Human-computer Guided Interaction Method, Device, Computer Equipment and Readable Storage Medium

Abstract

The present application provides a human-computer guided interaction method, apparatus, computer device, and readable storage medium. The method includes: in response to a player character in a game scene initiating a conversation with a non-player character and not detecting a task objective, obtaining the character conversation content; wherein the task objective includes a first objective set for the player character; if the character conversation content is not empty, determining a target decision space based on the character conversation content to obtain the first objective; wherein the target decision space includes target guidance information for guiding replies and a target judgment space for determining whether the first objective has been achieved; based on the target guidance information, generating and displaying a first reply message to obtain a second reply message fed back by the player character in response to the first reply message; and in response to detecting the first objective, determining whether the first objective has been achieved based on the second reply information and the target judgment space. This method can improve the activity of a player's game account.

Description

Technical Field

[0001] The present application relates to the field of game technology, and in particular to a human-computer guided interaction method, apparatus, computer equipment, and readable storage medium. Background Art

[0002] In a game scenario, when players have conversations with NPCs (non-player characters), they usually chat aimlessly and the content of the conversation is very random. However, in order to meet the needs of players, NPCs need to give answers that are in line with the current scenario.

[0003] However, the NPC's response method in existing game scenarios is usually based on preset formulas, which cannot fully understand the player's conversation intentions. This leads to the player receiving contradictory information during the conversation with the NPC, or the NPC's answer cannot meet the player's needs. In serious cases, the player will not continue to chat with the NPC, causing the player to give up the game.

[0004] Therefore, the human-computer interaction method in the existing game scene has the technical problem that the dialogue is random and uncontrollable, resulting in low activity of players' game accounts. Summary of the Invention

[0005] Based on this, it is necessary to provide a human-computer guided interaction method, device, computer equipment and readable storage medium to address the above technical problems, so as to improve the activity of players' game accounts by enhancing the controllability of human-computer dialogue.

[0006] In a first aspect, the present application provides a human-computer guided interaction method, comprising:

[0007] In response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, obtaining character conversation content; wherein the task objective includes a first objective set for the player character;

[0008] If the character dialogue content is not empty, then based on the character dialogue content, a target decision space is determined to obtain a first target; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for judging whether the first target is achieved;

[0009] Based on the target guidance information, first reply information is generated and displayed to obtain second reply information fed back by the player character in response to the first reply information.

[0010] In response to detecting the first goal, it is determined whether the first goal is achieved based on the second reply information and the goal judgment space.

[0011] In a second aspect, the present application provides a human-machine guided interaction device, comprising:

[0012] An information acquisition module, configured to, in response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, acquire the content of the character conversation; wherein the task objective includes a first objective set for the player character;

[0013] A target decision module is configured to determine a target decision space based on the character dialogue content to obtain a first target if the character dialogue content is not empty; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for determining whether the first target has been achieved;

[0014] A reply guidance module is used to generate and display first reply information based on the target guidance information to obtain second reply information fed back by the player character in response to the first reply information;

[0015] The target judgment module is used to determine whether the first target is achieved based on the second reply information and the target judgment space in response to detecting the first target.

[0016] In a third aspect, the present application further provides a computer device, comprising:

[0017] one or more processors;

[0018] a memory; and one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the processor to implement the above-mentioned human-computer guided interaction method.

[0019] In a fourth aspect, the present application also provides a computer-readable storage medium on which a computer program is stored, and the computer program is loaded by a processor to execute the above-mentioned human-computer guided interaction method.

[0020] In a fifth aspect, embodiments of the present application provide a computer program product or computer program, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the human-computer guided interaction method provided in the first aspect.

[0021] The above-mentioned human-computer guided interaction method, device, computer equipment and computer-readable storage medium, in response to the fact that the task target is not detected when the player character and the non-player character start a dialogue in the game scene, first obtain the character dialogue content, and then determine the target decision space based on the character dialogue content when it is determined that the character dialogue content is not empty, thereby obtaining the first target, and then based on the target guidance information included in the target decision space, generate and display the first reply information to obtain the second reply information fed back by the player character in response to the first reply information, so as to respond to the first target being detected in the game scene, based on the second reply information and the target judgment space included in the target decision space, determine whether the first target is achieved, and finally allocate certain game resources according to the achievement of the task target to improve the activity of the player's game account. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 This is an application scenario diagram of the human-computer guided interaction method provided in an embodiment of the present application;

[0024] Figure 2 Schematic diagram of the human-machine guided interaction method provided in the embodiment of the present application;

[0025] Figure 3 Schematic diagram of the structure of the human-machine guided interaction device provided in an embodiment of the present application;

[0026] Figure 4 It is a structural diagram of the computer device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0028] In the description of this application, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art will recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in this application.

[0029] See Figure 1 The human-computer guided interaction method can be implemented and executed based on a cloud interaction system, wherein the cloud interaction system includes a terminal device 102 and a server 104. The terminal device 102 can be a device that includes both receiving and transmitting hardware, that is, a device that has receiving and transmitting hardware capable of performing two-way communication over a two-way communication link. Such devices may include: cellular or other communication devices having a single-line display or a multi-line display. The terminal device 102 can specifically be a desktop terminal or a mobile terminal. The terminal device 102 can also specifically be a mobile phone, a tablet computer, a laptop computer, or a game console. The server 104 can be a standalone server or a server network or server cluster composed of servers, including but not limited to computers, network hosts, a single network server, an edge server, a collection of multiple network servers, or a cloud server composed of multiple servers. The cloud server is composed of a large number of computers or network servers based on cloud computing. In addition, a communication connection is established between the terminal device 102 and the server 104 via a network, which can specifically be any of a wide area network, a local area network, and a metropolitan area network.

[0030] In some embodiments of the present application, various cloud applications, such as cloud games, can be run under the cloud interaction system. Taking cloud games as an example, cloud games refer to a gaming method based on cloud computing. In the cloud gaming operation mode, the operating body of the game program and the main body of the game screen presentation are separated. The storage and operation of the human-computer guided interaction method are completed on the cloud gaming server. The role of the terminal device is to receive and send data and present the game screen. For example, the terminal device can be a display device with data transmission function close to the user side, such as a mobile terminal, TV, computer, PDA, game console, etc.; however, it is the cloud gaming server in the cloud that performs information processing. When playing the game, the player operates the terminal device to send operation instructions to the cloud gaming server. The cloud gaming server runs the game according to the operation instructions, encodes and compresses the game screen and other data, and returns it to the terminal device through the network. Finally, the terminal device decodes and outputs the game screen.

[0031] In some embodiments of the present application, the terminal device may be a local terminal device. Taking a game as an example, the local terminal device stores the game program and is used to present the game screen. The local terminal device is used to interact with the player through a graphical user interface, that is, conventionally downloading and installing the game program through an electronic device and running it. The local terminal device may provide the graphical user interface to the player in a variety of ways, for example, it may be rendered on the terminal's display screen, or provided to the player through a holographic projection. For example, the local terminal device may include a display screen and a processor, the display screen being used to present the graphical user interface, the graphical user interface including the game screen, and the processor being used to run the game, generate the graphical user interface, and control the display of the graphical user interface on the display screen.

[0032] A game scene (or virtual scene) is a virtual scene displayed (or provided) when an application is running on a terminal or server. Optionally, the virtual scene is a simulation of the real world, a virtual environment that is partially simulated and partially fictional, or a purely fictional virtual environment. The virtual scene can be either a two-dimensional virtual scene or a three-dimensional virtual scene. The virtual environment can be the sky, land, ocean, etc., where the land includes environmental elements such as deserts and cities. Among them, the virtual scene is a scene with complete game logic of virtual objects such as user control. For example, for sandbox 3D shooting games, the virtual scene is a 3D game world for players to control virtual objects to fight. Example virtual scenes may include: mountains, plains, rivers, lakes, oceans, deserts, skies, plants, buildings, and at least one of vehicles. For example, for 2D card games, the virtual scene is a scene for displaying released cards or displaying virtual objects corresponding to cards. Example virtual scenes may include: an arena, a decisive battlefield, or other "field" elements or other elements that can display the status of card battles. For 2D or 3D multiplayer online tactical competitive games, the virtual scene is a 2D or 3D terrain scene for virtual objects to fight. Example virtual scenes may include: canyon-style mountains, routes, rivers, classrooms, tables and chairs, podiums and other elements.

