Information Display Method, Device, Computer Equipment and Storage Medium

Abstract

The present application provides an information display method, apparatus, computer device, and storage medium, belonging to the field of artificial intelligence technology. The method includes: obtaining a plurality of scene features based on a target scene currently being displayed; respectively making predictions based on the plurality of scene features to obtain prediction information corresponding to each scene feature; determining weight information based on the plurality of scene features and the corresponding prediction information; and displaying prompt information according to the weight information. In the above technical solution, based on the target scene currently being displayed, a plurality of scene features and the probability of at least one target event occurring in the scene indicated by each scene feature are obtained, and the influence weights of each feature dimension in the target scene features corresponding to the target scene currently being displayed on each target event can be obtained. The influence weights of at least one feature dimension on at least one target event are prompted based on the prompt information, providing relevant information about various target events in the scene and increasing the amount of information.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and particularly relates to an information display method, device, computer device, and storage medium. Background Art

[0002] Multiplayer Online Battle Arena (MOBA) games have become one of the most popular game types. How to predict events that will occur later during the game process of MOBA games is a popular research direction.

[0003] Regarding the real-time prediction of related MOBA games, it is usually based on real-time game data during the game process to predict the real-time winning probabilities of both camps and display the real-time winning probabilities, but the amount of information displayed is less. Summary of the Invention

[0004] Embodiments of the present application provide an information display method, device, computer device, and storage medium, which can prompt the influence weights of at least one feature dimension on at least one target event based on prompt information, provide information related to various target events in a virtual scene, and increase the amount of information. The technical solutions are as follows:

[0005] On the one hand, an information display method is provided. The method includes:

[0006] Obtain multiple scene features based on the currently displayed target scene. Each scene feature includes features corresponding to at least two feature dimensions, and the multiple scene features include a target scene feature corresponding to the currently displayed target scene;

[0007] Perform predictions respectively based on the multiple scene features to obtain prediction information corresponding to each scene feature. The prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature;

[0008] Determine weight information based on the multiple scene features and the corresponding prediction information. The weight information represents the influence weights of each feature dimension in the target scene feature on each target event;

[0009] Display prompt information according to the weight information. The prompt information includes the influence weights of at least one feature dimension on at least one target event.

[0010] On the other hand, an information display device is provided. The device includes:

[0011] A feature acquisition module, configured to acquire multiple scene features based on a target scene currently being displayed. Each scene feature includes features corresponding to at least two feature dimensions, and the multiple scene features include a target scene feature corresponding to the target scene currently being displayed;

[0012] A prediction module, configured to perform predictions respectively based on the multiple scene features to obtain prediction information corresponding to each scene feature. The prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature;

[0013] A determination module, configured to determine weight information based on the multiple scene features and the corresponding prediction information. The weight information represents the influence weights of each feature dimension in the target scene feature on each target event;

[0014] A display module, configured to display prompt information according to the weight information. The prompt information includes the influence weights of at least one feature dimension on at least one target event.

[0015] In an optional implementation manner, the determination module is configured to, for any one of the at least two feature dimensions, respectively obtain first features corresponding to the feature dimension from the multiple scene features; determine the influence weight of the feature dimension on each target event based on the first features corresponding to the multiple scene features and the prediction information; and determine the influence weights of each feature dimension on each target event as the weight information.

[0016] In an optional implementation manner, the feature acquisition module includes:

[0017] A feature extraction unit, configured to perform feature extraction on target scene data corresponding to the target scene currently being displayed to obtain the target scene feature;

[0018] A feature acquisition unit, configured to acquire the multiple scene features based on the target scene feature. The distance between every two adjacent scene features in the multiple scene features is the same, or the difference between the features corresponding to the same feature dimension in every two adjacent scene features in the multiple scene features is the same.

[0019] In an optional implementation manner, the target scene data corresponding to the target scene currently being displayed includes data corresponding to each moment within a target duration before the current moment;

[0020] The feature extraction unit is configured to perform feature extraction on the data corresponding to each moment respectively to obtain target scene sub-features corresponding to each moment; and combine the target scene sub-features corresponding to each moment in chronological order to obtain the target scene feature.

[0021] In an alternative implementation, the target scenario data includes data corresponding to at least two feature dimensions, the at least two feature dimensions including a discrete dimension and a continuous dimension, wherein the data of the discrete dimension belongs to data with a discrete distribution, and the data of the continuous dimension belongs to data with a continuous distribution;

[0022] The feature extraction unit is configured to extract features from the data of the discrete dimension in the target scenario data to obtain a second feature; extract features from the data of the continuous dimension in the target scenario data to obtain a third feature; and splice the second feature and the third feature to obtain the target scenario feature.

[0023] In an alternative implementation, among the multiple scenario features, there are a reference scenario feature and an intermediate scenario feature. The feature acquisition unit is configured to determine a line segment formed by a reference feature point corresponding to the reference scenario feature and a target feature point corresponding to the target scenario feature; extract a target number of feature points from the line segment so that the reference feature point, the target number of feature points, and the target feature point are equally spaced; and determine the features corresponding to the target number of feature points as the intermediate scenario feature.

[0024] In an alternative implementation, the prediction module is configured to process each scenario feature based on a target prediction model to obtain prediction information corresponding to each scenario feature.

[0025] In an alternative implementation, the training process of the target prediction model includes:

[0026] Obtaining first training data based on a sample scenario, the first training data including at least one sample event that occurred in the sample scenario and multiple sample scenario features associated with the at least one sample event, and each sample scenario feature including features corresponding to at least two feature dimensions;

[0027] Training the target prediction model based on the first training data.

[0028] In an alternative implementation, the first training data includes first-class training data and second-class training data, and the at least one sample event includes a repeatable event and a non-repeatable event;

[0029] The obtaining first training data based on a sample scenario includes:

[0030] Obtaining the data at a sample moment before the occurrence of the repeatable event from the sample scenario data corresponding to the sample scenario to obtain the first-class training data;

[0031] Obtain the data within the sample time period before the occurrence of the non-repeatable event from the sample scenario data corresponding to the sample scenario, to obtain the second type of training data.

[0032] In an alternative implementation, the apparatus further includes:

[0033] A fidelity information acquisition module, configured to predict the first test data based on the target prediction model to obtain test prediction information corresponding to each test scenario feature, where the first test data includes multiple test scenario features associated with the at least one sample event, and each test scenario feature includes features corresponding to at least two feature dimensions; determine test weight information based on the multiple test scenario features and the corresponding test prediction information, where the test weight information represents the influence weight of each feature dimension in the multiple test scenario features on each sample event; set to zero the data in the first training data and the first test data that does not belong to the target feature dimension, to obtain second training data and second test data, where the target feature dimension represents at least one feature dimension with the highest influence weight on each sample event; train a target alternative model based on the second training data, where the target alternative model has the same structure as the target prediction model; test the target alternative model according to the second test data to obtain fidelity information, where the fidelity information represents the credibility of the test weight information.

[0034] On the other hand, a computer device is provided, where the computer device includes a processor and a memory, and the memory is used to store at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the operations performed by the information display method in the embodiments of the present application.

[0035] On the other hand, a computer-readable storage medium is provided, where at least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by a processor to implement the operations performed by the information display method in the embodiments of the present application.

[0036] On the other hand, a computer program product or a computer program is provided, where the computer program product or the computer program includes computer program code, and the computer program code is stored in a computer-readable storage medium. The processor of the computer device reads the computer program code from the computer-readable storage medium, and the processor executes the computer program code, so that the computer device executes the information display method provided in the above various alternative implementations.

