A game information processing method and device, computer equipment and storage medium

Abstract

Embodiments of the present application disclose a game information processing method and device, computer equipment and a storage medium. The present application obtains the to-be-processed text and the to-be-processed image in the to-be-processed game question information of a specified game, encodes the to-be-processed text and the to-be-processed image respectively to obtain text encoding information and image encoding information, and further determines target encoding information corresponding to the to-be-processed game question information based on the text encoding information and the image encoding information. The target probability information of the target encoding information is determined to belong to different processing personnel, and the target processing personnel who needs to process the game question information text is determined from the multiple processing personnel according to the target probability information, so that the processing personnel corresponding to the game question information text can be accurately located, and the processing efficiency of the game question information can be improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to a method, apparatus, computer device, and storage medium for processing game information. Background Technology

[0002] With the increasing popularity of the internet and the development of online games, the game player base is constantly expanding. To ensure game quality and meet player demands, game development teams are growing in size. These teams can utilize collaborative software to work together on game development. This software allows for the recording and management of project requirements and bug reports, enabling team members to easily and clearly access their assigned tasks and follow-up requests.

[0003] Game testing is a crucial part of the game development process. When testers discover bugs in the game, they need to submit the bugs to the relevant personnel in the game development team.

[0004] In related technologies, testers search for relevant bug reports in the historical bug database of multi-user collaborative software by using historical bug report numbers that are similar to the current bug, or by comparing the current bug content with historical bug reports. This allows them to further determine the person responsible for the current bug. However, both methods of searching by historical bug report number and by current bug content require testers to input accurate historical bug report numbers and current bug content. This places high demands on testers' input skills, leading to them spending considerable time submitting bug reports and thus impacting the efficiency of bug handling in the game. Summary of the Invention

[0005] This application provides a method, apparatus, computer device, and storage medium for processing game information, which can improve the efficiency of processing game problem information.

[0006] This application provides a method for processing game information, including:

[0007] Obtain pending game problem information for a specified game, wherein the pending game problem information includes pending text and pending image;

[0008] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed.

[0009] The image to be processed is encoded to obtain the image encoding information corresponding to the image to be processed;

[0010] Based on the text encoding information and the image encoding information, determine the target encoding information corresponding to the game problem information to be processed;

[0011] The target encoding information is determined to be categorized into target probability information of different processors, wherein the processors are used to process game problem information of the specified game;

[0012] Based on the target probability information, the target handler corresponding to the game problem information to be processed is determined from the different handlers.

[0013] Accordingly, embodiments of this application also provide a game information processing apparatus, including:

[0014] The first acquisition unit is used to acquire unprocessed game problem information for a specified game, wherein the unprocessed game problem information includes unprocessed text and unprocessed images;

[0015] The first processing unit is used to encode the text to be processed to obtain the text encoding information corresponding to the text to be processed.

[0016] The second processing unit is used to encode the image to be processed to obtain image encoding information corresponding to the image to be processed.

[0017] The first determining unit is configured to determine the target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information;

[0018] The second determining unit is used to determine the target encoding information as target probability information of different processors, wherein the processors are used to process the game problem information of the specified game;

[0019] The third determining unit is used to determine the target processing personnel corresponding to the game problem information to be processed from the different processing personnel based on the target probability information.

[0020] Accordingly, this application also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes any of the game information processing methods provided in this application.

[0021] Accordingly, this application also provides a storage medium that stores multiple instructions, which are adapted for a processor to load in order to execute the above-described method for processing game information.

[0022] This application embodiment obtains the text and image to be processed from the game problem information of a specified game, encodes the text and image to be processed respectively to obtain text encoding information and image encoding information, and further determines the target encoding information corresponding to the game problem information to be processed based on the text encoding information and image encoding information; determines the target probability information of the target encoding information to be classified into different processing personnel, and determines the target processing personnel who need to process the game problem information text from multiple processing personnel according to the target probability information, thereby accurately locating the processing personnel corresponding to the game problem information text and improving the processing efficiency of game problem information. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a method for processing game information provided in an embodiment of this application.

[0025] Figure 2 This is a flowchart illustrating another method for processing game information provided in an embodiment of this application.

[0026] Figure 3 This is a structural block diagram of a game information processing device provided in an embodiment of this application.

[0027] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] This application provides a method, apparatus, storage medium, and computer device for processing game information. Specifically, the method for processing game information in this application can be executed by a computer device, which can be a terminal or a server. The terminal can be a smartphone, tablet computer, laptop computer, touch screen, personal computer (PC), personal digital assistant (PDA), or other terminal device. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0030] For example, the computer device can be a server that can acquire unprocessed game problem information for a specified game, including unprocessed text and unprocessed images; encode the unprocessed text to obtain text encoding information corresponding to the unprocessed text; encode the unprocessed images to obtain image encoding information corresponding to the unprocessed images; determine target encoding information corresponding to the unprocessed game problem information based on the text encoding information and image encoding information; determine target probability information for classifying the target encoding information into different processing personnel, wherein the processing personnel are used to process the game problem information of the specified game; and determine the target processing personnel corresponding to the unprocessed game problem information from among the different processing personnel based on the target probability information.

[0031] Based on the above problems, this application provides a first method, apparatus, computer device, and storage medium for processing game information, which can improve the processing efficiency of game problem information.

[0032] The following sections provide detailed descriptions of each example. It should be noted that the order in which the embodiments are described is not intended to limit the preferred order of the embodiments.

[0033] This application provides a method for processing game information, which can be executed by a terminal or a server. This application example illustrates the method for processing game information by a server.

[0034] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for processing game information provided in an embodiment of this application. The specific flow of this game information processing method can be as follows:

[0035] 101. Obtain pending game issue information for a specified game.

[0036] In this embodiment, game problem information refers to game bug information, which can be game bugs detected by game testers during testing of a specified game. Game problem information can include at least the game bug content, trace information, and image information. Specifically, trace information refers to error messages from game testing, including the erroneous game code and the reported problem; image information refers to screenshots of the game bug during testing.

[0037] Specifically, "game issue information to be processed" refers to game issues in a specific game that require handling. This information may include text and images to be processed. The text and images can be used to find the appropriate personnel to handle the game issue, allowing them to fix bugs based on the provided information.

[0038] Among them, "text to be processed" refers to the text content of the game problem information to be processed, such as game bug content, trace information, etc. in the game problem information to be processed; "image to be processed" refers to the image content of the game problem information to be processed, such as the game bug screen in the game problem information to be processed.

[0039] In some embodiments, the text and image to be processed can be input by a user (such as a tester). Specifically, the user can input the text and image to be processed through a query client. The query client can send a query request to the server based on the text and image to be processed, so that after receiving the query request, the server can query the corresponding handler for the game problem information from all handlers of the specified game based on the text and image to be processed.

[0040] For example, if a game tester detects a bug 1 during testing of a specified game, the tester can query the client to input the content of the bug 1 (such as the bug title) and trace information as text to be processed, and input the game bug screen as an image to be processed, thereby triggering a query for the person responsible for handling the bug 1.

[0041] In some embodiments, in order to accurately identify the personnel responsible for handling pending game issues, the step "obtaining pending game issue information for a specified game" may include the following operations:

[0042] Obtain the user-inputted content of the game issue to be processed and the game issue screen;

[0043] Retrieve candidate personnel from the module leader information who correspond to the game issue content;

[0044] Based on the game problem content, game problem screen, and candidate personnel, information on the game problem to be processed is obtained.

[0045] Among them, the game issues to be processed are the game bugs and trace information detected by game testers when testing a specified game; the game bug screens to be processed are the game bug screens corresponding to the detected game bugs.

[0046] In this embodiment of the application, for a specific game with a large amount of game content, the specific game can be divided into multiple modules, and at least one person from the project team can be assigned to each module. The module manager information includes the personnel responsible for each game module of the specific game.

