Game report generation method, device, electronic device and storage medium

Abstract

The present application provides a method, device, electronic device and storage medium for generating a game report, and relates to the field of game technology. The method includes: in response to a request to generate a game report, obtaining game data of a target battle; performing natural language processing of the game data in the game field to obtain an initial game report; inputting the initial game report into a game report model, the game report model analyzes the player's behavior pattern and strategy selection, and annotates the result summary and optimization strategy of the target battle in the initial game report to obtain a game report of the target battle. According to the embodiment of the present application, players can directly obtain the result summary and optimization strategy from the game report, which greatly improves the application value, readability and convenience of the game report, helps players to quickly progress in the game, enhances the user experience, improves the adhesion between users and the game, and reduces the probability of losing game users.

Description

Technical Field

[0001] The present application relates to the field of game technology, and in particular to a method, device, electronic device and storage medium for generating a game report. Background Art

[0002] Game reports are an important basis for gamers to evaluate game performance and further formulate game strategies. At present, the content of game reports is generally the operational data of game battles, which truly records the various operations and game status of players during the battle. However, game reports have poor readability and low application value. They neither analyze game performance nor generate game strategies. It is difficult for players to dig out useful information from the complex information in the game reports to make game decisions, so players cannot make rapid progress in the game, which reduces the fun of the game, affects the user experience, and even leads to the loss of game users. Summary of the invention

[0003] The embodiments of the present application provide a game report generation method, device, electronic device and storage medium to improve the application value and readability of the game report and enhance the user experience.

[0004] In a first aspect, an embodiment of the present application provides a method for generating a game report, comprising:

[0005] In response to a request to generate a game battle report, obtaining game data of a target match;

[0006] Performing natural language processing on the game data to obtain an initial game report;

[0007] The initial game report is input into a game report model, and the game report model analyzes the player's behavior pattern and strategy selection, and annotates the result summary and optimization strategy of the target battle in the initial game report to obtain a game report of the target battle.

[0008] In a second aspect, an embodiment of the present application provides a game report generating device, including:

[0009] A response module, used to respond to a request to generate a game report and obtain game data of a target match;

[0010] A processing module, used for performing natural language processing of the game data in the game field to obtain an initial game battle report;

[0011] A generation module is used to input the initial game report into a game report model, the game report model analyzes the player's behavior pattern and strategy selection, marks the result summary and optimization strategy of the target battle in the initial game report, and obtains the game report of the target battle.

[0012] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor implements any of the above methods when executing the computer program.

[0013] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any of the methods described above is implemented.

[0014] Compared with the prior art, this application has the following advantages:

[0015] By responding to a request to generate a game report, obtaining the game data of the target match, performing natural language processing on the game data in the game field to obtain an initial game report and inputting it into a game report model, the game report model analyzes the player's behavior pattern and strategy selection, and annotates the result summary and optimization strategy of the target match in the initial game report to obtain the game report of the target match, so that players can directly obtain the result summary and optimization strategy from the game report, which greatly improves the application value, readability and convenience of the game report, helps players to make rapid progress in the game, enhances the user experience, improves the adhesion between users and the game, and reduces the probability of game user loss.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments according to the present application and should not be regarded as limiting the scope of the present application.

[0018] Figure 1 A schematic diagram of an application scenario of the game report generation method provided in this application;

[0019] Figure 2 This is a flow chart of a method for generating a game report according to an embodiment of the present application;

[0020] Figure 3 This is a flow chart of a method for generating a game report according to another embodiment of the present application;

[0021] Figure 4 is a flow chart of a model training process according to another embodiment of the present application;

[0022] Figure 5 This is a schematic diagram of a game report generation process according to another embodiment of the present application;

[0023] Figure 6 This is a structural block diagram of a game report generating device according to an embodiment of the present application;

[0024] Figure 7 A block diagram of an electronic device used to implement an embodiment of the present application. DETAILED DESCRIPTION

[0025] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the concept or scope of the present application. Therefore, the drawings and descriptions are considered to be exemplary in nature and not restrictive.

[0026] To facilitate understanding of the technical solutions of the embodiments of the present application, the following describes the related technologies of the embodiments of the present application. The following related technologies can be combined with the technical solutions of the embodiments of the present application as optional solutions, and they all belong to the protection scope of the embodiments of the present application.

