Appearance interaction method and device, electronic equipment and storage medium

Abstract

The appearance interaction method provided in the application provides a graphical user interface through a terminal device, acquires dialogue information in a social state in response to a controlled virtual character entering the social state, and controls the controlled virtual character through the terminal device; semantic analysis is performed based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intention; in response to the controlled virtual character having the appearance interaction intention, an appearance interaction control is displayed on the graphical user interface, and the appearance interaction control is used to provide content corresponding to the appearance interaction intention. Through the above steps, the appearance interaction method provided in the application can effectively solve the technical problems of fragmented user experience and complicated operation caused by the fact that the game social system and the appearance system are independent of each other and information is not communicated in the prior art.

Description

Technical Field

[0001] This application relates to the field of computers, and more specifically to an appearance interaction method, device, electronic device, and storage medium. Background Technology

[0002] In existing online games or social applications, the appearance system (such as fashion, skins, and accessories) and the social system (such as chat channels and private messages with friends) of virtual characters are usually two independent functional modules. When players are in a social scenario, such as chatting with friends and talking about an appearance they are interested in, the operation path to view the detailed information of that appearance or try it on is usually cumbersome and fragmented. To illustrate, imagine Player A sees Player B showcasing a new outfit in a chat and becomes interested. Player A must first interrupt or exit the current chat, then manually open the game's main menu, find the "Shop," "Wardrobe," or "Appearance Equipment" entry, and enter a completely new full-screen interface. Next, Player A must navigate through a vast array of items, filtering, searching, or flipping through pages to locate the desired outfit. Throughout this process, Player A is completely disconnected from the original social context, unable to simultaneously see their friend's chat content, resulting in a forced interruption of the social experience. Similarly, if a player wants to change their existing appearance during a chat based on the topic or their mood—for example, changing a chat bubble or avatar frame—they must follow a similar interrupted workflow.

[0003] The aforementioned existing technologies suffer from significant technical flaws: First, the social system and the appearance system suffer from severe information silos, lacking effective data exchange and linkage mechanisms, preventing the system from providing intelligent assistance to players based on social contexts. Second, the social system lacks context awareness, failing to capture and understand players' immediate needs or intentions regarding appearance revealed in conversations, thus hindering accurate recommendations and resulting in the loss of potential commercial conversion opportunities. This fragmented operational process greatly reduces the user experience, increases the difficulty and learning cost for users, and makes browsing and changing appearances a high-cost activity, thereby inhibiting players' willingness to spend and personalize their appearance. Summary of the Invention

[0004] This application provides an appearance interaction method, device, electronic device, and storage medium, which can effectively solve the technical problems of fragmented user experience and cumbersome operation caused by the independence and lack of information exchange between game social systems and appearance systems in the prior art.

[0005] In a first aspect, embodiments of this application provide an appearance interaction method, which provides a graphical user interface through a terminal device. The method includes: in response to a controlled virtual character entering a social state, acquiring dialogue information in the social state, and controlling the controlled virtual character through the terminal device; performing semantic analysis based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intention; and in response to the controlled virtual character having an appearance interaction intention, displaying appearance interaction controls on the graphical user interface, the appearance interaction controls being used to provide content corresponding to the appearance interaction intention.

[0006] Secondly, embodiments of this application provide an appearance interaction device that provides a graphical user interface through a terminal device. The device includes: an acquisition module, configured to acquire dialogue information in the social state in response to a controlled virtual character entering a social state, wherein the controlled virtual character is controlled through the terminal device; an analysis module, configured to perform semantic analysis based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intent; and a display module, configured to display appearance interaction controls on the graphical user interface in response to the controlled virtual character having an appearance interaction intent, wherein the appearance interaction controls are used to provide content corresponding to the appearance interaction intent.

[0007] Thirdly, embodiments of this application provide an electronic device including a processor and a memory. The memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the above-described appearance interaction method.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement the aforementioned appearance interaction method.

[0009] The appearance interaction method provided in this application offers a graphical user interface (GUI) via a terminal device. In response to a controlled virtual character entering a social state, it acquires dialogue information within that state, allowing the controlled virtual character to be controlled via the terminal device. Semantic analysis is performed based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intent. If the controlled virtual character has such an intent, appearance interaction controls are displayed on the GUI, providing content corresponding to that intent. Through these steps, the appearance interaction method provided in this application effectively solves the technical problems of fragmented user experience and cumbersome operation caused by the independence and lack of information exchange between game social systems and appearance systems in existing technologies.

[0010] Other features and advantages of this application will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the above-described techniques of this application.

[0011] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating the appearance interaction method provided in an embodiment of this application;

[0014] Figure 2 This is a schematic diagram of an interface of the appearance interaction method provided in an embodiment of this application; Figure 3 This is another schematic diagram of the appearance interaction method provided in the embodiments of this application; Figure 4 This is another schematic diagram of the appearance interaction method provided in the embodiments of this application; Figure 5 This is a schematic diagram of the appearance interaction device provided in an embodiment of this application; Figure 6 This is a structural block diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0015] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0016] It should be noted that the terms "first," "second," "third," etc., in the claims, specification, and drawings of this application are used to distinguish similar objects and are not used to describe a specific order or sequence. Such data are interchangeable where appropriate so that the embodiments of this application described herein can be implemented in a sequence other than that shown or described herein. Furthermore, the terms "comprising," "having," and their variations are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or apparatuses.

[0017] It should be understood that in the embodiments of this application, "at least one" refers to one or more, "several" refers to one or more, and "more than" refers to two or more. "And / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship. "Contains A, B, and / or C" means containing any one, two, or three of A, B, and C.

[0018] It should be understood that in the embodiments of this application, "B corresponding to A", "B corresponding to A", "A corresponds to B" or "B corresponds to A" means that B is associated with A, and B can be determined based on A. Determining B based on A does not mean that B is determined solely based on A; B can also be determined based on A and / or other information.

[0019] The appearance interaction method provided in this application can be executed by an electronic device, which can be a terminal or a server. The terminal can be a smartphone, tablet, laptop, or other terminal device. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can be hardware or software. When the server is hardware, it can be implemented as a distributed server cluster composed of multiple servers or as a single server. When the server is software, it can be implemented as multiple software programs or software modules (e.g., software or software modules used to provide distributed services) or as a single software program or software module. This application does not specifically limit this aspect.

[0020] In one alternative embodiment, taking a game as an example, when the appearance interaction method is run on a terminal device, the terminal device stores a game application and a virtual game scene. The terminal device interacts with the player through a sender's graphical user interface. The terminal device can provide the sender's graphical user interface to the player in various ways, such as rendering it on the terminal device's display screen or presenting the sender's graphical user interface through holographic projection.

[0021] In an optional embodiment, taking cloud gaming as an example, when the appearance interaction method runs on a server, the method can be implemented and executed based on a cloud gaming system. A cloud gaming system refers to a gaming method based on cloud computing. A cloud gaming system includes a server and client devices. The main body running the game application and the main body presenting the game screen are separate. The storage and operation of the methods for controlling vehicles in the game are completed on the server. The presentation of the game screen is completed on the client, which is mainly used for receiving and sending game data and presenting the game screen. For example, the client can be a display device with data transmission capabilities located close to the player, such as a mobile terminal, television, computer, PDA, personal digital assistant, head-mounted display device, etc. However, the terminal device for processing game data is the server in the cloud. When playing the game, the player operates the client to send instructions to the server. The server controls the game to run according to the instructions, encodes and compresses the game screen and other data, returns it to the client through the network, and finally, the client decodes and outputs the game screen.

[0022] This application provides a visual interaction method, apparatus, electronic device, and computer-readable storage medium. The visual interaction apparatus can be integrated into a computer device, and the electronic device can be a server, a terminal, or other similar device.

[0023] Figure 1 This is a flowchart illustrating the appearance interaction method provided in an embodiment of this application.

[0024] This application provides an interactive method for using items in a game, which provides a graphical user interface through a terminal device, such as... Figure 1 As shown, the method includes the following specific steps: Step 101: In response to the controlled virtual character entering a social state, obtain the dialogue information in the social state, and control the controlled virtual character through the terminal device; Step 102: Perform semantic analysis based on dialogue information to determine whether the controlled virtual character has an intention to interact with the appearance. Step 103: In response to the controlled virtual character having an appearance interaction intent, display appearance interaction controls on the graphical user interface. The appearance interaction controls are used to provide content corresponding to the appearance interaction intent.

[0025] Through the above steps, the appearance interaction method provided in this application accurately captures the user's appearance interaction intent by acquiring and analyzing dialogue information in real time in social scenarios, and immediately provides a lightweight interactive control on the current interface to satisfy the intent, avoiding the interruption and operational costs caused by scene switching. This solution greatly enhances the user's interactive experience by achieving a seamless and coherent operation process, integrating behaviors such as browsing, purchasing, and changing appearances into social contexts, creating new interactive gameplay and display opportunities, increasing the richness of the game, and effectively solving the technical problems of fragmented user experience and cumbersome operation caused by the independence and lack of information sharing between the game's social system and appearance system in existing technologies.

[0026] The steps described above are explained in detail below: Figure 2 This is a schematic diagram of an interface of the appearance interaction method provided in the embodiments of this application.

[0027] Step 101: In response to the controlled virtual character entering a social state, obtain the dialogue information in the social state, and control the controlled virtual character through the terminal device.

[0028] Among them, social status is a game scene state that indicates that the virtual character is currently in a state of communication and interaction with other virtual characters.

[0029] Optionally, social status can include scenarios such as virtual characters participating in chat channels, team communication, and private chats with friends. For example, when a player's virtual character joins a team and activates the team chat channel, the virtual character enters a social status.

[0030] Optionally, social status can include the virtual character being in large social scenarios such as guild chat or world chat. For example, when a player speaks or participates in a discussion in a guild channel, their virtual character is in a social status.

[0031] Optionally, social status can include the virtual character being in a multimedia social mode such as voice chat or video call. For example, when a player activates the voice chat function to communicate with other players in real time, the system simultaneously obtains the speech-to-text result as the dialogue information.

[0032] Dialogue information is a type of digitized text data that records the communication content generated by virtual characters in social situations.

[0033] Optionally, the dialogue information may include text chat messages, that is, text content entered by the player via keyboard or touch screen. For example, when a player types "This outfit is really beautiful" in the team channel, the system retrieves and records this text message.

[0034] Optionally, dialogue information may include textual descriptions of non-textual interactive information such as emoticons and stickers. For example, when a player sends a "thumbs up" emoticon, the system can convert it into a text description of "[thumbs up emoticon]" for recording.

[0035] Optionally, the dialogue information may include text information converted from voice chat, using speech recognition technology to convert the player's voice into text in real time. For example, when a player says "I want to buy a new outfit," the speech recognition system converts it into corresponding text and processes it as dialogue information.

