Program and information processing system

Abstract

To provide a program for improving amusement of service.SOLUTION: A non-transitory computer-readable storage medium storing a program for causing an event in a game to progress, the program causing the event to progress in a mode corresponding to generated information generated by a machine learning model before the event is started. Further, in the above configuration, the event includes a preceding event and a subsequent event that is started after the preceding event, and the control means progresses the subsequent event in a mode according to the generated information generated by the machine learning model from the information determined according to the selection of the user in the preceding event.SELECTED DRAWING: Figure 5

Description

[Technical Field]

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

[0002] Services using machine learning models have been provided for some time (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2023-023978 Summary of the Invention [Problem to be solved by the invention]

[0004] The present invention aims to improve the interest of the service. [Means for solving the problem]

[0005] In order to solve the above problem, the program of the present invention is a program that causes a computer to progress an event in a game, and causes the computer to function as a control means that progresses the event in a manner corresponding to generation information generated by a machine learning model before the event starts. [Effects of the Invention]

[0006] According to the present invention, the interest of the service is improved. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 2 is a diagram illustrating each configuration of the information processing system. [Figure 2] FIG. 1 is a hardware configuration diagram of an information processing system. [Figure 3] FIG. 2 is a functional block diagram of the information processing system. [Figure 4] FIG. 10 is a schematic diagram of a specific example of a map screen. [Figure 5] FIG. 1 is a diagram for explaining an outline of the operation of an information processing system. [Figure 6] FIG. 10 is a diagram illustrating a specific example of a battle event. [Figure 7] FIG. 10 is a diagram illustrating a specific example of a pre-prompt. [Figure 8] FIG. 10 is a diagram for explaining a specific example of a negotiation event. [Figure 9] FIG. 10 is a diagram for explaining details of an intermediate prompt. [Figure 10] FIG. 10 is a diagram illustrating a specific example of a reunion event. [Figure 11] FIG. 2 is a sequence diagram illustrating an operation of the information processing system. DETAILED DESCRIPTION OF THE INVENTION

[0008] First Embodiment Fig. 1 is a diagram illustrating the components of an information processing system 1000. The information processing system 1000 includes a terminal device 100, a server device 200, and an AI (Artificial Intelligence) system 300. As shown in Fig. 1, the terminal device 100, the server device 200, and the AI ​​(Artificial Intelligence) system 300 can communicate with each other via a network N. The network N may be, for example, the Internet.

[0009] The terminal device 100 may be, for example, a smartphone, a personal computer, or a portable game console. The terminal device 100 stores various programs including application programs. In reality, a plurality of terminal devices 100 communicate with the server device 200. However, for the sake of explanation, only one terminal device 100 is shown in FIG. 1. The terminal device 100 may be, for example, a desktop personal computer.

[0010] The server device 200 provides the terminal device 100 with various types of information that the terminal device 100 uses when executing an application program. Specifically, the terminal device 100 and the server device 200 work together to allow a game to progress on the terminal device 100. Note that while FIG. 1 shows an example in which the server device 200 is configured as a single server device, the server device 200 may also be configured as a plurality of server devices (systems).

[0011] The AI ​​system 300 functions as a machine-learned AI chatbot. Specifically, when a prompt (see FIGS. 7 and 9 described below) is input to the AI ​​system 300, an answer (generated information) to the prompt is generated. When a prompt is input from the server device 200 via an AIP (Application Programming Interface), the AI ​​system 300 of this embodiment replies with generated information corresponding to the prompt. The above AI system 300 may be configured to include multiple computers or a single computer.

[0012] Although details will be described later, the AI ​​system 300 receives, for example, a preliminary prompt DX (see FIG. 7) including information (such as the name) about an enemy character in the game. When the AI ​​system 300 receives the preliminary prompt DX, it generates a scenario (scenario Cx, described later) showing interactions between the enemy character and the player character. With the above configuration, the scenario for the game is automatically generated, which has the advantage that there is no need for a human to think up the scenario. Another advantage is that it becomes easier to generate scenarios that are more varied than those thought up by a human. Note that the server device 200 may be configured to have the functionality of an AT chatbot.

[0013] Fig. 2 is a hardware configuration diagram of an information processing system 1000. As described above, the information processing system 1000 includes a terminal device 100, a server device 200, and an AI system 300. As shown in Fig. 2, the terminal device 100 includes a processing device 101, a storage device 102, a communication device 103, a display device 104, a GPS receiving unit 105, and an acceleration and direction sensor 106. Each of the above components is connected to each other so as to be able to communicate with each other via a system bus.

[0014] The processing device 101 controls the entire terminal device 100. The processing device 101 may be configured with one or more processors. Specifically, the processing device 101 may be configured with one or more types of processors, such as a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC).

[0015] The storage device 102 stores various programs including a basic program and an application program PGx. Well-known storage media such as semiconductor storage media and magnetic storage media can be used as the storage device 102. The storage device 102 may be configured as a single storage medium or multiple storage media. The application program PGx is downloaded in advance and installed on the terminal device 100. The storage device 102 also stores various game data used in the game (such as data indicating items given to the user and data indicating the status of the user character).

[0016] The display device 104 includes a display panel 104a and a touch panel 104b. The display panel 104a is, for example, a flat display configured with organic EL (Electro Luminescence). The touch panel 104b is configured to be able to detect a touch operation by a user. Specifically, the touch panel 104b is provided over the display panel 104a and receives a touch operation on an image displayed on the display panel 104a.

[0017] The GPS receiver 105 receives GPS signals from GPS satellites. The acceleration and direction sensor 106 is configured by combining various sensors including a compass, an acceleration sensor, and a gyro sensor that detects direction. The communication device 103 communicates with the server device 200 via the network N.

[0018] The server device 200 includes a processing device 201, a storage device 202, and a communication device 203. These components are communicatively connected via a system bus. The processing device 201 controls the entire server device 200. The processing device 201 of the server device 200 may be configured with one or more processors, similar to the processing device 101 of the terminal device 100 described above. Specifically, the processing device 201 may be configured with one or more types of processors, such as a CPU, a GPU, a DSP, an FPGA, or an ASIC.

[0019] The storage device 202 stores various programs including a basic program and a game management program PGy. The storage device 202 of the server device 200 may be, for example, a known storage medium such as a semiconductor storage medium or a magnetic storage medium. The storage device 202 may be configured with a single storage medium or multiple storage media. The storage device 202 also stores various game data used in the game (such as data indicating items given to the user and data indicating the status of the user character) in association with the user's identifier. The communication device 203 communicates with the terminal device 100 via the network N.

[0020] The AI ​​system 300, like the server device 200 described above, has various components including a CPU, a storage device, and a communication device. The CPU of the AI ​​system 300 executes a program in the storage device to realize the function of an AI chatbot. The communication device in the AI ​​system 300 transmits and receives various information (prompts, generated information, etc.) to and from the communication device in the server device 203.