[0027] First, the terms involved in this application are explained.

[0028] LLM (Large Language Model): A deep learning-based natural language processing model that is trained using a large-scale corpus, typically has billions to trillions of parameters, and can handle a variety of natural language tasks such as natural language generation, text classification, text summarization, question answering, dialogue, machine translation, or speech recognition.

[0029] SFT (Supervised Fine-Tuning): A model optimization technique that uses a pre-trained model as the initial state and fine-tunes it on the training set of the target task to make the model better suited to the target task. SFT can improve the accuracy and generalization ability of the model and can greatly reduce the training time and computing resource consumption.

[0030] Game battle report: refers to the battle report generated by the battle between players in fighting games, which is used to record the various operations and game status of the players during the battle.

[0031] Figure 1The schematic diagram of an exemplary application scenario for implementing the method for generating a game report of an embodiment of the present application. Among them, the server in the cloud can provide game-related services for one or more clients, including a game report generation service. Users can experience the game on the client and initiate a request to generate a game report on the game interface. The server in the cloud can be deployed as needed, including centralized deployment or distributed deployment. For example, multiple servers can be deployed in the cloud, or one server can be deployed, without specific limitation. The figure only illustrates the deployment of one server as an example. There can be many types of clients, including but not limited to computers, mobile phones, tablet computers or laptops, etc.

[0032] The game report generation method and device provided in the embodiment of the present application can be applied to any game, such as fighting games or chess games, etc. The method and device can be applied in many specific scenarios, including but not limited to: game report analysis or game auxiliary decision-making scenarios.

[0033] The present application embodiment provides a method for generating a game report, which can be applied to a server. Figure 2 Shown is a flow chart of a method for generating a game report according to an embodiment of the present application. The method may include the following steps.

[0034] S201: In response to a request to generate a game report, obtain game data of a target game.

[0035] Among them, the request to generate a game report can be initiated by the client, such as the user triggering the generation entrance of the game report on the game interface of the client, thereby initiating the request to the server.

[0036] The target battle can be the current battle in the game execution process, such as the last battle that just ended; or it can be a historical battle of the game, such as a battle that ended yesterday, etc., which is not specifically limited in the present embodiment of the application. The client can reserve a button or a drop-down menu on the game interface as a request entry for the user to select a desired battle to initiate a request to generate a game battle report.

[0037] S202: Perform natural language processing on the game data in the game field to obtain an initial game battle report.

[0038] The game field in the embodiment of the present application can be understood as a field with a game worldview. The game worldview refers to the cognition of the world where the game story takes place, including but not limited to the natural environment and social environment. The natural environment refers to the mountains, rivers, vegetation and deserts in the game world, as well as the survival and distribution of various common creatures. The social environment includes the political, economic and cultural environment of the game world, as well as the lifestyles, living conditions, beliefs, cluster rules, power systems and camp distribution of various intelligent creatures.

[0039] The above-mentioned natural language processing in the game field refers to processing the game data in accordance with the game world view and in a manner consistent with the game language expression. The initial game report obtained after this processing includes one or more paragraphs of text, and the sentences in each paragraph of text are consistent with the game language expression, and the connotation of the sentence is consistent with the game world view. Those skilled in the art can clearly understand the meaning of the game field expressed by the sentence. This method of natural language processing can enable the game report model to better understand the report text, improve the accuracy of the model, and improve the convenience of the game report.

[0040] S203: Inputting the initial game battle report into the game battle report model, the game battle report model analyzes the player's behavior pattern and strategy selection, marks the result summary and optimization strategy of the target battle in the initial game battle report, and obtains the game battle report of the target battle.

[0041] In the embodiment of the present application, the game report model refers to a model capable of generating game reports, which can be implemented using a natural language processing model. In one embodiment, the game report model can be a large language model that has been fine-tuned using game data for supervised training, and the game report can be generated by the large language model using a prompt word function. In addition, illustratively, the large language model can also be trained using a few-shot learning method for game data during the training process, which is not specifically limited in the embodiment of the present application.

[0042] The above-mentioned optimization strategy may include at least one of the following: tactics, matching of generals, positioning of generals, order of releasing generals' skills, level of upgrade, improving the quality of weapons and equipment or the degree of attribute bonus, etc., and the embodiments of the present application do not make specific limitations on this.