[0047] For example, a designated game may include game module 1, game module 2, game module 3, etc. The personnel handling the game within the project team may include: personnel A, personnel B, personnel C, personnel D, etc. Game module 1 can be assigned to personnel A, game module 2 can be assigned to personnel B and personnel C, and game module 3 can be assigned to personnel D.

[0048] Specifically, the selection of candidate personnel for handling game issues can be based on the game bug content and trace information to determine the corresponding target game module, and then the personnel responsible for that target game module can be selected as candidate personnel for handling the game issue. Alternatively, the corresponding personnel can be retrieved from the module manager information by using keyword searches based on the game bug content and trace information, and these personnel can be selected as candidate personnel for handling the game issue.

[0049] Then, based on the game problem content, the game problem screenshots, and the candidate personnel for handling it, the information on the game problem to be processed is obtained, that is, the candidate personnel for handling it is added to the information on the game problem to be processed. The text to be processed can include the game bug content, trace information, and candidate personnel for handling it.

[0050] 102. Encode the text to be processed to obtain the text encoding information corresponding to the text to be processed.

[0051] Specifically, encoding the text to be processed refers to extracting features from the text to obtain its text features, which are then used as a digital representation of the text to obtain the corresponding text encoding information.

[0052] Since the text to be processed contains multiple types of text, in order to improve encoding efficiency, this scheme encodes different texts in the text to be processed using multiple encoding methods. This allows us to obtain the encoding information of different texts in the text to be processed under different encoding methods, thereby obtaining the text encoding information corresponding to the text to be processed.

[0053] In some embodiments, the game problem information to be processed may at least include game problem content, and the text to be processed includes the first text to be processed corresponding to the game problem content. For example, the game problem content may be game bug content and trace information, then the first text to be processed may be the text corresponding to the game bug content and trace information. In this case, the step "encode the text to be processed to obtain the text encoding information corresponding to the text to be processed" may include the following operations:

[0054] Input the first text to be processed into the trained text encoding model;

[0055] Based on the trained text encoding model, the keywords in the first text to be processed are encoded into word vectors to obtain the encoded word vectors corresponding to each keyword.

[0056] The encoded word vectors are weighted to determine the sentence vector corresponding to the first text to be processed, thus obtaining the text encoding information.

[0057] The trained text encoding model can be used to encode the first text to be processed corresponding to the game problem content, obtaining the first text encoding information corresponding to the game problem content. Specifically, the first text to be processed is input into the trained text encoding model, which performs word vector encoding on the text keywords in the first text to be processed, obtaining the encoded word vectors corresponding to each keyword. Then, based on the weighted processing result of each encoded word vector, the sentence vector corresponding to the first text to be processed is obtained, which can be used as the first text encoding information corresponding to the first text to be processed.

[0058] In this embodiment of the application, the trained text encoding model can be a word2vec model. Word2vec is a model that learns semantic knowledge from a large amount of text corpus in an unsupervised manner. Specifically, it learns the semantic information of words by using word vectors to represent the semantic information of words through an embedding space, that is, it makes semantically similar words very close in the space.

[0059] Specifically, word2vec models can include various methods, such as skip-gram or CBOW. Both CBOW and skip-gram are implementations in word2vec used to represent text as vectors. CBOW predicts the center word using surrounding words; during training, it learns information about surrounding words (embedding) from the output loss, but the intermediate layers show average predictions, resulting in a total of V predictions. Skip-gram predicts surrounding words using the center word, with K words as output for each center word, resulting in K predictions for a single word. This allows for more effective learning of information from the context, with a total of K*V predictions. Therefore, in this embodiment, skip-gram or CBOW can be used as a text encoding model to encode the first text to be processed.

[0060] In some embodiments, before inputting the first text to be processed into the trained text encoding model, keywords can be extracted from the first text to be processed to obtain words related to the specified game in the first text to be processed as keywords to be encoded. Then, the keywords in the first text to be processed are encoded to ensure the accuracy of the encoding.

[0061] Specifically, extracting keywords from the first text to be processed can include the following operations:

[0062] The first text to be processed is segmented into words to determine the vocabulary set of the first text to be processed;

[0063] The keywords are obtained by filtering out words related to a specific game from the vocabulary set.

[0064] The word segmentation of the first text to be processed can be based on a dictionary, which includes multiple words.

[0065] In this embodiment of the application, since the text related to a specific game is being segmented, and there are many game-specific terms in the specified game, in order to improve the accuracy of text segmentation, the dictionary can be expanded based on the vocabulary in the specified game, so that the expanded dictionary can include words related to the content of the specified game.

[0066] Specifically, a word segmentation algorithm can be used to segment the first text to be processed based on a dictionary. Word segmentation algorithms can include various types; for example, the jieba word segmentation algorithm can be used.

[0067] Jieba is an open-source Chinese word segmentation package. Essentially, the Jieba algorithm performs statistical segmentation on words in a dictionary. For words not found in the dictionary, it uses a pre-trained Hidden Markov Model (HMM) with initial states, transition probabilities, and observation probabilities to determine if it is a word. The Jieba algorithm includes new word recognition capabilities, and allows developers to add new words themselves, ensuring higher segmentation accuracy.

[0068] Specifically, the word segmentation process of the first text to be processed using the jieba word segmentation algorithm can include: dividing the first text to be processed into multiple words according to the dictionary to obtain the word set corresponding to the first text to be processed; and then selecting words related to the specified game from the word set to obtain the text keywords.

[0069] For example, the first text to be processed is segmented into words, and the keywords obtained may include: keyword 1, keyword 2, keyword 3, keyword 4, and keyword 5. Further, each keyword is encoded into a word vector using a trained text encoding model, resulting in encoded word vectors for each keyword. The encoded word vectors for each keyword have the same length. Then, the encoded word vectors of each keyword in the first text to be processed are weighted to obtain the sentence vector of the first text to be processed, which serves as the first text encoding information.

[0070] In some embodiments, to further improve the accuracy of encoding the first text to be processed, the following steps may be included before the step of "inputting the first text to be processed into the trained text encoding model":

[0071] Collect the text corresponding to the content of historical game questions from the historical game question information of a specified game;

[0072] Based on the text of each text type in the historical game problem content, the preset text encoding model corresponding to each text type is trained to obtain the trained text encoding model corresponding to each text type.

[0073] Historical game issue information refers to game bug information detected during testing of a specified game within a historical time period. This information includes historical game issue content, i.e., historical game bug content, and historical trace information. Specifically, the text corresponding to the historical game issue content is the same as the text corresponding to the historical game bug content and historical trace information.

[0074] In this embodiment of the application, the historical game problem information may include multiple historical game problem information. The historical game problem content in the multiple historical game problem information is used as training samples to train the preset text encoding model, so that the trained text encoding model can be obtained.

[0075] Since the first text to be processed can be the text corresponding to game bug content and trace information, and the game bug content and trace information belong to different types, the first text to be processed includes text of multiple text types.

[0076] For example, game bug information is mostly in Chinese, such as "Gardener Ghost Princess skin's handheld item is abnormally transparent," which is a Chinese type; trace information is mostly in English, such as "'set' object has no attribute 'items'," which is an English type. In other words, the first text to be processed can include at least both Chinese and English text.

[0077] In this embodiment of the application, considering different text types, corresponding text encoding models can be trained separately for different text types.

[0078] Specifically, based on the text of each text type in the historical game problem content, a pre-set text encoding model corresponding to each text type is trained, resulting in a trained text encoding model for each text type, which may include:

[0079] Based on the historical game bug information in the historical game problem content, a preset word2vec model is trained to obtain a trained text encoding model corresponding to the Chinese type, which can be used to encode the Chinese text in the first text to be processed; and based on the historical trace information in the historical game problem content, a preset word2vec model is trained to obtain a trained text encoding model corresponding to the English type, which can be used to encode the English text in the first text to be processed.

[0080] In some embodiments, the step "inputting the first text to be processed into the trained text encoding model" may include the following operations:

[0081] Input the text of each text type in the first text to be processed into the trained text model corresponding to each text type.