[0021] 3 is a functional block diagram of the information processing system 1. As shown in FIG. 3, the information processing system 1 of this embodiment is configured to include a terminal device 10, a management device 20, and a machine learning model 30. For example, the above-mentioned terminal device 100 executes an application program PGx to function as the terminal device 10. Furthermore, the above-mentioned server device 200 executes a game management program PGy to function as the management device 20. Each of the above components can communicate via a network N.

[0022] The terminal device 10 includes a control means 11 and an assignment means 12. However, the management device 20 may be configured to have the above configuration (functions).

[0023] The control means 11 progresses an event in the game (a negotiation event shown in FIG. 8 and a reunion event shown in FIG. 10) in a manner (in which scenario C and option string A are displayed) according to generation information (a scenario C and option string A described below) generated by the machine learning model 30 before the event starts. The events include a preceding event (the negotiation event) and a subsequent event (the reunion event) that starts after the preceding event, and the control means 11 progresses the subsequent event in a manner according to generation information (Cy1, Cy2, Ay1, Ay2 in FIG. 10) generated by the machine learning model 30 from information displayed by a user's selection in the preceding event (scenario Cx2 displayed in response to operation of selection button Bx1).

[0024] Specifically, the control means 11 progresses the preceding event (negotiation event) in a manner corresponding to the generation information (Cx1, Cx2, Ax1-3 in Figure 8) generated by the machine learning model 30 before the preceding event started, and progresses the subsequent event in a manner corresponding to the generation information (Cy1, Cy2, Ay1, Ay2 in Figure 10) generated by the machine learning model 30 from the information displayed by the user's selection in the preceding event (scenario Cx2 displayed in response to operation of selection button Bx1).

[0025] Furthermore, in this embodiment, the event (negotiation event) is composed of a pre-selection part (negotiation part) in which a selection area (button Bx1 shown in Figure 8) is presented to the user and a post-selection part that starts when the user selects the selection area, and the control means 11 enables the selection area to be displayed in a manner corresponding to the generation information (option string Ax1) generated by the machine learning model 30 based on the type of the post-selection part (see Dx51 in Figure 7).

[0026] The granting means 12 grants a predetermined benefit (a card Gc, described later) in accordance with the user's selection in the event (reunion event). The above configuration will be described in detail with reference to FIG. 10. Note that in this embodiment, the card Gc is used as the benefit, but the "benefit" of the present invention is not limited to the above example. For example, the right to recover the hit points of a player character Gp, described later, may be granted as the "benefit." Furthermore, the right to enable the start of an event that was previously impossible to start may be granted as the "benefit."

[0027] FIG. 4 is a schematic diagram of a specific example of the map screen Mp. The map screen Mp described above is displayed on the display device 104, for example. As shown in FIG. 4, the map screen Mp includes an event position Gi, a player character Gp, and a partner character Gq. A plurality of types of player characters Gp are provided, and the user can progress through the game (battle events, negotiation events, and reunion events, which will be described later) with any of the player characters Gp. Furthermore, in this embodiment, the user can select any of the partner characters Gq and progress through the game. The partner character Gq assists the player character Gp in the game. For example, the partner character Gq gives advice to help the player progress through the game to their advantage.

[0028] The user moves the player character Gp to one of the event positions Gi by appropriately operating the terminal device 10. Specifically, the user can select the event position Gi where the player character Gp is ​​currently located and one of the adjacent event positions Gi. The player character Gp moves to the event position Gi selected by the user. When the player character Gp moves to the event position Gi, an event corresponding to the event position Gi starts.

[0029] Specifically, the event position Gi includes a battle position Gia, a recovery position Gib, and a conversation position Gic. When the player character Gp moves to the battle position Gia, a battle event, which will be described later, is executed (see FIG. 6). In the battle event, the player character Gp battles one of multiple enemy characters Ge. In the above battle event, if the hit points of the enemy character Ge reach the value "0" before the hit points of the player character Gp, the player character Gp wins.

[0030] When the player character Gp moves to the recovery position Gib, the hit points of the player character Gp that were reduced in the battle event are restored. The current value of the hit points of the player character Gp is ​​displayed on the map screen Mp. In this embodiment, the type of enemy character Ge that will battle at the battle position Gia is not announced until the battle event actually begins. However, a configuration may be adopted in which the enemy character Ge that will battle at the battle position Gia is announced before moving to the battle position Gia. For example, a configuration may be adopted in which the enemy character Ge that will battle at the battle position Gia is displayed at the battle position Gia.

[0031] When the player character Gp moves to the conversation position Gic, one of the predetermined conversation events is started. For example, a reunion event (see FIG. 10) described below may start at the conversation position Gic. In the reunion event, the player reunites with the enemy character Ge who fought in the battle event before the reunion event. Specifically, in the battle event, a negotiation event (see FIG. 8) described below may occur. If the user makes a predetermined choice in the negotiation event with the enemy character Ge (if the user chooses to accept the enemy character Ge's plea for his life), the conditions for the occurrence of the reunion event in which the player reunites with the enemy character Ge are met. When the conditions for the occurrence of the reunion event are met and the player character Gp moves to the conversation position Gic, the reunion event may be executed.

[0032] Fig. 5 is a diagram for explaining an outline of the operation of the information processing system 1. Fig. 5 shows specific examples of events that progress in the game, and specific examples of information (DX, DY, Cx, Ax, Cy, Ay) that is used to progress the events.

[0033] When the terminal device 10 is operated appropriately, the player character Gp moves on the map screen Mp (see FIG. 4 above) (Sa1 in FIG. 5). When the player character Gp moves to the battle position Gia, a battle event starts (Sa2 in FIG. 5). In the specific example of FIG. 5, it is assumed that a negotiation event occurs in the battle event of step Sa2. In the negotiation event, a scenario Cx (a plea for one's life) and character strings Ax for each option ("help," "continue fighting," etc.) are displayed (see FIG. 8 described below).

[0034] The scenario Cx and the character strings Ax for each option displayed in the negotiation event are generated in advance by the machine learning model 30. Specifically, the management device 20 inputs a pre-prompt DX (text including the name of the enemy character Ge, etc.; see FIG. 7 described later) to the machine learning model 30 before the negotiation event starts (Sb1 in FIG. 5). The machine learning model 30 generates generation information (hereinafter, sometimes simply referred to as "Cx, Ax") including the scenario Cx and the character strings Ax for each option from the pre-prompt DX (Sc1 in FIG. 5). The management device 20 causes the terminal device 10 to display the generation information "Cx, Ax" generated by the machine learning model 30 in the negotiation event (Sb2 in FIG. 5).

[0035] The management device 20 uses (reuses) the scenario Cx and the selection string Ax displayed in the negotiation event to create an intermediate prompt DY (Sb3 in FIG. 5). The intermediate prompt DY is used to generate the scenario Cy (interaction at the time of reunion) and each selection string Ay ("Believe in fate," "See you again") displayed in the reunion event. Details of the intermediate prompt DY will be described later using FIG. 9.