[0033] The game interface refers to the interface corresponding to the application provided or displayed through a graphical user interface, which includes a UI interface and a game screen for players to interact. In an optional embodiment, the UI interface may include game controls (such as skill controls, movement controls, function controls, etc.), indicators (such as direction indicators, character indicators, etc.), information display areas (such as game time, etc.), or game setting controls (such as system settings, stores, gold coins, etc.). In an optional embodiment, the game screen is a display screen corresponding to the virtual scene displayed by the terminal device, and the game screen may include virtual objects such as game characters (also called "player characters"), NPC characters (also called "non-player characters"), and AI characters that execute game logic in the virtual scene.

[0034] It should be noted that the game scenario described in the embodiment of the present application is intended to more clearly illustrate the technical solution of the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Ordinary technicians in this field can know that with the emergence of abnormal new business scenarios, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.

[0035] See Figure 2 The present application embodiment provides a human-computer guided interaction method. The following embodiments will apply this method to the above Figure 1Taking the terminal device 102 in the example, the method includes steps S201 to S204, which are specifically as follows:

[0036] S201, in response to not detecting a task target when a player character initiates a conversation with a non-player character in a game scene, obtaining character conversation content; wherein the task target includes a first target set for the player character.

[0037] Non-player characters, also known as non-player characters (NPCs), refer to virtual characters in the game that are not controlled by the user. The types of NPCs present in actual game scenes may include at least one type: storyline NPCs, combat NPCs, and so on. Storyline NPCs are NPCs that provide storylines or information to the controlled virtual character. When the controlled virtual character performs tasks in the game scene, the tasks are accompanied by corresponding storylines. During actual gameplay, the controlled virtual character can interact with the storyline NPCs and receive corresponding interactive feedback from them, such as dropping virtual items, providing prompts, and issuing new tasks. Combat NPCs are NPCs that can engage in combat with the controlled virtual character. Combat NPCs can be on the same team as the controlled virtual character, or on a different team. During actual gameplay, the controlled virtual character can interact with the combat NPCs and receive corresponding interactive feedback from them, such as assisting with attacks, providing supplementary items, and providing instructions for attack targets.

[0038] Interaction with non-player characters (NPCs) serves as a medium of communication between players and the game world, providing a crucial way to understand relevant information within the game world. This information guides players through the completion of quests, the progression of game scenes, and the unfolding of the game story. Therefore, improving interaction with NPCs in the game, especially the relevance of dialogue, is crucial for increasing player account activity.

[0039] In a specific implementation, the terminal device 102 can display a graphical user interface (GUI) by running a game application. The GUI displays a game scene with at least two virtual characters, including at least one player character and at least one non-player character. However, if there are at least two player characters in the current game scene, a target player character can be determined based on the player character's game state (e.g., the state of a pending dialogue with a non-player character). The target player character and the non-player character can then be analyzed to achieve guided human-computer interaction. Optionally, the player character and non-player character in the game scene can be at least one of a human figure, an elf figure (e.g., a fairy), a monster figure (e.g., a man-eating beast), and other figures capable of dialogue.

[0040] Furthermore, non-player characters often require certain triggering conditions (such as dialogue content or actions by the player character) before they begin interacting. It should be noted that the player character engaging in a conversation with the non-player character can be a controlled virtual character currently controlled by the user. For example, in a massively multiplayer online role-playing game, when a controlled virtual character and a non-player character (e.g., a non-player character issuing a quest) meet (e.g., the facing distance between the two characters is less than a certain threshold), the system will determine that the controlled virtual character and the non-player character will initiate a conversation. At this point, the terminal device 102 can detect whether a first target and a second target are currently present, and then, if the first target and the second target are not detected, obtain the content of the character conversation.

[0041] It should be noted that the task objectives in the embodiments of the present application include at least one of the following: a first objective and a second objective; the first objective is a task objective set for the player character, and the executor of the goal after setting it should be the player character. For example, if the first objective is [Request - Candied Haws], it means that the NPC needs a candied haws. At this time, the player character must complete the action of delivering the candied haws to the NPC before the goal is determined to be achieved; the second objective is a task objective set for a non-player character, and the executor of the goal after setting it should be the NPC. For example, if the second objective is [Hold Hands], it means that the player character wants to hold hands with the NPC. At this time, the NPC must respond with actions or words to whether it is willing to hold hands with the player character before the goal is determined to be achieved. The character dialogue content can include historical dialogue information, and in addition, it can also include the trigger conditions mentioned above, that is, the dialogue content submitted by the player character to trigger the non-player character to start interaction.

[0042] For example, in a certain game scene, there is a player character "Shaoxia" and a non-player character "Achan". When Shaoxia says to Achan, "I have a lot of delicious food here", if it is detected that the historical dialogue information between the two characters is empty, the character dialogue content may include "I have a lot of delicious food here".

[0043] S202, if the character dialogue content is not empty, determine the target decision space based on the character dialogue content to obtain the first goal; wherein the target decision space includes target guidance information for guiding the response and a target judgment space for judging whether the first goal is achieved.

[0044] Among them, the decision space is a set of goals, including at least one task goal that can be used as the first goal or the second goal. For example, the decision space is: [Fight] "Achan wants to fight with the young man", [Leave] "Achan wants to leave", [Gift - Steamed Buns] "Achan plans to give a steamed bun to the young man", [Gift - Kite] "Achan plans to give a kite to the young man", [Request - Candied Haws] "Achan needs a Candied Haws at this time", etc.

[0045] Guidance information can be used to guide NPC responses with text messages that closely match the target decision space. This allows the player character to fully grasp the controllability of the conversation, preventing NPCs from answering irrelevant questions while enhancing feedback. This encourages players to actively control the direction of the conversation, thereby increasing the activity of their game account. Each decision space contains a corresponding guidance message, and the guidance information contained in the target decision space is the target guidance message. For example, if the target decision space is: [Request - Candied Haws] "Achan needs a candied haws at this time," the corresponding target guidance message could be: "Achan, thinking he needs a bunch of candied haws and asking the young man to give it to him, said: "

[0046] The judgment space can be a player action or semantic description used to determine whether the player character has completed the corresponding goal. Each decision space contains at least one judgment space, and the judgment space contained in the target decision space is the target judgment space. For example, if the target decision space is: [Request - Candied Haws] "Achan needs a Candied Haws at this time," the corresponding target judgment space can be: "Young Man - Gift - Candied Haws" and / or "Young Man - Gift - Other Items."

[0047] In a specific implementation, after the terminal device 102 obtains the character dialogue content, it can analyze the character dialogue content to filter out a target decision space that matches the character dialogue content from a preset decision space. The filtered target decision space can be used as the basis for determining the first target.

[0048] In one embodiment, the human-computer guided interaction method also includes: if the character dialogue content is empty, displaying at least one preset candidate target; determining a second target in response to a trigger operation submitted by the player character for the candidate target; obtaining a second target judgment result for indicating whether the second target is achieved, and determining second target judgment guidance information based on the second target judgment result and the second target; generating and displaying third reply information based on the second target judgment guidance information; determining that the second target has been achieved in response to receiving interactive information fed back by the player character for the third reply information; wherein the interactive information includes action information and / or text information.

[0049] In the specific implementation, the above embodiment describes the situation where the character dialogue content is not empty, that is, when it is detected that the character dialogue content is not empty, the target decision can be made based on the non-empty character dialogue content, that is, the first target is determined through context semantic analysis; and when it is detected that the character dialogue content is empty, the system cannot make a target decision, and can only rely on the player to specify the second target, or the system randomly specifies the first target. Specifically, this embodiment will explain in detail the execution logic when the player specifies the second target.