[0037] The beneficial effects brought by the technical solutions provided in the embodiments of the present application are:

[0038] The technical solution provided by the embodiments of the present application obtains multiple scene features based on the target scene currently being displayed, and respectively obtains the probabilities of at least one target event occurring in the scenes indicated by each scene feature. Thereby, it is possible to obtain the influence weights of each feature dimension in the target scene features corresponding to the target scene currently being displayed on each target event. Finally, based on the prompt information, the influence weights of at least one feature dimension on at least one target event are prompted, providing relevant information about various target events in the scene and improving the amount of information. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0040] Figure 1 is a schematic diagram of the implementation environment of an information display method provided by an embodiment of the present application;

[0041] Figure 2 is a flowchart of an information display method provided by an embodiment of the present application;

[0042] Figure 3 is a flowchart of another information display method provided by an embodiment of the present application;

[0043] Figure 4 is a schematic diagram of feature extraction provided by an embodiment of the present application;

[0044] Figure 5 is a schematic diagram of a target prediction model provided by an embodiment of the present application;

[0045] Figure 6 is a schematic diagram of displaying prompt information provided by an embodiment of the present application;

[0046] Figure 7 is a block diagram of an information display device provided by an embodiment of the present application;

[0047] Figure 8 is a block diagram of the structure of a terminal provided by an embodiment of the present application;

[0048] Figure 9 is a schematic diagram of the structure of a server provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] To make the objectives, technical solutions, and advantages of the present application clearer, the following will further describe the embodiments of the present application in detail with reference to the accompanying drawings.

[0050] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0051] The following briefly introduces the technologies used in the embodiments of the present application:

[0052] Artificial Intelligence (AI) is a theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a way similar to human intelligence. Artificial intelligence also studies the design principles and implementation methods of various intelligent machines to enable the machines to have the functions of perception, reasoning, and decision-making.

[0053] Artificial intelligence technology is an interdisciplinary subject with a wide range of fields, including both hardware-level and software-level technologies. Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, and mechatronics. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0054] Machine Learning (ML) is an interdisciplinary subject that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers simulate or implement human learning behaviors to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve their own performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications cover all fields of artificial intelligence. Machine learning and deep learning usually include technologies such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and learning from demonstration.

[0055] Long Short-Term Memory (LSTM) is a time-recurrent neural network designed to solve the long-term dependency problem of general RNN (Recurrent Neural Network). All RNNs have a chain form of repeated neural network modules. In the standard RNN, this repeated structural module has only a very simple structure, such as a tanh (hyperbolic tangent) layer.

[0056] The following is an explanation of some terms involved in the embodiments of the present application.

[0057] Virtual scene: a virtual scene displayed (or provided) when the application is running on the terminal. The virtual scene can be a simulation environment of the real world, a semi-simulation and semi-fictitious virtual environment, or a purely fictitious virtual environment. The virtual scene can be any one of a two-dimensional virtual scene, a 2.5-dimensional virtual scene or a three-dimensional virtual scene. The embodiment of the present application does not limit the dimension of the virtual scene. For example, the virtual scene includes the sky, land, ocean, etc., and the land includes environmental elements such as deserts and cities. The terminal user can control the virtual object to move in the virtual scene. Optionally, the virtual scene can also be used for a virtual scene battle between at least two virtual objects, and there are virtual resources available for at least two virtual objects in the virtual scene. Optionally, the virtual scene includes two symmetrical areas, and virtual objects belonging to two hostile camps occupy one of the areas respectively, and destroying the target building / stronghold / base / crystal deep in the opponent's area is used as the victory goal, wherein the symmetrical areas are such as the lower left corner area and the upper right corner area, and the left middle area and the right middle area. Optionally, the initial position of one camp in the MOBA game, that is, the position where the virtual objects belonging to the camp are born, is at the lower left corner of the virtual scene, while the initial position of the other camp is at the upper right corner of the virtual scene.

[0058] Virtual object: It refers to an object displayed in a virtual scene. This object can be a virtual character, virtual animal, anime character, etc. For example, the characters, animals, plants, oil drums, walls, stones, etc. displayed in the virtual scene. This virtual object can be a virtual avatar in the virtual scene that represents the user. The virtual scene can include multiple virtual objects, and each virtual object has its own shape and volume in the virtual scene, occupying a part of the space in the virtual scene. Optionally, when the virtual scene is a three-dimensional virtual scene, the virtual object can be a three-dimensional solid model, and this three-dimensional solid model can be a three-dimensional character constructed based on three-dimensional human bone technology. The same virtual object can display different external images by wearing different skins. In some embodiments, the virtual object can also be implemented using a 2.5D or 2D model, and the embodiments of the present application do not limit this. In a MOBA game, the virtual object can be called a hero.

[0059] Optionally, the virtual object is a user character controlled by operations on the client, or an artificial intelligence (AI) set in the virtual scene battle through training, or a non-player character (NPC) set in the virtual scene interaction. Optionally, the virtual object is a virtual character that conducts adversarial interactions in the virtual scene. Optionally, the number of virtual objects participating in the interaction in the virtual scene can be preset or dynamically determined according to the number of clients joining the interaction.

[0060] MOBA (Multiplayer Online Battle Arena) game: It is a game that provides several strongholds in a virtual scene. Users in different camps control virtual objects to battle in the virtual scene, aiming to occupy the strongholds or destroy the strongholds of the opposing camp. For example, a MOBA game can divide users into at least two opposing camps, and different virtual teams belonging to at least two opposing camps respectively occupy their own map areas and compete with a certain victory condition as the goal. The victory condition includes but is not limited to: occupying the stronghold or destroying the stronghold of the opposing camp, killing the virtual objects of the opposing camp, ensuring one's own survival within a specified scene and time, snatching a certain resource, and having an interaction score exceeding that of the other party within a specified time, etc. For example, a mobile MOBA game can divide users into two opposing camps, scatter the virtual objects controlled by users in the virtual scene to compete with each other, and take destroying or occupying all the strongholds of the enemy as the victory condition.

[0061] Optionally, each virtual team includes one or more virtual objects, such as 1, 2, 3, or 5. According to the number of virtual objects in each team participating in the tactical competition, the tactical competition is divided into 1V1 competition, 2V2 competition, 3V3 competition, 5V5 competition, etc. Among them, 1V1 means "one-on-one", which will not be elaborated here.

[0062] Optionally, the MOBA game is carried out in units of games (or rounds). The maps of each game of the tactical competition are the same or different. The duration of a game of the MOBA game is from the start time of the game to the time when the victory condition is achieved.

[0063] In the MOBA game, the user can control the virtual object to release skills to fight against other virtual objects. For example, the types of skills include attack skills, defense skills, healing skills, support skills, execution skills, etc. Each virtual object has its own fixed one or more skills, and different virtual objects usually have different skills, and different skills can produce different effects. For example, if the virtual object releases an attack skill and hits the hostile virtual object, it will cause certain damage to the hostile virtual object, usually manifested as deducting a part of the virtual health value of the hostile virtual object. Another example is that if the virtual object releases a healing skill and hits the friendly virtual object, it will produce certain healing for the friendly virtual object, usually manifested as restoring a part of the virtual health value of the friendly virtual object. The corresponding effects can be produced by other various skills, which will not be enumerated one by one here.

