data corresponding to the functional division are stored in the user terminals 1500, and the functions corresponding to the server processing unit 200s in the above embodiment are distributed and implemented by the user terminals 1500, which act as P2P nodes. The same effects as in the above embodiment can be obtained with this configuration as well.

[0181] (Variation 2) For example, the above embodiment may be modified to include a configuration that changes the amount of information in the comprehensive generation instruction information 730 based on the billing information of user 2. Specifically, as shown in Figure 36, when the server system 1100 creates the comprehensive generation instruction information 730, it refers to the billing information among the user information 600 of user 2 and determines the amount of information in the comprehensive generation instruction information 730 using a predefined information amount setting function 536. Then, it applies this to the amount of information in the newly created comprehensive generation instruction information 730.

[0182] The information quantity setting function 536 is a function that calculates information quantities that have a positive relationship with variables such as the total billing amount and billing frequency.

[0183] (Variation 3) In the above embodiment, regarding the communication control of individual communication, there were cases in which the control items included dialogue text and cases in which they did not. Specifically, an example of the latter case is the communication control of individual communication using action description data 754 (see Figure 14) that does not include dialogue.

[0184] However, it is also possible to omit the separation of the action description data 754 that does not include dialogue. In this configuration, communication control will be executed by including at least the dialogue of the target character 7 among the candidate control items. This configuration is suitable for providing content intended for users who cannot understand gesture-based communication.

[0185] (Variation 4) In the above embodiment, the expression form description 733 (see Figure 8) also serves as the item specification information 734, thereby specifying fixed control items related to the communication control of individual communications in the communication event, but the embodiment is not limited to this.

[0186] For example, the above embodiment may be modified to include a configuration in which the item specification information 734 is determined variably according to the number of target characters 7. Specifically, the number of opponent characters (8) plus the number of initiating target characters is considered to be the "number of target characters (7)" in this context. Then, in the expression form information 562 (see Figure 6), different control items (expression forms) are used depending on the range of the "number of target characters (7)". For example, if the "number of target characters (7)" is "5" or more, the description is limited to one control item (expression form), and if it is less than "5", the description uses multiple control items.

[0187] (Variation 5) Furthermore, in the above embodiment, the content provision system 1000 may be implemented not as a client-server type, but as a single computer system that was designated as the user terminal 1500 in the above embodiment.

[0188] Specifically, the user terminal 1500 stores all the data that the server system 1100 in the above embodiment is supposed to store (see Figure 27). However, instead of the server program 501 and the distribution client program 503, a content provision program is provided as an application program for the user terminal 1500.

[0189] In the above embodiment, the content provision program implements all of the functional units of the server system 1100 (see Figure 28) on the user terminal 1500. In this modified example, the content provision program is executed on the user terminal 1500. The processing flow in the above embodiment (see Figures 31 to 32) can be interpreted by replacing the execution entity from the server system 1100 to the user terminal 1500. [Explanation of symbols]

[0190] 5…NPC 7…Target Character 8... Opponent character 20…Server-side generated AI 21... Supervising AI 23... AI generated by control item 30…Device-side generated AI 200s... Server Processing Unit 210... Sufficient Monitoring Department 212...Transmission Information Setting Unit 214... Background Information Setting Section 216...Representation Information Acquisition Control Unit 218...Communication control execution unit 220...Control time length determination unit 230...Generation section 230t... User terminal side generation unit 232...Control Item Generation Unit 234...General Affairs Department 501…Server program 508... Pre-trained AI model 510...Content Initial Setup Data 520...Character initial settings data 530... Personality Information 535... CHARACTER RELATED INFORMATION 536... Information amount setting function 550…Communication event definition data 556...Event trigger requirements 557a...Pre-content status conditions 557b...Pre-target character status conditions 557c...Content status conditions 557d...Current target character status conditions 557f...Character Relationship Conditions 560…Prerequisite situation information 562...Expression form information 564... Information transmitted 566…Direction policy information 568... Information on the production policy 580... Situation history range setting function 700... Play data 706...Content Status Log 708... Character Status Log 720...Communication generation and management data 728…Background information 730... Comprehensive generation instruction information 734…Item specification information 735...Description of production policy 736…Reference Information 737…Settings transmission information 740…List of control items for generation order 742... Generation instruction information by control item 750...Scenario text data 762... Order of execution of individual communication 766...Expression information