[0043] In one implementation, the above step S201 may include:

[0044] In response to a request to generate a game report of a current match, game data of the current match is obtained as game data of a target match; or, in response to a request to generate a game report of a historical match, game data of a historical match is obtained as game data of the target match.

[0045] In one implementation, the above step S202 may include:

[0046] A pre-generated language template of the game domain is obtained, and the game data is filled into the language template according to the fields and natural language sentence patterns set in the language template to obtain an initial game battle report.

[0047] In one implementation, the above step S203 may include:

[0048] The initial game report is input into the game report model. The game report model analyzes the player's behavior data in the initial game report to obtain the player's behavior pattern, and generates a result summary based on the player's strategy selection, behavior pattern, and the victory or defeat of the target battle; generates the game's optimization strategy based on the strategy selection and result summary; annotates the result summary and optimization strategy in the initial game report to obtain the game report of the target battle.

[0049] In one implementation, the above method may further include:

[0050] Game data of fighting games are collected in advance as corpus, and the corpus is filtered according to the pre-generated language template of the game field to obtain training samples, and the result summary and optimization strategy for the annotation of the training samples are obtained. The training samples and the corresponding annotations are used to train the natural language processing model to obtain a game battle report model with the ability to generate game battle reports.

[0051] In one implementation, filtering the corpus according to the pre-generated language template of the game field to obtain the training sample may include:

[0052] The collected corpus is cleaned to remove invalid words, and the cleaned corpus is filtered according to the pre-generated language template of the game field to select the words that meet the preset fields in the language template, and the words are processed according to the natural language sentence pattern set by the language template to obtain training samples.

[0053] The above method provided in this embodiment obtains the game data of the target battle by responding to the request for generating the game report, performs natural language processing on the game data in the game field to obtain the initial game report and inputs it into the game report model, and the game report model analyzes the player's behavior pattern and strategy selection, and obtains the game report of the target battle after marking the result summary and optimization strategy of the target battle in the initial game report, so that the player can directly obtain the result summary and optimization strategy from the game report, which greatly improves the application value, readability and convenience of the game report, helps the player to make rapid progress in the game, enhances the user experience, improves the adhesion between the user and the game, and reduces the probability of losing game users.

[0054] Another embodiment of the present application provides a method for generating a game report, which can be applied to a server. Figure 3 Shown is a flow chart of a method for generating a game report according to another embodiment of the present application. The method may include the following steps.

[0055] S301: In response to a request to generate a game report, obtain game data of a target game.

[0056] In the embodiment of the present application, the above step S301 may specifically include:

[0057] In response to a request to generate a game report of a current match, game data of the current match is obtained as game data of a target match; or, in response to a request to generate a game report of a historical match, game data of a historical match is obtained as game data of the target match.

[0058] The current battle or the historical battle can be selected by the user of the client, that is, the player. The client can provide different buttons on the game interface, and the user can trigger the battle report generation request of the corresponding battle by selecting the required button. Alternatively, the client can also provide a drop-down menu on the game interface, and the user can trigger the battle report generation request of the corresponding battle by selecting different options in the drop-down menu. The specific implementation method is not limited in the embodiment of the present application.

[0059] S302: Acquire a pre-generated language template for the game domain.

[0060] Among them, the language template in the game field includes at least fields, and further may include sentence patterns. The fields in the language template may be one or more, and may be set in advance as needed. The fields in the language template are all fields that are meaningful for generating game battle reports, that is, fields that are not related to game battle reports can be ignored and not set in the language template. For example, the language template is set to include at least one of the following fields: enemy and friendly identification, tactics, generals, equipment, attribute bonus, strategy selection, or number of enemies killed, etc.

[0061] The sentence pattern in the above language template refers to the grammatical structure of a sentence, which can be set in a way that conforms to the game language expression. For example, the sentence pattern is a "subject + predicate + object" structure, etc. The setting of the sentence pattern can ensure that the meaning of the sentence is clear and unambiguous, thereby ensuring the accuracy of the game report.

[0062] In the embodiment of the present application, the game data includes the game operation and game status of the player. Among them, the game operation includes various operations performed by the player in the battle, and the game status includes various attribute information of the player. For example, the game data includes: how much the attribute is added, which tactics are launched or how much damage is caused, etc. Usually, the game data is relatively large and even disorganized. Therefore, filling it into the language template can obtain a text expression that conforms to the language template fields and sentence patterns, which is convenient for the game battle report model to understand and process.