[0064] argmax (arguments of the maxima, the set of independent variable points of the maximum value). When we have another function y = f(x), if the result is x0 = argmax(f(x)), it means that when the function f(x) takes x = x0, the maximum value in the value range of f(x) is obtained; if there are multiple points that make f(x) obtain the same maximum value, then the result of argmax(f(x)) is a set of points. In other words, argmax(f(x)) is the variable point x (or the set of x) corresponding to the maximum value of f(x).

[0065] The information display method provided in the embodiments of the present application can be applied to a computer device. Optionally, the computer device is a terminal or a server. First, taking the computer device as a server as an example, the implementation environment of the information display method provided in the embodiments of the present application will be introduced. Figure 1 It is a schematic diagram of the implementation environment of an information display method provided in the embodiments of the present application. Refer to Figure 1 This implementation environment includes a terminal 101 and a server 102.

[0066] The terminal 101 and the server 102 can be directly or indirectly connected through wired or wireless communication methods, which are not limited in this application.

[0067] Optionally, the terminal 101 is a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto.

[0068] Optionally, the server 102 is an independent physical server, and can also be a server cluster or a distributed system composed of multiple physical servers, and can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The server 102 is used to provide background services for the application program that supports information display. Optionally, the server 102 undertakes the main work and the terminal 101 undertakes the secondary work; or, the server 102 undertakes the secondary work and the terminal 101 undertakes the main work; or, the server 102 and the terminal 101 adopt a distributed computing architecture for collaborative computing.

[0069] In an alternative implementation, the application program that supports information display is a game application program.

[0070] For example, the game application program is a MOBA game program. During the running of the MOBA game on the terminal 101, the virtual scene data corresponding to the currently displayed virtual scene is obtained, and then the virtual scene data is sent to the server 102. After receiving the virtual scene data, the server 102 adopts the information display method provided in the embodiments of the present application to obtain prompt information based on the virtual scene data, and then sends the prompt information to the terminal 101, and the terminal 101 displays the prompt information, so as to be able to prompt which factors will affect the game result.

[0071] In another alternative implementation, the application program that supports information display is a multimedia application program.

[0072] For example, the multimedia application is a sports video playback program. During the operation of the sports video playback program on the terminal 101, the sports game scene data corresponding to the currently displayed sports game scene is obtained, and then the sports game scene data is sent to the server 102. Among them, the sports game scene can be a football game scene, a basketball game scene, a billiards game scene, a skiing game scene, a racing game scene, etc. After receiving the sports game scene data, the server 102 uses the information display method provided in the embodiment of the present application to obtain prompt information based on the sports game scene data, and then sends the prompt information to the terminal 101, and the terminal 101 displays the prompt information, so as to prompt which factors will affect the game result.

[0073] In another optional implementation manner, the application program that supports information display is a monitoring application program.

[0074] For example, the monitoring application program is a traffic monitoring program. During the operation of the traffic monitoring program on the terminal 101, the traffic monitoring data corresponding to the currently displayed traffic monitoring scene is obtained, and then the traffic monitoring data is sent to the server 102. After receiving the traffic monitoring scene data, the server 102 uses the information display method provided in the embodiment of the present application to obtain prompt information based on the traffic monitoring data, and then sends the prompt information to the terminal 101, and the terminal 101 displays the prompt information, so as to prompt which factors will cause traffic jams or traffic accidents, etc.

[0075] Optionally, the server 102 includes an access server, a prediction server, and a database server. The access server is used to provide access services for the terminal 101. The prediction server is used to provide prediction services. The prediction server can be one or more. When there are multiple prediction servers, at least two prediction servers are used to provide different services, and / or at least two prediction servers are used to provide the same service, such as providing the same service in a load balancing manner. The embodiment of the present application does not limit this. Optionally, a target prediction model is set in the prediction server, so as to obtain prediction information based on the target prediction model, and the prediction information is used to generate prompt information.

[0076] The terminal 101 can generally refer to one of multiple terminals. Only the terminal 101 is used as an example in this embodiment. Those skilled in the art can know that the number of the above terminals can be more or less. For example, the above terminal can be only one, or the above terminal can be dozens or hundreds, or more. The embodiment of the present application does not limit the number and device type of the terminals.

[0077] Figure 2 is a flowchart of an information display method provided according to an embodiment of the present application, asFigure 2 As shown, in the embodiment of the present application, the information display method is applied to a computer device as an example for illustration. The method includes the following steps:

[0078] 201. The computer device obtains multiple scene features based on the currently displayed target scene. Each scene feature includes features corresponding to at least two feature dimensions, and the multiple scene features include target scene features corresponding to the currently displayed target scene.

[0079] In the embodiment of the present application, the target scene is a virtual scene, a sports competition scene, a traffic monitoring scene, etc., and the embodiment of the present application does not limit this. The computer device can execute the information display method once every once in a while. Each time it is executed, it obtains the currently displayed target scene, and then obtains multiple scene features based on the target scene. Among them, the multiple scene features include target scene features corresponding to the currently displayed target scene, and scene features associated with the target scene features.

[0080] 202. The computer device makes predictions based on the multiple scene features respectively to obtain prediction information corresponding to each scene feature. The prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature.

[0081] In the implementation of the present application, among the above multiple scene features, each scene feature indicates a scene. The computer device can make predictions for each scene feature respectively to obtain the probability of at least one target event occurring in the scene indicated by each scene feature, that is, obtain the prediction information corresponding to each scene feature. For example, the at least one target event is an event occurring in a MOBA game, such as defeating a virtual object, defeating a neutral creature, destroying a defense tower, and any camp winning the victory, etc.

[0082] 203. The computer device determines weight information based on the multiple scene features and the corresponding prediction information. The weight information represents the influence weights of each feature dimension in the target scene feature on each target event.

[0083] In the implementation of the present application, for each scene feature, the computer device can determine the influence weights of each feature dimension in the target scene feature on the above each target event based on the scene feature and the corresponding prediction information. Among them, the feature dimensions included in each scene feature are the same.

[0084] 204. The computer device displays prompt information according to the weight information. The prompt information includes the influence weights of at least one feature dimension on at least one target event.

[0085] In an embodiment of the present application, after the computer device obtains the weight information, it obtains the influence weights of at least one feature dimension on at least one target event from the weight information as prompt information, and then displays the prompt information to display more information.

[0086] An embodiment of the present application provides an information display method. By obtaining multiple scene features based on the currently displayed target scene and respectively obtaining the probabilities of at least one target event occurring in the scenes indicated by each scene feature, it is possible to obtain the influence weights of each feature dimension on each target event in the target scene features corresponding to the currently displayed target scene. Finally, based on the prompt information, the influence weights of at least one feature dimension on at least one target event are prompted, providing relevant information about various target events in the scene and increasing the amount of information.

[0087] Figure 3 It is a flowchart of another information display method provided according to an embodiment of the present application. As Figure 3 shown, in an embodiment of the present application, taking the target scene as a virtual scene and the information display method being applied to a computer device as an example for illustration. The method includes the following steps:

[0088] 301. The computer device extracts features from the virtual scene data corresponding to the currently displayed virtual scene to obtain target scene features.

[0089] In an embodiment of the present application, taking the application in a MOBA game as an example, the currently displayed virtual scene is the virtual scene at any moment in the MOBA game. The computer device can obtain the virtual scene data corresponding to that moment and then extract the virtual scene data to obtain the target scene features corresponding to the currently displayed virtual scene. Among them, the target scene features include at least two feature dimensions.