[0082] For example, the Chinese text in the first text to be processed (such as the bug title) could be "Gardener Ghost Princess Skin Handheld Item Transparent Abnormal". Word segmentation of the Chinese text yields keywords including: 'gardener', 'Ghost Princess', 'skin', 'handheld item', 'transparent', and 'abnormal'. Then, using a text encoding model trained on the Chinese text type, word vectors are encoded for each keyword, resulting in encoded word vectors for each keyword. For example, the encoded word vector for 'gardener' could be: [0.96, 0.25, 0.41, 1.02, 0.11, -0.73, 0.05]. Finally, weighted processing is applied to each encoded word vector to obtain the sentence vector corresponding to the Chinese text, which serves as the encoding information for the Chinese text.

[0083] For example, the English text (such as the error code) in the first text to be processed can be "AttributeError: 'set' object has no attribute 'items'". The English text is segmented to obtain the following keywords: 'AttributeError', 'set', 'object', 'has', 'no', 'attribute', 'items'. Then, using a text encoding model trained according to the English type, each English keyword is encoded into a word vector, resulting in an encoded word vector for each keyword. For example, the encoded word vector for 'AttributeError' could be [0.13, 0.15, 0.17, -1.02, 0.11, -0.73, 0.07]. Finally, the encoded word vectors are weighted to obtain the sentence vector corresponding to the English text, which serves as the encoding information for the English text.

[0084] In some embodiments, the game problem information to be processed may include at least candidate handlers, and the text to be processed includes a second text to be processed corresponding to the candidate handler. The step "encoding the text to be processed to obtain text encoding information corresponding to the text to be processed" may include the following operations:

[0085] Perform a single-bit valid encoding on the second text to be processed to determine the encoding information corresponding to the second text to be processed.

[0086] One-hot encoding, also known as one-bit state register, uses an N-bit state register to encode N states. Each state has its own independent register bit, and only one bit is active at any given time. One-hot encoding represents categorical variables as binary vectors. This first requires mapping categorical values ​​to integer values. Then, each integer value is represented as a binary vector, with all integer values ​​except for the integer index being zero, which is marked as 1.

[0087] For example, the list of personnel to be processed includes: personnel A, personnel B, personnel C, personnel D, and personnel E. The candidate personnel to be processed can be personnel C. Then, one-hot encoding is performed on personnel C to obtain the encoding information corresponding to the second text to be processed, which can be [0, 0, 0, 0, 0, 0, 1].

[0088] In some embodiments, there may be multiple candidate processors. In this case, the one-hot encoding of each candidate processor can be weighted to obtain the encoding information corresponding to the second text to be processed. Then, the step "encode the text to be processed to obtain the text encoding information corresponding to the text to be processed" may include the following operations:

[0089] Each candidate processor in the second candidate text is encoded with a single valid bit to obtain the encoding information corresponding to each candidate processor.

[0090] The encoded information corresponding to each candidate processor is weighted to obtain the text encoded information.

[0091] For example, the list of personnel to be processed includes: Person A, Person B, Person C, Person D, and Person E. Candidate personnel to be processed can be Person A and Person B. Person A is one-hot encoded, resulting in the encoding information [0, 1, 0, 0, 0, 0, 0]. Person B is one-hot encoded, resulting in the encoding information [1, 0, 0, 0, 0, 0, 0]. Then, the encodings [0, 1, 0, 0, 0, 0, 0] and [1, 0, 0, 0, 0, 1] are weighted to obtain the weighted encoding [0.5, 0.5, 0, 0, 0, 0, 0]. This weighted encoding information is used as the text encoding information for the second text to be processed.

[0092] 103. Encode the image to be processed to obtain the image encoding information corresponding to the image to be processed.

[0093] Specifically, encoding the image to be processed refers to extracting features from the image to be processed, obtaining the image features of the image to be processed, which serve as the digital representation of the image to be processed, and obtaining the image encoding information corresponding to the image to be processed.

[0094] In some embodiments, the step "encoding the image to be processed to obtain image encoding information corresponding to the image to be processed" may include the following operations:

[0095] The first neural network model is used to extract features from the image to be processed, thereby obtaining the multidimensional image features of the image to be processed.

[0096] Multidimensional image features are flattened to obtain image encoding information.

[0097] The first neural network model can be a CNN (Convolutional Neural Networks) model, which can be used to extract features from images. Specifically, the image to be processed is input into the CNN model, which extracts features of the image in multiple dimensions to obtain the multi-dimensional image features of the image to be processed.

[0098] Furthermore, multidimensional image features can be flattened, for example, by using the flatten layer of the network model to compress multidimensional image features into one-dimensional image features. The flattened image features can then be used as image encoding information.

[0099] 104. Determine the target encoding information corresponding to the game problem information to be processed based on text encoding information and image encoding information.

[0100] The target encoding information refers to the encoding information obtained after encoding the text and images in the game problem information to be processed.

[0101] In some embodiments, the step of "determining the target encoding information corresponding to the game problem information to be processed based on text encoding information and image encoding information" may include the following operations:

[0102] The text encoding information is input into the second neural network for processing to obtain the processed text encoding information.

[0103] The processed text encoding information is concatenated with the image encoding information to obtain the target encoding information.

[0104] The second neural network model can be an ANN (Artificial Neural Network) model, which can be used to optimize text features of encoded text information. Specifically, the encoded text information is input into the ANN model, and the ANN model optimizes the text features in the encoded text information to obtain the processed encoded text information.

[0105] Furthermore, the processed text encoding information is concatenated with the image encoding information. For example, this can be done through the concat layer of the network model to obtain the target encoding information of the game problem information to be processed.

[0106] 105. Determine the target coding information and classify it into the target probability information of different processing personnel.

[0107] The processing personnel are responsible for handling game-related issues for a specific game. These personnel are also members of the project team involved in the game's development. The development process requires collaboration among multiple processing personnel, each responsible for different modules. After completion, game testers test the game to check for bugs. If a bug is detected, the corresponding handler for that bug is identified and assigned to them for modification.

[0108] Specifically, once the target coding information is determined, the probability information of classifying the target coding information into different processing personnel can be obtained by calculating the probability value of classifying the target coding information into different processing personnel through a classification model.

[0109] In this embodiment of the application, in order to accurately predict the processing personnel corresponding to the game problem information to be processed, a multi-stage classification model is designed. The step of "determining the target probability information of the target encoded information being classified into different processing personnel" may include the following operations:

[0110] Input the target encoded information into the trained first-order classification model;

[0111] The probability value of classifying the target encoded information into each person type is calculated based on the first-order classification model after training, and the first probability information is obtained.

[0112] Input the first probability information and the target encoding information into the trained second-order classification model;

[0113] Based on the trained second-order classification model and the first probability information, the probability value of classifying the target encoded information into each processor is calculated to obtain the target probability information.

[0114] In this embodiment, a designated game may include multiple game modules, and different game modules correspond to processing personnel. To facilitate querying the processing personnel of different game modules, the processing personnel can be classified by profession according to the function of the game module to determine the personnel type of each processing personnel. Therefore, in this solution, different processing personnel include processing personnel of different professional types.

[0115] For example, a specified game may include: a game art module, a game design module, and a game programming module. The person handling the game art module could be Person A, the person handling the game design module could be Person B, and the person handling the game programming module could be Person C. Furthermore, based on the function of each game module, the personnel can be professionally categorized: Person A could be categorized as an artist, Person B as a designer, and Person C as a programmer.

[0116] The first-order classification model after training can be used to calculate the probability value of the target encoding type being classified into different personnel types, thus obtaining the first probability information.

[0117] For example, the personnel involved in the production of a specific game can be categorized as artists, designers, or programmers. After training, a first-order classification model can be used to calculate the probability that the target encoded information is classified as an artist (0.2), a designer (0.7), and a programmer (0.1), thus obtaining the first probability information.

[0118] The trained second-order classification model can be used to calculate the probability value of classifying the target encoded information into each processing personnel. Specifically, the target encoded information and the first probability information are input into the trained second-order classification model. The trained second-order classification model calculates the probability value of classifying the target encoded information into each processing personnel based on the first probability information, thus obtaining the target probability information.