[0036] Among them, the controlled virtual character is a digital character in the game that receives control commands from the player through a terminal device and performs corresponding actions.

[0037] Optionally, the controlled virtual character can be the player's main character, that is, the character entity that the player primarily controls and develops. For example, in a role-playing game, the player creates main characters such as warriors, mages, and archers.

[0038] Optionally, the controlled virtual character can be the player's pet, follower, or secondary character—a support character entity that the player can switch between controlling. Examples include the player's pet mount, summoned beast, or alternate character.

[0039] Optionally, the controlled virtual character can be operated through various control methods, including keyboard and mouse, touch screen, gamepad, and voice commands. For example, players can move the character's position by tapping the screen or control the character to perform corresponding actions by using the voice command "open inventory".

[0040] Through the above steps, this embodiment of the application can monitor the player's social activity status in real time and accurately obtain all communication information generated in social scenarios. This real-time and comprehensive information acquisition lays a solid foundation for accurately judging the player's appearance interaction needs.

[0041] Step S102: Perform semantic analysis based on dialogue information to determine whether the controlled virtual character has an intention to interact with the appearance.

[0042] Semantic analysis is a natural language processing technique that can understand and interpret the meaning, emotion, and intention of human language.

[0043] Optionally, semantic analysis can be achieved by invoking a pre-trained natural language recognition model that can deeply understand the contextual meaning of language. For example, using a BERT model based on the Transformer architecture to perform semantic understanding of the dialogue content can identify the purchase intent implied in "This dress looks good, I want one too".

[0044] Optionally, semantic analysis can be aided by keyword matching technology, establishing appearance-related and behavior-related lexicons for precise matching. For example, the appearance-related lexicon contains words such as "fashion," "skin," and "avatar frame," while the behavior-related lexicon contains action words such as "buy," "change," and "try." When these word combinations are detected, intent recognition is triggered.

[0045] Optionally, semantic analysis can improve accuracy by fusing multiple recognition results, comprehensively considering both semantic understanding confidence and keyword matching confidence. For example, when the semantic model outputs an 85% confidence level of purchase intent, and keyword matching identifies three relevant words, the system calculates a combined confidence level exceeding a preset threshold using a weighted algorithm, thus confirming the existence of an interaction intent.

[0046] Among them, appearance interaction intent is a digital expression of user behavior tendencies, which represents the virtual character's needs and desires for appearance-related operations.

[0047] Optionally, appearance interaction intent can be divided into appearance recommendation intent, which is the user's need to receive recommendations for appearance items or purchase suggestions. For example, when a player says "Any good-looking new outfits to recommend?", the system recognizes this as an appearance recommendation intent.

[0048] Optionally, appearance interaction intents can be divided into appearance switching intents, which are the user's desire to change or manage their existing appearance. For example, when a player says "I want to change my avatar frame," the system recognizes it as an appearance switching intent.

[0049] Optionally, appearance interaction intents can include various subcategories, such as browsing intents, collection intents, and sharing intents—more specific behavioral tendencies. For example, when a player says, "How's the effect of this skin? I want to take a look first," the system recognizes it as a browsing intent.

[0050] Among them, the pre-trained natural language recognition model is a language understanding algorithm model trained based on deep learning technology, which has the ability to understand complex language structures and implicit meanings.

[0051] Optionally, the model can be a pre-trained model based on the Transformer architecture, such as BERT or GPT, which can handle contextual semantic relationships. For example, when a player says "This isn't the style I want," the model can understand from the preceding context that this is a negative evaluation of a certain appearance.

[0052] In an alternative implementation, the model can be a specialized model fine-tuned for a corpus specific to the gaming domain, better understanding gaming terminology and player expression habits. For example, the model can understand the true meaning of gaming slang such as "grind," "pay-to-win," and "farm," improving recognition accuracy.

[0053] Optionally, the model can support multilingual recognition and dialect understanding, adapting to the different expressions of players from different regions. For example, the model can understand different terms for appearance in various dialects, such as synonyms for "clothes," "outfit," and "equipment."

[0054] Through the above steps, this application embodiment can accurately identify players' potential needs in chat. For example, when a player types "What clothes, do you want to buy?" in the team channel, the system backend of this application will process it simultaneously. First, the keyword matching module will identify "fashion" (appearance-related vocabulary) and "buy" (behavioral vocabulary). At the same time, the natural language recognition model will analyze the structure and sentiment of the entire sentence, understanding that this is a positive evaluation with a strong desire to purchase. The system will fuse the high semantic confidence output of the model with the keyword matching results, and the resulting comprehensive confidence score will far exceed the threshold, thereby accurately determining that the player has generated an appearance interaction intent, and further characterizing this intent as an appearance recommendation intent, preparing for the subsequent accurate display of appearance recommendation controls.

[0055] Figure 3 This is a schematic diagram of an interface for the appearance interaction method provided in an embodiment of this application.

[0056] In step S103, in response to the controlled virtual character having an appearance interaction intent, an appearance interaction control is displayed on the graphical user interface, the appearance interaction control being used to provide content corresponding to the appearance interaction intent.

[0057] Among them, the appearance interaction control is a windowed UI element that is displayed on top of the current game interface, and is usually semi-transparent or draggable.

[0058] Optionally, the appearance interaction controls may include an appearance recommendation control 201, used to display recommended appearance items and their detailed information to the player. For example, when the system determines that the player's intention is "want to buy" or "what new items are available," indicating a tendency to browse and purchase, the appearance recommendation control 201 will be displayed. This control will show one or more recommended appearance item models, prices, and purchase buttons.

[0059] Optionally, the appearance interaction controls may include an appearance switching control 202, which displays a list of appearances the player already owns for quick switching. For example, when the system determines that the player's intention is to "change" or "match their outfit," which is more about managing existing items, the appearance switching control will be displayed. This control will display the appearances that the player already owns and that are available for switching in the form of a list or icon grid.

[0060] Optionally, the appearance and interaction controls can have interactive features such as draggability, scaling, and minimization to ensure that they do not affect the player's normal gameplay. For example, the control appears on the right side of the screen by default, and the player can drag it to any position or click the minimize button to collapse it, and expand it again when needed.

[0061] Among them, the content corresponding to the appearance interaction intent is the personalized service content provided according to different intent types.

[0062] Optionally, when the intent is recommendation-based, the corresponding content includes a list of recommended fashion items, product details, and purchase links. For example, for an intent to buy fashion items, the system will display currently popular fashion items, including 3D model previews, price information, and how to obtain them.

[0063] Optionally, when the intent is to switch types, the corresponding content includes the player's existing appearance inventory, category filtering options, and one-click switching functionality. For example, for the intent to change an avatar frame, the system will display all the avatar frames the player owns and provide sorting options by quality, acquisition time, and other dimensions.

[0064] Optionally, the content may also include social features, such as sharing, favorites, and gifting options. For example, in the recommended control 201, players can not only view and purchase appearances, but also share their favorite appearances with friends, or directly purchase and gift them to teammates.

[0065] The graphical user interface (GUI) is the visual interface through which players interact with the game; it contains the game screen and various control controls.

[0066] Optionally, the graphical user interface can be the main game screen, containing basic elements such as game scenes, character models, and control buttons. For example, the main game screen when the player is engaged in a dungeon battle includes elements such as a health bar, skill buttons, and a minimap.

[0067] Optionally, the graphical user interface could be a chat interface, a dedicated area for displaying and inputting social information. For example, a chat window at the bottom of the screen could display chat history and input boxes for various channels.

[0068] Optionally, the graphical user interface supports multi-level display, with visual interactive controls appearing as upper-level floating windows on top of the base interface without obscuring important information. For example, controls appear semi-transparently on one side of the interface, allowing players to see the game content behind them without interfering with normal operation.

[0069] Through the above steps, the embodiments of the present application can provide differentiated services according to the player's intention. Continuing with the above example, after the system determines that the player has an appearance recommendation intention, a small floating window of an appearance recommendation control will pop up on the player's game interface. This floating window will display the 3D model and price of XX fashion, and provide quick buttons such as try on, purchase, and gift. If another player says in the chat, "Just received a pattern, got it", the system recognizes the appearance switching intention and detects that the player has more than 10 pieces of fashion, and will pop up an appearance switching control for him. This control will display all the fashion icons he owns. The player only needs to click gently on the floating window, and the fashion on his character model will be replaced in real time, and the whole process does not require leaving the chat channel.

[0070] In an optional implementation manner, based on the dialogue information for semantic analysis to determine whether the controlled virtual character has an appearance interaction intention, including: calling a pre-trained natural language recognition model to perform in-depth semantic understanding on the dialogue information; performing keyword matching on the dialogue information based on a preset keyword library, the preset keyword library includes: an appearance category word library, the appearance category word library contains vocabulary related to appearance equipment in the game and a behavior category word library, the behavior category word library contains behavior vocabulary related to appearance interaction operations; performing fusion calculation on the in-depth semantic understanding result and the keyword matching result to obtain a comprehensive confidence level of the appearance interaction intention; when the comprehensive confidence level reaches the preset confidence threshold, it is determined that the controlled virtual character has an appearance interaction intention.

[0071] Among them, the pre-trained natural language recognition model is a language understanding system constructed based on deep learning technology, which can perform intelligent analysis and semantic understanding on in-game player conversations.

[0072] Optionally, the pre-trained natural language recognition model adopts a BERT model based on the Transformer architecture or its variant, and is pre-trained through a large amount of game dialogue data, and has the ability to deeply understand game-specific vocabulary and expressions. For example, when a player says in the chat, "This set of fashion looks really good, and I also want a set", the model can recognize key information such as "fashion", "good-looking", and "want", and understand that the player has the intention to obtain appearance equipment.

[0073] Optionally, the pre-trained natural language recognition model uses transfer learning technology to fine-tune on the basis of a general language model using game domain-specific data, improving the recognition accuracy of in-game jargon and players' habitual expressions. For example, the model can understand that "brocade" refers to high-level fashion, "bubble" refers to the chat bubble frame, and "mount" refers to the vehicle appearance and other game-specific nouns.

[0074] Optionally, the pre-trained natural language recognition model adopts a multi-task learning architecture and simultaneously trains multiple sub-tasks such as intent recognition, entity extraction, and sentiment analysis to provide a more comprehensive semantic understanding ability. For example, for a conversation like "Change to a new avatar frame. The previous one was too ordinary", the model can not only recognize the dressing intention but also extract the specific appearance category of the avatar frame and the player's dissatisfaction with the current equipment.

[0075] Among them, the appearance vocabulary library is a professional dictionary containing various appearance equipment-related vocabulary in the game, which is used to accurately match the appearance elements involved in the player's conversation.