[0090] For example, during the game process of a MOBA game, virtual scene data such as the levels of each hero, the positions of each hero in the virtual scene, the amount of gold each hero has, the equipment each hero has, the survival status of each neutral creature, and the status of each defense tower change continuously over time. For any moment during the game process, the computer device can obtain the virtual scene data corresponding to that moment in the virtual scene.

[0091] In an alternative implementation, the virtual scene data corresponding to the currently displayed virtual scene includes data corresponding to each moment within a target duration before the current moment, so that the obtained target scene features include the temporal relationship between the data of multiple consecutive moments. Correspondingly, this step is as follows: The computer device extracts features from the data corresponding to each moment respectively to obtain target scene sub-features corresponding to each moment, and then the computer device combines the target scene sub-features corresponding to each moment in chronological order to obtain the target scene features. By obtaining the data corresponding to multiple consecutive moments before the current moment, the temporal relationship between the data of each moment is included in the extracted target scene features.

[0092] For example, if the current moment is represented as t and the target duration is l consecutive seconds, then the target scene features are represented as X = [x t -l+l ,..., x t T , where x t-l+1 represents the target scene sub-feature corresponding to the l-th second before the current moment, x t represents the target scene sub-feature corresponding to the current moment, and T represents matrix transpose.

[0093] In an alternative implementation, the virtual scene data corresponding to the currently displayed virtual scene includes data corresponding to at least two feature dimensions, and the at least two feature dimensions include a discrete dimension and a continuous dimension, where the data in the discrete dimension belongs to discretely distributed data and the data in the continuous dimension belongs to continuously distributed data. Correspondingly, this step is as follows: The computer device extracts features from the data in the virtual scene data that belongs to the discrete dimension to obtain a second feature; and extracts features from the data in the virtual scene data that belongs to the continuous dimension to obtain a third feature. Then the computer device concatenates the second feature and the third feature to obtain the target scene features. By extracting features from the discretely distributed data and the continuously distributed data in the virtual scene data respectively and then concatenating them, the obtained target scene features can represent the data of the above two distributions.

[0094] For example, for the discretely distributed data in the virtual scene data such as hero ID (Identity Document, unique coding), skill ID, npc category, whether the hero is alive, whether the neutral creature is alive, and whether the defense tower is destroyed, the computer device can encode the discretely distributed data into a one-hot vector. For the continuously distributed data in the virtual scene data such as the economic difference, hero health value, and kill number difference, the computer device can normalize the discretely distributed data. The computer device concatenates the one-hot vector and the normalized data to obtain the target scene features. ​

[0095] In an alternative implementation, the computer device can also convert discretely distributed data and continuously distributed data into an embedding vector. Correspondingly, for discretely distributed data, the computer device processes the discretely distributed data through parallel fully-connected layers, with one fully-connected layer corresponding to one feature dimension. For continuously distributed data, the computer device processes the continuously distributed data through a normalization layer. Finally, the computer device concatenates the outputs of the fully-connected layer and the normalization layer to obtain an embedding vector, and uses this embedding vector as the target scene feature. By converting the virtual scene data into an embedding vector, the discretely distributed data is converted into the form of a continuous vector, which is convenient for subsequent processing.

[0096] For example, referring to Figure 4 as shown Figure 4 is a schematic diagram of feature extraction provided according to an embodiment of the present application. As Figure 4 shown, data in fields such as hero_unique encoding, skill_unique encoding, and non-user role_category are discretely distributed data. Among them, one field corresponds to one feature dimension and is processed by one fully-connected layer. The continuously distributed data is processed by the normalization layer. The computer device concatenates the outputs of the fully-connected layer and the normalization layer to obtain an embedding vector. Among them, when the computer device processes the discretely distributed data, it can play a role in compression. For example, when the discretely distributed data and the continuously distributed data have a total of 5885 dimensions before processing, after passing through the fully-connected layer and normalization processing, a 2001-dimensional embedding vector is finally obtained.

[0097] 302. Based on the target scene feature, the computer device obtains multiple scene features, and each scene feature includes features corresponding to at least two feature dimensions.

[0098] In the embodiment of the present application, after obtaining the target scene feature, the computer device can obtain multiple scene features that have the same feature dimensions as the target scene feature based on the target scene feature. The distance between every two adjacent scene features among the multiple scene features is the same, or the difference between the features corresponding to the same feature dimension in every two adjacent scene features among the multiple scene features is the same.

[0099] In an alternative implementation, the multiple scenario features include a reference scenario feature and an intermediate scenario feature. Correspondingly, this step is as follows: The computer device determines a line segment formed by a reference feature point corresponding to the reference scenario feature and a target feature point corresponding to the target scenario feature. Then the computer device extracts a target number of feature points from this line segment so that the reference feature point, the target number of feature points, and the target feature point are equally spaced. Finally, the computer device determines the features corresponding to the target number of feature points as the intermediate scenario features. By determining the line segment formed by the reference feature point corresponding to the reference scenario feature and the target feature point corresponding to the target scenario feature, and extracting the target number of feature points, the features corresponding to the target number of feature points can reflect the intermediate scenario features in the process of developing from the reference scenario feature to the target scenario feature.

[0100] For example, taking the reference scenario feature as a zero vector, the computer device obtains the reference feature point corresponding to the reference scenario feature and the target feature point corresponding to the target scenario feature. Then the reference feature point and the target feature point are connected by a line segment. Then the computer device divides the line segment equally into N parts, extracts N - 1 feature points, and then obtains the features corresponding to the N - 1 feature points to obtain N - 1 intermediate scenario features. Here, N is a positive integer greater than or equal to 1.

[0101] In an alternative implementation, the computer device can obtain the feature difference corresponding to each feature dimension. Then, taking the target scenario feature as a reference, based on the feature difference corresponding to each feature dimension, the computer device obtains the intermediate scenario features adjacent to the target scenario feature, and then, taking this intermediate scenario feature as a reference, based on the feature difference corresponding to each feature dimension, obtains another intermediate scenario feature adjacent to this intermediate scenario feature until the target number of intermediate scenario features is obtained. By taking the target scenario feature as a reference and performing interpolation calculations based on each feature dimension, the obtained intermediate scenario features can reflect the features in the process of developing from each feature dimension to the target scenario feature.

[0102] For example, taking the target number as 2 and the target scenario feature as {3, 6, 15, 28}, the feature difference corresponding to the first feature dimension is 1, the feature difference corresponding to the second feature dimension is 2, the feature difference corresponding to the third feature dimension is 5, and the feature difference corresponding to the fourth feature is 7. Then the intermediate scenario feature adjacent to the target scenario feature is {2, 4, 10, 21}, and the other intermediate scenario feature adjacent to this intermediate scenario feature is {1, 2, 5, 14}.

[0103] 303. The computer device makes predictions based on multiple scenario features respectively, and obtains prediction information corresponding to each scenario feature. This prediction information is used to represent the probability of at least one target event occurring in the virtual scenario indicated by the corresponding scenario feature.

[0104] In the embodiments of the present application, for each scene feature, the computer device can predict the probability of at least one target event occurring in the virtual scene indicated by the scene feature based on the scene feature. Among them, the at least one target event is an event that occurs in a MOBA game, such as defeating a virtual object, defeating a neutral creature, destroying a defense tower, and any camp winning the victory, etc.