[0063] Furthermore, the text in the initial game report can be optimized according to the rules of natural language to make it more concise, accurate and easy to understand.

[0064] The initial game report in the embodiment of the present application refers to a text that is compiled based on game data and is in line with the game language expression, which can truly reflect various operations and states in the game, but does not have a battle result analysis and summary and further optimization strategy, so it is called an initial game report. The initial game report is used as the input of the game report model. After being processed by the game report model, a game report with a result summary and optimization strategy can be output.

[0065] S304: Input the initial game report into the game report model. The game report model analyzes the player's behavior data in the initial game report to obtain the player's behavior pattern, and generates a result summary based on the player's strategy selection, behavior pattern, and the victory or defeat of the target battle.

[0066] Among them, the player's behavior pattern can be understood as some operating habits of the user during the game, such as the generals that the user often chooses, the strategies that the user often uses, or the weapons and equipment that the user often uses. The behavior pattern can be used to analyze the impact of the user's operation on the battle results, such as whether the probability of leading to a victory in the battle is high, or the probability of leading to a failure in the battle is high.

[0067] The result summary in the embodiment of the present application refers to the conclusion drawn by analyzing the winning and losing results of the target battle. The user can learn how the target battle performed through the conclusion, such as where the performance was better, where mistakes occurred, etc., so as to have a clear understanding of the target battle.

[0068] In the embodiment of the present application, the game report model can be an LLM trained by SFT using game data. Furthermore, the LLM trained by SFT can also generate game reports using the prompt word function, thereby obtaining game reports that better meet the needs of players, and outputting reasonable report analysis and strategy optimization decisions for players.

[0069] In one implementation, the large language model can be trained using a few-shot learning method for game data during the training process. The few-shot learning method can solve the problem of scarce data samples. By learning a small amount of data samples, the game report model can have a deeper understanding and grasp of the language rules and characteristics of the game field, and thus can generate more accurate game reports.

[0070] S305: Generate an optimization strategy for the game based on the strategy selection and result summary.

[0071] In the embodiment of the present application, the above-mentioned optimization strategy refers to the strategy obtained after optimizing and adjusting the strategy currently used by the player, which can be used as an auxiliary decision-making in the player's subsequent game process, thereby providing a game experience and gameplay that better meets the player's needs, helping the player to make rapid progress and improvement in the game.

[0072] In one implementation, the above optimization strategy may include at least one of the following: tactics, general collocation, general movement, order of general skill release, upgrade level, improving weapon equipment quality or attribute addition, etc., which is not specifically limited in the present application embodiment. For example, the optimization strategy in the game battle report includes: how to better match tactics, insufficient equipment attribute addition, and general collocation problems.

[0073] S306: Annotate the result summary and optimization strategy in the initial game battle report to obtain the game battle report of the target match.

[0074] In the embodiment of the present application, the initial game report includes the game data of the target battle, and after natural language processing, a paragraph or paragraphs of text are obtained. When annotating, the result summary and optimization strategy can be annotated for each paragraph to obtain the game report. For example, the target battle includes 10 rounds, corresponding to 10 paragraphs of text in the initial game report, and the result summary and optimization strategy are annotated for each paragraph of text, that is, the result summary and optimization strategy are annotated for each round, so as to obtain the final game report.

[0075] The above method provided in this embodiment generates a game report so that players can directly obtain result summaries and optimization strategies from the game report, solving the problem that game players cannot understand the report, and greatly improving the application value, readability and convenience of the game report. Providing game report analysis and game decision-making for game players, providing users with a convenient way to analyze the report, helping players to quickly improve in the game, and better improving the player's game level. It not only enhances the user experience and improves the adhesion between users and games, but also reduces the probability of losing game users.

[0076] In any of the above embodiments of the present application, a model training step may be included before using the game report model. Figure 4 Shown is a flowchart of a model training process according to another embodiment of the present application, which may include the following steps.

[0077] S401: Collect game data of battle games in advance as corpus.

[0078] Specifically, game data including players' operation behaviors and strategy selections may be collected.

[0079] S402: Clean the collected corpus and remove invalid words.

[0080] Among them, invalid words can be modifiers or interjections that are meaningless to the game, etc., and the embodiments of the present application do not make specific limitations on this.