[0050] That is, when the terminal device 102 detects that the character's conversation content is empty, it first displays at least one preset candidate goal, including but not limited to a drop-down bar, pop-up window, or drawer display. Then, based on the player's selected information, a second goal (e.g., "I want to hold hands") is determined. However, due to the aforementioned situation where the character's conversation content is empty, the terminal device 102 cannot determine whether the NPC satisfies the player based on the current information. Therefore, the second goal judgment result should be "uncertain" rather than "achieved" or "not achieved." Based on the second goal judgment result and the second goal, a second goal judgment guidance message (e.g., "Achan hesitated about whether to hold hands with the young man, so he said:") can be generated, which in turn guides the generation and display of a third response message encouraging the player character to continue interacting (e.g., "This is our first meeting, this isn't a good idea"). At this point, if the player character receives an interactive response to the third response message (e.g., "How about I take you to buy some delicious food?"), the second goal can be determined to have been achieved based on actual business needs. For example, if the business requirement is set to consider the goal achieved as long as interactive information is received, the second goal can be determined to be achieved; if the business requirement is set to determine the goal achievement based on the semantics of the interactive information / interaction count, the semantics of the interactive information and the number of interactions can be further analyzed to see whether the subject matter and the number of interactions match the preset conditions. If they do, the second goal is determined to be achieved. It should be noted that after the second goal is achieved, the terminal device 102 can issue a task reward to the corresponding player character, or unlock a new game scene for the corresponding player character, which is not limited in the specific embodiment of the present application.

[0051] In one embodiment, step S202 includes: if the character dialogue content is not empty, obtaining scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; through the trained decision model, analyzing the scene description information and the character dialogue content to determine the target decision space and obtain the first goal.

[0052] The scene description information includes character introduction information and / or character status information. For example, character introduction information could be "Young man, you have just come down from your journey" or "Achan is a young monk in the temple at the foot of the mountain"; character status information could be "Achan is cleaning outside the temple, and the player encounters him at the temple entrance"; character dialogue content could include historical dialogue content and the trigger conditions described above. For example, if historical dialogue content is empty, the trigger condition could be "I have a lot of delicious food here."

[0053] The trained decision model is based on the "Roberta-Large" model trained on the training corpus. The training method is to give the previous context and a sentence, and then the model scores the sentence. The obtained score represents the probability of whether the sentence can be inferred based on the previous context.

[0054] In a specific implementation, the model training process can be set up to be executed in the server 104. After the model training is completed and a trained decision model is obtained, the terminal device 102 can call the trained decision model to determine the target decision space. Specifically, before model training, the corpus required for model training must be constructed. The corpus preparation method can be: using a sliding window to extract the long text "sentence1" and the following short text "sentence2" as a positive sample sentence pair (sentence1, sentence2), and also randomly extract a short sentence "sentence3" and the long text "sentence1" to construct a negative sample sentence pair (sentence1, sentence3). And so on, a large amount of training corpus can be obtained.

[0055] Furthermore, to better align the model with the target decision scenario, we annotate data from actual conversational corpora. For example, in the conversation above, "sentence 1" is: [The young man had just come down the mountain for a tour and ran into Achan, a young monk who had just left the temple. The young man said to Achan, "I have a lot of delicious food here."] Then, we issue an annotation task and select the most appropriate sentence from the pre-set decision space, such as [Request for a candied haws]. The corresponding text for this target is "sentence 2": So Achan asked for a string of candied haws. This constructs a positive sentence pair (sentence 1, sentence 2). For other targets, such as "sentence 3": So Achan wanted to fight the young man, we construct a negative sentence pair (sentence 1, sentence 3). By mixing these two parts of the corpus together, we train the model using a binary classification strategy, enabling the model to classify each sentence pair as either a positive or negative example, thus completing the model training task.

[0056] Furthermore, after model training is complete, a trained decision model can be obtained. The terminal device 102 then uses the scene description information and character dialogue content as input, feeds this model into the model for analysis, and thereby determines the target decision space. The target decision space determination scheme involved in this embodiment will be described in detail below.

[0057] In one embodiment, a trained decision model is used to analyze scene description information and character dialogue content to determine a target decision space and obtain a first target, including: inputting the scene description information and character dialogue content into the trained decision model so that the trained decision model scores multiple pre-stored decision spaces and outputs score information for each decision space; screening out decision spaces whose corresponding score information is greater than or equal to a preset score threshold as target decision spaces; wherein each decision space includes guidance information, judgment space, and descriptive content; determining the descriptive content of the target decision space as the first target; wherein the target guidance information is the guidance information included in the target decision space, and the target judgment space is the judgment space included in the target decision space.

[0058] The description content may be text description information of the target decision space. For example, if the target decision space is: [Battle] "Achan wants to fight with the young man", the description content may be "{npc} wants to fight with {user}".

[0059] In a specific implementation, the decision model's input includes not only scene description information and character dialogue content, but also multiple pre-stored decision spaces. This information and character dialogue content constitute the "previous context," with each decision space serving as a candidate, forming sentence pairs. In this case, inputting multiple sentence pairs [(previous context, candidate 1), (previous context, candidate 2), (previous context, candidate 3)...] into the trained decision model yields a score [0.9, 0.3, 0.6,...] for each sentence pair, with higher scores indicating a higher likelihood of positive examples. A threshold (e.g., "0.8") is then set, and all candidates above this threshold are considered to meet the target of the current context. During actual output, the candidate with the highest score, or a random candidate with a score exceeding the threshold, can be directly selected. Of course, if all candidate scores are below the threshold, then no target exists. This allows the target decision space to be selected, and the first target set for the player character to be determined based on the target decision space. Of course, if all scores are below the threshold, the target decision space is empty.

[0060] It should be noted that since the embodiments of this application focus on how to determine and judge a target, the solution provided in step S203 actually applies to the case where the target decision space is not empty. In actual application scenarios, if the target decision space is empty, the terminal device 102 can obtain the second target specified by the player character, or obtain the first target specified by the NPC, and then execute subsequent steps based on the specified target.

[0061] S203: Generate and display first reply information based on the target guidance information to obtain second reply information fed back by the player character in response to the first reply information.

[0062] The first reply information may be a text message used to reply to the player character.

[0063] In a specific implementation, after the terminal device 102 determines the target decision space based on the character's dialogue content analysis, it can then determine target guidance information. This target guidance information then guides the response module to generate information that matches the player character's request or question, serving as a first response. At this point, the first response information is displayed on the terminal device 102 for the player character to review. Simultaneously, after obtaining the player character's second response information in response to the first response information, the terminal device 102 can analyze whether the second response information satisfies the first target, thereby determining whether the task corresponding to the first target has been achieved.

[0064] In one embodiment, the target decision space also includes target description content, and step S203 includes: inputting the target guidance information and the character dialogue content into the trained response model, and outputting the first response information; displaying the first response information to obtain the interactive information of the player character in response to the first response information as the second response information; wherein the interactive information includes action information and / or text information.

[0065] In a specific implementation, in order to obtain the first reply information, the terminal device 102 may input the target guidance information and the character dialogue content into the trained reply model after analyzing the target guidance information, so that the model outputs the first reply information. Specifically, the trained reply model may adopt any open source reply generation model, such as the General Pre-Training (GPT) language model, which is a pre-trained language model that uses "Transformer" as a feature extractor and is trained based on a language model. This model is highly dependent on the above content. If it is allowed to continue naturally, it is usually difficult to control the direction of continuation. Therefore, the embodiment of the present application proposes to introduce target guidance information to indicate the subsequent generation direction of the model, so that the reply content related to the player character dialogue content can be obtained with a high probability.

[0066] Furthermore, by displaying the first reply information, the first reply information can be provided to the player character for review, and then the action information and / or text information fed back by the player character in response to the first reply information can be obtained as the second reply information, so as to subsequently analyze whether the player character has completed the first goal based on the second reply information.

[0067] In one embodiment, target guidance information and character dialogue content are input into a trained response model, and first response information is output, including: obtaining scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; inputting the scene description information, target guidance information, and character dialogue content into the trained response model, and outputting the first response information.

[0068] In the specific implementation, in order to improve the accuracy of the model response and thus enhance the activity of the player's game account, the embodiment of the present application proposes that the scene description information can be combined with the target guidance information and the character dialogue content to output the first reply information. The first reply information that refers to the scene description information is more in line with the dialogue intention proposed by the player character than the first reply information that does not refer to the scene description information.