[0105] In an alternative implementation, the computer device can process each scene feature based on the target prediction model respectively to obtain prediction information corresponding to each scene feature. Among them, the target prediction model is a machine learning deep model, such as an LSTM model, a fully connected application network model, and a Transformer (deep self-attention transformation network), etc., and the embodiments of the present application do not limit this.

[0106] For example, taking the target prediction model as an LSTM model and the input as the target scene feature as an example, see Figure 5 as shown. Figure 5 is a schematic diagram of a target prediction model provided according to the embodiments of the present application. As Figure 5 shown, the target scene feature is X = [x t-l+1 ,..., x t T , where x t-l+1 represents the target scene sub-feature corresponding to the second before the current moment, x t-1 represents the target scene sub-feature corresponding to the 1st second before the current moment, x t represents the target scene sub-feature corresponding to the current moment, and T represents matrix transpose. Each target scene sub-feature is obtained by the computer device extracting features from discrete-distributed data and continuous-distributed data. The computer device processes the target scene sub-features in the above target scene feature in chronological order based on the LSTM model, and then obtains the prediction information corresponding to the target scene feature through activation functions such as Tanh or Sigmoid, denoted as P(y|X). Optionally, the computer device can also obtain the maximum value of P(y|X) based on the argmax() function in the fully understanding neural network layer to obtain a further prediction result. For example, for the target event of killing a neutral creature in a MOBA game, P(y|X) represents the probability distribution of this target event occurring: the probability that the red camp kills this neutral creature is 20%, and the probability that the blue camp kills this neutral creature is 80%. The final prediction result F(X) obtained by the computer device based on the argmax() function is that the blue camp kills this neutral creature.

[0107] ​In an alternative implementation, taking a sample virtual scene as an example of the sample scenario, the training process of the target prediction model includes: The computer device obtains first training data based on the sample virtual scene. The first training data includes at least one sample event that occurs in the sample virtual scene and multiple sample scenario features associated with the at least one sample event. Each sample scenario feature includes features corresponding to at least two feature dimensions. The computer device trains the target prediction model based on the first training data.

[0108] In an alternative implementation, the first training data includes first-class training data and second-class training data. The at least one sample event includes repeatable events and non-repeatable events. Among them, repeatable events include killing virtual objects and killing neutral creatures, etc., and non-repeatable events include destroying defense towers and winning, etc. Correspondingly, the step of the computer device obtaining the first training data based on the sample virtual scene includes: The computer device obtains the data at the sample moment before the occurrence of the repeatable event from the sample scenario data corresponding to the sample virtual scene to obtain the first-class training data; and obtains the data within the sample time period before the occurrence of the non-repeatable event from the sample scenario data corresponding to the sample virtual scene to obtain the second-class training data.

[0109] 304. For any one of the at least two feature dimensions, the computer device respectively obtains the first feature corresponding to this feature dimension from the multiple scenario features.

[0110] In the embodiments of the present application, each scenario feature includes features of at least two feature dimensions. For any one feature dimension, the computer device can obtain the first feature corresponding to this feature dimension from each scenario feature to obtain multiple first features. By respectively extracting the first features corresponding to each feature dimension in each intermediate scenario feature, the computer device can determine the influence weight of each feature dimension on each target event based on the development trend and differences between the first features.

[0111] For example, the multiple scenario features include a reference scenario feature, a target scenario feature, and N - 1 intermediate scenario features. Where N is a positive integer greater than or equal to 1. Then for any one feature dimension, the computer device obtains N + 1 first features corresponding to this feature dimension from the reference scenario feature, the target scenario feature, and the N - 1 intermediate scenario features.

[0112] 305. The computer device determines the influence weight of this feature dimension on each target event based on the first features corresponding to the multiple scenario features and the prediction information.

[0113] In an embodiment of the present application, for any feature dimension, the computer device can determine the influence weight of each feature dimension on each target event based on the multiple first features corresponding to the feature dimension in the multiple scenario features and the prediction information of the multiple scenario features. Among them, the computer device can calculate the influence weight of the j-th dimensional feature on each target event according to formula (1).

[0114]

[0115] Among them, IG j represents the influence weight of the j-th feature dimension on each target event; X j represents the first feature corresponding to the j-th feature dimension in the target scenario feature; X′ j represents the first feature corresponding to the j-th feature dimension in the reference scenario feature; steps represents the number of equal divisions between the reference scenario feature and the target scenario feature, that is, the sum of the number of intermediate scenario features and the target scenario feature; represents the k-th intermediate scenario feature, and the steps-th intermediate scenario feature is the target scenario feature; represents the first feature corresponding to the j-th feature dimension in the k-th intermediate scenario feature; represents the prediction information corresponding to the k-th intermediate scenario feature; represents taking the partial derivative. The value of steps is positively correlated with the accuracy of IG j , that is, the larger the value of steps, the more accurate the IG j . Experiments show that good results can be obtained when the value ranges from 100 to 300.

[0116] In an alternative implementation, the computer device can also use the Integrated Gradients (IG) method to obtain the influence weight of each feature dimension on each target event. In this integrated gradient method, represents the scenario sub-feature corresponding to the t-th moment, n represents the n-th feature dimension, and n is a positive integer greater than 1. The computer device combines the scenario sub-features at the t-th moment and the continuous l seconds before the t-th moment into the target scenario feature X = [x t -l+1 ,..., x t T , where x t-l+1 represents the target scenario sub-feature corresponding to the l-th second before the current moment, and x t ​Denote the target scenario sub-feature corresponding to time t. P(y|X) represents the prediction information corresponding to the target scenario feature. The computer obtains a all-zero vector with the same dimension as the target scenario feature as the reference scenario feature X′, and the line integral of the gradient from X′ is the integral gradient of X. The computer device calculates the integral gradient of the target scenario feature according to the following formula (2).

[0117]

[0118] Among them, IG i，j represents the influence weight of the j-th feature dimension of the scenario sub-feature corresponding to the i-th moment in the target scenario feature on each target event; X i，j represents the scenario sub-feature corresponding to the j-th feature dimension at the i-th moment in the target scenario feature; X′ i，j represents the scenario sub-feature corresponding to the j-th feature dimension at the i-th moment in the reference scenario feature; represents that the integral gradient method uses linear interpolation as the variational path; represents the intermediate scenario feature on the variational path; represents the scenario sub-feature corresponding to the j-th feature dimension at the i-th moment in the intermediate scenario feature; represents the prediction information corresponding to the intermediate scenario feature; represents taking the partial derivative.

[0119] It should be noted that since formula (2) is a theoretical formula and the theoretical value cannot be calculated in actual applications, an approximate method is used for calculation. The approximate formula is shown in formula (1) and will not be elaborated here.

[0120] 306. The computer device determines the influence weight of each feature dimension on each target event as the weight information.

[0121] In the embodiment of the present application, the computer device can obtain the influence weight of each feature dimension on each target event, so as to obtain the weight information.

[0122] For example, taking the influence weights of 5 feature dimensions on 4 target events in a MOBA game as an example. The 4 target events are: 1. Which camp wins the game; 2. Which camp kills large jungle monsters; 3. Which hero can kill other heroes; 4. Which hero will be killed. The 5 feature dimensions are: 1. Distance in the virtual scene; 2. Whether the hero is alive; 3. Economy difference; 4. Kill count difference; 5. Skill level difference. Correspondingly, the computer device obtains a total of 20 influence weights and determines these 20 influence weights as the weight information.

[0123] 307. The computer device displays prompt information according to the weight information, and the prompt information includes the influence weight of at least one feature dimension on at least one target event.