[0081] S403: According to the pre-generated language template of the game field, the cleaned corpus is filtered to select the words that meet the preset fields in the language template, and the words are processed according to the natural language sentence pattern set by the language template to obtain training samples.

[0082] Among them, data cleaning can achieve the effect of retaining the words that conform to the preset fields in the language template, and the words that are not related to the preset fields are deemed to be irrelevant to the game battle report and do not need to be retained. In addition, the filtered data can also be normalized as needed to facilitate statistical analysis, etc., which is not specifically limited in the embodiments of the present application.

[0083] S404: Obtain a result summary and optimization strategy for training sample annotation.

[0084] In the embodiment of the present application, in order to improve the accuracy of training, manual annotation can be used, such as the method of summarizing the annotation results and optimizing strategies by senior players. Therefore, the above step S404 can be specifically to obtain the summary of the results and optimization strategies of manual annotation of training samples. For example, the obtained annotations include: "Zhou Yu's tactics selection can be optimized to XXX, and the generals' collocation can be changed to YYY", "The level gap is too large, please increase the level", or "The equipment gap is too large, please improve the equipment quality", etc.

[0085] S405: Using the training samples and the corresponding annotations, a natural language processing model is trained to obtain a game report model capable of generating game reports.

[0086] The training samples and corresponding annotations are used as inputs of the natural language processing model, and the natural language processing model is trained so that the trained model can output game reports.

[0087] The above-mentioned model trained using corpus from the vertical field of games can overcome the defect that LLM cannot fully cover the specific language rules and semantic features in the field of games. The LLM trained with SFT can adapt LLM to specific vertical fields of games and improve its performance in this field. Not only can it more accurately perform semantic analysis on game reports and deeply explore the behavior patterns and strategic choices of game players, but it can also provide battle report analysis output in the game, provide game players with better combat decisions and auxiliary decisions, and thus enhance the player's gaming experience.

[0088] Figure 5 FIG. 1 is a schematic diagram of a game report generation process according to another embodiment of the present application. Figure 5 As shown, the LLM is first trained and a language template is generated during initialization. Among them, the model training can use natural language historical battle reports and game-specific data to perform vertical domain training on the LLM, and then perform SFT training to obtain a large language model SFT-LLM after supervised fine-tuning training. The design of the language template can be based on the characteristics of the game field. First, the content of the battle report is defined, such as defining multiple fields, and then the sentence pattern is defined to obtain the language template. An entrance can be reserved on the game interface. When the game player triggers the entrance to initiate a request to generate a game battle report, the corresponding game data is obtained according to the target battle targeted by the request, and the battle report is processed in natural language. The natural language processing includes filling the acquired game data into the language template and optimizing the expression processing, so that the initial game battle report can be obtained. Then the initial game battle report is input into the trained model SFT-LLM, combined with the prompt word Prompt function for analysis, and the result summary and optimization strategy of the target battle are marked to obtain the game battle report of the target battle, so as to assist players in optimizing game strategies and improving the game experience.

[0089] Corresponding to the application scenario and method of the method provided in the embodiment of the present application, the embodiment of the present application also provides a game report generation device. Figure 6 The figure is a structural block diagram of a game report generating device according to an embodiment of the present application, and the device may include:

[0090] A response module 601 is used to obtain game data of a target game in response to a request to generate a game report;

[0091] Processing module 602, used to perform natural language processing on the game data in the game field to obtain an initial game battle report;

[0092] The generation module 603 is used to input the initial game report into the game report model. The game report model analyzes the player's behavior pattern and strategy selection, marks the result summary and optimization strategy of the target battle in the initial game report, and obtains the game report of the target battle.

[0093] In one implementation, the processing module may be specifically used to: obtain a pre-generated language template for the game domain, fill the game data into the language template according to the fields and natural language sentences set in the language template, and obtain an initial game battle report.

[0094] In one implementation, the generation module can be specifically used to: input the initial game report into the game report model, the game report model analyzes the player's behavior data in the initial game report to obtain the player's behavior pattern, and generate a result summary based on the player's strategy selection, behavior pattern, and the victory or defeat of the target battle; generate an optimized strategy for the game based on the strategy selection and result summary; mark the result summary and optimization strategy in the initial game report to obtain a game report of the target battle.