[0036] If the conditions for the reunion event to occur are met in the negotiation event (if the enemy character Ge's plea for his life is accepted), the enemy character Ge will flee and the battle event will end. In the specific example of FIG. 5, it is assumed that the conditions for the reunion event to occur are met in the negotiation event. In this case, the battle event (negotiation event) ends and the screen switches to the map screen Mp. When the conditions for the reunion event to occur are met and the player character Gp moves to the conversation position Gic (Sa3 in FIG. 5), the reunion event will start (Sa4 in FIG. 5).

[0037] The management device 20 inputs the above-mentioned mid-course prompt DY to the machine learning model 30 before the reunion event starts (Sb4 in FIG. 5). The machine learning model 30 generates generation information "Cy, Ay" including a scenario Cy (interaction at the time of reunion) and character strings Ay for each option ("Believe in fate," "See you again") from the mid-course prompt DY (Sc2 in FIG. 5). The management device 20 displays the generation information "Cy, Ay" generated by the machine learning model 30 on the terminal device 10 during the reunion event (Sb5 in FIG. 5).

[0038] In the reunion event, the user selects whether or not to make the enemy character Ge their ally (see FIG. 10 described below). If the user selects to make the enemy character Ge their ally, the user is given a card Gc corresponding to that enemy character Ge. On the other hand, if the user selects not to make the enemy character Ge their ally, the user is not given the card Gc. In the specific example of FIG. 5, it is assumed that the card Gc is given in the reunion event.

[0039] FIG. 6 is a schematic diagram of a specific example of a battle event screen Mb. The battle event screen Mb is displayed on the terminal device 10 (display device 104) during a battle event. As described above, during a battle event, a battle occurs between the player character Gp and the enemy character Ge. Specifically, during a battle event, one or more random enemy characters Ge appear. The specific example of FIG. 6 assumes that, of the enemy characters Ge, enemy characters Ge1 and Ge2 appear during the battle event.

[0040] As shown in FIG. 6, the battle event screen Mb displays each enemy character Ge, as well as the name of the enemy character Ge ("xxx" and "yyy" in the specific example of FIG. 6), and the remaining hit points of the enemy character Ge. Two cards Gc are also displayed on the battle event screen Mb. Specifically, in this embodiment, the user is given multiple cards Gc in advance. The effect of using the cards Gc in a battle event differs depending on the type of card Gc. For example, the user may be given a card Gc in the reunion event described above.

[0041] In a battle event, two of the cards Gc possessed by the user are displayed in sequence. The user can progress through the battle event by selecting one of the displayed cards Gc. For example, in the specific example of FIG. 6, assume that cards Gc1 and Gc2 are displayed. If card Gc1 is selected, the hit points of the enemy character Ge can be reduced by the value "100" (inflicting damage of the value "100"). On the other hand, if card Gc2 is selected, damage to the player character Gp caused by the enemy character Ge's attack can be reduced by the value "50". Note that battle events are not limited to the above examples and can be modified as appropriate.

[0042] As described above, a negotiation event may occur in a battle event. Specifically, if multiple enemy characters Ge appear in a battle event and all but one enemy character Ge are defeated, a negotiation event with the remaining enemy character Ge may occur. For example, in the specific example of FIG. 6, assume that enemy character Ge2 of the enemy characters Ge (1, 2) is defeated. In the above case, a negotiation event with the remaining enemy character Ge1 may occur.

[0043] The conditions for generating a negotiation event can be set as appropriate. For example, a configuration may be adopted in which a negotiation event does not occur even when there is only one enemy character Ge remaining. Alternatively, a configuration may be adopted in which a negotiation event occurs only when a predetermined lottery is won. Furthermore, a configuration may be adopted in which the type of enemy character Ge for which a negotiation event occurs is predetermined.

[0044] 7 is a diagram for explaining a specific example of a pre-prompt DX. The pre-prompt DX is stored, for example, in the management device 20. As described above, when the pre-prompt DX is input into the machine learning model 30, generation information "Cx1-4 (scenario), Ax1-3 (string for choice)" to be displayed in the negotiation event is generated.

[0045] The pre-prompt DX includes input texts Dx0 to Dx6. Of the pre-prompt DX, input text Dx0 explains the worldview of the game. For example, input text Dx0 includes an explanation that the game of this embodiment is a game in which the protagonist (player character Gp) explores a dungeon called □□ with partner character Gq. Input text Dx1 explains a situation in which a negotiation event (begging for one's life) occurs in a battle event.

[0046] The input text Dx2 (21-23) includes input text Dx21, input text Dx22, and input text Dx23. The input text Dx21 is a description of the player character Gp (protagonist). For example, the input text Dx21 includes the name, gender, personality, and background (family structure, etc.) of the player character Gp. The input text Dx22 is a description of the partner character Gq. The above input text Dx22 includes the name, gender, personality, and background (relationship with the protagonist, etc.) of the partner character Gq. The input text Dx23 is a description of the enemy character Ge. The above input text Dx23 includes the name, personality, and background (reason for fighting, etc.) of the enemy character Ge.

[0047] As described above, in this embodiment, multiple types of player characters Gp, multiple types of partner characters Gq, and multiple types of enemy characters Ge are provided. The management device 20 stores combinations of player characters Gp, partner characters Gq, and enemy characters Ge (hereinafter, sometimes referred to as "negotiation groups"). For example, assume that there are N1 types of player characters Gp, N2 types of partner characters Gq, and N3 types of enemy characters Ge. In the above case, there will be at least N1 x N2 x N3 types of negotiation groups. A pre-prompt DX is provided for each negotiation group. The generation information "Cx, Ax" displayed in a negotiation event in which a negotiation group negotiates is generated from the pre-prompt DX of the negotiation group.

[0048] The input text Dx3 of the pre-prompt DX is a description of a topic that is currently being discussed in the real world. For example, in the specific example of FIG. 7, the input text Dx3 includes the name of a popular song. However, the input text Dx3 is not limited to the above specific example. For example, the input text Dx3 may include a popular sport or a popular phrase. It is preferable to periodically update the input text Dx3. With the above configuration, the latest topics can be reflected in the generation information "Cx1-4 (scenarios), Ax1-3 (strings for options)" generated by the machine learning model 30.

[0049] FIG. 7 shows the display position and display time in a negotiation event of the generation information "Cx1-4 (scenarios), Ax1-3 (strings for options)" generated by the pre-prompt DX. As will be described in detail later with reference to FIG. 8, the negotiation event includes a negotiation part in which multiple selection buttons Gx (1-3) are displayed, and a post-selection part that starts after the selection button Gx is selected. In addition, the scenarios Cx generated by the pre-prompt DX include scenarios Cx1, Cx2, Cx3, and Cx4. For example, scenario Cx1 is displayed in the scenario display area Rx1 in the negotiation part, as shown in FIG. 7.