[0124] In the embodiments of the present application, for any target event, the computer device can obtain, from the weight information, the influence weights of each feature dimension on the any target event, then select at least one feature dimension with the highest influence weight from the influence weights of each feature dimension on the any target event, and then determine the influence weight of the at least one feature dimension on the any target event as the prompt information. Of course, the computer device can also fill the influence weight of the at least one feature dimension on the any target event into a text template and then display it. The embodiments of the present application do not limit this.

[0125] For example, referring to Figure 6 shown Figure 6 is a schematic diagram showing prompt information provided according to the embodiments of the present application. As Figure 6 shown, for the target event of which hero should kill the large wild monster "Baron Nashor". The computer device obtains from the weight information the influence weights of the distance between "Baron Nashor" and each hero, the survival status of Hero 3, the economic gap, the kill count gap, and the skill levels of each hero on killing "Baron Nashor". The displayed prompt information is as follows: First reason: The heroes of the red camp are closer to Baron Nashor and thus more likely to kill it; Second reason: Hero 3 has been killed. Hero 3 belongs to the blue team, so the blue team is at a disadvantage. Third reason: The blue camp is at a disadvantage in terms of economy. Fourth reason: The blue camp is at a disadvantage in terms of kill count. Fifth reason: The skill level of Skill 3 of Hero 1 is low. This skill is very important for team battles, so the team of Hero 1 (blue team) is difficult to kill "Baron Nashor". Optionally, the audience can, based on the displayed prompt information, understand the events that may occur in the MOBA game and the reasons for their occurrence. Optionally, the commentator can, based on the displayed prompt information, commentate on the MOBA game.

[0126] It should be noted that, in order to ensure the credibility of the weight information determined by the computer device, after the computer device trains the target prediction model, it also evaluates the target prediction model. The computer device uses a fidelity evaluation method for evaluation. Correspondingly, the computer device predicts the first test data based on the target prediction model to obtain test prediction information corresponding to each test scenario feature. The first test data includes multiple test scenario features associated with at least one sample event, and each test scenario feature includes features corresponding to at least two feature dimensions. Then, the computer device determines test weight information based on the multiple test scenario features and the corresponding test prediction information. The test weight information represents the influence weights of each feature dimension in the multiple test scenario features on each sample event. Then, the computer device sets to zero the data in the first training data and the first test data that do not belong to the target feature dimension, where the target feature dimension represents at least one feature dimension with the highest influence weight on each sample event. Then, the computer device trains a target replacement model based on the second training data. The target replacement model has the same structure as the target prediction model. Finally, the computer device tests the target replacement model according to the second test data to obtain fidelity information, which represents the credibility of the test weight information.

[0127] For example, the algorithm of the Fidelity evaluation method is as follows:

[0128] Input: target prediction model F; target replacement model Q; way φ to obtain weight information; feature dimension m to be selected; first training data T r , first test data T;

[0129] Output: Fidelity;

[0130]

[0131]

[0132] The embodiments of the present application provide an information display method. By obtaining multiple scenario features based on the currently displayed virtual scenario and respectively obtaining the probability of at least one target event occurring in the virtual scenario indicated by each scenario feature, it is possible to obtain the influence weights of each feature dimension in the target scenario features corresponding to the currently displayed virtual scenario on each target event. Finally, based on the prompt information, the influence weights of at least one feature dimension on at least one target event are prompted, providing relevant information about various target events in the virtual scenario and improving the amount of information.

[0133] Moreover, by separately extracting the first feature corresponding to each feature dimension in each intermediate scene feature, the computer device can determine the influence weight of each feature dimension on each target event based on the development trend and differences among the first features.

[0134] Moreover, by obtaining data corresponding to multiple consecutive moments before the current moment, the obtained target scene features include the temporal connection between the data of each moment.

[0135] Moreover, by separately performing feature extraction on the discretely distributed data and continuously distributed data in the virtual scene data, and then splicing them, the obtained target scene features can represent the data of the above two distributions.

[0136] Moreover, by determining the line segment formed by the reference feature points corresponding to the reference scene feature and the target feature points corresponding to the target scene feature, and extracting the target number of feature points, the features corresponding to the target number of feature points can reflect the intermediate scene features in the process of developing from the reference scene feature to the target scene feature.

[0137] Moreover, by taking the target scene feature as a benchmark and performing interpolation calculation based on each feature dimension, the obtained intermediate scene features can reflect the features in the process of each feature dimension developing to the target scene feature.

[0138] Moreover, by using the fidelity evaluation method for evaluation, the target prediction model can be effectively evaluated, and it can be determined that the solution provided by this application can accurately predict the target events in the virtual scene and determine the influence degree of each feature dimension on each target event.

[0139] Figure 7 It is a block diagram of an information display device provided by an embodiment of this application. The device is used to execute the steps in the above information display method. Refer to Figure 7 The device includes: a feature acquisition module 701, a prediction module 702, a determination module 703, and a display module 704.

[0140] The feature acquisition module 701 is used to obtain multiple scene features based on the target scene currently being displayed. Each scene feature includes features corresponding to at least two feature dimensions, and the multiple scene features include the target scene feature corresponding to the target scene currently being displayed.

[0141] The prediction module 702 is used to perform predictions respectively based on the multiple scene features to obtain prediction information corresponding to each scene feature. The prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature.

[0142] A determination module 703, configured to determine weight information based on the multiple scenario features and the corresponding prediction information, where the weight information represents the influence weights of the respective feature dimensions in the target scenario feature on the respective target events;

[0143] A display module 704, configured to display a prompt message according to the weight information, where the prompt message includes the influence weights of at least one feature dimension on at least one target event.

[0144] In an optional implementation manner, the determination module 703 is configured to, for any one of the at least two feature dimensions, respectively obtain a first feature corresponding to the feature dimension from the multiple scenario features; determine the influence weight of the feature dimension on the respective target events based on the first features corresponding to the multiple scenario features and the prediction information; and determine the influence weights of the respective feature dimensions on the respective target events as the weight information.

[0145] In an optional implementation manner, the feature acquisition module 701 includes:

[0146] A feature extraction unit, configured to perform feature extraction on the target scenario data corresponding to the currently displayed target scenario to obtain the target scenario feature;

[0147] A feature acquisition unit, configured to obtain the multiple scenario features based on the target scenario feature, where the distance between every two adjacent scenario features in the multiple scenario features is the same, or the difference between the features corresponding to the same feature dimension in every two adjacent scenario features in the multiple scenario features is the same.

[0148] In an optional implementation manner, the target scenario data corresponding to the currently displayed target scenario includes data corresponding to each moment within a target duration before the current moment;

[0149] The feature extraction unit is configured to perform feature extraction on the data corresponding to each moment respectively to obtain target scenario sub-features corresponding to each moment; and combine the target scenario sub-features corresponding to each moment in chronological order to obtain the target scenario feature.

[0150] In an optional implementation manner, the target scenario data includes data corresponding to at least two feature dimensions, where the at least two feature dimensions include a discrete dimension and a continuous dimension, and the data of the discrete dimension belongs to data with a discrete distribution, and the data of the continuous dimension belongs to data with a continuous distribution;

[0151] The feature extraction unit is configured to extract features from the data belonging to the discrete dimension in the target scenario data to obtain a second feature; extract features from the data belonging to the continuous dimension in the target scenario data to obtain a third feature; and splice the second feature and the third feature to obtain the target scenario feature.