[0119] For example, the personnel involved in creating a specific game may include: Person A, Person B, Person C, Person D, etc. The first probability information and the target encoding information are input into the trained second-order classification model. The probability of the target encoding information being classified as Person A is calculated to be 0.3, the probability of being classified as Person B is 0.2, the probability of being classified as Person C is 0.11, and the probability of being classified as Person D is 0.39, thus obtaining the target probability information.

[0120] In some embodiments, to improve the computational accuracy of the trained first-order classification model, the following steps may be included before the step "inputting the target encoding information into the trained first-order classification model":

[0121] Retrieve the historical encoding information corresponding to the historical game issue information of a specified game;

[0122] Obtain the actual personnel type corresponding to the personnel who handled each historical game issue;

[0123] The pre-set first-order classification model is trained based on historical coding information and actual personnel types to obtain the trained first-order classification model.

[0124] The historical encoding information can be obtained by encoding the text and images in the historical game question information separately.

[0125] Among them, the actual personnel type refers to the occupational category of the personnel who actually handled the historical game issue information.

[0126] Specifically, a pre-defined first-order classification model is trained using historical coding information corresponding to historical game problem information and actual personnel types to obtain a trained first-order classification model. This allows the trained first-order classification model to output the probability value of the coding information of the input game problem information under each personnel type.

[0127] In some embodiments, the step "training a preset first-order classification model based on historical coding information and actual personnel types to obtain a trained first-order classification model" may include the following operations:

[0128] The first training sample pair is generated based on historical coding information and actual personnel types;

[0129] The first training sample pair is used to train the preset first-order classification model to obtain the first predicted probability value of the historical encoded information in the first training sample pair being classified into each personnel type.

[0130] Obtain the first actual probability value of classifying the historical encoded information of the first training sample pair into each processing personnel;

[0131] The model parameters of the preset first-order classification model are adjusted by using the first predicted probability value and the first actual probability value until the preset first-order classification model converges, thus obtaining the trained first-order classification model.

[0132] In some embodiments, to improve the computational accuracy of the trained second-order classification model, the following steps may be included before the step "inputting the first probability information and target encoding information into the trained second-order classification model":

[0133] The pre-set second-order classification model is trained based on historical coding information, actual processing personnel, and actual personnel types to obtain the trained second-order classification model.

[0134] Specifically, the pre-set second-order classification model is trained by using historical coding information corresponding to historical problem information, actual personnel type, and actual processing personnel to obtain a trained second-order classification model. This allows the trained second-order classification model to output the probability value of the coded information under each processing personnel based on the coded information and first probability information of the input game problem information.

[0135] In some embodiments, the step "training a preset second-order classification model based on historical coding information, actual processing personnel, and actual personnel type to obtain a trained second-order classification model" may include the following operations:

[0136] A second training sample pair is generated based on historical coding information, actual personnel types, and actual processing personnel.

[0137] The preset second-order classification model is trained based on the second training sample pair to obtain the second predicted probability value of the historical encoded information in the second training sample pair being classified into each processing personnel.

[0138] Obtain the second actual probability value of classifying the historical encoded information in the second training sample pair for each processor;

[0139] The model parameters of the preset second-order classification model are adjusted by using the second predicted probability value and the second actual probability value until the preset second-order classification model converges, thus obtaining the trained second-order classification model.

[0140] 106. Based on the target probability information, determine the target personnel corresponding to the game problem information to be processed from different personnel.

[0141] The target probability information includes the probability value of classifying the target encoding information into each processor. The target processor is the predicted personnel used to process the game problem information.

[0142] In some embodiments, the step of "determining the target handler corresponding to the game problem information to be processed from different handlers based on target probability information" may include the following operations:

[0143] Determine the highest probability value from the target probability information;

[0144] The target personnel are determined based on the personnel corresponding to the highest probability value.

[0145] For example, the target probability information includes: the probability of the target code information being classified as person A is 0.3, the probability of being classified as person B is 0.2, the probability of being classified as person C is 0.11, and the probability of being classified as person D is 0.39. Selecting the highest probability value of 0.39, person D can be chosen as the target person for processing.

[0146] Furthermore, the target personnel are provided to the user as query results, allowing the user to assign the game issue information to that target personnel.

[0147] This application discloses a method for processing game information. The method includes: acquiring game problem information to be processed for a specified game, the game problem information to be processed including text to be processed and image to be processed; encoding the text to be processed to obtain text encoding information corresponding to the text to be processed; encoding the image to be processed to obtain image encoding information corresponding to the image to be processed; determining target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information; determining target probability information for classifying the target encoding information into different processing personnel, wherein the processing personnel are used to process the game problem information of the specified game; and determining the target processing personnel corresponding to the game problem information to be processed from different processing personnel based on the target probability information, thereby improving the processing efficiency of game problem information.

[0148] Based on the above description, the following examples will further illustrate the game information processing method of this application. Please refer to... Figure 2 , Figure 2 The following is a flowchart illustrating another method for processing game information provided in this application embodiment. Taking the application of this game information processing method to a terminal as an example, the specific process can be as follows:

[0149] 201. The terminal obtains game problem information input by the user.

[0150] In this embodiment, game problem information refers to content related to game bugs detected by testers during testing of a specified game. Game problem information may include: bug title, trace information, and game bug images.

[0151] 202. The terminal retrieves information on candidate personnel for handling game-related issues based on the game problem information.

[0152] First, obtain the personnel assignment table for the specified game. The personnel assignment table records the game modules that each person is responsible for. For example, the personnel assignment table may include: "Priest - Person A", "Mechanic - Person B", "Gardener - Person C", etc., which means that the "Priest" module in the specified game is the responsibility of Person A, the "Mechanic" module is the responsibility of Person B, and the "Gardener" module is the responsibility of Person C.

[0153] Specifically, based on the game problem information, a keyword search method can be used, that is, to search for whether keywords such as "Priest," "Mechanic," and "Gardener," which correspond to the game modules, appear in the game problem information.

[0154] For example, the bug title in the game issue information is: "Gardener Ghost Princess skin handheld item transparency abnormality", which contains the keyword "gardener". Then, the person in charge of "gardener" is found in the personnel assignment table: Personnel C, as a candidate to handle the issue.

[0155] 203. The terminal performs text encoding on the text content and candidate personnel information in the game problem information to obtain text-encoded information.

[0156] The text content in the game issue information includes the bug title and trace information.

[0157] In this embodiment of the application, the text content in the game problem information can be encoded using word vectors through a trained word2vec model, and the candidate processing personnel information can be encoded using one-hot encoding.

[0158] To avoid different bug titles for different games, the word2vec model can set the length of the output word vectors. This ensures that word vector encoding using the trained word2vec model guarantees consistent encoding length.

[0159] For example, the bug title could be "Gardener Ghost Princess skin's handheld item is abnormally transparent". Word segmentation of "Gardener Ghost Princess skin's handheld item is abnormally transparent" yields multiple keywords including: 'gardener', 'Ghost Princess', 'skin', 'handheld item', 'transparent', and 'abnormal'. Then, word vector encoding is performed using a word2vec model trained on Chinese text to obtain encoded word vectors for each keyword. Weighted word vectors are then used to generate sentence vectors, thus obtaining the encoded information corresponding to the bug title.

[0160] For example, the trace information could be "AttributeError("'set'object has no attribute'items'")". Segmenting "AttributeError("'set'object has no attribute'items'")" yields multiple keywords including 'AttributeError', 'set', 'object', 'has', 'no', 'attribute', and 'items'. Then, a word vector encoding is performed using a word2vec model trained on English text, resulting in encoded word vectors for each keyword. Weighted word vectors are then used to generate sentence vectors, thus obtaining the encoded information corresponding to the trace information.