[0050] The input text Dx4 of the pre-prompt DX is an instruction for generating the scenario Cx1. The above scenario Cx1 is displayed in the negotiation part as described above, and shows an exchange in which the enemy character Ge begs for his life from the player character Gp. Specifically, the scenario Cx1 is generated based on the input text Dx0 to Dx4 of the pre-prompt DX. According to the above configuration, the scenario Cx1 (the exchange of begging for his life) reflects, for example, the worldview of the game, the personality of the player character Gp, the personality of the partner character Gq, the personality of the enemy character Ge, and topics of conversation in the real world.

[0051] The input text Dx5 (51-53) of the pre-prompt DX is an instruction for generating the option character strings Ax (1-3). Specifically, the option character strings Ax (1-3) displayed in the negotiation event include option character strings Ax1, Ax2, and Ax3. As described above, in the negotiation part of the negotiation event, the selection buttons Bx (1-3) are displayed in conjunction with the scenario Cx1 (a plea for one's life) (see FIG. 8). The selection buttons Bx displayed in the negotiation event include the selection button Bx1, the selection button Bx2, and the selection button Bx3. When the selection button Bx1 of the selection buttons Bx is selected by the user, the enemy character Ge is rescued and the battle event ends. On the other hand, when the selection button Bx2 or the selection button Bx3 of the selection buttons Bx is selected, the battle event continues.

[0052] Of the option strings Ax (1-3), option string Ax1 represents the option of accepting the enemy character Ge's plea for his life. Of the input text Dx5 of the pre-prompt DX, input text Dx51 instructs the generation of option string Ax1. The option string Ax1 generated by the above input text Dx51 is displayed on the selection button Bx1, as shown in FIG. 7. Furthermore, the option string Ax1 is generated based on the input text Dx0-3, 51 of the pre-prompt DX. According to the above configuration, the option string Ax1 reflects, for example, the worldview of the game, the personality of the player character Gp, the personality of the partner character Gq, the personality of the enemy character Ge, and topics of conversation in the real world.

[0053] Of the option strings Ax (1 to 3), option string Ax2 represents the option of not accepting the enemy character Ge's plea for his life. Input text Dx52 of the input text Dx5 of the pre-prompt DX instructs the generation of option string Ax2. The option string Ax2 generated by the above input text Dx52 is displayed on the selection button Bx2, as shown in FIG. 7. Furthermore, the option string Ax2 is generated based on the input text Dx0 to 3, 52 of the pre-prompt DX. According to the above configuration, the option string Ax2 reflects, for example, the worldview of the game, the personality of the player character Gp, the personality of the partner character Gq, the personality of the enemy character Ge, and topics of conversation in the real world.

[0054] Of the option strings Ax (1-3), option string Ax3 represents an ambiguous option as to whether to accept the enemy character Ge's plea for his life. Input text Dx53 of the input text Dx5 of the pre-prompt DX instructs the generation of option string Ax3. The option string Ax3 generated by the above input text Dx53 is displayed on the selection button Bx3, as shown in FIG. 7. Furthermore, the option string Ax3 is generated based on the input text Dx0-3, 53 of the pre-prompt DX. According to the above configuration, option string Ax3 reflects, for example, the worldview of the game, the personality of the player character Gp, the personality of the partner character Gq, the personality of the enemy character Ge, and topics of conversation in the real world.

[0055] As described above, when a selection button Bx is selected in the negotiation part of a negotiation event, the post-selection part corresponding to the selection button Bx begins. In addition, a scenario Cx (2 to 4) corresponding to the post-selection part is displayed in the scenario display area Rx2. For example, in the post-selection part after the selection button Bx1 (help) is selected, scenario Cx2 is displayed. The above scenario Cx2 shows the exchange that takes place after the enemy character Ge's plea for his life is accepted. On the other hand, in the post-selection part after the selection button Bx2 (continue) is selected, scenario Cx3 is displayed. The above scenario Cx3 shows the exchange that takes place after the enemy character Ge's plea for his life is clearly rejected. Similarly, in the post-selection part after the selection button Bx3 (ambiguous) is selected, scenario Cx4 is displayed. The above scenario Cx4 shows the exchange that takes place after the enemy character Ge's plea for his life is ambiguously rejected.

[0056] The input text Dx6 (61-63) of the pre-prompt DX is an instruction to generate scenarios Cx2-4 to be displayed in the post-selection part of the negotiation event. Specifically, the input text Dx6 includes input text Dx61, input text Dx62, and input text Dx63. The input text Dx61 is an instruction to generate scenario Cx2 to be displayed in the post-selection part after the selection button Bx1 (Help) is selected. The input text Dx62 is an instruction to generate scenario Cx3 to be displayed in the post-selection part after the selection button Bx2 (Continue) is selected. The input text Dx63 is an instruction to generate scenario Cx4 to be displayed in the post-selection part after the selection button Bx3 (Ambiguous) is selected.

[0057] In the above configuration, scenario Cx2, like the above-mentioned scenario Cx1, is generated based on the input text Dx0-3, 62 of the pre-prompt DX. Similarly, scenario Cx3 is generated based on the input text Dx0-3, 63 of the pre-prompt DX, and scenario Cx4 is generated based on the input text Dx0-3, 64 of the pre-prompt DX. According to the above configuration, scenarios Cx2-4 reflect, for example, the worldview of the game, the personality of player character Gp, the personality of partner character Gq, the personality of enemy character Ge, and topics of conversation in the real world.

[0058] FIG. 8 is a diagram for explaining the details of a negotiation event. The left part of FIG. 8 shows a negotiation event screen Mk1 that is displayed in the negotiation part immediately after the negotiation event starts. As shown in FIG. 8, the negotiation event screen Mk1 displays an enemy character Ge with which to negotiate in the negotiation event, as well as a scenario display area Rx1 and each selection button Bx (1 to 3). The scenario display area Rx1 displays a scenario Cx1 (a plea for one's life) generated by the machine learning model 30 from the above-mentioned advance prompt DX.

[0059] As described above, the selection buttons Bx on the negotiation event screen Mk1 include selection button Bx1, selection button Bx2, and selection button Bx3. When the user selects selection button Bx1 from among the selection buttons Bx, the post-selection part begins, in which the enemy character Ge accepts his plea for his life. In the above cases, the conditions for the reunion event with the enemy character Ge are met, and the battle event ends after the post-selection part. On the other hand, when selection button Bx2 or selection button Bx3 from among the selection buttons Bx is selected, the conditions for the reunion event are not met, and the battle event continues after the post-selection part.

[0060] In the above configuration, a user who wishes to attend a reunion event must select the selection button Bx1. However, the position (top, center, bottom) where the selection button Bx1 is displayed is random for each negotiation event. Therefore, depending on the display position of each selection button Bx, the user cannot determine whether the selection button Bx is the selection button Bx1.

[0061] However, in this embodiment, it is possible to estimate whether or not a selection button Bx is the selection button Bx1 based on each option character string Ax displayed on each selection button Bx. Specifically, option character string Ax1 is displayed on selection button Bx1, option character string Ax2 is displayed on selection button Bx2, and option character string Ax3 is displayed on selection button Bx3. Of the above option character strings Ax, option character string Ax1 displayed on selection button Bx1 is a character string that suggests that the option is to accept the enemy monster Ge's plea for its life.