[0076] Optionally, the appearance vocabulary library includes the standard names and common names of various appearance equipment such as fashion, headgear, mounts, weapon appearances, avatar frames, chat bubbles, background cards, and flat shows. For example, the vocabulary library contains words such as "brocade", "fine clothes", and "ancient costumes" that refer to high-level fashion, and words such as "bubble" and "bubble frame" that refer to chat decorations.

[0077] Optionally, the appearance vocabulary library automatically extracts high-frequency appearance-related vocabulary from a large number of player conversations through data mining technology and is regularly updated in combination with game operation data to ensure the timeliness and integrity of the vocabulary library. For example, when a new appearance type is launched in the game or players create new common names, the system can automatically identify and add them to the vocabulary library.

[0078] Optionally, the appearance vocabulary library adopts a hierarchical structure design and is classified and managed according to dimensions such as appearance type, quality level, and acquisition method, supporting fuzzy matching and synonym recognition. For example, when a player mentions "skin", the system can understand that it may refer to various appearance types such as fashion, mount appearance, or weapon skin.

[0079] Among them, the behavior vocabulary library is a professional dictionary containing behavior vocabulary related to appearance interaction operations, which is used to identify the player's specific operation intentions.

[0080] Optionally, the behavior vocabulary library includes purchase-related vocabulary such as "buy", "purchase", "get", and "spend money", switching-related vocabulary such as "change", "modify", "try", and "match", and appreciation-related vocabulary such as "look at", "take a look", and "display".

[0081] Optionally, the behavior vocabulary library is personalized and extended according to player behavior data and language habits, including informal expressions such as Internet buzzwords and game jargon. For example, the vocabulary library includes "planting grass" to indicate the desire to purchase, "pulling the grass" to indicate completing a purchase, and "charging money" to indicate paying to purchase and other commonly used expressions by players.

[0082] Optionally, the behavioral lexicon employs semantic similarity clustering technology to categorize and manage behavioral terms with similar meanings, thereby improving the accuracy and coverage of recognition. For example, terms such as "buy," "purchase," "acquire," and "take down" are categorized as purchasing behaviors, while terms such as "change," "replace," "substitute," and "change" are categorized as switching behaviors.

[0083] Among them, the semantic deep understanding result is the structured information output by the natural language recognition model after performing deep semantic analysis on the dialogue content.

[0084] Optionally, the semantic deep understanding results include multi-dimensional data such as intent classification probability, entity recognition results, sentiment analysis, and contextual information. For example, for the dialogue "This new brocade fashion is so beautiful, how much does it cost?", the understanding results include a 90% probability of purchase intent, entity annotation of "brocade fashion", positive sentiment, and intent to inquire about price information.

[0085] Optionally, the semantic deep understanding results are labeled with the importance weights of keywords through an attention mechanism, providing accurate semantic contribution information for subsequent fusion calculations. For example, in the sentence "I want to buy a nice profile picture frame," the weight of "buy" is 0.8, the weight of "nice" is 0.6, and the weight of "profile picture frame" is 0.9.

[0086] Optionally, the semantic deep understanding results include a confidence assessment, reflecting the model's reliability of the current understanding results, for risk control in subsequent decisions. For example, for the clearly expressed "I want to buy this outfit," the confidence level can reach 95%; for the vaguely expressed "This is nice," the confidence level may only be 60%.

[0087] The keyword matching result is the matching information obtained by accurately matching the dialogue content based on a preset thesaurus.

[0088] Optionally, the keyword matching results include a list of matched keywords, the matching position of each keyword, keyword type tags, and a match strength score. For example, for the dialogue "Change the bubble", the matching results include the keywords "change" (behavioral category, match strength 0.9) and "bubble" (appearance category, match strength 1.0).

[0089] Optionally, keyword matching results support fuzzy matching and synonym expansion, and can recognize various expressions such as word variations, abbreviations, and homophones to improve the flexibility and accuracy of matching.

[0090] Optionally, the keyword matching results are evaluated by using the TF-IDF algorithm to calculate the importance weight of the keywords, and combining the frequency of the words in the dialogue with their rarity in the vocabulary.

[0091] Among them, the overall confidence score is the final confidence score obtained by combining two analysis methods: semantic understanding and keyword matching.

[0092] Optionally, the overall confidence score is calculated using a weighted average algorithm, where the confidence score for semantic understanding has a weight of 0.7 and the confidence score for keyword matching has a weight of 0.3, with the final score ranging from 0 to 1. For example, when the confidence score for semantic understanding is 0.8 and the confidence score for keyword matching is 0.9, the overall confidence score is 0.8 × 0.7 + 0.9 × 0.3 = 0.83.

[0093] Optionally, the overall confidence score employs a dynamic weighting adjustment mechanism, adaptively adjusting the weight ratio of the two methods based on factors such as dialogue complexity, contextual completeness, and user historical behavior. For example, for short and clear dialogues, the weight of keyword matching is increased; for complex dialogues with long sentences, the weight of semantic understanding is increased.

[0094] Optionally, combining confidence scores with Bayesian inference methods can incorporate prior knowledge (such as user historical preferences and the current game scenario) as additional factors in the calculation, thereby improving prediction accuracy. For example, if a player has frequently purchased fashion items in the past, the system will be more inclined to identify fashion-related interaction intents in cases of ambiguous expression.

[0095] Through the above steps, this application embodiment can intelligently identify the player's appearance interaction needs in social dialogues, and then trigger the corresponding appearance recommendation function to seamlessly display relevant appearance options in the chat interface, realizing an intelligent transition from dialogue to appearance display.

[0096] In an optional implementation, the semantic deep understanding result and the keyword matching result are fused and calculated, including: outputting the semantic understanding confidence score through a natural language recognition model; calculating the keyword matching confidence score based on the number and weight of matched keywords; and using a weighted fusion algorithm to fuse the semantic understanding confidence score and the keyword matching confidence score to obtain a comprehensive confidence score.

[0097] Among them, semantic understanding confidence is a quantitative assessment of the reliability of the natural language recognition model's understanding results.

[0098] Optionally, the semantic understanding confidence score is calculated by taking the maximum value of the probability distribution output by the softmax layer inside the model, reflecting the model's confidence in the current prediction result. For example, when the model predicts that a sentence has a probability of 0.85 for appearance purchase intent, a probability of 0.10 for appearance switching intent, and a probability of 0.05 for other intents, the semantic understanding confidence score is 0.85.

[0099] Optionally, semantic understanding confidence is combined with the model's multi-layer feature representations, and the stability of understanding is evaluated by calculating the consistency of attention weights across layers. For example, if different layers of the model assign high attention weights to the word "fashion," it indicates that the model has a consistent understanding of the importance of the word, and the confidence score is correspondingly improved.

[0100] Optionally, semantic understanding confidence is assessed using an ensemble learning approach, evaluating the consistency of predictions from multiple models. Higher confidence is achieved when multiple models produce similar results. For example, if three different language models all identify a sentence as an appearance purchase intent with probabilities of 0.82, 0.87, and 0.85 respectively, the overall confidence is the weighted average of these results, which is 0.85.

[0101] The keyword matching confidence score is a confidence score calculated based on the number, type, and weight of the matched keywords.

[0102] Optionally, keyword match confidence is calculated using the formula: "Match Confidence = (Number of Appearance Keywords × Appearance Weight + Number of Behavioral Keywords × Behavioral Weight) / Total Keyword Threshold", where the appearance weight is 0.6 and the behavioral weight is 0.4. For example, when two appearance keywords and one behavioral keyword are matched, the confidence is (2 × 0.6 + 1 × 0.4) / 3 = 0.53.

[0103] Optionally, keyword matching confidence score considers the importance weight of keywords. High-value keywords such as "brocade" and "legend" have a higher contribution score, while general terms such as "good-looking" and "not bad" have a lower contribution score. For example, matching "brocade fashion" (weight 0.9) will get a higher confidence score than matching "good-looking clothes" (weight 0.5).

[0104] Optionally, keyword matching confidence uses a positional decay mechanism, where closely related keyword combinations receive additional confidence bonuses. For example, in the phrase "buy new fashion," the words "buy" and "fashion" are close together, resulting in a 1.2x weighting bonus; while in the phrase "fashion...buy," keyword combinations that are far apart receive a smaller weighting bonus.

[0105] The weighted fusion algorithm is a calculation method that mathematically combines multiple confidence scores.

[0106] Optionally, the weighted fusion algorithm employs a linear weighted average method, summing the confidence scores from different sources using preset weight parameters. For example, the formula "Overall Confidence = α × Semantic Understanding Confidence + β × Keyword Matching Confidence" can be used, where α = 0.7 and β = 0.3, ensuring that semantic understanding dominates the fusion process.

[0107] Optionally, the weighted fusion algorithm employs a dynamic weight adjustment strategy, adaptively adjusting the weight ratios based on the features of the input text. For example, for long, complex dialogues, the weight of semantic understanding is increased to 0.8; for short, simple dialogues, the weight of keyword matching is increased to 0.5, to adapt to the characteristics of different text types.

[0108] Optionally, the weighted fusion algorithm incorporates variance information from the confidence scores. When the confidence scores of the two methods differ significantly, the overall confidence score is reduced to reflect uncertainty. For example, when the semantic understanding confidence score is 0.9 and the keyword matching confidence score is 0.3, the system detects high variance and adjusts the overall confidence score to 0.55 instead of the directly calculated 0.72, avoiding overconfident judgments.

[0109] Through the above steps, this embodiment of the application achieves multi-dimensional and high-precision recognition of appearance interaction intent. The comprehensive score accurately reflects the player's true intent, providing a reliable decision-making basis for subsequent appearance recommendations and ensuring a high degree of match between recommended content and player needs.

[0110] In an optional implementation, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to determining that the controlled virtual character has an appearance interaction intent, the appearance interaction control is displayed on the graphical user interface, including: when the appearance interaction intent of the controlled virtual character is determined to be an appearance recommendation intent based on semantic analysis results, displaying the appearance recommendation control on the graphical user interface; or when the appearance interaction intent of the controlled virtual character is determined to be an appearance switching intent, displaying the appearance switching control on the graphical user interface.

[0111] Among them, the appearance recommendation is intended to meet the interactive needs of players who want to acquire new appearance equipment or learn about appearance items.

[0112] Optionally, the intention to recommend appearances can be determined by analyzing behavioral words such as "buy," "acquire," and "learn" in the dialogue, including expressions like "buy," "purchase," "want," and "how to obtain." For example, when a player says, "How do I get this new mount appearance?" the system recognizes the behavioral word "acquire" and determines it as an intention to recommend appearances.