[0069] For example, the input of the trained response model is: [The young man just came down the mountain to travel and met A Chan, a young monk who had just left the temple. The young man said to A Chan: "I have a lot of delicious food here"; A Chan thought he needed a string of candied haws and wanted the young man to give him one, so he said:], then the output can be: [Young man, can you give me a string of candied haws? I really want to eat some]. Among them, "The young man just came down the mountain to travel and met A Chan, a young monk who had just left the temple" is the scene description information, "The young man said to A Chan: "I have a lot of delicious food here" is the character dialogue content, "A Chan thought he needed a string of candied haws and wanted the young man to give him one, so he said:" is the target guidance information, and "Young man, can you give me a string of candied haws? I really want to eat some" is the first response information.

[0070] S204 : In response to detecting the first target, determining whether the first target is achieved based on the second reply information and the target judgment space.

[0071] In a specific implementation, after detecting the presence of a first goal in the current game scenario, the terminal device 102 can compare and analyze the second reply information fed back by the player character in response to the first reply information with the target judgment space to determine whether the first goal has been achieved. If it is determined that the first goal has been achieved, the player character is configured with preset game resources, such as issuing a preset task reward corresponding to the first goal to the player character to increase the player's game account activity. It is understandable that each decision space can be bound to a certain number of task rewards. After a decision space is determined to be a target decision space, the first goal has been determined. At this time, if it is determined that the goal has been achieved, a reward can be issued. The game reward can be game equipment, game cards, game lives, etc.

[0072] In one embodiment, the target decision space also includes a target completion plan, and the target judgment space includes first target judgment guidance information. After step S204, it also includes: obtaining a first target judgment result used to indicate whether the first target has been achieved; analyzing the second reply information to determine whether the player character has completed the target completion plan and obtain a plan judgment result; generating and displaying a fourth reply information based on the first target judgment result, the plan judgment result and the first target judgment guidance information, and issuing a task reward to the player character when it is determined that the first target has been achieved.

[0073] A goal completion plan can be a sequence of actions that a player character or non-player character must perform to complete a goal. For example, the plan for the goal "Get Warm" could be: {user} - Give Gift - Money -> {NPC} - Buy - Hot Water, or: {user} - Give Gift - Hot Water. For most simple goals, the plan defaults to a single action for the player character. For example, the plan for the goal "Request - Candied Haws" could be: {user} - Give Gift - Candied Haws. "User" refers to the player character, and "NPC" refers to the non-player character.

[0074] Among them, the essence of the first target judgment guidance information is the same as the target guidance information. The only difference is that the target guidance information is associated with the target decision space, while the first target judgment guidance information is associated with the target judgment space. The function of both is to guide non-player characters to reply to information related to the dialogue content of the player characters.

[0075] In a specific implementation, after detecting the existence of the first goal, the terminal device 102 may analyze the second reply information provided by the player character and the degree of match between the goal judgment space, thereby determining whether the first goal has been achieved by the player character, thereby obtaining a first goal judgment result indicating the progress of the goal achievement. Simultaneously, the second reply information may be further analyzed separately to obtain a solution judgment result, and then, based on the first goal judgment result, the solution judgment result, and the first goal judgment guidance information, a fourth reply information may be generated and displayed, and a task reward may be issued to the player character if it is determined that the first goal has been achieved.

[0076] In one embodiment, the target judgment space includes at least one set of text collections, which are composed of at least one role subject, at least one action predicate and at least one object object. Obtaining a first target judgment result for indicating whether the first target is achieved includes: analyzing the second reply information and the text collection based on a preset hard matching algorithm to determine whether the first target is achieved; if the second reply information is action information, and the action information matches the text collection, then determining that the first target judgment result is that the target has been achieved; if the second reply information is action information, and the action information matches the text collection inconsistently, then determining that the first target judgment result is that the target has not been achieved; if the second reply information is text information, then determining that the first target judgment result is that the target has not been achieved.

[0077] Among them, the hard matching algorithm is a pragmatic level method based on keywords or restriction rules.

[0078] In a specific implementation, the target judgment space in this embodiment includes at least one set of text sets, and the text set is composed of at least one role subject, at least one action predicate, and at least one object object. For example, the target judgment space includes: a. Young man - gift giving - candied haws; b. Young man - gift giving - other items. The role subject is "young man", the action predicate is "gift giving", and the object objects are "candied haws" and "other items". Analysis shows that the target judgment space at this time is "player action", so if the second reply information fed back by the player character is action information, and the action information matches one of the above-mentioned target judgment spaces, then it can be determined that the first target judgment result is that the target has been achieved, otherwise the target has not been achieved. However, if the second reply information is text information, then the first target judgment result can be determined to be that the target has not been achieved without analyzing the text information.

[0079] In one embodiment, the target judgment space includes at least one set of text collections, which are composed of at least one role subject, at least one action predicate and at least one object object. Obtaining a first target judgment result for indicating whether the first target is achieved includes: analyzing the second reply information and the text collection based on a preset semantic similarity algorithm to determine whether the first target is achieved; if the second reply information is action information, obtaining first action description information corresponding to the action information and second action description information corresponding to the text collection to obtain a first similarity between the first action description information and each second action description information, and when detecting that the first similarity is greater than or equal to a preset similarity threshold, determining that the first target judgment result is that the target has been achieved; if the first similarity is less than the preset similarity threshold, determining that the first target judgment result is that the target has not been achieved; if the second reply information is text information, determining that the first target judgment result is that the target has not been achieved.

[0080] The statistical method for calculating word semantic similarity is an empirical approach that bases the study of word similarity on observable linguistic facts. It is based on the assumption that two words are semantically similar if and only if they are in similar contexts. It utilizes large-scale corpora and uses word contextual information as a reference for calculating semantic similarity.

[0081] In a specific implementation, the target judgment space in this embodiment also includes at least one set of text collections, and the text collections are also composed of at least one character subject, at least one action predicate, and at least one object. For example, the target judgment space includes: a. Young man - gift - candied haws; b. Young man - gift - other items. Unlike the previous embodiment, the algorithm used in this embodiment is a semantic similarity algorithm, that is, the terminal device 102 can analyze the semantic similarity between the second reply information fed back by the player character and one of the above target judgment spaces. If, based on the analysis to determine that the information type is consistent, and the similarity value is measured to be higher than a threshold, the first target judgment result can be determined to be the goal achieved, otherwise the goal is not achieved.

[0082] In one embodiment, the target judgment space includes at least one piece of semantic description information, and obtaining a first target judgment result for indicating whether the first target is achieved includes: based on a preset semantic similarity algorithm, analyzing the second reply information and the semantic description information to determine whether the first target is achieved; if the second reply information is text information, obtaining a second similarity between the text information and the semantic description information, and when detecting that the second similarity is greater than or equal to a preset similarity threshold, determining that the first target judgment result is that the target has been achieved; if the second similarity is less than the preset similarity threshold, determining that the first target judgment result is that the target has not been achieved; if the second reply information is action information, determining that the first target judgment result is that the target has not been achieved.

[0083] In specific implementations, the target judgment space in this embodiment includes at least one semantic description. For example, the target judgment space includes: a. A Chan understands the meaning of life, b. A Chan does not understand the true meaning of life from the young man. Analysis shows that the target judgment space in this case is a "semantic description." Therefore, if the second reply provided by the player character is textual, and the semantic similarity between the textual information and one of the target judgment spaces is above a preset threshold, the first target judgment result can be determined as goal achieved; otherwise, the goal is not achieved. However, if the second reply is action information, the first target judgment result can be determined as goal not achieved without analyzing the action information.

[0084] In one embodiment, the second reply information is analyzed to determine whether the player character has completed the target completion plan to obtain a plan judgment result, including: analyzing the second reply information to determine whether the player character has completed the target completion plan; if the second reply information is action information, and it is detected that the player character has executed the completion action information, then the plan judgment result is determined to be that the plan has ended; if the second reply information is action information, and it is detected that the player character has not executed the completion action information, then the plan judgment result is determined to be that the plan has not ended; if the second reply information is text information, and the text information is displayed, then the plan judgment result is determined to be that the plan has ended; if the second reply information is text information, and the text information is not displayed, then the plan judgment result is determined to be that the plan has not ended.