[0062] For example, in the specific example of FIG. 8, it is assumed that an option string Ax1 of "Stop the fight" has been generated. With the above configuration, a user who wishes to hold a reunion event can infer the selection button Bx1 from the option string Ax1. Meanwhile, the option string Ax2 displayed on the selection button Bx2 among each option string Ax is a string that suggests that this is an option that does not accept the enemy monster Ge's plea for his life. For example, in the specific example of FIG. 8, it is assumed that an option string Ax2 of "Continue the fight" has been generated. With the above configuration, a user who wishes to continue the battle event can infer the selection button Bx2 from the option string Ax2.

[0063] However, if it is too easy to select the selection button Bx1 (accept the plea for help) from the selection buttons Bx, some users may find the game less interesting. In consideration of the above, in this embodiment, a configuration is adopted that makes it difficult to guess the selection button Bx1.

[0064] Specifically, as described above, when the selection button Bx3 is selected, the condition for the reunion event to occur is not satisfied, just as when the selection button Bx2 is selected. In the above configuration, this embodiment employs a configuration in which the selection button Bx3 displays the option character string Ax3, which is ambiguous as to whether or not it is an option to help the enemy monster Ge. For example, in the specific example of FIG. 8, assume that the option character string Ax3 "remain silent" is displayed on the selection button Bx3. In the above case, some users may assume (mistakenly recognize) that the selection button Bx3 displaying the option character string Ax3 is the selection button Bx1. This increases the sense of tension when selecting the selection button Bx, improving the enjoyment of the game.

[0065] The right side of FIG. 8 shows a negotiation event screen Mk2 that is displayed in the post-selection part after a selection button Bx is selected. As shown in FIG. 8, the negotiation event screen Mk2 displays the enemy character Ge as well as a scenario display area Rx2. Furthermore, the selection button Bx selected in the negotiation part continues to be displayed in the post-selection part, while the other selection buttons Bx are hidden. For example, the specific example in FIG. 8 assumes a post-selection part in which selection button Bx1 is selected among the selection buttons Bx. In the post-selection part described above, selection button Bx1 continues to be displayed.

[0066] In the scenario display area Rx2 of the post-selection part, one of the scenarios Cx2 to Cx4 is displayed. As described above, the scenarios Cx2 to Cx4 are generated in advance by the machine learning model 30 from the pre-prompt DX. Specifically, if the selection button Bx1 (Help) is selected, the scenario Cx2 is displayed in the post-selection part. Also, if the selection button Bx2 (Continue) is selected, the scenario Cx3 is displayed in the post-selection part. Similarly, if the selection button Bx3 (Ambiguous) is selected, the scenario Cx4 is displayed in the post-selection part.

[0067] In this embodiment, all of the scenarios Cx2 to Cx4 are generated before the battle event begins. In the above configuration, two of the three scenarios Cx2 to Cx4 generated in advance are discarded without being displayed. Instead of the above configuration, one of the scenarios Cx2 to Cx4 may be generated in accordance with the selection button Bx immediately after the selection button Bx is selected. The above configuration prevents a scenario generated by the machine learning model 30 from being discarded without being displayed.

[0068] However, there is a situation in which a waiting time of approximately 10 seconds or more may occur before the machine learning model 30 generates the generation information (including scenarios Cx2 to Cx4). Therefore, in a configuration in which generation of the scenario Cx starts immediately after the selection button Bx is selected, an inconvenience may arise in that the waiting time from when the selection button Bx is selected until the post-selection part can start becomes excessively long. According to this embodiment, the above inconvenience is suppressed.

[0069] 8, an intermediate prompt DY is created from the scenario Cx1 (interaction of pleading for one's life), scenario Cx2 (interaction after accepting the pleading for one's life), and option string Ax1 (help) among the generation information "Cx1-4, Ax1-3" generated by the machine learning model 30 for the negotiation event. The intermediate prompt DY is used to generate the scenario Cy and option string Ay that are displayed in the reunion event.

[0070] 9 is a diagram illustrating a specific example of the mid-game prompt DY. The mid-game prompt DY is stored, for example, in the management device 20. As shown in FIG. 9, the mid-game prompt DY includes input text Dx0-3 (explanations of the game worldview, etc.) similar to the above-mentioned pre-game prompt DX. However, the mid-game prompt DY may be configured such that the input text Dx0-3 is omitted.

[0071] The midway prompt DY includes input texts Dy1 to Dy5. Of the above input texts Dy1 to Dy5, input text Dy1 is an explanation of the situation in the reunion event. Specifically, input text Dy1 is an explanation that the player has reunited with the enemy character Ge that he helped in the battle event (negotiation event) in which the conditions for the reunion event were met. Input text Dy2 of the midway prompt DY is an explanation of the interaction (conversation history) with the enemy character Ge in the battle event (negotiation event) in which the conditions for the reunion event were met. Specifically, input text Dy2 includes input text Dy21, input text Dy22, and input text Dy23.

[0072] For example, input text Dy21 is a description of the interaction in the negotiation part of a negotiation event in which the conditions for the occurrence of a reunion event are met. Specifically, as shown in FIG. 9, input text Dy21 includes scenario Cx1 (the interaction of pleading for one's life) displayed in the negotiation part. Furthermore, input text Dy22 is a description of the option selected by the user in the negotiation part of the negotiation event. Specifically, input text Dy22 includes option string Ax1 (help) selected in the negotiation part. Input text Dy23 is a description of the interaction in the post-selection part of the negotiation event. Specifically, input text Dy23 includes scenario Cx2 (the interaction after accepting the pleading for one's life) displayed in the post-selection part of the negotiation event.

[0073] FIG. 9 shows the display position and display time of the generation information "Cy1-3 (scenarios), Ay1, 2 (strings for selection options)" generated by the mid-event prompt DY in a reunion event. Details will be described later using FIG. 10, but the reunion event includes a reunion part in which multiple selection buttons By (1, 2) are displayed, and a post-selection part that starts after the selection button By is selected. Furthermore, the scenario Cy generated by the mid-event prompt DY includes scenario Cy1, scenario Cy2, and scenario Cy3. For example, scenario Cy1 is displayed in the scenario display area Ry1 in the reunion part, as shown in FIG. 9.

[0074] The scenario Cy1 displayed in the reunion part shows the interaction that occurs when the player character Gp and the enemy character Ge meet again. The generation of the above scenario Cy1 is instructed by the input text Dy3 of the midway prompt DY. Specifically, the scenario Cy1 is generated based on the input text Dx0-3 of the midway prompt DY and the input text Dy1-3. The above scenario Cy1 (interaction at the time of reunion) reflects the interaction (reused Cx1, Cx2, Ax1) in the negotiation event in which the conditions for the occurrence of this reunion event were met by the above-mentioned input text Dy2 (21-23).