[0113] Optionally, the intention to recommend appearances is determined by considering the player's current appearance ownership status. When semantic analysis indicates that a player is interested in a certain type of appearance, but the player does not yet own that appearance, the system tends to classify it as a recommendation intention. For example, when a new player asks "What is the purpose of an avatar frame?", the system detects that the player does not yet own any avatar frames, and classifies it as a recommendation intention.

[0114] Optionally, the determination of appearance recommendation intent can be aided by sentiment analysis. When positive emotions such as envy or liking towards others' appearances are detected, the weight of the recommendation intent is increased. For example, the emotional words "wow" and "so beautiful" in "Wow, your outfit is so beautiful, I want one too" strengthen the determination of recommendation intent.

[0115] Among them, the appearance switching is intended to meet the interactive needs of players who want to change the appearance equipment they currently own or to customize their appearance.

[0116] Optionally, the intention to switch appearances can be determined by recognizing behavioral words such as "changing," "matching," and "replacing," like "matching," "trying," and "replacing." For example, when a player says, "Try changing my avatar frame," the system recognizes the switching behavioral words such as "change" and "try," and determines that it is an intention to switch appearances.

[0117] Optionally, the intent to switch appearances needs to be verified in conjunction with the player's existing appearance inventory. Only when a player possesses a certain number of appearance items of the relevant type can it be considered an intent to switch. For example, if the system detects that a player has three different weapon appearances, and the player expresses an intention to change weapon skins, it is confirmed as an intent to switch.

[0118] Among them, the appearance recommendation control 201 is a user interface element that displays appearance product recommendation information to players and provides related operations.

[0119] Optionally, the appearance recommendation control 201 adopts a card layout design. Each card displays a preview image of a recommended appearance. By clicking to view details, users can obtain core information such as the appearance's name, quality, price, and acquisition method.

[0120] Optionally, the appearance recommendation control 201 provides quick operation buttons, including functions such as preview, favorite, share, and purchase, allowing players to complete appearance-related operations without leaving the current chat interface. For example, the "Buy Now" button in the control supports one-click ordering, and the "Preview / Try On" button allows users to view the appearance effect on the current interface.

[0121] Optionally, the appearance recommendation control 201 supports a personalized recommendation algorithm, which intelligently filters recommended content based on factors such as the player's historical preferences, friends' ownership status, and current trends. For example, for players who prefer ancient Chinese style appearances, the control prioritizes recommending Hanfu and ancient costumes; for players who are actively social, it recommends appearances that are popular among their friends.

[0122] Among them, the appearance switching control 202 is a user interface element that displays the appearance equipment that the player already owns and provides the switching operation.

[0123] Optionally, the appearance switching control 202 uses a grid layout to display the appearance equipment that the player already owns, and supports sorting by dimensions such as quality, acquisition time, and usage frequency. For example, the control displays 12 owned avatar frames in a 3×4 grid, arranged from high to low rarity, allowing players to intuitively compare the visual effects of different appearances.

[0124] Optionally, the appearance switching control 202 provides a real-time preview function. Players can click on any appearance equipment to display the wearing effect on the current character model in real time, and view the combination result without confirmation. For example, when a player clicks on different weapon appearances, the weapon in the character's hand is immediately switched, helping the player to make a quick decision.

[0125] Optionally, the appearance switching control 202 supports intelligent matching recommendations. When a player selects a certain appearance item, the system automatically recommends other types of appearances that best match it. For example, when a player selects an ancient style outfit, the control automatically recommends matching classical headwear, traditional weapon appearances, etc., forming a complete theme matching scheme.

[0126] Through the above steps, this application embodiment can provide a precise appearance interaction interface according to different player intentions. Players can directly browse appearance details, view the try-on effect, complete the purchase operation, or share their favorite appearance with friends in the chat, realizing a seamless connection from interest generation to purchase decision, greatly improving the conversion efficiency of appearance commercialization.

[0127] In an optional implementation, before determining that the appearance interaction intent of the controlled virtual character is an appearance switching intent based on the semantic analysis result, the method further includes: determining that the number of appearances possessed by the controlled virtual character reaches a preset number threshold and that the dialogue information is issued by the controlled virtual character.

[0128] The preset quantity threshold is the minimum number of appearances a player must possess to determine whether they are eligible to switch appearances.

[0129] Optionally, preset quantity thresholds can be set separately for different appearance types: 2 for fashion appearances, 3 for headwear appearances, and 2 for weapon appearances, ensuring players have sufficient choices for switching. For example, the fashion switching function can only be triggered when a player owns 3 outfits, avoiding meaningless switching operations when only 1 outfit is owned.

[0130] Optionally, the preset quantity threshold adopts a dynamic adjustment mechanism, which can be personalized based on factors such as the player's game level, activity level, and payment status. For example, the threshold for high-level players may be 3 items, while the threshold for novice players may be reduced to 2 items, in order to accommodate the appearance ownership of different player groups.

[0131] Optionally, a preset quantity threshold is used to calculate the weight based on appearance quality. Rare quality appearances have a higher weight value, and one legendary quality appearance may be equivalent to two common quality appearances. For example, if a player owns one legendary costume and one common costume, and the system calculates that the total weight reaches the threshold, the switching function can be triggered.

[0132] The number of appearance items is the total number of various appearance items currently owned by the player, as counted by the system.

[0133] Optionally, the number of appearance items owned is calculated in real time by querying the player's item inventory and equipped appearance database, including all appearance items that are worn and not worn. For example, the system query may find that the player has 5 sets of costumes (2 of which are worn and 3 are in the inventory) and 4 avatar frames (1 in use and 3 as backups).

[0134] Optionally, the appearance ownership statistics support separate calculations for temporary and permanent appearances. Only permanently owned appearances are included in the switching criteria to avoid misjudgments caused by appearances during trial periods. For example, limited-time appearances obtained by players through events for a 7-day trial are not counted in the ownership count; only permanent appearances obtained through purchase or as a reward for completing tasks are included in the statistics.

[0135] Optionally, the number of appearances can be quickly retrieved using a caching mechanism. The cached data is updated in real-time whenever a player gains or loses an appearance, avoiding frequent database queries. For example, when a player purchases a new appearance, the system immediately updates the appearance quantity cache from 3 to 4, ensuring that subsequent checks use the latest data.

[0136] In this context, the sender of the dialogue message is a controlled virtual character who actively sends messages in a social setting.

[0137] Optionally, the system identifies the sender of the conversation through message header information, ensuring that the appearance switching control is only displayed to the player who actively expresses the intention to switch, thus avoiding interference with other players who passively receive the information. For example, when player A says, "I want to change to a new avatar frame," only player A will see the appearance switching control, and other chat participants will not be affected.

[0138] Optionally, the authentication of the message sender includes anti-cheating mechanisms to prevent malicious players from triggering others' appearance features by spoofing the message source. For example, the system verifies the message's digital signature and session token to ensure that the message truly comes from the claimed player account.

[0139] Optionally, the system records the historical behavior patterns of those sending dialogue messages. When abnormally high-frequency triggering behavior is detected, a cooling-off mechanism is activated to prevent the abuse of system resources. For example, if a player sends multiple appearance-related messages in a short period of time, the system may temporarily reduce their trigger priority or increase the trigger interval.

[0140] Through the above steps, this application embodiment ensures the accurate triggering and reasonable use of the appearance switching function. In specific game scenarios, players can quickly browse the matching effects of different outfits and choose the most suitable look for the current activity without having to exit the chat interface to access the complex appearance system. This achieves a perfect integration of social scenarios and appearance management, greatly improving user convenience and gaming experience.

[0141] In an optional implementation, the method further includes: an appearance recommendation control determining recommended appearances in a first priority order, wherein the first priority order is positively correlated with the number of recommended appearances that satisfy a first preset condition, and the first preset condition includes at least one of the following: target appearances and similar appearances determined by semantic analysis results; appearances not owned by the controlled virtual character; appearances owned by other virtual characters in the largest number; appearances that can be directly obtained; and the latest released appearances.

[0142] Among them, the appearance recommendation control 201 is a user interface element that displays appearance product recommendation information to players and provides related operations.

[0143] Optionally, the appearance recommendation control 201 is a standalone floating window module in the game interface that can analyze player needs based on intelligent algorithms and display relevant appearance items. For example, when a player mentions in chat that they "want a nice speech bubble," the appearance recommendation control 201 will automatically pop up and display speech bubble appearance options that meet the criteria.

[0144] Optionally, the appearance recommendation control 201 features dynamic content updates, enabling it to adjust recommended content in real time to match the player's current appearance needs and preferences. For example, the control dynamically adjusts the recommended appearance types and priority order based on factors such as the player's level, activity level, and social circle preferences.

[0145] Optionally, the appearance recommendation control 201 offers a variety of interactive functions, including one-stop services such as browsing details, adding to favorites, sharing, and purchasing. For example, players can directly complete the entire process of previewing and trying on appearances, checking prices, and making purchases within the recommendation control without having to navigate to other interfaces.

[0146] The first priority order is the order of priority of recommended appearances in the list.

[0147] Optionally, the first priority order is an intelligent sorting algorithm based on the number of condition matches, where appearances that meet more conditions are ranked higher.

[0148] Optionally, the first priority order is calculated using a weighted average method, with different conditions having different importance weight values. For example, semantic matching might account for 40% of the weight, friend ownership might account for 30%, posting time might account for 20%, and availability might account for 10%.

[0149] Optionally, the first priority order can be personalized based on the player's historical behavior and preferences. For example, if a player frequently purchases fashion items, then fashion-related recommendations will receive a higher priority weight.

[0150] The recommended appearance items are those selected by the system based on an algorithm that are suitable for the current player.

[0151] Optionally, recommended appearances are a collection of personalized items selected from the game's appearance shop using a multi-dimensional filtering algorithm. For example, the system will select 2-6 appearances from thousands of options that best suit the player's current needs and spending power for display.

[0152] Optionally, the recommended appearances are dynamically updated, adjusting the recommendation strategy based on real-time player feedback and behavior. For example, if a player repeatedly skips a certain type of appearance recommendation, the system will reduce the frequency of recommending that type of appearance.

[0153] Optionally, recommended appearances include complete product information and acquisition paths, providing players with comprehensive decision support. For example, each recommended appearance will display detailed attribute descriptions, price information, acquisition methods, and usage effect previews.

[0154] The first preset condition is a set of judgment criteria used to filter and sort recommended appearances.

[0155] Optionally, the first preset condition provides a comprehensive matching score for each recommended appearance by combining different evaluation dimensions, ensuring that the recommendation results not only meet the user's personalized needs, but also take into account the popularity and accessibility of the appearance.

[0156] Optionally, the first preset condition adopts a configurable weight system, and the importance of each condition can be adjusted for different application scenarios to achieve a personalized recommendation strategy.