Claims

1. A computer system that performs communication control when a target character communicates with a given other character in content featuring multiple characters, A means for setting communication information, which is the information to be transmitted in the aforementioned communication, Background information setting means for setting background information based on at least one of the following: the status of the content, the status of the target character, the status of the opponent character, and the relationship between the target character and the opponent character. An expression information acquisition control means that generates generation instruction information based on the transmission information and the background information, and controls the acquisition of expression information when the target character communicates by providing the generation instruction information to a given generation unit, A communication control execution means that performs the communication control based on the aforementioned representation information, A computer system equipped with the following features.

2. The communication information setting means sets the communication information when the conditions for the occurrence of the communication based on the progress status, which is the status of the content and / or the status of the target character, are met. The background information setting means sets the background information when the occurrence requirements are met. The computer system according to claim 1.

3. Prior conditions are set for satisfying the aforementioned conditions for occurrence, A satisfaction monitoring means that monitors whether the conditions for occurrence are met when the aforementioned preconditions are met. The computer system according to claim 2, further comprising:

4. The aforementioned expression information acquisition control means controls the amount of information in the generation instruction information in a variable manner based on the billing information of the user playing the content. The computer system according to claim 1.

5. The aforementioned background information includes at least the history of the situation of the content and / or the history of the situation of the target character (hereinafter collectively referred to as "situation history"), The computer system according to claim 1.

6. The background information setting means determines the scope of the circumstances to be included in the background information. The computer system according to claim 5.

7. Candidate control items for the aforementioned communication control include the target character's lines, the target character's spoken voice, the target character's motion, and the effects during communication. The computer system according to claim 1.

8. The communication control execution means executes the communication control by including at least the lines of the target character among the candidate control items as control items. The computer system according to claim 7.

9. The representation information acquisition control means acquires the representation information for the control item by generating the generation instruction information, which includes item specification information specifying the candidate control item to be the control item, and providing it to the generation unit. The communication control execution means performs the communication control by controlling the control items based on the representation information. The computer system according to claim 7.

10. The representation information acquisition control means variably determines the item specification information according to the number of target characters. The computer system according to claim 9.

11. The generating unit is The control item generation unit for each candidate control item, A control unit that causes the control item generation unit to generate representation information related to the control item based on the generation instruction information, Having, The computer system according to claim 7.

12. The control unit determines the generation order variably and causes the control item generation unit to generate representation information related to the control item according to the said generation order. The computer system according to claim 11.

13. Control time length determination means for determining the control time length, which is the length of the communication control, Furthermore, The representation information acquisition control means provides the generation unit with the control time length included in the generation instruction information. The computer system according to claim 1.

14. Control time length determination means for determining the control time length, which is the length of the communication control, Furthermore, The representation information acquisition control means provides the control time length to the control unit, including the generation instruction information. The control unit determines candidate control items to be used as control items based on the control time length, and causes the control item generation unit corresponding to the candidate control item to generate representation information related to that control item. The computer system according to claim 11.

15. The control unit, based on the representation information relating to the control item generated by the control item generation unit, causes the control item generation unit to regenerate the representation information in order to modify it. The computer system according to claim 11.

16. The generating unit is The server system that provides the aforementioned content includes a generation unit, The content generation unit on the user terminal side, Having, The computer system according to claim 1.

17. The generation unit on the server system side has a control unit that determines, based on the generation instruction information, whether to generate the representation information on the server system side or on the user terminal side. The computer system according to claim 16.

18. A program for causing a computer system to perform communication control when a target character communicates with a given other character in content featuring multiple characters, A means for setting communication information, which is the information to be transmitted in the aforementioned communication. Background information setting means for setting background information based on at least one of the following: the status of the content, the status of the target character, the status of the opponent character, and the relationship between the target character and the opponent character. Expression information acquisition control means that generates generation instruction information based on the transmission information and the background information, and provides the generation instruction information to a given generation unit to control the acquisition of expression information when the target character communicates, Communication control execution means that executes the communication control based on the aforementioned representation information, A program for causing the aforementioned computer system to function.

Images