[0095] In one implementation, the game report model may be a large language model that has been fine-tuned using game data for supervised training, and the game report is generated by the large language model using a prompt word function.

[0096] In one implementation, the large language model may be trained using a few-shot learning approach on game data during the training process.

[0097] In one embodiment, the above device further includes:

[0098] The collection module is used to collect game data of battle games in advance as corpus;

[0099] A filtering module is used to filter the corpus according to a pre-generated language template in the gaming field to obtain training samples;

[0100] The acquisition module is used to obtain the result summary and optimization strategy for the training sample annotation;

[0101] The training module is used to use the training samples and corresponding annotations to train the natural language processing model to obtain a game report model capable of generating game reports.

[0102] In one implementation, the filtering module can be specifically used to: perform data cleaning on the collected corpus, remove invalid words, filter the cleaned corpus according to a pre-generated language template for the gaming field, screen out words that conform to the preset fields in the language template, and process the words according to the natural language sentence patterns set by the language template to obtain training samples.

[0103] In the embodiment of the present application, the optimization strategy may include at least one of the following: tactics, matching of generals, positioning of generals, order of releasing generals' skills, level of upgrade, and improving the quality of weapons and equipment or the degree of attribute bonus.

[0104] In one embodiment, the response module can be specifically used to: respond to a request to generate a game report of a current battle, obtain the game data of the current battle as the game data of the target battle; or, respond to a request to generate a game report of a historical battle, obtain the game data of the historical battle as the game data of the target battle.

[0105] The functions of each module in each device in the embodiments of the present application can be found in the corresponding description in the above method, and have corresponding beneficial effects, which will not be repeated here.

[0106] The above-mentioned device provided in this embodiment obtains the game data of the target battle by responding to the request for generating the game battle report, performs natural language processing on the game data in the game field to obtain the initial game battle report and inputs it into the game battle report model, and the game battle report model analyzes the player's behavior pattern and strategy selection, and obtains the game battle report of the target battle after marking the result summary and optimization strategy of the target battle in the initial game battle report, so that the player can directly obtain the result summary and optimization strategy from the game battle report, which greatly improves the application value, readability and convenience of the game battle report, helps the player to make rapid progress in the game, enhances the user experience, improves the adhesion between the user and the game, and reduces the probability of losing game users.

[0107] Figure 7 is a block diagram of an electronic device used to implement an embodiment of the present application. Figure 7 As shown, the electronic device includes: a memory 710 and a processor 720. The memory 710 stores a computer program that can be run on the processor 720. When the processor 720 executes the computer program, the method in the above embodiment is implemented. The number of the memory 710 and the processor 720 can be one or more.

[0108] The electronic device also includes: a communication interface 730, which is used to communicate with external devices and perform data exchange transmission.

[0109] If the memory 710, the processor 720 and the communication interface 730 are implemented independently, the memory 710, the processor 720 and the communication interface 730 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0110] Optionally, in a specific implementation, if the memory 710, the processor 720 and the communication interface 730 are integrated on a chip, the memory 710, the processor 720 and the communication interface 730 can communicate with each other through an internal interface.

[0111] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which implements the method provided in the embodiment of the present application when the program is executed by a processor.

[0112] An embodiment of the present application also provides a chip, which includes a processor for calling and executing instructions stored in the memory from the memory, so that a communication device equipped with the chip executes the method provided by the embodiment of the present application.

[0113] An embodiment of the present application also provides a chip, including: an input interface, an output interface, a processor and a memory, wherein the input interface, the output interface, the processor and the memory are connected via an internal connection path, and the processor is used to execute the code in the memory. When the code is executed, the processor is used to execute the method provided in the embodiment of the application.

[0114] It should be understood that the above processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0115] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of exemplary but not limiting description, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (DR RAM).

[0116] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0117] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0118] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0119] Any process or method described in the flow chart or otherwise described herein can be understood as a module, fragment or portion of a code representing one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.

[0120] The logic and / or steps described in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with these instruction execution systems, apparatuses or devices.

[0121] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0122] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.

[0123] It should be noted that the embodiments of the present application may involve the use of user data. In actual applications, user-specific personal data can be used in the scheme described in this article within the scope permitted by applicable laws and regulations, subject to the requirements of applicable laws and regulations of the country where the user is located (for example, with the user's explicit consent, effective notification to the user, etc.).

