Wargame data annotation method, system, storage medium and electronic device

Abstract

The present invention relates to a method, system, storage medium and electronic device for marking war game data, including: obtaining intention description information recorded in war game report data, and labeling based on the intention description information; collecting real-time war game data and voice data, and labeling the voice data based on a label recognition strategy; obtaining the start time and end time of the voice data corresponding to the label, obtaining the start frame and end frame in the real-time war game data corresponding to the voice data based on the voice data between the start time and the end time, and labeling the real-time war game data between the start frame and the end frame as the label; replaying the labeled real-time war game data, receiving the label information input according to the replayed labeled real-time war game data and the label recognition strategy displayed on the replay interface, and re-labeling the labeled real-time war game data. The labeling efficiency and labeling accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of computer tag annotation, and in particular to a method, system, storage medium and electronic device for annotating war game data. Background Art

[0002] Wargame simulation is a means of war simulation. With the development of computer technology, computer wargames generated by computer technology have been used more and more widely for simulation. Computer wargames are composed of wargame elements such as chessboard, chess pieces, rules, and scenarios. Among them, the chessboard is an abstraction of the combat space, including information such as terrain and objects; chess pieces (wargame operators) are abstractions of combat forces, including information such as defense and health; rules are abstractions of combat actions, including information such as line-of-sight rules and concealment rules; scenarios are the settings of combat backgrounds, including information such as the combat intentions of both sides and the conditions for winning or losing. Limited by the line-of-sight rules and concealment rules, computer wargames are a complex decision-making problem with imperfect information, that is, no participant (wargame operator) can obtain all the action information of other participants.

[0003] In computer war games, how to learn the diverse behaviors and strategies of participants from replay data is an important basis for improving the game decision-making ability of computer war games. To learn the diverse behaviors and strategies of war game participants, it is necessary to understand the long-term and short-term action sequences in the game from the different levels of the actions, behaviors, and tactical intentions of the war game participants, and accurately identify the intention labels to take countermeasures. However, since tactical intentions are domain knowledge with an overall concept, statistical learning technology based on computer language can only achieve the understanding and recognition of specific actions in a short period of time, and cannot achieve the understanding of intentions at the tactical level. Therefore, human experts are needed to assist in understanding, that is, to provide learning resources for computers by labeling labels, but manual labeling of computer labels is usually inefficient and labor costs are expensive. Summary of the invention

[0004] In view of this, the present invention provides a method, system, storage medium and electronic device for marking war game data.

[0005] Specifically, the present invention is achieved through the following technical solutions:

[0006] According to a first aspect of the present invention, a wargame data annotation system is provided, the wargame data annotation system comprising:

[0007] The war game simulation client is used to parse and display real-time simulation data to players, send the received action instructions issued by players and / or agents based on the real-time simulation data to the war game simulation engine module, and send the pre-acquired players' simulation battle report data to the simulation battle report collection module;

[0008] The war game report collection module is used to collect and receive the war game report data through the war game client and send it to the label extraction module;

[0009] A label extraction module is used to obtain the intention description information recorded in the deduction battle report data based on a preset label recognition strategy, label the deduction battle report data based on the intention description information, and send the label and the corresponding deduction battle report data to the label database;

[0010] The label database is used to store the mapping relationship between the labels sent by the label extraction module and the corresponding deduction battle report data;

[0011] The war game simulation engine is used to receive action instructions issued by players and / or agents, update the environmental state and battlefield situation of the simulated environment of the war game according to the action instructions, generate real-time simulation data based on the updated simulation environment, and send it to the war game simulation client and the data acquisition module respectively;

[0012] A recorder is used to collect the voice data of players during the war game and send it to the data collection module;

[0013] A data collection module is used to extract the description information of the war game in the voice data, label the voice data based on the description information of the war game and the label recognition strategy, and send the labeled voice data and the real-time deduction data sent by the war game deduction engine to the first labeling module;

[0014] A first labeling module is used to extract a label from the labeled voice data using the labeled voice data, and for each label, obtain a start time and an end time of the voice data corresponding to the label; based on the voice data between the start time and the end time, obtain a start frame and an end frame in the real-time deduction data corresponding to the voice data, label the real-time deduction data between the start frame and the end frame as the label, and send the labeled real-time deduction data to a pre-labeling database;

[0015] Pre-annotated database, used to store annotated real-time deduction data;

[0016] A visualization module is used to receive the real-time deduction data of the annotations sent from the pre-annotation database and display them visually on the annotation terminal;

[0017] The annotation terminal is used to visualize the annotated real-time deduction data, receive the label information input by the annotator based on the visualized real-time deduction data, and send it to the annotation module;

[0018] The second labeling module is used to label the real-time deduction data of the visual display label based on the received label information if it is determined that the received label information is one of the labels stored in the label database.