[0161] For example, candidate information may include one candidate, such as employee C. Employee C is one-hot encoded according to the list of personnel being processed to obtain the corresponding encoding information for the candidate. Alternatively, candidate information may include multiple candidates, such as employee A and employee B. Employee A and employee B are one-hot encoded separately according to the list of personnel being processed to obtain the encoding for employee A and the encoding for employee B. Then, the encoding for employee A and the encoding for employee B are weighted to obtain the corresponding encoding information for the candidate.

[0162] Specifically, the encoding information corresponding to the bug title, the trace information, and the candidate information will be used as text encoding information.

[0163] 204. The terminal performs image encoding on the image content in the game problem information to obtain image encoding information.

[0164] The images in the game issue information include images related to game bugs.

[0165] In this embodiment, game bug images can be encoded using a CNN model. Specifically, the game bug image is input into the CNN model, and the CNN model extracts image features as a digital representation of the game bug image, thus obtaining the image encoding information.

[0166] 205. The terminal concatenates the text encoding information with the image encoding information to obtain the target encoding information.

[0167] Specifically, text encoding information and image encoding information can be concatenated, for example, through the concat layer of a network model, to serve as the target encoding information for game problem information.

[0168] 206. The terminal inputs the target encoding information into the trained first-order classification model and calculates the first probability information of classifying the target encoding information into different personnel types.

[0169] In this embodiment, a trained first-order classification model is used to calculate the probability values ​​of the target encoded information under different personnel types. Since the word vector length is set in the above word vector encoding mode, the input to the first-order classification model is of fixed length. The output of the first-order classification model is a softmax three-class classification mode, with the three categories representing preset personnel types: programmer, planner, and artist.

[0170] Specifically, the target encoding information is input into the trained first-order classification model. The probability value of the target encoding information under different personnel types is calculated by the trained first-order classification model, and the probabilities of the programmer, planner, and artist are output respectively to obtain the first probability information.

[0171] When training a first-order classification model, labels can be assigned based on the personnel type (i.e., occupational category) of the label handlers. Taking historical game problem information as an example, each historical game problem problem corresponds to an actual handler. The labels are assigned based on the occupational category of the handler. If the actual handler for this historical game problem information is employee C, and employee C's occupational category is game planner, then the labels for the first-order classification model can be 0, 1, 0. These labels represent the one-hot encoding of the game planner. During actual scenario reasoning, the probabilities of the three occupational categories are directly output based on the input of the trained first-order classification model.

[0172] 207. The terminal inputs the target encoding information and the first probability information into the trained second-order classification model to calculate the second probability information of classifying the target encoding information into different processing personnel.

[0173] In this embodiment, a trained second-order classification model is used to calculate the probability values ​​of the target encoded information under different processing personnel. The input to the trained second-order classification model includes the output of the trained first-order classification model and the target encoded information. The output is a softmax multi-class classification model, with multiple categories representing the various processing personnel.

[0174] Specifically, the target encoding information and the first probability information are input into the trained second-order classification model. The probability value of the target encoding information under different processing personnel is calculated through the trained second-order classification model to obtain the second probability information.

[0175] When training a second-order classification model, labels can be assigned to the model based on the personnel handling the issues. Taking historical game problem information as an example, each historical game problem corresponds to an actual personnel handling it. If the actual personnel handling this historical game problem information is employee C, then the label of the second-order classification model can be: [0, 0, 0, 0, ..., 1]. This label is also the one-hot encoding of employee C. During inference in a real-world scenario, the probability of each personnel is directly output based on the input of the trained second-order classification model.

[0176] 208. The terminal determines the target handler who needs to process the game problem information from multiple handlers based on the second probability information.

[0177] Specifically, the highest probability is determined from the second probability information, and the personnel corresponding to the highest probability are selected as the target personnel for processing game problem information.

[0178] This application discloses a method for processing game information. The method includes: a terminal acquiring user-input game question information; retrieving candidate personnel information based on the game question information; text encoding the text content of the game question information and the candidate personnel information to obtain text-encoded information; image encoding the image content of the game question information to obtain image-encoded information; concatenating the text-encoded information and the image-encoded information to obtain target-encoded information; inputting the target-encoded information into a trained first-order classification model to calculate a first probability information for classifying the target-encoded information into different personnel types; inputting the target-encoded information and the first probability information into a trained second-order classification model to calculate a second probability information for classifying the target-encoded information into different personnel; and determining the target personnel from multiple personnel who need to process the game question information based on the second probability information. This improves the efficiency of processing game question information.

[0179] To facilitate better implementation of the game information processing method provided in this application embodiment, this application embodiment also provides a game information processing apparatus based on the above-described game information processing method. The meanings of the terms used are the same as in the game information processing method described above, and specific implementation details can be found in the descriptions within the method embodiments.

[0180] Please see Figure 3 , Figure 3 This application provides a structural block diagram of a game information processing apparatus, which includes:

[0181] The first acquisition unit 301 is used to acquire unprocessed game problem information of a specified game, wherein the unprocessed game problem information includes unprocessed text and unprocessed image;

[0182] The first processing unit 302 is used to encode the text to be processed to obtain the text encoding information corresponding to the text to be processed.

[0183] The second processing unit 303 is used to encode the image to be processed to obtain image encoding information corresponding to the image to be processed.

[0184] The first determining unit 304 is used to determine the target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information;

[0185] The second determining unit 305 is used to determine the target encoding information as target probability information of different processors, wherein the processors are used to process the game problem information of the specified game;

[0186] The third determining unit 306 is used to determine the target processing personnel corresponding to the game problem information to be processed from the different processing personnel based on the target probability information.

[0187] In some embodiments, the first processing unit 302 may include:

[0188] The first input subunit is used to input the first text to be processed into the trained text encoding model;

[0189] The first encoding subunit is used to encode the keywords in the first text to be processed based on the trained text encoding model, so as to obtain the encoded word vectors corresponding to each keyword.

[0190] The first processing subunit is used to perform weighted processing on the encoded word vectors to determine the sentence vector corresponding to the first text to be processed, and to obtain the text encoding information.

[0191] In some embodiments, the device may further include:

[0192] The data acquisition unit is used to acquire the text corresponding to the content of historical game questions in the historical game question information of the specified game;

[0193] The first training unit is used to train the preset text encoding model corresponding to each text type based on the text of each text type in the content of the historical game problem, so as to obtain the trained text encoding model corresponding to each text type.

[0194] In some embodiments, the first input subunit may specifically be used for:

[0195] Input the text of each text type in the first text to be processed into the trained text model corresponding to each text type.

[0196] In some embodiments, the first processing unit 302 may include:

[0197] The second encoding subunit is used to perform a one-bit valid encoding on the second text to be processed, and to determine the encoding information corresponding to the second text to be processed.

[0198] In some embodiments, the first processing unit 302 may include:

[0199] The third encoding subunit is used to encode each candidate processor in the second candidate text with one effective bit to obtain the encoding information corresponding to each candidate processor.

[0200] The second processing subunit is used to perform weighted processing on the encoding information corresponding to each candidate processor to obtain the text encoding information.

[0201] In some embodiments, the second processing unit 303 may include:

[0202] An extraction subunit is used to extract features from the image to be processed using a first neural network model to obtain multidimensional image features of the image to be processed.

[0203] The flattening subunit is used to flatten the multidimensional image features to obtain the image encoding information.

[0204] In some embodiments, the first determining unit 304 may include:

[0205] The second input subunit is used to input the text encoding information into the second neural network for processing to obtain the processed text encoding information.

[0206] The splicing subunit is used to splice the processed text encoding information with the image encoding information to obtain the target encoding information.

[0207] In some embodiments, the second determining unit 305 may include:

[0208] The third input subunit is used to input the target encoded information into the trained first-order classification model;

[0209] The first calculation subunit is used to calculate the probability value of the target encoded information being classified into each person type based on the trained first-order classification model, and to obtain the first probability information. The different processing personnel include processing personnel of different personnel types.