[0124] The above is only an exemplary embodiment of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope recorded in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for generating a game report, characterized in that: The method comprises: In response to a request to generate a game battle report, obtaining game data of a target match; The game data is subjected to natural language processing in the field of game to obtain an initial game report; the natural language processing in the field of game refers to processing the game data in accordance with the game world view and in a manner consistent with the game language expression; the game world view includes the cognition of the natural environment and social environment where the game story takes place, the initial game report includes one or more paragraphs of text, the sentences in each paragraph of the text conform to the manner of the game language expression, and the connotation of the sentences conforms to the game world view; Inputting the initial game battle report into a game battle report model, the game battle report model analyzes the player's behavior pattern and strategy selection, annotates the result summary and optimization strategy of the target battle in the initial game battle report, and obtains the game battle report of the target battle; The game data is processed into a natural language in the game field to obtain an initial game report, including: Get pre-generated game domain language templates; According to the fields and natural language sentences set in the language template, the game data is filled into the language template to obtain an initial game report; The training samples used by the game report model in the training process are obtained by filtering the corpus using a pre-generated language template in the game field.

2. The method according to claim 1, characterized in that The game battle report model analyzes the player's behavior patterns and strategy choices, marks the result summary and optimization strategy of the target battle in the initial game battle report, and obtains the game battle report of the target battle, including: The game report model analyzes the player's behavior data in the initial game report to obtain the player's behavior pattern, and generates a result summary based on the player's strategy selection, the behavior pattern, and the victory or defeat of the target battle; Generating an optimized strategy for the game based on the strategy selection and the result summary; The result summary and optimization strategy are marked in the initial game report to obtain a game report of the target match.

3. The method according to claim 2, characterized in that The game report model is a large language model that is fine-tuned using game data for supervised training, and the game report is generated by the large language model using a prompt word function.

4. The method according to claim 3, characterized in that During the training process, the large language model is trained using a few-shot learning method for the game data.

5. The method according to claim 1, characterized in that Also includes: Collect game data of battle games in advance as corpus; According to a pre-generated language template of the game field, the corpus is filtered to obtain training samples; Obtaining a result summary and optimization strategy for the training sample annotation; Using the training samples and corresponding annotations, a natural language processing model is trained to obtain a game report model capable of generating game reports.

6. The method according to claim 5, characterized in that According to the pre-generated language template of the game field, the corpus is filtered to obtain training samples, including: Clean the collected corpus and remove invalid words; According to the pre-generated language template of the game field, the cleaned corpus is filtered to select the words that meet the preset fields in the language template, and the words are processed according to the natural language sentence pattern set by the language template to obtain training samples.

7. The method according to claim 1, characterized in that The optimization strategy includes at least one of the following: tactics, military commander collocation, military commander positioning, order of military commander skill release, upgrade level, and improvement of weapon equipment quality or attribute bonus.

8. The method according to claim 1, characterized in that In response to a request to generate a game report, game data of the target game is obtained, including: In response to a request to generate a game report of a current match, obtaining game data of the current match as game data of the target match; or, In response to a request to generate a game report of a historical match, game data of the historical match is acquired as game data of the target match.

9. A game report generating device, characterized in that: The device comprises: A response module, used to respond to a request to generate a game report and obtain game data of a target match; a processing module, for performing natural language processing in the field of gaming on the game data to obtain an initial game report; the natural language processing in the field of gaming refers to processing the game data in accordance with the game world view and in a manner consistent with the game language expression; the game world view includes the cognition of the natural environment and social environment where the game story takes place, the initial game report includes one or more paragraphs of text, the sentences in each paragraph of the text conform to the manner of the game language expression, and the connotation of the sentences conforms to the game world view; A generation module, used for inputting the initial game battle report into a game battle report model, wherein the game battle report model analyzes the behavior patterns and strategy selections of the players, marks the result summary and optimization strategy of the target battle in the initial game battle report, and obtains the game battle report of the target battle; The processing module is further used to: obtain a pre-generated language template for the game field; fill the game data into the language template according to the fields and natural language sentences set in the language template to obtain an initial game battle report; The training samples used by the game report model in the training process are obtained by filtering the corpus using a pre-generated language template in the game field.

10. An electronic device comprising a memory, a processor and a computer program stored in the memory, wherein the processor implements the method according to any one of claims 1 to 8 when executing the computer program.

11. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images