[0075] The input text Dy4 (41, 42) of the mid-game prompt DY is an instruction for generating the option character string Ay (1, 2). Specifically, the option character string Ay includes an option character string Ay1 and an option character string Ay2. Furthermore, each selection button By displayed in the reunion part includes a selection button By1 and a selection button By2. When the selection button By1 of each selection button By is selected, the player becomes friends with the enemy character Ge and is granted a benefit (card Gc). On the other hand, when the selection button By2 is selected, the player breaks up with the enemy character Ge and is not granted a benefit. In the above configuration, the option character string Ay1 represents the option of becoming friends with the enemy character Ge and is displayed on the selection button By1. Furthermore, the option character string Ay2 represents the option of breaking up with the enemy character Ge and is displayed on the selection button By2.

[0076] Of the input text Dx4, input text Dy41 instructs the generation of an option string Ay1. Specifically, the option string Ay1 is generated based on the input text Dx0-3 of the intermediate prompt DY and the input text Dy1, 2, 41. The above option string Ay1 reflects the exchange (reused Cx1, Cx2, Ax1) in the negotiation event in which the conditions for the occurrence of this reunion event were met, using the above-mentioned input text Dy2 (21-23).

[0077] Furthermore, input text Dy42 of input text Dx4 instructs the generation of option string Ay2. Specifically, option string Ay2, like option string Ay1, is generated based on input text Dx0-3 of intermediate prompt DY and input text Dy1, 2, 42. The above option string Ay2 reflects the exchange (reused Cx1, Cx2, Ax1) in the negotiation event in which the conditions for generating this reunion event were met by the above-mentioned input text Dy2 (21-23).

[0078] The input text Dx5 (51, 52) of the mid-game prompt DY is an instruction to generate scenarios Cy2 and Cy3 to be displayed in the post-selection part of the reunion event. As shown in FIG. 9, each of the above scenarios Cy2 and Cy3 is displayed in the scenario display area Ry2 in the post-selection part. Specifically, scenario Cy2 is displayed in the post-selection part after the selection button By1 (make friend) is selected, and shows the interaction that occurs after the selection to make enemy character Ge a friend. Furthermore, scenario Cy3 is displayed in the post-selection part after the selection button By2 (break up) is selected, and shows the interaction that occurs after the selection to break up with enemy character Ge.

[0079] As shown in FIG. 9, input text Dy5 includes input text Dy51 and input text Dy52. Input text Dy51 is an instruction to generate scenario Cy2. Input text Dy52 is an instruction to generate scenario Cy3. Each of the above scenarios Cy2 and Cy3 is generated based on input text Dx0-3 of the intermediate prompt DY and input text Dy1, 2, 51 or 52, just like the above scenario Cy1. Each of the above scenarios Cy2 and Cy3 reflects the exchanges (reused Cx1, Cx2, Ax1) in the negotiation event in which the conditions for the occurrence of this reunion event were met by the above input text Dy2 (21-23).

[0080] As can be understood from the above explanation, the mid-point prompts DY (Dy21 to Dy23) of this embodiment are created by reusing the generation information "Cx1, Cx2, Ax1" (generation information already used in the negotiation event) generated using the pre-prompt DX. The above mid-point prompts DY make it possible to display scenarios Cy (1 to 3) and option strings Ay (1, 2) that reflect the exchanges (conversation history) in the negotiation event in the reunion event. Furthermore, compared to a configuration in which the mid-point prompt DY is created without reusing the generation information "Cx1, Cx2, Ax1" used in the negotiation event, there is an advantage in that the amount of work required to create the mid-point prompt DY is reduced.

[0081] 10 is a diagram for explaining the details of the reunion event. The reunion event is divided into a reunion part and a post-selection part, similar to the negotiation event described above.

[0082] The left side of FIG. 10 shows a reunion event screen Ms1 that is displayed in the reunion part immediately after the start of the reunion event. As shown in FIG. 10, the reunion event screen Ms1 displays the reunited enemy character Ge as well as a scenario display area Ry1. The scenario display area Ry1 displays a scenario Cy1. As described above, the scenario Cy1 is generated in advance from the mid-game prompt DY by the machine learning model 30. The scenario Cy1 shows the interaction between the player character Gp and the enemy character Ge when they meet again.

[0083] As shown in FIG. 10, the reunion event screen Ms1 displays each selection button By. As described above, each of the selection buttons By includes the selection button By1 and the selection button By2. When the selection button By1 of the selection buttons By is selected, the post-selection part begins in which the enemy character Ge becomes a companion. As will be described in detail later, in the post-selection part, a card Gc representing the reunited enemy character Ge is granted as a bonus. The selection button By1 displays the above-mentioned option character string Ay1. The above-mentioned option character string Ay1 is, for example, a character string that suggests that the enemy character Ge will be accepted as a companion. The specific example in FIG. 10 assumes that the option character string Ay1, "Believe in fate," is displayed on the selection button By1.

[0084] On the other hand, when the selection button By2 is selected, the post-selection part in which the player separates from the enemy character Ge begins. In the post-selection part, the card Gc is not awarded. The selection button By2 displays the above-mentioned option character string Ay2. The option character string Ay2 is, for example, a character string that suggests separation from the enemy character Ge. The specific example in FIG. 10 assumes that the option character string Ay2 "See you again" is displayed on the selection button By2.

[0085] In this embodiment, when a reunion event is executed, the occurrence condition of the reunion event is invalidated regardless of whether or not a card Gc is granted in the reunion event. If the occurrence condition of the reunion event is invalidated, the reunion event will not start even if the player character Gp stops at the conversation position Gic on the map screen Mp thereafter. However, the occurrence condition of the reunion event may be configured not to be invalidated until a card Gc is granted in the reunion event.

[0086] The right side of FIG. 10 shows a reunion event screen Ms2 that is displayed in the post-selection part after the selection button By is selected. As shown in FIG. 10, the reunion event screen Ms2 displays the enemy character Ge as well as a scenario display area Ry2. Furthermore, the selection button By selected on the reunion event screen Ms1 continues to be displayed in the post-selection part, while the other selection buttons By are hidden. For example, the specific example of FIG. 10 assumes the post-selection part after the selection button By1 of the selection buttons By is selected.

[0087] In the scenario display area Ry2 of the post-selection part, scenario Cy2 (interactions when joining a party) or scenario Cy3 (interactions when parting) is displayed. For example, in the post-selection part when selection button By1 (join a party) is selected, scenario Cy2 is displayed. On the other hand, in the post-selection part when selection button By2 (parting) is selected, scenario Cy3 is displayed. The specific example in Figure 10 assumes the post-selection part when selection button By1 (join a party) is selected.

[0088] As shown in FIG. 10, in a reunion event, if an enemy character Ge is selected to be added to the team, the user is given a card Gc representing the enemy character Ge. Specifically, a card Gc of a type corresponding to the enemy character Ge added to the team in the reunion event is given to the user in the reunion event. The effect of using a card Gc in a battle event (such as damage inflicted on the enemy character Ge) varies depending on the type of card Gc. For example, the more difficult it is for an enemy character Ge to fulfill the conditions for the reunion event, the more advantageous a card Gc is given to the enemy character Ge when added to the team in the reunion event.