[0019] The wargame data annotation system in the present technical solution can intuitively present the wargame simulation process through a visualization module, and by adding modules such as a simulation battle report collection module and a first annotation module in which players participate, it can realize label extraction and automatic label annotation functions through interaction with player voice data, thereby effectively improving label annotation efficiency and accuracy.

[0020] According to a second aspect of the present invention, a method for marking wargame data is provided, the method for marking wargame data comprising:

[0021] Obtain the war game's battle report data, obtain the intention description information recorded in the war game's battle report data based on a pre-set label recognition strategy, and label the war game's battle report data based on the intention description information;

[0022] Collecting real-time wargame data and voice data, extracting description information of the wargame in the voice data based on the label recognition strategy, and labeling the voice data based on the description information of the wargame and the label recognition strategy;

[0023] For each label annotated with the voice data, obtain the start time and end time of the voice data corresponding to the label, obtain the start frame and end frame in the real-time deduction data corresponding to the voice data based on the voice data between the start time and the end time, and annotate the real-time deduction data between the start frame and the end frame as the label;

[0024] Replay the annotated real-time deduction data, receive label information input according to the annotated real-time deduction data replayed and the label identification strategy displayed on the playback interface, and re-label the annotated real-time deduction data.

[0025] According to a third aspect of the present invention, there is provided a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for marking wargame data in any possible implementation of the first aspect.

[0026] According to a fourth aspect of the present invention, there is provided an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for marking war game data in any possible implementation of the first aspect are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0029] Figure 1 A schematic diagram of a war game data annotation system provided by an embodiment of the present invention;

[0030] Figure 2 A schematic diagram of a flow chart of a method for marking war game data provided by an embodiment of the present invention;

[0031] Figure 3 A schematic diagram of a user interface provided by an embodiment of the present invention;

[0032] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] In related technologies, statistical learning technology based on computer language can only learn the diverse behaviors and strategies of war game decision makers from replay data, but can only understand and recognize specific actions in a short period of time, and cannot understand the intentions at the tactical level. Since tactical intentions are domain knowledge with an overall concept, human experts are needed to provide learning resources for computers by labeling them, so as to learn the diverse behaviors and strategies of war game decision makers. However, the manual labeling method for computer labels is usually inefficient and labor costs are expensive.

[0035] The game data (replay data) of computer war games is a kind of spatiotemporal data with both temporal and spatial attributes, which is usually stored in a structured form as a replay file. At present, the annotation of game data of computer war games is usually based on pure data tables. For example, the pure data table can be a comma-separated values ​​(CSV) file, or the replay file is made into a video and annotated according to the method and process of behavior video data annotation. Among them, the behavior video data annotation method is to watch the behavior video and annotate the frames and labels of the label. For example, the behavior video data is divided into several short videos, and the short videos are annotated with the annotated labels to obtain the image sequence data corresponding to the label. The image sequence data includes the video frames obtained frame by frame from the short video and the corresponding behavior labels. However, compared with other behavior annotation tasks, the computer war game annotation is more professional and requires more information. For example, for other human behavior labeling tasks, one only needs to watch the behavior video to determine the type of labeled behavior, without having to worry about the age, health status and other information of the behavior subject. However, when labeling war game labels, in addition to watching the game process, it is often necessary to make a comprehensive judgment based on information such as operator health, weapon status, terrain, visibility, etc. The current video playback area does not provide relevant auxiliary analysis functions, making it difficult to present all labeling-related information in a concise and intuitive manner, resulting in low labeling efficiency and low labeling accuracy.

[0036] Furthermore, the semi-automatic behavior labeling results are used to establish a behavior recognition model through data mining or deep learning methods, and then the behavior is automatically labeled using the classification results of the behavior recognition model. The prerequisite is that high-accuracy labeling results are required. However, due to the low accuracy of the current war game data labeling, the constructed behavior recognition model has a low recognition accuracy for labels.

[0037] In this embodiment, based on the war game data for computer war games, a semi-automatic annotation system in the cloud data set is constructed in which players and annotators participate together. A semi-automatic computer war game data annotation method based on 4W1H is proposed to achieve human-computer interactive and visualized semi-automatic label annotation, so as to simplify the annotation process, improve the annotation efficiency, and provide high-quality annotation data for computer war game research.