[0152] In an alternative implementation, among the multiple scenario features, there are a reference scenario feature and an intermediate scenario feature. The feature acquisition unit is configured to determine a line segment formed by a reference feature point corresponding to the reference scenario feature and a target feature point corresponding to the target scenario feature; extract a target number of feature points from the line segment so that the reference feature point, the target number of feature points, and the target feature point are equidistantly distributed; and determine the features corresponding to the target number of feature points as the intermediate scenario feature.

[0153] In an alternative implementation, the prediction module 702 is configured to process each scenario feature based on a target prediction model to obtain prediction information corresponding to each scenario feature.

[0154] In an alternative implementation, the training process of the target prediction model includes:

[0155] Obtaining first training data based on a sample scenario, where the first training data includes at least one sample event that occurs in the sample scenario and multiple sample scenario features associated with the at least one sample event, and each sample scenario feature includes features corresponding to at least two feature dimensions;

[0156] Training the target prediction model based on the first training data.

[0157] In an alternative implementation, the first training data includes first-class training data and second-class training data, and the at least one sample event includes repeatable events and non-repeatable events;

[0158] The obtaining of the first training data based on the sample scenario includes:

[0159] Obtaining the data at the sample moment before the occurrence of the repeatable event from the sample scenario data corresponding to the sample scenario to obtain the first-class training data;

[0160] Obtaining the data within the sample time period before the occurrence of the non-repeatable event from the sample scenario data corresponding to the sample scenario to obtain the second-class training data.

[0161] In an alternative implementation, the apparatus further includes:

[0162] The fidelity information acquisition module is used to predict the first test data based on the target prediction model to obtain test prediction information corresponding to each test scenario feature. The first test data includes multiple test scenario features associated with the at least one sample event, and each test scenario feature includes features corresponding to at least two feature dimensions. Based on the multiple test scenario features and the corresponding test prediction information, test weight information is determined. The test weight information represents the influence weight of each feature dimension in the multiple test scenario features on each sample event. Set the data in the first training data and the first test data that do not belong to the target feature dimension to zero to obtain the second training data and the second test data. The target feature dimension represents at least one feature dimension with the highest influence weight on each sample event. A target alternative model is trained based on the second training data. The target alternative model has the same structure as the target prediction model. The target alternative model is tested according to the second test data to obtain the fidelity information, and the fidelity information represents the credibility of the test weight information.

[0163] An information display device provided by an embodiment of the present application can obtain multiple scenario features based on the currently displayed virtual scenario, and respectively obtain the probability of at least one target event occurring in the virtual scenario indicated by each scenario feature. Thus, it can obtain the influence weight of each feature dimension in the target scenario features corresponding to the currently displayed virtual scenario on each target event. Finally, based on the prompt information, the influence weight of at least one feature dimension on at least one target event is prompted, providing relevant information about various target events in the virtual scenario and increasing the amount of information.

[0164] It should be noted that when the information display device provided in the above embodiment displays information, only the above-mentioned division of each functional module is used for illustration. In actual application, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the information display device provided in the above embodiment and the embodiment of the information display method belong to the same concept. For the specific implementation process, refer to the method embodiment, which will not be elaborated here.

[0165] In the embodiment of the present application, the computer device can be configured as a terminal or a server. When the computer device is configured as a terminal, the terminal can be used as the execution subject to implement the technical solution provided by the embodiment of the present application. When the computer device is configured as a server, the server can be used as the execution subject to implement the technical solution provided by the embodiment of the present application, or the technical solution provided by the present application can be implemented through the interaction between the terminal and the server. The embodiment of the present application does not make any limitation on this.

[0166] Figure 8It is a structural block diagram of a terminal 800 provided according to an embodiment of the present application. The terminal 800 may be a portable mobile terminal, such as: a smart phone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer or a desktop computer. The terminal 800 may also be referred to by other names such as user equipment, portable terminal, laptop terminal, desktop terminal, etc.

[0167] Generally, the terminal 800 includes: a processor 801 and a memory 802.

[0168] The processor 801 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 801 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), PLA (Programmable Logic Array). The processor 801 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the wake state, also known as the CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 801 may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 801 may also include an AI (Artificial Intelligence) processor, and the AI processor is used to process computational operations related to machine learning.

[0169] The memory 802 may include one or more computer-readable storage media, and the computer-readable storage media may be non-transitory. The memory 802 may also include high-speed random access memory, as well as non-volatile memory, such as one or more disk storage devices, flash storage devices. In some embodiments, the non-transitory computer-readable storage media in the memory 802 is used to store at least one computer program, and the at least one computer program is used to be executed by the processor 801 to implement the information display method provided in the method embodiments of the present application.

[0170] In some embodiments, the terminal 800 may further optionally include: a peripheral device interface 803 and at least one peripheral device. The processor 801, the memory 802, and the peripheral device interface 803 may be connected through a bus or signal lines. Each peripheral device may be connected to the peripheral device interface 803 through a bus, signal lines, or a circuit board. Specifically, the peripheral device includes at least one of a display screen 804, an audio circuit 805, and a power supply 806.

[0171] The peripheral device interface 803 can be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 801 and the memory 802. In some embodiments, the processor 801, the memory 802, and the peripheral device interface 803 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 801, the memory 802, and the peripheral device interface 803 can be implemented on a separate chip or circuit board, and this embodiment does not limit this.

[0172] The display screen 804 is used to display a UI (User Interface). The UI may include graphics, text, icons, videos, and any combination thereof. When the display screen 804 is a touch display screen, the display screen 804 also has the ability to collect touch signals on or above the surface of the display screen 804. The touch signals can be input to the processor 801 as control signals for processing. At this time, the display screen 804 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 804, which is set on the front panel of the terminal 800; in some other embodiments, there may be at least two display screens 804, which are respectively set on different surfaces of the terminal 800 or in a folding design; in some other embodiments, the display screen 804 may be a flexible display screen, which is set on the curved surface or folding surface of the terminal 800. Even, the display screen 804 can also be set to an irregular non-rectangular shape, that is, a special-shaped screen. The display screen 804 can be prepared using materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0173] The audio circuit 805 may include a microphone and a speaker. The microphone is used to collect sound waves of the user and the environment, and convert the sound waves into electrical signals for input to the processor 801 for processing, or input to the radio frequency circuit 804 to enable voice communication. For the purpose of stereo collection or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the terminal 800. The microphone may also be an array microphone or an omnidirectional collection microphone. The speaker is used to convert the electrical signal from the processor 801 or the radio frequency circuit 804 into sound waves. The speaker may be a traditional thin film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for uses such as ranging. In some embodiments, the audio circuit 805 may further include a headphone jack.

[0174] The power supply 806 is used to supply power to each component in the terminal 800. The power supply 806 may be alternating current, direct current, a disposable battery or a rechargeable battery. When the power supply 806 includes a rechargeable battery, the rechargeable battery may be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery may also be used to support fast charging technology.

[0175] Those skilled in the art can understand that Figure 8 the structure shown in

[0176] Figure 9 is a schematic structural diagram of a server provided by an embodiment of the present application. The server 900 may vary greatly due to different configurations or performances, and may include one or more processors (Central Processing Units, CPUs) 901 and one or more memories 902. Among them, at least one computer program is stored in the memory 902, and the at least one computer program is loaded and executed by the processor 901 to implement the information display method provided by each of the above method embodiments. Of course, the server may also have components such as a wired or wireless network interface, a keyboard, and an input / output interface for input / output. The server may also include other components for implementing the functions of the device, which will not be elaborated here.