[0085] In a specific implementation, the terminal device 102 may separately analyze the second reply information to determine whether the player character has completed the target completion plan. That is, when the second reply information is determined to be an action information type, the terminal device 102 may further detect whether the player character has executed the instructions of the action information. If so, the plan determination result may be determined to be that the plan has been completed; otherwise, the plan has not been completed. If the second reply information is determined to be a text message, the terminal device 102 may further detect whether the text message has been displayed in the interactive interface contained in the terminal device 102. If so, the plan determination result may be determined to be that the plan has been completed; otherwise, the plan has not been completed.

[0086] In one embodiment, a fourth reply message is generated and displayed based on the first target judgment result, the solution judgment result, and the first target judgment guidance information, and a task reward is issued to the player character when it is determined that the first target has been achieved, including: inputting the first target judgment result, the solution judgment result, and the first target judgment guidance information into the trained reply model, outputting the fourth reply message; displaying the fourth reply message; analyzing the first target judgment result and the solution judgment result, so as to issue the task reward to the player character when it is determined that the first target has been achieved.

[0087] In the specific implementation, the operating mechanism of the reply model has been explained above and will not be repeated here.

[0088] In one embodiment, the target decision space also includes a target completion plan, analyzing the first target judgment result and the plan judgment result to issue a task reward to the player character when it is determined that the first target has been achieved, including: if the first target judgment result is that the target has been achieved and the plan judgment result is that the plan has ended, then the first target is cleared and the task reward is issued to the player character; if the first target judgment result is that the target has been achieved and the plan judgment result is that the plan has not ended, then the plan completion progress information of the target completion plan is obtained and the plan completion progress information is displayed until the plan judgment result is that the plan has ended, and the task reward is issued to the player character; if the first target judgment result is that the target has not been achieved, then the preset number of interaction rounds is increased by one, and the interaction information fed back by the player character in response to the fourth reply information is obtained as the fifth reply information, and the fifth reply information is analyzed with the target judgment space until the number of interaction rounds reaches the preset upper limit of the number of interaction rounds, then the first target is cleared and the task reward is issued to the player character.

[0089] In the specific implementation, this embodiment illustrates that the terminal device 102 can analyze whether to clear the first target of the current analysis and whether to issue the bound task reward to the player character according to the first target judgment result and the plan judgment result according to the established strategy, so as to stimulate the player character's interest in the game and improve its account activity.

[0090] The human-computer guided interaction method in the above embodiment, in response to the fact that the task target is not detected when the player character and the non-player character start a dialogue in the game scene, first obtains the character dialogue content, and then determines the target decision space based on the character dialogue content when it is determined that the character dialogue content is not empty, thereby obtaining the first target, and then generates and displays the first reply information based on the target guidance information included in the target decision space to obtain the second reply information fed back by the player character in response to the first reply information, so as to respond to the first target being detected in the game scene, based on the second reply information and the target judgment space included in the target decision space, determine whether the first target is achieved, and finally allocate certain game resources according to the achievement of the task target to improve the activity of the player's game account.

[0091] It should be understood that although Figure 2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 2At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0092] In order to better implement the human-machine guided interaction method provided in the embodiment of the present application, on the basis of the human-machine guided interaction method proposed in the embodiment of the present application, the embodiment of the present application also provides a human-machine guided interaction device, such as Figure 3 As shown, the human-machine guided interaction device 300 includes:

[0093] The information acquisition module 310 is configured to, in response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, acquire the content of the character conversation; wherein the task objective includes a first objective set for the player character;

[0094] A target decision module 320 is configured to determine a target decision space based on the character dialogue content to obtain a first target if the character dialogue content is not empty; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for determining whether the first target has been achieved;

[0095] The reply guidance module 330 is used to generate and display first reply information based on the target guidance information to obtain second reply information fed back by the player character in response to the first reply information;

[0096] The target judgment module 340 is configured to, in response to detecting the first target, determine whether the first target is achieved based on the second reply information and the target judgment space.

[0097] In one embodiment, the task goal also includes a second goal set for a non-player character, and the human-computer guidance interaction device 300 also includes a goal designation module, which is used to display at least one preset candidate goal if the character dialogue content is empty; determine the second goal in response to a trigger operation submitted by the player character for the candidate goal; obtain a second goal judgment result for indicating whether the second goal is achieved, and determine second goal judgment guidance information based on the second goal judgment result and the second goal; generate and display third reply information based on the second goal judgment guidance information; determine that the second goal has been achieved in response to receiving interactive information fed back by the player character in response to the third reply information; wherein the interactive information includes action information and / or text information.

[0098] In one embodiment, the target decision module 320 is also used to obtain scene description information of the game scene if the character dialogue content is not empty; wherein the scene description information includes character introduction information and / or character status information; through the trained decision model, the scene description information and the character dialogue content are analyzed to determine the target decision space and obtain the first target.

[0099] In one embodiment, the target decision module 320 is also used to input the scene description information and the character dialogue content into the trained decision model, so that the trained decision model scores the pre-stored multiple decision spaces and outputs the score information of each decision space; screens out the decision spaces whose corresponding score information is greater than or equal to the preset score threshold as the target decision space; wherein each decision space includes guidance information, judgment space and description content; determines the description content of the target decision space as the first target; wherein the target guidance information is the guidance information included in the target decision space, and the target judgment space is the judgment space included in the target decision space.

[0100] In one embodiment, the target decision space also includes target description content, and the reply guidance module 330 is further used to input the target guidance information and the character dialogue content into the trained reply model to output the first reply information; display the first reply information to obtain the interactive information of the player character in response to the first reply information as the second reply information; wherein the interactive information includes action information and / or text information.

[0101] In one embodiment, the reply guidance module 330 is also used to obtain scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; the scene description information, target guidance information and character dialogue content are input into the trained reply model, and the first reply information is output.

[0102] In one embodiment, the target decision space also includes a target completion plan, and the target judgment space includes first target judgment guidance information. The target judgment module 340 is also used to obtain a first target judgment result used to indicate whether the first target has been achieved; analyze the second reply information to determine whether the player character has completed the target completion plan and obtain a plan judgment result; generate and display a fourth reply information based on the first target judgment result, the plan judgment result and the first target judgment guidance information, and issue a task reward to the player character when it is determined that the first target has been achieved.

[0103] In one embodiment, the target judgment space includes at least one set of text collections, which are composed of at least one role subject, at least one action predicate and at least one object object. The target judgment module 340 is also used to analyze the second reply information and the text collection based on a preset hard matching algorithm to determine whether the first target is achieved; if the second reply information is action information, and the action information matches the text collection, then the first target judgment result is determined to be that the target has been achieved; if the second reply information is action information, and the action information does not match the text collection, then the first target judgment result is determined to be that the target has not been achieved; if the second reply information is text information, then the first target judgment result is determined to be that the target has not been achieved.

[0104] In one embodiment, the target judgment space includes at least one set of text collections, which are composed of at least one role subject, at least one action predicate and at least one object object. The target judgment module 340 is also used to analyze the second reply information and the text collection based on a preset semantic similarity algorithm to determine whether the first goal is achieved; if the second reply information is action information, the first action description information corresponding to the action information and the second action description information corresponding to the text collection are obtained to obtain the first similarity between the first action description information and each second action description information, and when it is detected that the first similarity is greater than or equal to the preset similarity threshold, the first target judgment result is determined to be that the goal has been achieved; if the first similarity is less than the preset similarity threshold, the first target judgment result is determined to be that the goal has not been achieved; if the second reply information is text information, the first target judgment result is determined to be that the goal has not been achieved.

[0105] In one embodiment, the target judgment space includes at least one piece of semantic description information, and the target judgment module 340 is also used to analyze the second reply information and the semantic description information based on a preset semantic similarity algorithm to determine whether the first target is achieved; if the second reply information is text information, the second similarity between the text information and the semantic description information is obtained, and when it is detected that the second similarity is greater than or equal to the preset similarity threshold, the first target judgment result is determined to be that the target has been achieved; if the second similarity is less than the preset similarity threshold, the first target judgment result is determined to be that the target has not been achieved; if the second reply information is action information, the first target judgment result is determined to be that the target has not been achieved.