[0210] The fourth input subunit is used to input the first probability information and the target encoding information into the trained second-order classification model;

[0211] The second calculation subunit is used to calculate the probability value of classifying the target encoded information into each processor based on the trained second-order classification model and the first probability information, thereby obtaining the target probability information.

[0212] In some embodiments, the device may further include:

[0213] The second acquisition unit is used to acquire historical encoding information corresponding to the historical game problem information of the specified game.

[0214] The third acquisition unit is used to acquire the actual personnel type corresponding to the actual personnel handling each historical game problem information;

[0215] The second training unit is used to train a preset first-order classification model based on the historical encoding information and the actual personnel type to obtain the trained first-order classification model.

[0216] In some embodiments, the device may further include:

[0217] The third training unit is used to train a preset second-order classification model based on the historical encoding information, the actual processing personnel, and the actual personnel type, so as to obtain the trained second-order classification model.

[0218] In some embodiments, the third determining unit 306 may include:

[0219] The first determining subunit is used to determine the highest probability value from the target probability information;

[0220] The second determining subunit is used to determine the target processing personnel based on the processing personnel corresponding to the highest probability value.

[0221] In some embodiments, the first acquisition unit 301 may include:

[0222] The first acquisition subunit is used to acquire the content of the game problem to be processed and the game problem screen input by the user;

[0223] The third determining subunit is used to retrieve candidate personnel corresponding to the game problem content from the module leader information, wherein the module leader information includes the personnel corresponding to each game module of the specified game.

[0224] The fourth determining subunit is used to obtain the game problem information to be processed based on the game problem content, the game problem screen, and the candidate processing personnel.

[0225] This application discloses a game information processing device. A first acquisition unit 301 acquires unprocessed game problem information for a specified game, including unprocessed text and unprocessed images. A first processing unit 302 encodes the unprocessed text to obtain text encoding information. A second processing unit 303 encodes the unprocessed image to obtain image encoding information. A first determining unit 304 determines target encoding information corresponding to the unprocessed game problem information based on the text encoding information and the image encoding information. A second determining unit 305 determines the target encoding information as target probability information for different processing personnel, wherein the processing personnel are used to process the game problem information of the specified game. A third determining unit 306 determines the target processing personnel corresponding to the unprocessed game problem information from among the different processing personnel based on the target probability information. This improves the processing efficiency of game problem information.

[0226] Accordingly, embodiments of this application also provide a computer device, which can be a terminal. For example... Figure 4 As shown, Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. The computer device 500 includes a processor 501 with one or more processing cores, a memory 502 with one or more computer-readable storage media, and a computer program stored on the memory 502 and executable on the processor. The processor 501 and the memory 502 are electrically connected. Those skilled in the art will understand that the computer device structure shown in the figure does not constitute a limitation on the computer device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0227] The processor 501 is the control center of the computer device 500. It connects various parts of the computer device 500 through various interfaces and lines. By running or loading software programs and / or modules stored in the memory 502, and calling data stored in the memory 502, it performs various functions of the computer device 500 and processes data, thereby monitoring the computer device 500 as a whole.

[0228] In this embodiment, the processor 501 in the computer device 500 loads the instructions corresponding to the processes of one or more applications into the memory 502 according to the following steps, and the processor 501 runs the applications stored in the memory 502 to achieve various functions:

[0229] Retrieve pending game issue information for a specified game, which includes pending text and pending images;

[0230] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed.

[0231] The image to be processed is encoded to obtain the image encoding information corresponding to the image to be processed;

[0232] The target encoding information corresponding to the game problem information to be processed is determined based on text encoding information and image encoding information;

[0233] The target coding information is classified into target probability information for different processing personnel, where the processing personnel are used to process game problem information for a specific game;

[0234] Based on the target probability information, the target personnel corresponding to the game problem information to be processed are determined from different personnel.

[0235] In some embodiments, the game problem information to be processed includes at least the game problem content, and the text to be processed includes the first text to be processed corresponding to the game problem content;

[0236] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including:

[0237] Input the first text to be processed into the trained text encoding model;

[0238] Based on the trained text encoding model, the keywords in the first text to be processed are encoded into word vectors to obtain the encoded word vectors corresponding to each keyword.

[0239] The encoded word vectors are weighted to determine the sentence vector corresponding to the first text to be processed, thus obtaining the text encoding information.

[0240] In some embodiments, the first text to be processed includes text of multiple text types;

[0241] Before inputting the first text to be processed into the trained text encoding model, the following steps are also included:

[0242] Collect the text corresponding to the content of historical game questions from the historical game question information of a specified game;

[0243] Based on the text of each text type in the historical game problem content, the preset text encoding model corresponding to each text type is trained to obtain the trained text encoding model corresponding to each text type.

[0244] The first text to be processed is input into the trained text encoding model, including:

[0245] Input the text of each text type in the first text to be processed into the trained text model corresponding to each text type.

[0246] In some embodiments, the game problem information to be processed includes at least candidate handlers, and the text to be processed includes a second text to be processed corresponding to the candidate handlers.

[0247] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including:

[0248] Perform a single-bit valid encoding on the second text to be processed to determine the encoding information corresponding to the second text to be processed.

[0249] In some embodiments, the number of candidate processors is multiple;

[0250] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including...

[0251] Each candidate processor in the second candidate text is encoded with a single valid bit to obtain the encoding information corresponding to each candidate processor.

[0252] The encoded information corresponding to each candidate processor is weighted to obtain the text encoded information.

[0253] In some embodiments, the image to be processed is encoded to obtain image encoding information corresponding to the image to be processed, including:

[0254] The first neural network model is used to extract features from the image to be processed, thereby obtaining the multidimensional image features of the image to be processed.

[0255] Multidimensional image features are flattened to obtain image encoding information.

[0256] In some embodiments, determining the target encoding information corresponding to the game problem information to be processed based on text encoding information and image encoding information includes:

[0257] The text encoding information is input into the second neural network for processing to obtain the processed text encoding information.

[0258] The processed text encoding information is concatenated with the image encoding information to obtain the target encoding information.

[0259] In some embodiments, determining the target encoding information to be categorized into target probability information for different processors includes:

[0260] Input the target encoded information into the trained first-order classification model;

[0261] Based on the first-order classification model after training, the probability value of the target encoded information being classified into each personnel type is calculated to obtain the first probability information. Different processing personnel include processing personnel of different personnel types.

[0262] Input the first probability information and the target encoding information into the trained second-order classification model;

[0263] Based on the trained second-order classification model and the first probability information, the probability value of classifying the target encoded information into each processor is calculated to obtain the target probability information.

[0264] In some embodiments, before inputting the target encoded information into the trained first-order classification model, the method further includes:

[0265] Retrieve the historical encoding information corresponding to the historical game issue information of a specified game;

[0266] Obtain the actual personnel type corresponding to the personnel who handled each historical game issue;

[0267] The pre-set first-order classification model is trained based on historical coding information and actual personnel types to obtain the trained first-order classification model.

[0268] In some embodiments, before inputting the first probability information and the target encoding information into the trained second-order classification model, the method further includes:

[0269] The pre-set second-order classification model is trained based on historical coding information, actual processing personnel, and actual personnel types to obtain the trained second-order classification model.

[0270] In some embodiments, the target probability information includes the probability value of the target encoding information being classified for each processor;

[0271] Based on target probability information, the target personnel corresponding to the game problem information to be processed are determined from different personnel, including:

[0272] Determine the highest probability value from the target probability information;

[0273] The target personnel are determined based on the personnel corresponding to the highest probability value.

[0274] In some embodiments, obtaining pending game issue information for a specified game includes:

[0275] Obtain the user-inputted content of the game issue to be processed and the game issue screen;

[0276] Retrieve candidate personnel corresponding to the game issue from the module leader information, which includes the personnel responsible for handling each game module of the specified game.

[0277] Based on the game problem content, game problem screen, and candidate personnel, information on the game problem to be processed is obtained.