[0177] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium is applied to a computer device. At least one segment of computer program is stored in the computer-readable storage medium, and the at least one segment of computer program is loaded and executed by a processor to implement the operations performed by the computer device in the information display method of the above embodiment.

[0178] The embodiments of the present application also provide a computer program product or a computer program. The computer program product or the computer program includes computer program code, and the computer program code is stored in a computer-readable storage medium. The processor of the terminal reads the computer program code from the computer-readable storage medium, and the processor executes the computer program code, so that the computer device executes the information display method provided in the above various optional implementation manners.

[0179] Those of ordinary skill in the art can understand that all or part of the steps to implement the above embodiments can be completed by hardware, or can be completed by a program instructing relevant hardware. The program can be stored in a computer-readable storage medium, and the above-mentioned storage medium can be a read-only memory, a disk or an optical disc, etc.

[0180] The above are only the optional embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An information display method, characterized in that, The method includes: Obtaining a plurality of scene features based on a target scene currently being displayed, each scene feature including features corresponding to at least two feature dimensions, and the plurality of scene features including a target scene feature corresponding to the target scene currently being displayed; Performing predictions respectively based on the plurality of scene features to obtain prediction information corresponding to each scene feature, where the prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature; For any one of the at least two feature dimensions, respectively obtaining first features corresponding to the feature dimension from the plurality of scene features; determining the influence weight of the feature dimension on each target event based on the first features corresponding to the plurality of scene features and the prediction information; determining the influence weights of each feature dimension on each target event as weight information; Displaying a prompt message according to the weight information, where the prompt message includes the influence weights of at least one feature dimension on at least one target event.

2. The method according to claim 1, wherein The obtaining a plurality of scene features based on a target scene currently being displayed includes: Performing feature extraction on target scene data corresponding to the target scene currently being displayed to obtain the target scene feature; Based on the target scene feature, obtaining the plurality of scene features, where the distance between every two adjacent scene features in the plurality of scene features is the same, or the difference between the features corresponding to the same feature dimension in every two adjacent scene features in the plurality of scene features is the same.

3. The method according to claim 2, characterized in that, The target scene data corresponding to the target scene currently being displayed includes data corresponding to each moment within a target duration before the current moment; The performing feature extraction on target scene data corresponding to the target scene currently being displayed to obtain the target scene feature includes: Performing feature extraction on the data corresponding to each moment respectively to obtain target scene sub-features corresponding to each moment; Combining the target scene sub-features corresponding to each moment in chronological order to obtain the target scene feature.

4. The method according to claim 2, characterized in that, The target scene data includes data corresponding to at least two feature dimensions, where the at least two feature dimensions include a discrete dimension and a continuous dimension, and the data in the discrete dimension belongs to data with a discrete distribution, and the data in the continuous dimension belongs to data with a continuous distribution; The performing feature extraction on target scene data corresponding to the target scene currently being displayed to obtain the target scene feature includes: Performing feature extraction on the data in the target scene data that belongs to the discrete dimension to obtain a second feature; Performing feature extraction on the data in the target scene data that belongs to the continuous dimension to obtain a third feature; Concatenating the second feature and the third feature to obtain the target scene feature.

5. The method according to claim 2, characterized in that, The plurality of scene features include a reference scene feature and intermediate scene features, and the obtaining the plurality of scene features based on the target scene feature includes: Determining a line segment formed by a reference feature point corresponding to the reference scene feature and a target feature point corresponding to the target scene feature; Extract a target number of feature points from the line segment so that the reference feature points, the target number of feature points, and the target feature points are equally spaced; Determine the features corresponding to the target number of feature points as the intermediate scene features.

6. The method according to claim 1, characterized in that, The performing predictions based on the multiple scene features respectively to obtain prediction information corresponding to each scene feature includes: Process each scene feature respectively based on a target prediction model to obtain prediction information corresponding to each scene feature.

7. The method according to claim 6, characterized in that, The training process of the target prediction model includes: Obtain first training data based on a sample scene, where the first training data includes at least one sample event that occurs in the sample scene and multiple sample scene features associated with the at least one sample event, and each sample scene feature includes features corresponding to at least two feature dimensions; Train the target prediction model based on the first training data.

8. The method according to claim 7, wherein The first training data includes first-class training data and second-class training data, and the at least one sample event includes repeatable events and non-repeatable events; The obtaining first training data based on a sample scene includes: Obtain data at a sample moment before the occurrence of the repeatable event from sample scene data corresponding to the sample scene to obtain the first-class training data; Obtain data within a sample time period before the occurrence of the non-repeatable event from sample scene data corresponding to the sample scene to obtain the second-class training data.

9. The method according to claim 7, wherein The method further includes: Perform predictions on first test data based on the target prediction model to obtain test prediction information corresponding to each test scene feature, where the first test data includes multiple test scene features associated with the at least one sample event, and each test scene feature includes features corresponding to at least two feature dimensions; Determine test weight information based on the multiple test scene features and the corresponding test prediction information, where the test weight information represents the influence weights of each feature dimension in the multiple test scene features on each sample event; Set data that does not belong to the target feature dimension in the first training data and the first test data to zero to obtain second training data and second test data, where the target feature dimension represents at least one feature dimension with the highest influence weight on each sample event; Train a target alternative model based on the second training data, where the target alternative model has the same structure as the target prediction model; Test the target alternative model according to the second test data to obtain fidelity information, where the fidelity information represents the credibility of the test weight information.

10. An information display device, characterized in that, The apparatus includes: A feature acquisition module, configured to obtain multiple scene features based on a target scene currently displayed, where each scene feature includes features corresponding to at least two feature dimensions, and the multiple scene features include a target scene feature corresponding to the target scene currently displayed; A prediction module, configured to perform predictions based on the multiple scene features respectively to obtain prediction information corresponding to each scene feature, where the prediction information is used to represent the probability of at least one target event occurring in the scene indicated by the corresponding scene feature; A determination module, configured to, for any one of the at least two feature dimensions, respectively obtain a first feature corresponding to the feature dimension from the multiple scenario features; determine an influence weight of the feature dimension on each target event based on the first feature corresponding to the multiple scenario features and the prediction information; and determine the influence weights of each feature dimension on each target event as weight information. A display module, configured to display a prompt message according to the weight information, where the prompt message includes the influence weights of at least one feature dimension on at least one target event.

11. The device according to claim 10, characterized in that, The feature acquisition module includes: A feature extraction unit, configured to perform feature extraction on target scenario data corresponding to the currently displayed target scenario to obtain the target scenario features. A feature acquisition unit, configured to obtain the multiple scenario features based on the target scenario features, where the distance between every two adjacent scenario features in the multiple scenario features is the same, or the difference between the features corresponding to the same feature dimension in every two adjacent scenario features in the multiple scenario features is the same.

12. A computer device, characterized in that, The computer device includes a processor and a memory, where the memory is used to store at least one segment of computer program, and the at least one segment of computer program is loaded and executed by the processor to perform the information display method according to any one of claims 1 to 9.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store at least one segment of computer program, and the at least one segment of computer program is used to perform the information display method according to any one of claims 1 to 9.

14. A computer program product, characterized in that, The computer program product includes computer program code, the computer program code is stored in a computer-readable storage medium, a processor of a computer device reads the computer program code from the computer-readable storage medium, and the processor executes the computer program code to enable the computer device to perform the information display method according to any one of claims 1 to 9.

Images