[0106] In one embodiment, the target judgment module 340 is also used to analyze the second reply information to determine whether the player character has completed the target completion plan; if the second reply information is action information, and it is detected that the player character has executed the completion action information, then the plan judgment result is determined to be that the plan has ended; if the second reply information is action information, and it is detected that the player character has not executed the completion action information, then the plan judgment result is determined to be that the plan has not ended; if the second reply information is text information, and the text information is displayed, then the plan judgment result is determined to be that the plan has ended; if the second reply information is text information, and the text information is not displayed, then the plan judgment result is determined to be that the plan has not ended.

[0107] In one embodiment, the target judgment module 340 is also used to input the first target judgment result, the solution judgment result and the first target judgment guidance information into the trained response model, output the fourth response information; display the fourth response information; analyze the first target judgment result and the solution judgment result, so as to issue a task reward to the player character when it is determined that the first target has been achieved.

[0108] In one embodiment, the target judgment module 340 is also used to clear the first target and issue a task reward to the player character if the first target judgment result is that the target has been achieved and the plan judgment result is that the plan has ended; if the first target judgment result is that the target has been achieved and the plan judgment result is that the plan has not ended, then obtain the plan completion progress information of the target completion plan and display the plan completion progress information until the plan judgment result is that the plan has ended, and then issue a task reward to the player character; if the first target judgment result is that the target has not been achieved, then add one to the preset number of interaction rounds, and obtain the interactive information of the player character in response to the fourth reply information as the fifth reply information, so as to analyze the fifth reply information and the target judgment space, until the number of interaction rounds reaches the preset upper limit of the number of interaction rounds, then clear the first target and issue a task reward to the player character.

[0109] In the above embodiment, in response to the fact that the task target is not detected when the player character and the non-player character start a dialogue in the game scene, the character dialogue content is first obtained, and then the target decision space is determined based on the character dialogue content when it is determined that the character dialogue content is not empty, so as to obtain the first target, and then based on the target guidance information included in the target decision space, a first reply message is generated and displayed to obtain the second reply message fed back by the player character in response to the first reply message, so as to respond to the first target being detected in the game scene, based on the second reply message and the target judgment space included in the target decision space, it is determined whether the first target is achieved, and finally certain game resources are allocated according to the achievement of the task target to improve the activity of the player's game account.

[0110] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a human-computer guided interaction method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0111] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0112] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0113] In response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, obtaining character conversation content; wherein the task objective includes a first objective set for the player character;

[0114] If the character dialogue content is not empty, then based on the character dialogue content, a target decision space is determined to obtain a first target; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for judging whether the first target is achieved;

[0115] Based on the target guidance information, first reply information is generated and displayed to obtain second reply information fed back by the player character in response to the first reply information;

[0116] In response to detecting the first goal, it is determined whether the first goal is achieved based on the second reply information and the goal judgment space.

[0117] In one embodiment, the mission objective further includes a second objective set for a non-player character, and when the processor executes the computer program, the following steps are further implemented:

[0118] If the character dialogue content is empty, at least one preset candidate target is displayed; in response to the trigger operation submitted by the player character for the candidate target, a second target is determined; a second target judgment result for indicating whether the second target is achieved is obtained, and second target judgment guidance information is determined based on the second target judgment result and the second target; based on the second target judgment guidance information, a third reply information is generated and displayed; in response to receiving the interactive information fed back by the player character for the third reply information, it is determined that the second target has been achieved; wherein the interactive information includes action information and / or text information.

[0119] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0120] If the character dialogue content is not empty, the scene description information of the game scene is obtained; wherein the scene description information includes character introduction information and / or character status information; through the trained decision model, the scene description information and the character dialogue content are analyzed to determine the target decision space and obtain the first goal.

[0121] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0122] The scenario description information and character dialogue content are input into a trained decision model so that the trained decision model scores multiple pre-stored decision spaces and outputs score information for each decision space; the decision spaces whose corresponding score information is greater than or equal to a preset score threshold are screened out as target decision spaces; wherein each decision space includes guidance information, a judgment space, and descriptive content; the descriptive content of the target decision space is determined as a first target; wherein the target guidance information is the guidance information included in the target decision space, and the target judgment space is the judgment space included in the target decision space.

[0123] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0124] The target guidance information and the character dialogue content are input into the trained response model to output the first response information; the first response information is displayed to obtain the interactive information of the player character in response to the first response information as the second response information; wherein the interactive information includes action information and / or text information.

[0125] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0126] Obtain scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; input the scene description information, target guidance information and character dialogue content into the trained response model, and output first response information.

[0127] In one embodiment, the target decision space further includes a target completion plan; the target judgment space includes first target judgment guidance information, and when the processor executes the computer program, the processor further implements the following steps:

[0128] Obtain a first target judgment result for indicating whether the first target has been achieved; analyze the second reply information to determine whether the player character has completed the target completion plan and obtain a plan judgment result; generate and display a fourth reply information based on the first target judgment result, the plan judgment result and the first target judgment guidance information, and issue a task reward to the player character when it is determined that the first target has been achieved.

[0129] In one embodiment, the target judgment space includes at least one set of texts, the text set consisting of at least one role subject, at least one action predicate, and at least one object. When the processor executes the computer program, the processor further implements the following steps:

[0130] Based on the preset hard matching algorithm, the second reply information and the text set are analyzed to determine whether the first goal is achieved; if the second reply information is action information, and the action information matches the text set, then the first goal judgment result is determined to be that the goal has been achieved; if the second reply information is action information, and the action information does not match the text set, then the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is text information, then the first goal judgment result is determined to be that the goal has not been achieved.

[0131] In one embodiment, the target judgment space includes at least one set of texts, the text set consisting of at least one role subject, at least one action predicate, and at least one object. When the processor executes the computer program, the processor further implements the following steps:

[0132] Based on a preset semantic similarity algorithm, the second reply information and the text set are analyzed to determine whether the first goal is achieved; if the second reply information is action information, the first action description information corresponding to the action information and the second action description information corresponding to the text set are obtained to obtain the first similarity between the first action description information and each second action description information, and when it is detected that the first similarity is greater than or equal to the preset similarity threshold, the first goal judgment result is determined to be that the goal has been achieved; if the first similarity is less than the preset similarity threshold, the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is text information, the first goal judgment result is determined to be that the goal has not been achieved.

[0133] In one embodiment, the target judgment space includes at least one piece of semantic description information, and when the processor executes the computer program, the processor further implements the following steps:

[0134] Based on a preset semantic similarity algorithm, the second reply information and the semantic description information are analyzed to determine whether the first goal is achieved; if the second reply information is text information, the second similarity between the text information and the semantic description information is obtained, and when it is detected that the second similarity is greater than or equal to the preset similarity threshold, the first goal judgment result is determined to be that the goal has been achieved; if the second similarity is less than the preset similarity threshold, the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is action information, the first goal judgment result is determined to be that the goal has not been achieved.

[0135] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0136] Analyze the second reply information to determine whether the player character has completed the target completion plan; if the second reply information is action information, and it is detected that the player character has executed the completion action information, then the plan judgment result is determined to be that the plan has ended; if the second reply information is action information, and it is detected that the player character has not executed the completion action information, then the plan judgment result is determined to be that the plan has not ended; if the second reply information is text information, and the text information is displayed, then the plan judgment result is determined to be that the plan has ended; if the second reply information is text information, and the text information is not displayed, then the plan judgment result is determined to be that the plan has not ended.

[0137] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0138] The first target judgment result, the solution judgment result and the first target judgment guidance information are input into the trained response model, and the fourth response information is output; the fourth response information is displayed; the first target judgment result and the solution judgment result are analyzed to issue a task reward to the player character when it is determined that the first target has been achieved.

[0139] In one embodiment, the goal decision space further includes a goal completion plan, and when the processor executes the computer program, the processor further implements the following steps:

[0140] If the first target judgment result is that the target has been achieved, and the plan judgment result is that the plan has ended, the first target is cleared and the task reward is issued to the player character; if the first target judgment result is that the target has been achieved, and the plan judgment result is that the plan has not ended, the plan completion progress information of the target completion plan is obtained, and the plan completion progress information is displayed until the plan judgment result is that the plan has ended, and the task reward is issued to the player character; if the first target judgment result is that the target has not been achieved, the preset number of interaction rounds is increased by one, and the interaction information of the player character in response to the fourth reply information is obtained as the fifth reply information, so as to analyze the fifth reply information and the target judgment space, until the number of interaction rounds reaches the preset upper limit of the number of interaction rounds, the first target is cleared and the task reward is issued to the player character.