[0038] In this embodiment, 4W1H refers to When, Who, Where, What, and How, which are used to describe the annotation framework of the war game. Among them, When refers to the start and end time of the label, that is, the start frame and the end frame; Who refers to the participating operator; Where refers to the coordinate sequence of the participating operator in the war game data, that is, the operator trajectory; What refers to the label; How refers to the confrontation situation related to the label, such as the state information and action information of the participating operator.

[0039] See also Figure 1 The embodiment of the present invention provides a wargame data annotation system, which may include the following modules:

[0040] The war game simulation client 11 is used to parse and display real-time simulation data to players, send the received action instructions issued by players and / or agents according to the real-time simulation data to the war game simulation engine 15, and send the pre-acquired players' simulation battle report data to the simulation battle report collection module 12;

[0041] In this embodiment, the war game report data is the analysis data of each player on the war game, including but not limited to: war game response strategy data, attack and defense data at each stage, etc. Real-time game data is real-time data collected during the war game game process.

[0042] A war game report collection module 12 is used to collect and receive war game report data through a war game client and send it to a tag extraction module 13;

[0043] The label extraction module 13 is used to obtain the intention description information recorded in the deduction battle report data based on the preset label recognition strategy, label the deduction battle report data based on the intention description information, and send the label and the corresponding deduction battle report data to the label database 14;

[0044] In this embodiment, as an optional embodiment, the pre-set tag identification strategy can be to pre-set the mapping relationship between each tag and the keyword group, extract the keywords contained in the paragraph data or chapter data in the deduced battle report data, calculate the similarity between the extracted keywords and the keyword group of each tag, select the tag corresponding to the highest similarity as the tag of the paragraph data or chapter data, and mark the corresponding tag.

[0045] In this embodiment, as an optional embodiment, continuous paragraph data or chapter data with the same tag are merged, and the tag is set for the merged data.

[0046] In this embodiment, as another optional embodiment, the pre-set tag identification strategy can be to perform cluster analysis on each received deduction battle report data to obtain multiple cluster centers, each cluster center corresponds to a tag, and extract the keyword group of the cluster center as the keyword group of the corresponding tag. When new deduction battle report data is input later, tagging is performed based on the keyword group of each tag.

[0047] In this embodiment, as an optional embodiment, the tag identification strategy can be stored in a tag database, that is, the tag database stores the mapping relationship between each tag and the keyword group.

[0048] In this embodiment, as an optional embodiment, the tags include but are not limited to: attack, concealment, charge, seizure, reconnaissance, troop transport, and mine laying.

[0049] The tag database 14 is used to store the mapping relationship between the tags sent by the tag extraction module and the corresponding deduction battle report data;

[0050] The war game simulation engine 15 is used to receive action instructions issued by players and / or agents, update the environmental state and battlefield situation of the simulated environment of the war game according to the action instructions, generate real-time simulation data based on the updated simulation environment, and send it to the war game simulation client 11 and the data acquisition module 17 respectively;

[0051] A recorder 16 is used to collect voice data of players during the war game and send it to the data collection module 17;

[0052] The data collection module 17 is used to extract the description information in the voice data, label the voice data based on the description information and the label recognition strategy, and send the labeled voice data and the real-time deduction data sent by the war game deduction engine 15 to the first labeling module 18;

[0053] In this embodiment, as an optional embodiment, the mapping relationship between each tag and the keyword group stored in the tag database can be used to extract the description information (keywords) in the voice data, and the tag of the voice data can be determined based on the similarity between the description information and each tag.

[0054] In this embodiment, as an optional embodiment, the voice data corresponds to one or more tags.

[0055] The first labeling module 18 is used to extract the label in the labeled voice data by using the labeled voice data, and for each label, obtain the start time and end time of the voice data corresponding to the label; based on the voice data between the start time and the end time, obtain the start frame and the end frame in the real-time deduction data corresponding to the voice data, label the real-time deduction data between the start frame and the end frame as the label, and send the labeled real-time deduction data to the pre-labeling database 19;

[0056] A pre-annotated database 19, used to store annotated real-time deduction data;

[0057] In this embodiment, as an optional embodiment, the annotated real-time deduction data includes the starting position and the ending position of the intention label in the real-time deduction data, and the label of the real-time deduction data is annotated based on the 4W1H labeling framework.