[0278] This application embodiment obtains the text and image to be processed from the game problem information to be processed in a specified game, encodes the text and image to be processed respectively to obtain text encoding information and image encoding information, and further determines the target encoding information corresponding to the game problem information to be processed based on the text encoding information and image encoding information; determines the target probability information of the target encoding information to be categorized into different processing personnel, so as to determine the target processing personnel who need to process the game problem information text from multiple processing personnel according to the target probability information, and accurately locates the processing personnel corresponding to the game problem information text, which can improve the processing efficiency of game problem information.

[0279] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0280] Optional, such as Figure 4As shown, the computer device 500 also includes: a touch screen display 503, a radio frequency circuit 504, an audio circuit 505, an input unit 506, and a power supply 507. The processor 501 is electrically connected to the touch screen display 503, the radio frequency circuit 504, the audio circuit 505, the input unit 506, and the power supply 507. Those skilled in the art will understand that... Figure 4 The computer device structure shown does not constitute a limitation on the computer device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0281] The touch display screen 503 can be used to display a graphical user interface (GUI) and receive operation commands generated by the user interacting with the GUI. The touch display screen 503 may include a display panel and a touch panel. The display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the computer device. These graphical user interfaces can be composed of graphics, guidance information, icons, videos, and any combination thereof. Optionally, the display panel can be configured using a liquid crystal display (LCD), an organic light-emitting diode (OLED), or other similar devices. The touch panel can be used to collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel), generate corresponding operation commands, and execute the corresponding program according to the operation commands. Optionally, the touch panel may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch location and the signal generated by the touch operation, transmitting the signal to the touch controller. The touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to the processor 501. It can also receive and execute commands from the processor 501. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it transmits the information to the processor 501 to determine the game information type of the touch event. Subsequently, the processor 501 provides corresponding visual output on the display panel based on the game information type of the touch event. In this embodiment, the touch panel and the display panel can be integrated into the touch display screen 503 to achieve input and output functions. However, in some embodiments, the touch panel and the touch display screen 503 can be implemented as two independent components to achieve input and output functions. That is, the touch display screen 503 can also be used as part of the input unit 506 to achieve input functions.

[0282] The radio frequency circuit 504 can be used to transmit and receive radio frequency signals to establish wireless communication with network devices or other computer devices, and to transmit and receive signals with network devices or other computer devices.

[0283] Audio circuitry 505 can be used to provide an audio interface between a user and a computer device via a speaker and a microphone. Audio circuitry 505 converts received audio data into electrical signals, transmits them to the speaker, and the speaker converts them into sound signals for output. Conversely, the microphone converts collected sound signals into electrical signals, which are then received by audio circuitry 505, converted back into audio data, and output to processor 501 for processing. The audio data is then transmitted via radio frequency circuitry 504 to, for example, another computer device, or output to memory 502 for further processing. Audio circuitry 505 may also include an earphone jack to facilitate communication between peripheral headphones and the computer device.

[0284] The input unit 506 can be used to receive input numbers, characters, or user characteristic information (such as fingerprints, iris, facial information, etc.), and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control.

[0285] Power supply 507 is used to supply power to various components of computer device 500. Optionally, power supply 507 can be logically connected to processor 501 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. Power supply 507 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0286] although Figure 4 As not shown in the diagram, the computer device 500 may also include a camera, sensor, wireless fidelity module, Bluetooth module, etc., which will not be described in detail here.

[0287] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0288] As can be seen from the above, the computer device provided in this embodiment can acquire game problem information to be processed for a specified game, which includes text to be processed and image to be processed; encode the text to be processed to obtain text encoding information corresponding to the text to be processed; encode the image to be processed to obtain image encoding information corresponding to the image to be processed; determine the target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information; determine the target probability information for classifying the target encoding information into different processing personnel, wherein the processing personnel are used to process the game problem information of the specified game; and determine the target processing personnel corresponding to the game problem information to be processed from different processing personnel based on the target probability information.

[0289] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0290] Therefore, embodiments of this application provide a computer-readable storage medium storing multiple computer programs that can be loaded by a processor to execute steps in any of the game information processing methods provided in embodiments of this application. For example, the computer program can execute the following steps:

[0291] Retrieve pending game issue information for a specified game, which includes pending text and pending images;

[0292] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed.

[0293] The image to be processed is encoded to obtain the image encoding information corresponding to the image to be processed;

[0294] The target encoding information corresponding to the game problem information to be processed is determined based on text encoding information and image encoding information;

[0295] The target coding information is classified into target probability information for different processing personnel, where the processing personnel are used to process game problem information for a specific game;

[0296] Based on the target probability information, the target personnel corresponding to the game problem information to be processed are determined from different personnel.

[0297] In some embodiments, the game problem information to be processed includes at least the game problem content, and the text to be processed includes the first text to be processed corresponding to the game problem content;

[0298] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including:

[0299] Input the first text to be processed into the trained text encoding model;

[0300] Based on the trained text encoding model, the keywords in the first text to be processed are encoded into word vectors to obtain the encoded word vectors corresponding to each keyword.

[0301] The encoded word vectors are weighted to determine the sentence vector corresponding to the first text to be processed, thus obtaining the text encoding information.

[0302] In some embodiments, the first text to be processed includes text of multiple text types;

[0303] Before inputting the first text to be processed into the trained text encoding model, the following steps are also included:

[0304] Collect the text corresponding to the content of historical game questions from the historical game question information of a specified game;

[0305] Based on the text of each text type in the historical game problem content, the preset text encoding model corresponding to each text type is trained to obtain the trained text encoding model corresponding to each text type.

[0306] The first text to be processed is input into the trained text encoding model, including:

[0307] Input the text of each text type in the first text to be processed into the trained text model corresponding to each text type.

[0308] In some embodiments, the game problem information to be processed includes at least candidate handlers, and the text to be processed includes a second text to be processed corresponding to the candidate handlers.

[0309] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including:

[0310] Perform a single-bit valid encoding on the second text to be processed to determine the encoding information corresponding to the second text to be processed.

[0311] In some embodiments, the number of candidate processors is multiple;

[0312] The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed, including...

[0313] Each candidate processor in the second candidate text is encoded with a single valid bit to obtain the encoding information corresponding to each candidate processor.

[0314] The encoded information corresponding to each candidate processor is weighted to obtain the text encoded information.

[0315] In some embodiments, the image to be processed is encoded to obtain image encoding information corresponding to the image to be processed, including:

[0316] The first neural network model is used to extract features from the image to be processed, thereby obtaining the multidimensional image features of the image to be processed.

[0317] Multidimensional image features are flattened to obtain image encoding information.

[0318] In some embodiments, determining the target encoding information corresponding to the game problem information to be processed based on text encoding information and image encoding information includes:

[0319] The text encoding information is input into the second neural network for processing to obtain the processed text encoding information.

[0320] The processed text encoding information is concatenated with the image encoding information to obtain the target encoding information.

[0321] In some embodiments, determining the target encoding information to be categorized into target probability information for different processors includes:

[0322] Input the target encoded information into the trained first-order classification model;

[0323] Based on the first-order classification model after training, the probability value of the target encoded information being classified into each personnel type is calculated to obtain the first probability information. Different processing personnel include processing personnel of different personnel types.

[0324] Input the first probability information and the target encoding information into the trained second-order classification model;

[0325] Based on the trained second-order classification model and the first probability information, the probability value of classifying the target encoded information into each processor is calculated to obtain the target probability information.

[0326] In some embodiments, before inputting the target encoded information into the trained first-order classification model, the method further includes:

[0327] Retrieve the historical encoding information corresponding to the historical game issue information of a specified game;

[0328] Obtain the actual personnel type corresponding to the personnel who handled each historical game issue;

[0329] The pre-set first-order classification model is trained based on historical coding information and actual personnel types to obtain the trained first-order classification model.

[0330] In some embodiments, before inputting the first probability information and the target encoding information into the trained second-order classification model, the method further includes:

[0331] The pre-set second-order classification model is trained based on historical coding information, actual processing personnel, and actual personnel types to obtain the trained second-order classification model.