[0141] In the above embodiment, in response to the fact that the task target is not detected when the player character and the non-player character start a dialogue in the game scene, the character dialogue content is first obtained, and then the target decision space is determined based on the character dialogue content when it is determined that the character dialogue content is not empty, so as to obtain the first target, and then based on the target guidance information included in the target decision space, a first reply message is generated and displayed to obtain the second reply message fed back by the player character in response to the first reply message, so as to respond to the first target being detected in the game scene, based on the second reply message and the target judgment space included in the target decision space, it is determined whether the first target is achieved, and finally certain game resources are allocated according to the achievement of the task target to improve the activity of the player's game account.

[0142] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0143] In response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, obtaining character conversation content; wherein the task objective includes a first objective set for the player character;

[0144] If the character dialogue content is not empty, then based on the character dialogue content, a target decision space is determined to obtain a first target; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for judging whether the first target is achieved;

[0145] Based on the target guidance information, first reply information is generated and displayed to obtain second reply information fed back by the player character in response to the first reply information;

[0146] In response to detecting the first goal, it is determined whether the first goal is achieved based on the second reply information and the goal judgment space.

[0147] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0148] If the character dialogue content is empty, at least one preset candidate target is displayed; in response to the trigger operation submitted by the player character for the candidate target, a second target is determined; a second target judgment result for indicating whether the second target is achieved is obtained, and second target judgment guidance information is determined based on the second target judgment result and the second target; based on the second target judgment guidance information, a third reply information is generated and displayed; in response to receiving the interactive information fed back by the player character for the third reply information, it is determined that the second target has been achieved; wherein the interactive information includes action information and / or text information.

[0149] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0150] The scenario description information and character dialogue content are input into a trained decision model so that the trained decision model scores multiple pre-stored decision spaces and outputs score information for each decision space; the decision spaces whose corresponding score information is greater than or equal to a preset score threshold are screened out as target decision spaces; wherein each decision space includes guidance information, a judgment space, and descriptive content; the descriptive content of the target decision space is determined as a first target; wherein the target guidance information is the guidance information included in the target decision space, and the target judgment space is the judgment space included in the target decision space.

[0151] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0152] The target guidance information and the character dialogue content are input into the trained response model to output the first response information; the first response information is displayed to obtain the interactive information of the player character in response to the first response information as the second response information; wherein the interactive information includes action information and / or text information.

[0153] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0154] Obtain scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; input the scene description information, target guidance information and character dialogue content into the trained response model, and output first response information.

[0155] In one embodiment, the target decision space further includes a target completion plan; the target judgment space includes first target judgment guidance information, and when the computer program is executed by the processor, the following steps are further implemented:

[0156] Obtain a first target judgment result for indicating whether the first target has been achieved; analyze the second reply information to determine whether the player character has completed the target completion plan and obtain a plan judgment result; generate and display a fourth reply information based on the first target judgment result, the plan judgment result and the first target judgment guidance information, and issue a task reward to the player character when it is determined that the first target has been achieved.

[0157] In one embodiment, the target judgment space includes at least one set of texts, the text set consisting of at least one role subject, at least one action predicate, and at least one object. When the computer program is executed by a processor, the following steps are further implemented:

[0158] Based on the preset hard matching algorithm, the second reply information and the text set are analyzed to determine whether the first goal is achieved; if the second reply information is action information, and the action information matches the text set, then the first goal judgment result is determined to be that the goal has been achieved; if the second reply information is action information, and the action information does not match the text set, then the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is text information, then the first goal judgment result is determined to be that the goal has not been achieved.

[0159] In one embodiment, the target judgment space includes at least one set of texts, the text set consisting of at least one role subject, at least one action predicate, and at least one object. When the computer program is executed by a processor, the following steps are further implemented:

[0160] Based on a preset semantic similarity algorithm, the second reply information and the text set are analyzed to determine whether the first goal is achieved; if the second reply information is action information, the first action description information corresponding to the action information and the second action description information corresponding to the text set are obtained to obtain the first similarity between the first action description information and each second action description information, and when it is detected that the first similarity is greater than or equal to the preset similarity threshold, the first goal judgment result is determined to be that the goal has been achieved; if the first similarity is less than the preset similarity threshold, the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is text information, the first goal judgment result is determined to be that the goal has not been achieved.

[0161] In one embodiment, the target judgment space includes at least one piece of semantic description information, and when the computer program is executed by a processor, the following steps are further implemented:

[0162] Based on a preset semantic similarity algorithm, the second reply information and the semantic description information are analyzed to determine whether the first goal is achieved; if the second reply information is text information, the second similarity between the text information and the semantic description information is obtained, and when it is detected that the second similarity is greater than or equal to the preset similarity threshold, the first goal judgment result is determined to be that the goal has been achieved; if the second similarity is less than the preset similarity threshold, the first goal judgment result is determined to be that the goal has not been achieved; if the second reply information is action information, the first goal judgment result is determined to be that the goal has not been achieved.

[0163] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0164] Analyze the second reply information to determine whether the player character has completed the target completion plan; if the second reply information is action information, and it is detected that the player character has executed the completion action information, then the plan judgment result is determined to be that the plan has ended; if the second reply information is action information, and it is detected that the player character has not executed the completion action information, then the plan judgment result is determined to be that the plan has not ended; if the second reply information is text information, and the text information is displayed, then the plan judgment result is determined to be that the plan has ended; if the second reply information is text information, and the text information is not displayed, then the plan judgment result is determined to be that the plan has not ended.

[0165] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0166] The first target judgment result, the solution judgment result and the first target judgment guidance information are input into the trained response model, and the fourth response information is output; the fourth response information is displayed; the first target judgment result and the solution judgment result are analyzed to issue a task reward to the player character when it is determined that the first target has been achieved.

[0167] In one embodiment, the goal decision space further includes a goal completion plan, and when the computer program is executed by the processor, the following steps are further implemented:

[0168] If the first target judgment result is that the target has been achieved, and the plan judgment result is that the plan has ended, the first target is cleared and the task reward is issued to the player character; if the first target judgment result is that the target has been achieved, and the plan judgment result is that the plan has not ended, the plan completion progress information of the target completion plan is obtained, and the plan completion progress information is displayed until the plan judgment result is that the plan has ended, and the task reward is issued to the player character; if the first target judgment result is that the target has not been achieved, the preset number of interaction rounds is increased by one, and the interaction information of the player character in response to the fourth reply information is obtained as the fifth reply information, so as to analyze the fifth reply information and the target judgment space, until the number of interaction rounds reaches the preset upper limit of the number of interaction rounds, the first target is cleared and the task reward is issued to the player character.

[0169] In the above embodiment, in response to the fact that the task target is not detected when the player character and the non-player character start a dialogue in the game scene, the character dialogue content is first obtained, and then the target decision space is determined based on the character dialogue content when it is determined that the character dialogue content is not empty, so as to obtain the first target, and then based on the target guidance information included in the target decision space, a first reply message is generated and displayed to obtain the second reply message fed back by the player character in response to the first reply message, so as to respond to the first target being detected in the game scene, based on the second reply message and the target judgment space included in the target decision space, it is determined whether the first target is achieved, and finally certain game resources are allocated according to the achievement of the task target to improve the activity of the player's game account.

[0170] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Any reference to memory, storage, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0171] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0172] The above is a detailed introduction to a human-computer guided interaction method, device, computer equipment and computer-readable storage medium provided in the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A human-computer guided interaction method, characterized in that: include: In response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, obtaining character conversation content; wherein the task objective includes a first objective set for the player character; If the character dialogue content is not empty, determining a target decision space based on the character dialogue content to obtain the first target; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for judging whether the first target is achieved; Based on the target guidance information, first reply information is generated and displayed to obtain second reply information fed back by the player character in response to the first reply information; In response to detecting the first goal, determining whether the first goal is achieved based on the second reply information and the goal judgment space; The target decision space further includes a target completion plan; the target judgment space includes first target judgment guidance information; after determining whether the first target is achieved based on the second reply information and the target judgment space in response to detecting the first target, the method further includes: Obtaining a first goal determination result for indicating whether the first goal has been achieved; Analyzing the second reply information to determine whether the player character has completed the goal completion plan, and obtaining a plan determination result; According to the first target judgment result, the solution judgment result and the first target judgment guidance information, a fourth reply message is generated and displayed, and a task reward is issued to the player character when it is determined that the first target has been achieved.