[0089] 11 is a system sequence diagram illustrating a specific example of the operation of the information processing system 1. Note that the entity that executes each step described below is not limited to the following example. For example, all or part of each step executed by the terminal device 10 may be executed on the management device 20 side or the machine learning model 30 side. Also, all or part of each step executed by the management device 20 may be executed on the terminal device 10 side or the machine learning model 30 side.

[0090] As shown in FIG. 11, the management device 20 transmits a pre-prompt DX to the machine learning model 30 (S101). Upon receiving the pre-prompt DX, the machine learning model 30 generates generation information "Cx, Ax" from the pre-prompt DX (S102) and returns the generation information "Cx, Ax" (S103). The management device 20 also creates an intermediate prompt DY from the generation information "Cx, Ax" received in step S103 and transmits the intermediate prompt DY to the machine learning model 30 (S104). Upon receiving the intermediate prompt DY, the machine learning model 30 generates generation information "Cy, Ay" from the intermediate prompt DY (S105) and returns the generation information "Cy, Ay" (S106).

[0091] The above steps S101 to S106 are repeatedly executed by changing the input text Dx2 (combination of player character Gp, partner character Gq, and enemy character Ge) in the pre-prompt DX. For example, steps S101 to S106 are repeatedly executed until all possible types of input text Dx2 have been used to generate the generation information. The above repeated processing is executed, for example, before the game is released.

[0092] The generated generation information "Cx, Ax" and the generated generation information "Cy, Ay" are accumulated and stored, for example, in the management device 20. According to the above configuration, it is sufficient to use any of the accumulated generation information "Cx, Ax" in the negotiation event, and there is no need to generate new generation information "Cx, Ax" immediately before the negotiation event starts. Similarly, it is sufficient to use any of the accumulated generation information "Cy, Ay" in the reunion event, and there is no need to generate new generation information "Cy, Ay" immediately before the reunion event starts. However, a configuration may be adopted in which the generation information "Cx, Ax" is generated in real time immediately before the negotiation event starts. Also, a configuration may be adopted in which the generation information "Cy, Ay" is generated in real time immediately before the reunion event starts.

[0093] When a negotiation event is about to start, the terminal device 10 notifies the management device 20 of this (S107). When the management device 20 receives the notification that a negotiation event is about to start, it transmits various pieces of information required to display a negotiation event screen Mk to the terminal device 10 (S108). Specifically, when the negotiation event is about to start, the management device 20 is notified of the player character Gp (protagonist), partner character Gq, and enemy character Ge in the negotiation event. The management device 20 identifies the notified player character Gp (protagonist), partner character Gq, and enemy character Ge from the generation information "Cx, Ax" stored in advance, which is generated from the advance prompt DX included in the input text Dx2, and transmits the generation information "Cx, Ax" to the terminal device 10.

[0094] In the specific example of FIG. 11, it is assumed that the selection button Bx1 (help) is selected in the negotiation part of the negotiation event. In the above case, the management device 20 is notified that the selection button Bx1 has been operated (S109). Upon receiving the above notification, the management device 20 stores information that the conditions for the reunion event to occur have been met for the negotiation group in the negotiation event. Thereafter, upon receiving a notification from the terminal device 10 that the reunion event will start (S110), the management device 20 transmits various information (e.g., generation information "Cy, Ay") required to display the reunion event screen Ms for the reunion event to the terminal device 10 (S111).

[0095] <Modification> The above embodiments can be modified in various ways. Specific modified embodiments are exemplified below. Two or more embodiments selected from the following examples can be combined as appropriate.

[0096] (1) In each embodiment, the generation information generated by the machine learning model 30 is not limited to the above examples. For example, an image may be generated by the machine learning model 30. Specifically, a configuration may be adopted in which the pre-prompt DX generates a background image to be displayed in the negotiation event in addition to the above-mentioned generation information "Cx, Ax." Also, a mid-prompt DY including a background image to be displayed in the negotiation event is created, and generation information including an image is created from the mid-prompt DY. The image generated from the mid-prompt DY is displayed in the reunion event.

[0097] Furthermore, in a reunion event, a selection image generated by the mid-event prompt DY may be displayed in place of the selection button By. For example, assume a configuration in which a first treasure image representing a treasure and a second treasure image representing another treasure are generated by the mid-event prompt DY. In the reunion part of the reunion event, a scenario Cy1 is displayed in which the player is prompted to select either the first treasure image or the second treasure image. Furthermore, if the first treasure image is selected, a benefit (card Gc) is awarded in the reunion event. On the other hand, if the second treasure image is selected, a benefit (card Gc) is not awarded in the reunion event. Furthermore, each image generated by the pre-event prompt DX may be displayed in place of each selection button Bx, and the conditions for the reunion event to occur may be met depending on the image selected.

[0098] Furthermore, a configuration may be adopted in which text other than the scenario C(x, y) and the character strings for options can be generated by the machine learning model 30. For example, a title to be displayed in a negotiation event ("Negotiation with xxx" in the example of FIG. 8) may be generated by the machine learning model 30 from a pre-prompt DX. Also, a title to be displayed in a reunion event ("A Chance Reunion" in the example of FIG. 10) may be generated by the machine learning model 30 from an interim prompt DY.

[0099] (2) In each embodiment, the generated information generated by the machine learning model 30 may be modified and then displayed on the terminal device 10. For example, the developer can review and modify the generated information (Cx, Ax, Cy, Ay) generated by the machine learning model 30 on the management device 20. The developer also reviews the generated information generated by the machine learning model 30 and modifies inappropriate sentences (e.g., sentences that may slander others). The generated information modified by the developer can be used in each event of the game (negotiation event, reunion event). As can be understood from the above explanation, the configuration of the present invention of "progressing an event in a manner according to the generated information generated by the machine learning model" includes a configuration of displaying the generated information generated by the machine learning model 30 as is, as well as a configuration of displaying modified generated information in an event.

[0100] (3) In each embodiment, the conditions for the reunion event are always met by selecting the appropriate selection button Bx (Bx1) in the negotiation event. However, the conditions for the reunion event may not be met in the negotiation event regardless of the type of selection button Bx selected in the negotiation event.

[0101] For example, in a negotiation event, a configuration may be adopted in which a selection button Bx4 is displayed instead of the selection button Bx1 for which the conditions for a reunion event are satisfied. An option character string Ax4 is displayed on the selection button Bx4. The option character string Ax4 is generated by the machine learning model 30 as a character string representing an option for a surprise attack on the enemy character Ge. In this modification, when the selection button Bx4 is selected, a scenario showing an exchange in which the enemy character Ge is a surprise attack is displayed, and the battle event ends. The above scenario is generated by the machine learning model 30, just like other scenarios.