[0332] In some embodiments, the target probability information includes the probability value of the target encoding information being classified for each processor;

[0333] Based on target probability information, the target personnel corresponding to the game problem information to be processed are determined from different personnel, including:

[0334] Determine the highest probability value from the target probability information;

[0335] The target personnel are determined based on the personnel corresponding to the highest probability value.

[0336] In some embodiments, obtaining pending game issue information for a specified game includes:

[0337] Obtain the user-inputted content of the game issue to be processed and the game issue screen;

[0338] Retrieve candidate personnel corresponding to the game issue from the module leader information, which includes the personnel responsible for handling each game module of the specified game.

[0339] Based on the game problem content, game problem screen, and candidate personnel, information on the game problem to be processed is obtained.

[0340] This application embodiment obtains the text and image to be processed from the game problem information to be processed in a specified game, encodes the text and image to be processed respectively to obtain text encoding information and image encoding information, and further determines the target encoding information corresponding to the game problem information to be processed based on the text encoding information and image encoding information; determines the target probability information of the target encoding information to be categorized into different processing personnel, so as to determine the target processing personnel who need to process the game problem information text from multiple processing personnel according to the target probability information, and accurately locates the processing personnel corresponding to the game problem information text, which can improve the processing efficiency of game problem information.

[0341] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0342] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0343] Since the computer program stored in the storage medium can execute the steps in any of the game information processing methods provided in the embodiments of this application, the beneficial effects that any of the game information processing methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0344] The foregoing has provided a detailed description of a method, apparatus, storage medium, and computer device for processing game information provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for processing game information, characterized in that, The method includes: Obtain the content of the game problem to be processed and the game problem screen input by the user for a specified game; Candidate personnel corresponding to the game problem content are retrieved from the module manager information. The module manager information includes the personnel corresponding to each game module of the specified game. The personnel type of the personnel is determined based on the function of the game module. Based on the game problem content, the game problem screen, and the candidate processing personnel, game problem information to be processed is obtained, which includes text to be processed and image to be processed. The text to be processed is encoded to obtain the text encoding information corresponding to the text to be processed. The image to be processed is encoded to obtain the image encoding information corresponding to the image to be processed; Based on the text encoding information and the image encoding information, determine the target encoding information corresponding to the game problem information to be processed; Based on the first-order classification model after training, the first probability information of classifying the target encoded information into different personnel types is determined; Based on the trained second-order classification model and the first probability information, the target encoding information is determined to be classified into the target probability information of different processors, wherein the different processors include processors of different types, and the processors are used to process the game problem information of the specified game. Based on the target probability information, the target handler corresponding to the game problem information to be processed is determined from the different handlers.

2. The method according to claim 1, characterized in that, The game problem information to be processed includes at least the game problem content, and the text to be processed includes the first text to be processed corresponding to the game problem content; The process of encoding the text to be processed to obtain the text encoding information corresponding to the text to be processed includes: Input the first text to be processed into the trained text encoding model; Based on the trained text encoding model, the keywords in the first text to be processed are encoded into word vectors to obtain the encoded word vectors corresponding to each keyword. The encoded word vectors are weighted to determine the sentence vector corresponding to the first text to be processed, thus obtaining the text encoding information.

3. The method according to claim 2, characterized in that, The first text to be processed includes text of multiple text types; Before inputting the first text to be processed into the trained text encoding model, the method further includes: Collect the text corresponding to the content of historical game questions from the historical game question information of the specified game; Based on the text of each text type in the historical game problem content, the preset text encoding model corresponding to each text type is trained to obtain the trained text encoding model corresponding to each text type. The step of inputting the first text to be processed into the trained text encoding model includes: Input the text of each text type in the first text to be processed into the trained text model corresponding to each text type.

4. The method according to claim 1, characterized in that, The game problem information to be processed includes at least candidate handlers, and the text to be processed includes a second text to be processed corresponding to the candidate handlers; The process of encoding the text to be processed to obtain the text encoding information corresponding to the text to be processed includes: Perform a single-bit valid encoding on the second text to be processed to determine the encoding information corresponding to the second text to be processed.

5. The method according to claim 4, characterized in that, The number of candidate processing personnel is multiple; The process of encoding the text to be processed to obtain the text encoding information corresponding to the text to be processed includes: Each candidate processor in the second candidate text is encoded with a single valid bit to obtain the encoding information corresponding to each candidate processor. The text encoding information is obtained by weighting the encoding information corresponding to each candidate processor.

6. The method according to claim 1, characterized in that, The process of encoding the image to be processed to obtain the image encoding information corresponding to the image to be processed includes: The image to be processed is subjected to feature extraction using a first neural network model to obtain multidimensional image features of the image to be processed. The multidimensional image features are flattened to obtain the image encoding information.

7. The method according to claim 1, characterized in that, The step of determining the target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information includes: The text encoding information is input into a second neural network for processing to obtain processed text encoding information. The processed text encoding information is concatenated with the image encoding information to obtain the target encoding information.

8. The method according to claim 1, characterized in that, The first probability information for determining whether the target encoded information belongs to different personnel types based on the trained first-order classification model includes: The target encoded information is input into the trained first-order classification model; Based on the trained first-order classification model, the probability value of classifying the target encoded information into each of the personnel types is calculated to obtain the first probability information; The step of determining the target encoded information to be classified into target probability information of different processors based on the trained second-order classification model and the first probability information includes: The first probability information and the target encoding information are input into the trained second-order classification model; Based on the trained second-order classification model and the first probability information, the probability value of classifying the target encoded information into each processor is calculated to obtain the target probability information.

9. The method according to claim 8, characterized in that, Before inputting the target encoded information into the trained first-order classification model, the method further includes: Obtain the historical encoding information corresponding to the historical game problem information of the specified game; Obtain the actual personnel type corresponding to the personnel who handled each historical game issue; The preset first-order classification model is trained based on the historical encoding information and the actual personnel type to obtain the trained first-order classification model.

10. The method according to claim 9, characterized in that, Before inputting the first probability information and the target encoding information into the trained second-order classification model, the method further includes: The preset second-order classification model is trained based on the historical encoding information, the actual processing personnel, and the actual personnel type to obtain the trained second-order classification model.

11. The method according to claim 1, characterized in that, The target probability information includes the probability value of classifying the target encoding information for each processing personnel; The step of determining the target handler corresponding to the game problem information to be processed from the different handlers based on the target probability information includes: Determine the highest probability value from the target probability information; Based on the personnel corresponding to the highest probability value, the target personnel are determined.

12. A device for processing game information, characterized in that, The device includes: The first acquisition unit is used to acquire the game problem content and game problem screen input by the user for a specified game; retrieve candidate processing personnel corresponding to the game problem content from the module manager information, the module manager information including the processing personnel corresponding to each game module of the specified game, and the personnel type of the processing personnel is determined based on the function of the game module; and obtain the game problem information to be processed based on the game problem content, the game problem screen, and the candidate processing personnel, the game problem information to be processed including the text to be processed and the image to be processed. The first processing unit is used to encode the text to be processed to obtain the text encoding information corresponding to the text to be processed. The second processing unit is used to encode the image to be processed to obtain image encoding information corresponding to the image to be processed. The first determining unit is configured to determine the target encoding information corresponding to the game problem information to be processed based on the text encoding information and the image encoding information; The second determining unit is used to determine, based on the first-order classification model after training, the first probability information of classifying the target encoded information into different personnel types; and to determine, based on the second-order classification model after training and the first probability information, the target probability information of classifying the target encoded information into different processing personnel, wherein the different processing personnel include processing personnel of the different personnel types, and the processing personnel are used to process the game problem information of the specified game. The third determining unit is used to determine the target processing personnel corresponding to the game problem information to be processed from the different processing personnel based on the target probability information.

13. A computer device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein, When the processor executes the program, it implements the game information processing method as described in any one of claims 1 to 11.

14. A storage medium, characterized in that, The storage medium stores multiple instructions, which are adapted for loading by a processor to execute the game information processing method according to any one of claims 1 to 11.

Images