2. The method according to claim 1, wherein The mission objective further includes a second objective set for the non-player character, and the method further includes: If the character dialogue content is empty, display at least one preset candidate target; In response to a trigger operation submitted by the player character for the candidate target, determining the second target; Obtaining a second target determination result indicating whether the second target has been achieved, and determining second target determination guidance information based on the second target determination result and the second target; generating and displaying a third reply message based on the second target judgment guidance information; In response to receiving interactive information fed back by the player character in response to the third reply information, it is determined that the second goal has been achieved; wherein the interactive information includes action information and / or text information.

3. The method according to claim 1, wherein If the character dialogue content is not empty, determining a target decision space based on the character dialogue content to obtain the first goal includes: If the character dialogue content is not empty, obtaining scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; The trained decision model is used to analyze the scene description information and the character dialogue content to determine the target decision space and obtain the first goal.

4. The method according to claim 3, wherein The step of analyzing the scene description information and the character dialogue content using the trained decision model to determine the target decision space and obtain the first goal includes: Inputting the scene description information and the character dialogue content into a trained decision model, so that the trained decision model scores a plurality of pre-stored decision spaces and outputs score information of each decision space; Filtering out decision spaces corresponding to the score information that is greater than or equal to a preset score threshold as the target decision space; wherein each decision space includes guidance information, a judgment space, and a description content; Determining a description of the target decision space as the first target; The target guidance information is the guidance information included in the target decision space, and the target judgment space is the judgment space included in the target decision space.

5. The method according to claim 1, wherein The generating and displaying first reply information based on the target guidance information to obtain second reply information fed back by the player character in response to the first reply information includes: Inputting the target guidance information and the character dialogue content into a trained response model, and outputting the first response information; The first reply information is displayed to obtain interactive information fed back by the player character in response to the first reply information as the second reply information; wherein the interactive information includes action information and / or text information.

6. The method according to claim 5, wherein inputting the target guidance information and the character dialogue content into a trained response model and outputting the first response information comprises: Obtaining scene description information of the game scene; wherein the scene description information includes character introduction information and / or character status information; The scene description information, the target guidance information and the character dialogue content are input into the trained response model, and the first response information is output.

7. The method according to claim 1, wherein The target judgment space includes at least one set of text collections, wherein the text collections are composed of at least one role subject, at least one action predicate, and at least one object; The obtaining of a first goal determination result indicating whether the first goal has been achieved includes: Analyzing the second reply information and the text set based on a preset hard matching algorithm to determine whether the first goal is achieved; If the second reply information is action information, and the action information matches the text set, determining that the first goal judgment result is that the goal has been achieved; If the second reply information is action information, and the action information does not match the text set, determining that the first goal judgment result is that the goal is not achieved; If the second reply information is a text message, it is determined that the first target judgment result is that the target is not achieved.

8. The method according to claim 1, wherein The target judgment space includes at least one set of text collections, wherein the text collections are composed of at least one role subject, at least one action predicate, and at least one object; The obtaining of a first goal determination result indicating whether the first goal has been achieved includes: Analyzing the second reply information and the text set based on a preset semantic similarity algorithm to determine whether the first goal has been achieved; If the second reply information is action information, obtaining first action description information corresponding to the action information and second action description information corresponding to the text set to obtain a first similarity between the first action description information and each of the second action description information, and determining that the first goal judgment result is that the goal has been achieved when it is detected that the first similarity is greater than or equal to a preset similarity threshold; If the first similarity is less than the preset similarity threshold, determining that the first target judgment result is that the target is not achieved; If the second reply information is a text message, it is determined that the first target judgment result is that the target is not achieved.

9. The method according to claim 1, wherein The target judgment space includes at least one piece of semantic description information; The obtaining of a first goal determination result indicating whether the first goal has been achieved includes: Analyzing the second reply information and the semantic description information based on a preset semantic similarity algorithm to determine whether the first goal has been achieved; If the second reply information is a text message, obtaining a second similarity between the text message and the semantic description information, and determining that the first goal judgment result is that the goal has been achieved when detecting that the second similarity is greater than or equal to a preset similarity threshold; If the second similarity is less than the preset similarity threshold, determining that the first target judgment result is that the target is not achieved; If the second reply information is action information, it is determined that the first target judgment result is that the target is not achieved.

10. The method according to claim 1, wherein The analyzing the second reply information to determine whether the player character has completed the goal completion plan and obtaining a plan determination result includes: analyzing the second reply information to determine whether the player character has completed the goal completion plan; If the second reply information is action information, and it is detected that the player character has completed the action information, then determining that the solution judgment result is that the solution has ended; If the second reply information is action information, and it is detected that the player character has not completed the action information, then determining that the solution judgment result is that the solution is not completed; If the second reply information is a text message, and the text message has been displayed, determining that the solution judgment result is that the solution has ended; If the second reply information is a text message, and the text message is not displayed, it is determined that the solution judgment result is that the solution is not completed.

11. The method according to claim 1, wherein The generating and displaying fourth reply information according to the first goal judgment result, the solution judgment result, and the first goal judgment guidance information, and issuing a task reward to the player character when it is determined that the first goal has been achieved, includes: Inputting the first target judgment result, the solution judgment result, and the first target judgment guidance information into the trained response model, and outputting fourth response information; displaying the fourth reply information; The first goal determination result and the solution determination result are analyzed to issue a task reward to the player character when it is determined that the first goal has been achieved.

12. The method according to claim 11, wherein The target decision space also includes a target completion plan; The analyzing the first goal determination result and the solution determination result to issue a task reward to the player character when it is determined that the first goal has been achieved includes: If the first goal is determined to have been achieved and the plan is determined to have ended, the first goal is cleared and a task reward is issued to the player character; If the first goal is determined to have been achieved, and the plan is determined to have not been completed, then obtaining the plan completion progress information of the goal completion plan and displaying the plan completion progress information until the plan is determined to have been completed, and then issuing the task reward to the player character; If the first target judgment result is that the target is not achieved, the preset number of interaction rounds is increased by one, and the interactive information fed back by the player character in response to the fourth reply information is obtained as the fifth reply information, so as to analyze the fifth reply information and the target judgment space until the number of interaction rounds reaches the preset upper limit of the number of interaction rounds, clear the first target and issue a task reward to the player character.

13. A human-machine guided interaction device, characterized in that: include: An information acquisition module, configured to, in response to not detecting a task objective when a player character initiates a conversation with a non-player character in a game scene, acquire content of the character conversation; wherein the task objective includes a first objective set for the player character; a target decision module, configured to determine a target decision space based on the character dialogue content to obtain the first target if the character dialogue content is not empty; wherein the target decision space includes target guidance information for guiding responses and a target judgment space for determining whether the first target is achieved; A reply guidance module, configured to generate and display first reply information based on the target guidance information, so as to obtain second reply information fed back by the player character in response to the first reply information; a target judgment module, configured to, in response to detecting the first target, determine whether the first target is achieved based on the second reply information and the target judgment space; The target decision space also includes a target completion plan; the target judgment space includes first target judgment guidance information; and the device also includes a reward module, which is used to: Obtaining a first goal determination result for indicating whether the first goal has been achieved; Analyzing the second reply information to determine whether the player character has completed the goal completion plan, and obtaining a plan determination result; According to the first target judgment result, the solution judgment result and the first target judgment guidance information, a fourth reply message is generated and displayed, and a task reward is issued to the player character when it is determined that the first target has been achieved.

14. A computer device, characterized in that: include: one or more processors; Memory; and one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the processor to implement the human-computer guided interaction method according to any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that A computer program is stored thereon, and the computer program is loaded by a processor to execute the steps of the human-computer guided interaction method according to any one of claims 1 to 12.

Images