[0102] Furthermore, in the above modified example, if the enemy character Ge is defeated by a surprise attack, when the player character Gp stops at the conversation position Gic thereafter, a revenge event may occur by the allies of the enemy character Ge. Each scenario displayed in the above revenge event is generated from the conversation history in the negotiation event (the intermediate prompt DY created from the scenario displayed in the negotiation event), similar to the scenario Cy in the reunion event.

[0103] (4) In each embodiment, a configuration is adopted in which character strings devised by the game developer and character strings generated by the machine learning model 30 are displayed without distinction. However, a configuration may be adopted in which character strings devised by the game developer and character strings generated by the machine learning model 30 are displayed in a distinguishable manner. For example, a configuration may be adopted in which a message saying "Generated by AI" is displayed near a character string (scenario C, choice character string A) generated by the machine learning model 30 among the character strings displayed in the game. This configuration has the advantage that even if the character string generated by the machine learning model 30 feels strange, the user can understand the reason for the strangeness. Note that, even in the above-described modified example in which the generation information generated by the machine learning model 30 includes an image, a configuration in which the message "Generated by AI" is displayed near the image is preferable.

[0104] (5) In each mode, a negotiation event with another enemy character Ge may not be started during a period in which the conditions for a reunion event in which the enemy character Ge is reunited are satisfied. In the above configuration, during a period in which the conditions for a reunion event in which the enemy character Ge is reunited are satisfied, the conditions for a reunion event in which the enemy character Ge is reunited with another enemy character Ge are not established.

[0105] However, a configuration may be adopted in which, during the period in which the conditions for the occurrence of a reunion event with an enemy character Ge are satisfied, the conditions for the occurrence of a reunion event with another enemy character Ge may also be satisfied. In the above configuration, when the player character Gp moves to the conversation position Gic, a reunion event with one of the enemy characters Ge for which the conditions for the occurrence of a reunion event are satisfied is executed. For example, a configuration may be considered in which a reunion event with an enemy character Ge for which the conditions for the occurrence of a reunion event were satisfied earlier is executed before reunion events with other enemy characters Ge. However, a configuration may be adopted in which the order in which the enemy characters Ge for which the reunion events are executed is random.

[0106] (6) In each embodiment, even if the conditions for generating a reunion event are satisfied in a negotiation event, the reunion event may not start when the player character Gp subsequently stops at the conversation position Gic. For example, consider a case where the player character Gp stops at the conversation position Gic before the generation information "Cy, Ay" displayed in the reunion event is generated. In this case, a conversation event other than the reunion event may start when the player character Gp stops at the conversation position Gic. Furthermore, if the generation information "Cy, Ay" displayed in the reunion event is generated during a period after the end of a conversation event other than the reunion event, the reunion event may start when the player character Gp subsequently stops at the conversation position Gic again.

[0107] (7) In each embodiment, a specific example of the present invention applied to a card battle game has been described. However, the present invention may be applied to games other than card battle games. For example, the present invention can be used in a location-based game in which the game progresses as the user moves around the real world, a rhythm game in which the user performs predetermined operations in time with music, a puzzle game in which predetermined puzzles can be completed on the terminal device 10, a quiz game in which predetermined quizzes are given, and the like.

[0108] <Summary of the functions and effects of the exemplary embodiments of the present invention> <First aspect> The programs (PGx, PGy) of this aspect are programs that cause a computer (terminal device 10, management device 20) to progress an event in a game, and cause the computer to function as control means (11) that progresses the event (negotiation event, reunion event) in a manner (in which Cx, Ax are displayed) according to generation information (Cx, Ax) generated by the machine learning model (30) before the event starts. According to this aspect, the interest of the service is improved.

[0109] <Second mode> In the program of this aspect, the events include a preceding event (a negotiation event) and a subsequent event (a reunion event) that starts after the preceding event, and the control means progresses the subsequent event in a manner according to generated information (Cy, Ay) generated by the machine learning model from information (scenario Cx2) displayed by a user's selection (operation of selection button Bx1) in the preceding event. According to this aspect, the interest of the service is improved.

[0110] <Third aspect> In the program of this aspect, the events include a preceding event (negotiation event) and a subsequent event (reunion event) that starts after the preceding event, and the control means progresses the preceding event in a manner according to generated information (Cx, Ax) generated by the machine learning model before the preceding event starts, and progresses the subsequent event in a manner according to generated information (Cy, Ay) generated by the machine learning model from information displayed by a user's selection in the preceding event. According to this aspect, the interest of the service is improved.

[0111] <Fourth aspect> In the program of this aspect, the event is configured to include a pre-selection part (negotiation part, reunion part) in which a selection area (selection button Bx, By) is presented to the user, and a post-selection part that starts when the user selects the selection area, and the control means makes it possible to display the selection area in a mode according to generation information (option string Ay1, option string Ay2) generated by the machine learning model based on the type of the post-selection part (Dx51 in FIG. 7, Dy41 in FIG. 9). According to this aspect, the interest of the service is improved.

[0112] <Fifth aspect> The program of this aspect causes a computer to function as a granting means (12) that grants a predetermined benefit (card Gc) in accordance with a user's selection in an event (reunion event) (see FIG. 10). According to this aspect, the interest of the service is improved.

[0113] <Sixth aspect> The program of this aspect is an information processing system (1) that progresses an event in a game, and includes control means that progresses the event using generation information generated by a machine learning model before the event starts. According to this aspect, the same effects as those of the first aspect can be achieved.

[0114] The problem-solving means configured by the above-mentioned program (for example, each configuration described in the appendix) can be appropriately diverted to an apparatus, a system, a method, an information recording medium, etc. [Explanation of symbols]

[0115] 1...information processing system, 10...terminal device, 11...control means, 12...assignment means, 20...management device, machine learning model 30.

Claims

1. A program that causes a computer to progress events in a game, A control means for progressing the event in a manner according to generation information generated by a machine learning model before the event is started. A program that causes the computer to function as a

2. The events include a preceding event and a subsequent event that starts after the preceding event, The control means progresses the subsequent event in a manner according to generated information generated by a machine learning model from information displayed in response to a user selection in the preceding event. The program according to claim 1.

3. The events include a preceding event and a subsequent event that starts after the preceding event, The control means causes the preceding event to proceed in a manner corresponding to generated information generated by a machine learning model before the preceding event is started, and causes the subsequent event to proceed in a manner corresponding to generated information generated by the machine learning model from information displayed in response to a user selection in the preceding event. The program according to claim 1.

4. the event includes a pre-selection part in which a selection area is presented to a user and a post-selection part that starts when the user selects the selection area; The control means enables the selection area to be displayed in a manner according to generation information generated by a machine learning model based on the type of the selected part. The program according to claim 1.

5. A granting means for granting a predetermined benefit in accordance with a user's selection in the event.

2. The program according to claim 1, which causes the computer to function as:

6. An information processing system for progressing an event in a game, A control means for progressing the event using information generated by a machine learning model before the event is started. An information processing system comprising:

Images