[0157] Optionally, the first preset condition supports dynamic learning and optimization. The system automatically adjusts the judgment threshold and weight allocation of each preset condition by analyzing the user's selection behavior and satisfaction feedback.

[0158] Among them, the target appearance and similar appearance determined by the semantic analysis results are appearance types that are matched by identifying user intent through natural language processing technology.

[0159] Optionally, the appearance types involved in the semantic analysis results are parsed using natural language processing technology to understand the user's desired switching needs and identify the user's desired appearance style, type, or specific attributes. For example, if a user expresses "I want a cooler look," the system identifies keywords such as "cool" and "fashionable," and marks cool appearance types such as "punk warrior suit" and "mechanical modification suit" as semantic matches.

[0160] Optionally, the appearance types involved in the semantic analysis results support fuzzy matching and semantic expansion, enabling the system to understand the user's implicit intent and synonymous expressions, thus expanding the range of matching appearances. For example, if a user says "I want more imposing equipment," the system will not only match appearances tagged with "imposing," but also appearances with related semantics such as "domineering," "majestic," and "majestic."

[0161] Optionally, the appearance types involved in the semantic analysis results can be combined with the user's role, class, and level information to provide more accurate type matching, ensuring that the recommended appearances not only meet semantic requirements but also suit the character's characteristics. For example, when a mage user expresses a desire for a more powerful appearance, the system will prioritize matching advanced mage appearances rather than powerful warrior appearances.

[0162] Among them, appearances that the controlled virtual character does not possess are appearance types that do not exist in the current character appearance library.

[0163] Optionally, for controlled virtual characters that they do not yet own, the system identifies appearances that the character has not yet acquired by comparing the character's current appearance list with the full appearance library, and prioritizes recommending these appearances to enrich the character's appearance collection. For example, if a character already owns a "Light Armor Set" or "Light Armor Set" but lacks a "Heavy Armor Set", the system will prioritize recommending various heavy armor appearances.

[0164] Optionally, for appearances that the controlled virtual character does not possess, the system prioritizes items that can form a complete set or create a good synergy with existing appearances, considering the completeness of the appearance series and compatibility. For example, if the character possesses "Knight Helmet" and "Knight Leggings," the system will prioritize recommending "Knight Breastplate" to complete the full knight outfit.

[0165] Optionally, appearances that the controlled virtual character does not yet own are recommended through tiered rarity recommendations, ensuring that users can gradually progress from common appearances to rare ones, enhancing the sense of accomplishment in collecting them. For example, for characters lacking legendary appearances, the system prioritizes recommending epic appearances, which are of moderate difficulty to obtain, as a transitional measure.

[0166] Among them, the appearance type with the most appearances for other virtual characters is the most popular appearance type among all users.

[0167] Alternatively, the most popular appearances for other virtual characters are determined through server-wide data statistics, reflecting their popularity among players. Prioritizing these appearances allows users to keep up with trends. For example, statistics show that the "Dragon Scale Armor" is owned by 85% of high-level players, so the system marks it as a popular appearance and prioritizes its recommendation.

[0168] Optionally, considering the different preferences of various player groups, the most popular appearances for other virtual characters are analyzed separately for novice, intermediate, and advanced players, providing targeted recommendations for users of different skill levels. For example, the "Adventurer Set" has the highest ownership rate among novice players, while the "Mythical Set" has the highest ownership rate among advanced players.

[0169] Optionally, other virtual characters with the most appearances can be analyzed over time to identify short-term popular and long-term classic appearances, providing users with recommendations that are both timely and have lasting value. For example, the ownership rate of the Lunar New Year limited set surges during the holiday, while the classic warrior set maintains a high ownership rate over a long period.

[0170] Among them, the appearance that can be obtained directly is the appearance type that users can obtain immediately through simple operations.

[0171] Optional, directly obtainable appearances include those sold directly in the store, those claimed for free, and those obtained as rewards for completing tasks. These appearances do not require a complicated acquisition process, allowing users to quickly own and experience them. For example, some outfits in the store can be purchased directly with in-game currency, and can be claimed directly after completing the tutorial.

[0172] Optional, directly obtainable appearances are filtered based on the user's current resource status and permission level to ensure that the recommended appearances are those the user can actually acquire. For example, for users with insufficient gold coins, the system prioritizes recommending free or low-priced appearances; for VIP users, the system recommends VIP-exclusive appearances.

[0173] Optionally, directly obtainable appearances are ranked by ease of access, with the easiest-to-obtain appearances appearing first to improve user acquisition success rate and satisfaction. For example, appearances that can be purchased with a single click receive a higher convenience score than those requiring multiple confirmation steps.

[0174] The newly released appearances are new cosmetic content recently added by the game operator.

[0175] Optionally, the latest releases are identified by timestamps and version control, and the system prioritizes recommending the most recently released releases so that users can experience the latest content as soon as possible.

[0176] Optionally, newly released outfits can be filtered and recommended based on user interests and preferences, avoiding recommendations of new outfits that don't match the user's taste and improving the accuracy of recommendations. For example, users who prefer a classic style will not receive a "futuristic sci-fi outfit" recommendation, even if it is the latest release.

[0177] Optionally, newly released appearances are managed using a freshness decay algorithm, which gradually reduces the recommendation weight of new appearances over time to maintain the balance of the recommendation system. For example, appearances released more than 30 days ago no longer receive the "newest release" weight bonus.

[0178] Through the above steps, this application embodiment can establish a multi-dimensional, intelligent appearance recommendation priority system. The system scores appearances based on the number of times they meet a first preset condition: appearances that meet four conditions receive the highest priority, appearances that meet three conditions receive the next highest priority, and so on, forming the final recommendation ranking displayed to the player.

[0179] In an optional implementation, the method further includes: an appearance switching control displays appearances to be switched in a second priority order, the second priority order being positively correlated with the number of appearances to be switched that satisfy a second preset condition, the second preset condition including at least one of the following: the target appearance and similar appearances determined by semantic analysis results; the appearance with the most appearances owned by the controlled virtual character; the type to which the high-value appearances belong; and the appearance type that the controlled virtual character has not changed for the longest time.

[0180] Among them, the appearance switching control 202 is an interactive interface component specifically used to manage and switch the appearances that players already own.

[0181] Optionally, the appearance switching control 202 has intelligent display logic, which can prioritize displaying the most relevant appearance options based on the current context and player habits. For example, when a player says "change avatar frame", the control will automatically locate the avatar frame category and prioritize displaying the player's most frequently used or most recently acquired avatar frames.

[0182] Optionally, the appearance switching control 202 provides real-time preview and comparison functions to help players intuitively choose the appropriate appearance combination. For example, when a player selects an appearance, the character model will update and display the wearing effect in real time, and can also compare it with the appearance of the currently equipped gear.

[0183] The second priority order is the arrangement priority rule of the appearance to be switched in the control.

[0184] Optionally, the second priority order is based on a comprehensive scoring algorithm that considers condition matching degree and usage frequency, ensuring that the appearance that best meets the current needs is displayed first. For example, the system will comprehensively consider factors such as semantic relevance, number of appearances owned, value level, and usage time to calculate the priority score for each appearance.

[0185] Optionally, the second priority order employs a dynamic weighting adjustment mechanism, adjusting the sorting strategy based on players' immediate needs and historical preferences. For example, if players frequently change their appearance before battle, battle-related appearances will receive a higher priority weight.

[0186] Optionally, the second priority order will be optimized in conjunction with social factors, taking into account the usage and popularity trends of appearances among friends. For example, if a player's friends have a high usage rate for a certain type of appearance, that type of appearance will be ranked higher in the toggle control.

[0187] The appearance to be switched is a collection of appearance items that the player already owns but is not currently equipped.

[0188] Optionally, the appearances to be switched include all appearance items that players obtain through various means such as purchases, event rewards, and quest completion. For example, a player's appearance inventory may contain 20 outfits, 15 avatar frames, 8 speech bubbles, 5 mounts, and other different types of appearances.

[0189] Optionally, the appearance to be switched will be intelligently filtered based on the current game scene and character status, displaying only appearance options suitable for the current situation. For example, in a combat scene, the system may prioritize displaying appearances with combat bonuses, while in a social scene, it may prioritize displaying appearances with higher aesthetic appeal.

[0190] Optionally, the appearance to be switched supports multi-dimensional combination and matching functions, allowing players to preview the combined effects of multiple appearance components at the same time. For example, players can select headwear, clothing, and weapon appearances simultaneously for combination preview to ensure overall coordination and unity.

[0191] The second preset condition is a set of judgment criteria used to filter and sort the appearances to be switched.

[0192] Optionally, the second preset condition is a set of evaluation rules based on player behavior and appearance attributes to determine the priority of appearance in the toggle list.

[0193] Optionally, the second preset condition adopts a tiered screening mechanism, first performing basic applicability screening, and then prioritizing. For example, the basic screening might be "applicable to the current character's class" and "not currently equipped," while the priority ranking is based on factors such as usage frequency, acquisition difficulty, and appearance quality.

[0194] Among them, the most numerous appearances for controlled virtual characters are the appearance types that dominate the user collection.

[0195] Optionally, the system identifies the user's collection preferences and main interests by statistically analyzing the distribution of the number of appearances of each type in the user's appearance library. For example, if a user owns 8 mage appearances, 3 warrior appearances, and 2 assassin appearances, the system identifies the mage appearance as the user's main collection type.

[0196] Optionally, the number of appearances possessed by a controlled virtual character reflects the user's aesthetic preferences and character positioning. Prioritizing the recommendation of appearances of this type can help improve user engagement and satisfaction. For example, users who prefer collecting classic-style appearances should have their various classic outfits displayed first when switching recommendations.

[0197] Optionally, the most numerous appearances for controlled virtual characters are analyzed through detailed category segmentation. This analysis considers not only the distribution of major types (such as mages and warriors) but also the preference trends of subtypes (such as ice mages and fire mages). For example, if a user has 5 out of 8 mage appearances that are ice-themed, the system can further identify the user's specific preference for ice mage appearances.

[0198] Among them, high-value appearances belong to the categories of appearances that are rare, costly to acquire, or have special significance.

[0199] Alternatively, the type of high-value appearances is determined by a comprehensive evaluation of factors such as the appearance's rarity level, difficulty of acquisition, and market value. Prioritizing these high-value appearances allows users to fully showcase their collecting achievements.

[0200] Optionally, the type of high-value appearance item is assessed based on market popularity and player awareness. Appearance types with high reputation and recognition within the player community are marked as high-value. For example, the "Mythical Guild Robe" enjoys extremely high reputation among players and is listed as a high-value appearance type even though it is relatively easy to obtain.

[0201] Among them, the appearance type of the controlled virtual character that has not been changed for the longest time is the appearance category that the user has not used or activated for a long time.

[0202] Optionally, the system identifies the least-used appearance type for controlled virtual characters by recording the last time each appearance was used, thus recognizing forgotten or neglected appearances and reminding users to rediscover these idle resources. For example, if a user's "Elf Ranger Set" has not been used for 3 months, the system marks it as the least-used appearance and increases its recommendation weight for switching.

[0203] Optionally, the longest-unused appearance type for controlled virtual characters is considered in terms of seasonality and scene applicability, with additional recommendation weight given to long-unused appearances that are suitable for the current time or scene. For example, during the Spring Festival, if the system finds that a user has not used the "New Year Celebration Set" for a year, it will immediately list it as a high-priority switch recommendation.

[0204] Optionally, the most recently used appearance type for controlled virtual characters can be restored through a freshness recovery mechanism, allowing users to relive appearances they once loved but have now forgotten, increasing the diversity and fun of appearance usage. For example, the system can automatically recommend a set of appearances that a user hasn't used for more than 30 days each week, accompanied by the prompt "Rediscover your collection".

[0205] Through the above steps, this embodiment of the application constructs a dual-priority appearance recommendation system and adopts a multi-dimensional preset condition evaluation mechanism to achieve personalized and intelligent appearance recommendation functions. The diverse recommendation strategies meet the personalized needs of different types of players, effectively responding to everything from the basic needs of novice players to the collection and display needs of experienced players, significantly improving the overall richness of the game's content and player engagement.

[0206] In an optional implementation, the method further includes: in response to the controlled virtual character entering a non-social state, detecting whether the controlled virtual character meets a third preset condition, the third preset condition including at least one of the controlled virtual character level, number of consecutive active days, and the controlled virtual character being online and not in a game state; when the third preset condition is met, providing appearance interaction controls on a graphical user interface.

[0207] Among them, the controlled virtual character entering a non-social state is the state transition process of the player character leaving the social function interface such as chat, team, guild, etc.

[0208] Optionally, a non-social state refers to a player not currently engaging in any form of real-time social interaction and being in single-player mode or operating independently. For example, actions such as closing the chat window, leaving the team interface, or leaving the guild lobby will trigger the determination of a non-social state.

[0209] Optionally, the detection of a non-social state is achieved by monitoring the player's interface status and operational behavior. The system will track the currently opened functional modules in real time. For example, when the system detects that the player has closed all social-related interfaces and has not performed any social actions for more than 5 seconds, it determines that the player has entered a non-social state.

[0210] Optionally, non-social states are prioritized, and different non-social scenarios have different appearance recommendation triggering strategies. For example, a non-social state of strolling around a town is more suitable for triggering the appearance recommendation function than a non-social state of fighting in the wild.

[0211] Among them, detecting whether the controlled virtual character meets the third preset condition is the judgment process of the system to comprehensively evaluate the player's status.

[0212] Optionally, the detection process involves comprehensive judgment based on multi-dimensional information such as player character data, online behavior, and game progress in real time. For example, the system will query the player's character level database records, count consecutive login days, and monitor current online status and operational behavior.

[0213] Optionally, the detection mechanism employs a layered verification approach, first filtering based on basic conditions, and then verifying detailed conditions. For example, the system first quickly checks whether the player's level meets the requirements, then queries activity data, and finally verifies whether the current state is suitable for triggering a recommendation.

[0214] The third preset condition is a set of player status standards that must be met to trigger the appearance recommendation function.

[0215] Optionally, the third preset condition is a set of quantitative indicators used to identify the appropriate timing for recommendations, ensuring that the recommendation function is triggered at the right time.

[0216] Optionally, the third preset condition adopts a strict judgment mode of full condition satisfaction. The recommendation will only be triggered when the player meets all the set conditions at the same time. For example, even if the player's level is high but the number of consecutive active days is insufficient, or the activity level meets the standard but the player is currently in combat, the appearance recommendation function will not be triggered.

[0217] Optionally, the third preset condition will be dynamically adjusted based on the game's operational strategy and player feedback. For example, during an event promoting a new appearance, the level requirement may be temporarily lowered to level 60, or during periods of low player activity, the requirement for consecutive active days may be relaxed to 5 days.

[0218] Among them, the controlled virtual character level is an indicator of the player character's growth progress in the game.

[0219] Optionally, the controlled virtual character level is an important indicator of a player's gaming experience and level of investment. Higher levels usually mean a stronger willingness to spend and a greater desire for cosmetic items. For example, players above level 80 have typically passed the beginner stage, have a deeper understanding of the game, and are more willing to invest in cosmetic items.

[0220] Optionally, level requirements can be set based on player behavior data analysis to determine the most effective timing for triggering recommendations. For example, data might show that level 80 is a key point where players start paying attention to cosmetic items; at this point, players are already familiar with the basic gameplay and begin to pursue personalized expression.

[0221] Optionally, the level requirement will be adjusted to suit the player's class characteristics and game progress. For example, the level requirement may be higher for classes that level up quickly, and lower for newly opened servers to fit the overall progress.

[0222] Among them, the number of consecutive active days is a time indicator for measuring players' engagement and stability in the game.

[0223] Optionally, the number of consecutive active days is calculated by tracking players' daily logins and effective playtime, reflecting their loyalty and engagement with the game. For example, seven consecutive days of activity suggests that players have developed stable gaming habits and are more likely to respond positively to cosmetic recommendations.

[0224] Optionally, the calculation of active days uses an intelligent recognition mechanism to distinguish between genuine activity and AFK behavior. For example, the system will monitor indicators such as the player's operation frequency, movement distance, and function usage, and only logins that reach the minimum activity threshold will be counted as consecutive active days.

[0225] Optionally, the number of consecutive active days can be correlated with a player's spending history and social activity to improve the accuracy of recommendations. For example, players who are consistently active and have a history of spending on cosmetic items will receive more personalized recommendation strategies.

[0226] Among them, "controlled virtual character is online but not in the game" is a description of the player's current game participation status.

[0227] Optionally, "online but not playing" refers to a state where a player is connected to the game server but is not currently engaged in core gameplay. For example, a player might be wandering around a town, checking their inventory, or resting in a safe area—all of which are suitable states for receiving cosmetic recommendations.

[0228] Optionally, the determination of non-game states is based on a comprehensive analysis of the player's current action type and the surrounding scenario. For example, the system will detect whether the player is in combat, performing a quest, or engaging in trading, excluding these scenarios that are not suitable for recommendation.

[0229] Optionally, the online but non-game status includes an assessment of player attention and receptiveness to ensure that recommended information receives effective attention. For example, ideal times to push appearance recommendations are when a player is resting after completing a difficult dungeon or waiting for a friend in town.

[0230] Through the above steps, this disclosure realizes an intelligent non-social appearance recommendation function. The entire recommendation process will not interfere with the player's normal game rhythm, but will be triggered when the player pays attention to the appearance during their idle time.

[0231] In an optional implementation, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to the controlled virtual character entering a non-social state, it detects whether the controlled virtual character meets a third preset condition. It also includes: the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to the controlled virtual character entering a non-social state, when it is determined that the number of appearances owned by the controlled virtual character reaches a preset number threshold, it detects whether the state of the controlled virtual character meets the third preset condition.

[0232] Among them, determining whether a player has a need to switch appearances is based on the condition that the number of appearances possessed by the controlled virtual character reaches a preset threshold.

[0233] Optionally, the appearance quantity statistics include all appearance items that players have acquired through various means, counted separately by type. For example, the system will count the number of costumes, mounts, weapon appearances, etc., and when a player has two or more of a certain type, it is considered to be worth switching.

[0234] Optionally, the preset quantity threshold can be set with different standards for different appearance types, taking into account the differences in the difficulty and importance of obtaining various appearances. For example, the threshold for common appearances might be 2, while the threshold for rare appearances might be 1, which is enough to trigger a switching suggestion.

[0235] Optionally, the statistics will exclude expired, damaged, or unsuitable appearances from the current character to ensure the accuracy and usability of the results. For example, expired limited-time appearances will not be included in the statistics, and class-specific appearances will only be counted for the corresponding class character.

[0236] The preset quantity threshold is the minimum number of appearances required to trigger the appearance switching function.

[0237] Optionally, the preset quantity threshold is a reasonable value determined based on user experience research, ensuring that players have sufficient choice to experience the switching function. For example, research shows that the switching function only has practical meaning and value when players have at least two of the same appearance.

[0238] Optionally, the preset quantity threshold can be personalized based on the player's game progress and spending power. For example, for high-spending players, the threshold may be set higher to match their more extensive cosmetic collections; for average players, the threshold remains at a basic level.

[0239] Optionally, the preset quantity threshold has a dynamic adjustment mechanism that optimizes based on the overall distribution of in-game cosmetics and player feedback. For example, if there are many cosmetic events during a certain period and players generally own more cosmetics, the threshold may be increased accordingly.

[0240] In an optional implementation, the method further includes: synchronizing appearance interaction information to other virtual characters in the game scene.

[0241] Among them, the appearance interaction operation information is a data record of various operations performed by players in the appearance system.

[0242] Optionally, the appearance interaction information includes all appearance-related behavioral data of the player, such as browsing history, purchase behavior, switching operations, and collection actions. For example, when a player views the details of an appearance, adds an appearance to their favorites, completes an appearance purchase, or switches between appearances, the system will generate corresponding operation information records.

[0243] Optionally, the appearance interaction operation information adopts a structured data format, including key elements such as operation type, operation object, operation time, and operation result. For example, an operation information may be recorded as "Player A purchased the 'Fiery Armor' costume at 15:30:25, spending 500 game coins, and the purchase was successful."

[0244] Optionally, appearance interaction information will be filtered based on privacy settings and social needs, synchronizing only operations suitable for public sharing. For example, appearance display and switching operations will be synchronized, but sensitive information such as specific consumption amounts will be filtered out.

[0245] Synchronization refers to the data transmission process of sending operation information to the relevant player's client in real time.

[0246] Optionally, the synchronization mechanism is implemented through the game server's message push system to ensure that information is delivered to the target recipient in a timely and accurate manner. For example, when a player completes a cosmetic action, the server will immediately encapsulate the action information into a message packet and push it to all relevant online players.

[0247] Optionally, the synchronization process employs intelligent filtering and prioritization to prevent information overload from negatively impacting the player experience. For example, the system may determine whether to push notifications and the priority of those notifications based on factors such as the closeness of the relationship between the player and the operator, and the importance of the information.

[0248] Optionally, the synchronization mechanism includes an offline caching function. For players who are temporarily offline, important operation information will be cached and resent when they log back in. For example, important appearance updates from friends will be cached for 24 hours to ensure that important social interactions are not lost due to time differences.

[0249] Among them, the other virtual characters in the game scene are a group of related characters who have social connections with the player who controls the game.

[0250] Optionally, other virtual characters include all characters who are in the same game scene, social circle, or have an interactive relationship with the player. For example, this includes players in the same chat room, members of a fixed team, guild members, and online friends in the friend list.

[0251] Optionally, the determination of relevant roles is based on multi-dimensional social relationship analysis, including factors such as geographical location, social networks, and interaction frequency. For example, the system will prioritize pushing operation information to players who are nearby, friends with whom they have recently interacted, and active members of their respective guilds.

[0252] Optionally, the scope of other virtual characters will be dynamically adjusted based on the type and importance of the action to ensure the effectiveness and relevance of information dissemination. For example, a regular appearance change might only notify the closest friends, while the acquisition of a rare appearance might be announced to a wider social circle.

[0253] Through the above steps, the synchronization mechanism of operation information in this embodiment of the application realizes the social effect of "one person operates, multiple points spread", which not only enhances the interactivity between players, but also creates more business conversion opportunities for the appearance mall, forming a virtuous cycle of social communication.

[0254] Figure 4 This is a schematic diagram of an interface for the appearance interaction method provided in an embodiment of this application.

[0255] In an optional implementation, synchronizing appearance operation interaction information to other virtual characters in the game scene includes: synchronizing appearance operation interaction information to other virtual characters in the game scene, including: converting appearance interaction operation results into broadcast information; pushing broadcast information to other virtual characters in the game scene so that other virtual characters can trigger their own appearance interaction controls based on the broadcast information.

[0256] Among them, the appearance interaction operation result is the final state or effect produced after the player completes the appearance-related operation.

[0257] Optionally, the appearance interaction operation result is a complete description of the final result of the player's appearance operation behavior, including information such as the success status of the operation, the change effect, and the scope of influence.

[0258] Optionally, the appearance interaction results include before-and-after comparison information to help other players understand the specific changes. For example, the result information might show "Player A switched their mount from a white horse to a fire dragon, increasing their movement speed from 120% to 150%".

[0259] Optionally, the results of appearance interaction operations will be categorized and marked according to the importance and impact of the operation to facilitate subsequent information processing and dissemination. For example, switching to a normal appearance is marked as a general level, obtaining a rare appearance is marked as an important level, and obtaining an epic appearance for the first time is marked as a special level.

[0260] The process of converting the data into broadcast information involves processing the operation results into a message format suitable for social dissemination.

[0261] Optionally, the conversion process uses natural language generation technology to transform structured operational data into easily understandable textual descriptions.

[0262] Optionally, the conversion process may adjust the level of detail and expression of the broadcast message based on the recipient's preferences and the closeness of the relationship. For example, the broadcast to a friend may be more detailed and friendly, while the broadcast to a stranger may be more concise and formal.

[0263] Optionally, the conversion process includes the integration of multimedia elements, generating not only text descriptions but also adding relevant icons, animations, or sound effects. For example, announcements featuring rare appearances might be accompanied by flashing icons and unique sound effects.

[0264] Among them, broadcast message 204 is a formatted message that can be received and understood by other players.

[0265] Optionally, the broadcast message 204 adopts a unified message format standard, including standard fields such as sender information, operation description, timestamp, and related links.

[0266] Optionally, the broadcast message 204 features an interactive design, allowing recipients to obtain more information or perform related functions through clicks, long presses, or other actions. For example, players can click on the broadcast message to directly view appearance details or open their own appearance recommendation control with a single click.

[0267] Optionally, the broadcast information supports personalized display, adjusting the display style based on factors such as the recipient's interface settings and language preferences. For example, in concise mode, only core information is displayed, while in detailed mode, a complete operation description and related recommendations are shown.

[0268] Among them, pushing broadcast information to other virtual characters in the game scene is the distribution process of delivering messages to the target recipient.

[0269] Optionally, the push notification process is executed through the game's real-time communication system, sending broadcast information to the corresponding recipient groups according to predetermined distribution rules. For example, the system may determine the push targets based on factors such as social relationship graphs, geographical location, and online status, and then distribute the information in batches through a message queue system.

[0270] Optionally, the push mechanism employs a tiered, incremental dissemination strategy to ensure information spreads effectively across the social network without causing information overload. For example, core friends receive full and detailed broadcasts, ordinary friends receive simplified versions, and unfamiliar players only receive important broadcasts under specific conditions.

[0271] Optionally, the push notification process includes intelligent deduplication and frequency control to avoid interfering with players by using the same or similar announcements. For example, if the same player changes their appearance multiple times in a short period of time, the system will merge related announcements or only push the most important changes.

[0272] Among them, the interactive process of triggering its own appearance control based on the broadcast information is the process of responding to the broadcast content and starting related functions for the receiver.

[0273] Optionally, the triggering mechanism is implemented through interactive elements in the broadcast information, allowing players to directly activate relevant functions through simple clicks or gestures. For example, clicking the "→" in the broadcast information will directly open the appearance recommendation control and locate the relevant product; clicking "View Details" will display the complete information of the appearance.

[0274] Optionally, the triggering process can be adapted to the recipient's current state and permissions to ensure the availability and relevance of the function. For example, if the recipient already has the same appearance, the appearance switching control may be triggered; if the recipient does not have purchasing permissions, a viewing mode or guidance on how to obtain it will be provided.

[0275] Optionally, the appearance interaction controls triggered by the broadcast will retain the context information of the broadcast, providing more accurate recommendations and operation options.

[0276] Through the above steps, the broadcast-based triggering mechanism of this application not only enables the rapid dissemination of appearance information on social networks, but also provides each recipient with a personalized interactive experience, achieving a seamless connection from information reception to action conversion, and greatly improving the social dissemination effect and commercial conversion efficiency of the appearance system.

[0277] Figure 5 This is a structural schematic diagram of the appearance interaction device provided in the embodiments of this application.

[0278] This application provides a graphical user interface (GUI) device that provides a graphical user interface via a terminal device. The device 300 includes: The acquisition module 301 is used to acquire dialogue information in the social state in response to the controlled virtual character entering the social state, and the controlled virtual character is controlled through the terminal device; Analysis module 302 is used to perform semantic analysis based on dialogue information to determine whether the controlled virtual character has an appearance interaction intent; The display module 303 is used to display appearance interaction controls on the graphical user interface in response to the appearance interaction intent of the controlled virtual character. The appearance interaction controls are used to provide content corresponding to the appearance interaction intent.

[0279] This application provides an appearance interaction device that accurately captures the user's appearance interaction intent by acquiring and analyzing dialogue information in real time in social scenarios. It then provides a lightweight interactive control on the current interface to satisfy this intent, avoiding the interruption and operational costs caused by scene switching. This solution significantly enhances the user's interactive experience by achieving a seamless and coherent operation flow. It integrates behaviors such as browsing, purchasing, and changing appearances into social contexts, creating new interactive gameplay and display opportunities, enriching the game, and effectively solving the technical problems of fragmented user experience and cumbersome operation caused by the independence and lack of information sharing between game social systems and appearance systems in existing technologies.

[0280] The appearance interaction device provided in this application embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the embodiment of the display device can be referred to the corresponding content in the aforementioned display method embodiment.

[0281] Figure 6 This is a structural block diagram of the electronic device provided in the embodiments of this application.

[0282] This application also provides an electronic device, such as... Figure 6 The diagram shows the structure of the electronic device, which includes a processor 111 and a memory 110. The memory 110 stores computer-executable instructions that can be executed by the processor 111. The processor 111 executes the computer-executable instructions to implement the following steps of the appearance interaction method: providing a graphical user interface through a terminal device; in response to a controlled virtual character entering a social state, acquiring dialogue information in the social state; controlling the controlled virtual character through the terminal device; performing semantic analysis based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intention; and in response to the controlled virtual character having an appearance interaction intention, displaying appearance interaction controls on the graphical user interface, the appearance interaction controls being used to provide content corresponding to the appearance interaction intention.

[0283] Optionally, semantic analysis is performed based on dialogue information to determine whether the controlled virtual character has an appearance interaction intent. This includes: calling a pre-trained natural language recognition model to perform deep semantic understanding of the dialogue information; performing keyword matching on the dialogue information based on a preset keyword library, which includes: an appearance-related lexicon containing words related to appearance equipment in the game, and a behavior-related lexicon containing behavioral words related to appearance interaction operations; fusing the deep semantic understanding results with the keyword matching results to obtain a comprehensive confidence score of the appearance interaction intent; and determining that the controlled virtual character has an appearance interaction intent when the comprehensive confidence score reaches a preset confidence threshold.

[0284] Optionally, the semantic deep understanding results and keyword matching results are fused and calculated, including: outputting the semantic understanding confidence score through the natural language recognition model; calculating the keyword matching confidence score based on the number and weight of matched keywords; and using a weighted fusion algorithm to fuse the semantic understanding confidence score and the keyword matching confidence score to obtain a comprehensive confidence score.

[0285] Optionally, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to determining that the controlled virtual character has an appearance interaction intent, the appearance interaction control is displayed on the graphical user interface, including: when the appearance interaction intent of the controlled virtual character is determined to be an appearance recommendation intent based on the semantic analysis results, displaying the appearance recommendation control on the graphical user interface; or when the appearance interaction intent of the controlled virtual character is determined to be an appearance switching intent, displaying the appearance switching control on the graphical user interface.

[0286] Optionally, before determining that the appearance interaction intent of the controlled virtual character is an appearance switching intent based on the semantic analysis results, the method further includes: determining that the number of appearances possessed by the controlled virtual character reaches a preset number threshold and that the dialogue information is issued by the controlled virtual character.

[0287] Optionally, the method further includes: an appearance recommendation control determines recommended appearances in a first priority order, wherein the first priority order is positively correlated with the number of recommended appearances that satisfy a first preset condition, and the first preset condition includes at least one of the following: target appearances and similar appearances determined by semantic analysis results; appearances not owned by the controlled virtual character; appearances owned by other virtual characters in the largest number; appearances that can be directly obtained; and the latest released appearances.

[0288] Optionally, the method further includes: an appearance switching control displays appearances to be switched in a second priority order, wherein the second priority order is positively correlated with the number of appearances to be switched that meet a second preset condition, and the second preset condition includes at least one of the following: the target appearance and similar appearances determined by semantic analysis results; the appearance with the most appearances owned by the controlled virtual character; the type to which the high-value appearances belong; and the appearance type that the controlled virtual character has not changed for the longest time.

[0289] Optionally, the method further includes: in response to the controlled virtual character entering a non-social state, detecting whether the controlled virtual character meets a third preset condition, the third preset condition including at least one of the controlled virtual character's level, number of consecutive active days, and the controlled virtual character being online and not in a game state; when the third preset condition is met, providing appearance interaction controls on the graphical user interface.

[0290] Optionally, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to the controlled virtual character entering a non-social state, it detects whether the controlled virtual character meets a third preset condition. It also includes: in response to the controlled virtual character entering a non-social state, when it is determined that the number of appearances owned by the controlled virtual character has reached a preset number threshold, it detects whether the state of the controlled virtual character meets the third preset condition.

[0291] Optionally, the method also includes: synchronizing appearance interaction information to other virtual characters in the game scene.

[0292] Optionally, the appearance operation interaction information can be synchronized to other virtual characters in the game scene, including: converting the appearance interaction operation results into broadcast information; and pushing the broadcast information to other virtual characters in the game scene so that other virtual characters can trigger their own appearance interaction controls based on the broadcast information.

[0293] exist Figure 6 In the illustrated embodiment, the electronic device further includes a bus 112 and a communication interface 113, wherein the processor 111, the communication interface 113, and the memory 110 are connected via the bus 112.

[0294] The memory 110 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 113 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 112 may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 112 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0295] Processor 111 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 111 or by software instructions. The processor 111 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory. The processor 111 reads the information in the memory and, in conjunction with its hardware, completes the steps of the appearance interaction method of the aforementioned embodiment.

[0296] This application also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement a visual interaction method. The method includes: providing a graphical user interface through a terminal device; in response to a controlled virtual character entering a social state, acquiring dialogue information in the social state, wherein the controlled virtual character is controlled through the terminal device; performing semantic analysis based on the dialogue information to determine whether the controlled virtual character has a visual interaction intent; and in response to the controlled virtual character having a visual interaction intent, displaying visual interaction controls on the graphical user interface, wherein the visual interaction controls provide content corresponding to the visual interaction intent.

[0297] Optionally, semantic analysis is performed based on dialogue information to determine whether the controlled virtual character has an appearance interaction intent. This includes: calling a pre-trained natural language recognition model to perform deep semantic understanding of the dialogue information; performing keyword matching on the dialogue information based on a preset keyword library, which includes: an appearance-related lexicon containing words related to appearance equipment in the game, and a behavior-related lexicon containing behavioral words related to appearance interaction operations; fusing the deep semantic understanding results with the keyword matching results to obtain a comprehensive confidence score of the appearance interaction intent; and determining that the controlled virtual character has an appearance interaction intent when the comprehensive confidence score reaches a preset confidence threshold.

[0298] Optionally, the semantic deep understanding results and keyword matching results are fused and calculated, including: outputting the semantic understanding confidence score through the natural language recognition model; calculating the keyword matching confidence score based on the number and weight of matched keywords; and using a weighted fusion algorithm to fuse the semantic understanding confidence score and the keyword matching confidence score to obtain a comprehensive confidence score.

[0299] Optionally, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to determining that the controlled virtual character has an appearance interaction intent, the appearance interaction control is displayed on the graphical user interface, including: when the appearance interaction intent of the controlled virtual character is determined to be an appearance recommendation intent based on the semantic analysis results, displaying the appearance recommendation control on the graphical user interface; or when the appearance interaction intent of the controlled virtual character is determined to be an appearance switching intent, displaying the appearance switching control on the graphical user interface.

[0300] Optionally, before determining that the appearance interaction intent of the controlled virtual character is an appearance switching intent based on the semantic analysis results, the method further includes: determining that the number of appearances possessed by the controlled virtual character reaches a preset number threshold and that the dialogue information is issued by the controlled virtual character.

[0301] Optionally, the method further includes: an appearance recommendation control determines recommended appearances in a first priority order, wherein the first priority order is positively correlated with the number of recommended appearances that satisfy a first preset condition, and the first preset condition includes at least one of the following: target appearances and similar appearances determined by semantic analysis results; appearances not owned by the controlled virtual character; appearances owned by other virtual characters in the largest number; appearances that can be directly obtained; and the latest released appearances.

[0302] Optionally, the method further includes: an appearance switching control displays appearances to be switched in a second priority order, wherein the second priority order is positively correlated with the number of appearances to be switched that meet a second preset condition, and the second preset condition includes at least one of the following: the target appearance and similar appearances determined by semantic analysis results; the appearance with the most appearances owned by the controlled virtual character; the type to which the high-value appearances belong; and the appearance type that the controlled virtual character has not changed for the longest time.

[0303] Optionally, the method further includes: in response to the controlled virtual character entering a non-social state, detecting whether the controlled virtual character meets a third preset condition, the third preset condition including at least one of the controlled virtual character's level, number of consecutive active days, and the controlled virtual character being online and not in a game state; when the third preset condition is met, providing appearance interaction controls on the graphical user interface.

[0304] Optionally, the appearance interaction control includes an appearance recommendation control or an appearance switching control. In response to the controlled virtual character entering a non-social state, it detects whether the controlled virtual character meets a third preset condition. It also includes: in response to the controlled virtual character entering a non-social state, when it is determined that the number of appearances owned by the controlled virtual character has reached a preset number threshold, it detects whether the state of the controlled virtual character meets the third preset condition.

[0305] Optionally, the method also includes: synchronizing appearance interaction information to other virtual characters in the game scene.

[0306] Optionally, the appearance operation interaction information can be synchronized to other virtual characters in the game scene, including: converting the appearance interaction operation results into broadcast information; and pushing the broadcast information to other virtual characters in the game scene so that other virtual characters can trigger their own appearance interaction controls based on the broadcast information.

[0307] Unless otherwise specifically stated, the relative steps, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0308] Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0309] In the description of this application, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0310] Finally, it should be noted that the above embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in this application, or make equivalent substitutions for some of the technical features. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the protection scope of this application.

Claims

1. A visual interaction method, providing a graphical user interface through a terminal device, characterized in that, The method includes: In response to a controlled virtual character entering a social state, the system acquires dialogue information within that social state, and the controlled virtual character is controlled via the terminal device. Based on the dialogue information, semantic analysis is performed to determine whether the controlled virtual character has an appearance interaction intention. In response to the controlled virtual character having an appearance interaction intention, an appearance interaction control is displayed on the graphical user interface, the appearance interaction control being used to provide content corresponding to the appearance interaction intention.

2. The method according to claim 1, characterized in that, The step of performing semantic analysis based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intent includes: The pre-trained natural language recognition model is invoked to perform semantic deep understanding of the dialogue information; The dialogue information is matched with keywords based on a preset keyword library. The preset keyword library includes: an appearance-related keyword library, which contains words related to appearance equipment in the game, and a behavior-related keyword library, which contains behavioral words related to appearance interaction operations. The semantic deep understanding results and the keyword matching results are fused together to obtain the comprehensive confidence level of the appearance interaction intent; When the overall confidence level reaches a preset confidence threshold, it is determined that the controlled virtual character has an intention to interact with the appearance.

3. The method according to claim 2, characterized in that, The process of fusing semantic deep understanding results with keyword matching results includes: The natural language recognition model outputs a semantic understanding confidence score. The keyword matching confidence score is calculated based on the number and weight of the matched keywords. The semantic understanding confidence and the keyword matching confidence are fused using a weighted fusion algorithm to obtain the comprehensive confidence.

4. The method according to claim 1, characterized in that, The appearance interaction controls include appearance recommendation controls or appearance switching controls. The step of displaying the appearance interaction controls on the graphical user interface in response to determining that the controlled virtual character has an appearance interaction intent includes: Based on the semantic analysis results, if the appearance interaction intent of the controlled virtual character is determined to be an appearance recommendation intent, an appearance recommendation control is displayed on the graphical user interface; or if the appearance interaction intent of the controlled virtual character is determined to be an appearance switching intent, an appearance switching control is displayed on the graphical user interface.

5. The method according to claim 4, characterized in that, Before determining, based on the semantic analysis results, that the appearance interaction intent of the controlled virtual character is an appearance switching intent, the method further includes: It is determined that the number of appearances possessed by the controlled virtual character reaches a preset threshold and that the dialogue information is sent by the controlled virtual character.

6. The method according to claim 4, characterized in that, The method further includes: the appearance recommendation control determines recommended appearances according to a first priority order, the first priority order being positively correlated with the number of recommended appearances that satisfy a first preset condition, the first preset condition including at least one of the following: The semantic analysis results determine the target appearance and similar appearances; The controlled virtual character does not possess an appearance; Other virtual characters have the most appearances; Appearance that can be obtained directly; The latest design.

7. The method according to claim 4, characterized in that, The method further includes: the appearance switching control displays appearances to be switched in a second priority order, the second priority order being positively correlated with the number of appearances to be switched that satisfy a second preset condition, the second preset condition including at least one of the following: The semantic analysis results determine the target appearance and similar appearances; The controlled virtual character has the largest number of appearances; The type to which a high-value appearance belongs; The appearance type of the controlled virtual character that has not been changed for the longest time.

8. The method according to claim 1, characterized in that, The method further includes: In response to a controlled virtual character entering a non-social state, it is detected whether the controlled virtual character meets a third preset condition. The third preset condition includes at least one of the following: the controlled virtual character's level, the number of consecutive active days, and the controlled virtual character being online but not in a game state. When the third preset condition is met, the appearance interaction control is provided on the graphical user interface.

9. The method according to claim 8, characterized in that, The appearance interaction controls include appearance recommendation controls or appearance switching controls. The step of detecting whether the controlled virtual character meets a third preset condition in response to the controlled virtual character entering a non-social state further includes: In response to a controlled virtual character entering a non-social state, when it is determined that the number of appearances possessed by the controlled virtual character has reached a preset threshold, it is detected whether the state of the controlled virtual character meets a third preset condition.

10. The method according to claim 1, characterized in that, The method further includes: synchronizing the appearance interaction information to other virtual characters in the game scene.

11. The method according to claim 1, characterized in that, The step of synchronizing appearance operation and interaction information to other virtual characters in the game scene includes: The results of the appearance interaction operation are converted into broadcast information; The broadcast information is pushed to other virtual characters in the game scene so that the other virtual characters can trigger their own appearance interaction controls based on the broadcast information.

12. An interactive device for appearance, providing a graphical user interface via a terminal device, characterized in that, The device includes: The acquisition module is used to acquire dialogue information in the social state in response to the controlled virtual character entering a social state, wherein the controlled virtual character is controlled through the terminal device; The analysis module is used to perform semantic analysis based on the dialogue information to determine whether the controlled virtual character has an appearance interaction intention; The display module is used to display appearance interaction controls on the graphical user interface in response to the controlled virtual character having an appearance interaction intention. The appearance interaction controls are used to provide content corresponding to the appearance interaction intention.

13. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor to implement the method of any one of claims 1 to 12.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to implement the method of any one of claims 1 to 12.

Images