[0058] A visualization module 20 is used to receive the annotated real-time deduction data sent from the pre-annotated database and perform visual display on the annotation terminal 21;

[0059] In this embodiment, the visualization module is used to enable the annotator to obtain information such as the health of the displayed operator, weapon status, terrain, visibility, etc., so as to make a comprehensive judgment.

[0060] The annotation terminal 21 is used to visually display the annotated real-time deduction data, receive the label information input by the annotator based on the visually displayed real-time deduction data, and send it to the second annotation module 22;

[0061] The second labeling module 22 is used to label the real-time deduction data of the visual display label based on the received label information if it is determined that the received label information is one of the labels stored in the label database.

[0062] In this embodiment, as an optional embodiment, the annotator makes a comprehensive judgment based on the visual replay of the war game simulation process (annotated real-time simulation data) according to the information such as the operator health, weapon status, terrain, visibility, etc. presented in the replay, so as to perform secondary annotation of the intention label, wherein the secondary annotation includes but is not limited to: viewing completed annotations, modifying annotations, adding annotations, deleting annotations, and saving annotations. For example, for the data that has been annotated with the intention label in the annotated real-time simulation data, the label is modified, deleted, or a new label is added, and for the data that has not been annotated with the intention label in the annotated real-time simulation data, the label is annotated.

[0063] In this embodiment, as an optional embodiment, the system further includes:

[0064] The user interface 23 is used to provide an interface for human-computer interaction with the player in the annotation terminal 21, so that the annotator can perform annotation according to the interface.

[0065] In this embodiment, as an optional embodiment, human-computer interaction includes but is not limited to: displaying a visualized deduction process, using an annotation function, and saving annotations.

[0066] In this embodiment, as another optional embodiment, the system further includes:

[0067] The annotation database 24 is used to store the mapping relationship between the intention label sent by the second annotation module 22 for annotating the annotated real-time deduction data and the corresponding real-time deduction data.

[0068] In this embodiment, as another optional embodiment, the system further includes:

[0069] The cloud service engine 25 is used for the mapping relationships stored in the cloud storage annotation database 24 , the mapping relationships stored in the tag database 14 , and the mapping relationships stored in the pre-annotation database 19 .

[0070] In this embodiment, the data in the database are uniformly stored and managed in the cloud, wherein the database includes but is not limited to: a marked database, a pre-marked database, and a label database.

[0071] In this embodiment, as another optional embodiment, the system further includes:

[0072] The model training module (not shown in the figure) is used to build a war game simulation model based on the labeled simulation battle report data and the re-labeled real-time simulation data using a pre-set neural network algorithm.

[0073] Figure 2 The following is a flow chart of a method for marking war game data provided by an embodiment of the present invention. Figure 2 As shown, the method includes:

[0074] S101, obtaining war game war report data, obtaining intention description information recorded in the war report data based on a preset tag recognition strategy, and labeling the war report data based on the intention description information;

[0075] In this embodiment, the war game simulation client provides a data collection portal for players. Players submit the simulation battle report data including analysis and strategy about the simulation before the war game simulation to the simulation battle report collection module. The simulation battle report collection module sends the received simulation battle report data to the label extraction module.

[0076] In this embodiment, the label extraction module performs semantic analysis on the received war game report data, extracts the descriptive information about the war game in the war game report, and abstracts the label.

[0077] In this embodiment, as an optional embodiment, the intention description information recorded in the deduced battle report data is obtained based on a preset label recognition strategy, and the label of the deduced battle report data is marked based on the intention description information, including:

[0078] Splitting the deduction battle report data into unit data according to the paragraphs or chapters contained in the deduction battle report data;

[0079] For each unit of data, extract keywords from the unit of data;

[0080] According to the preset mapping relationship between each tag and the keyword group, the similarity between the keywords extracted from the unit data and the keyword group of each tag is calculated, the tag corresponding to the highest similarity is selected, and the unit data is marked with the corresponding tag.

[0081] In this embodiment, as an optional embodiment, the method further includes:

[0082] Traverse each unit data in turn. If the label of the current unit data is the same as the label of the previous unit data, merge the current unit data with the previous unit data to obtain the merged unit data, and set the label for the merged unit data.

[0083] In this embodiment, continuous paragraph data or chapter data with the same tag are merged, and the tag is set for the merged data.

[0084] In this embodiment, as another optional embodiment, the intention description information recorded in the deduced battle report data is obtained based on a preset label recognition strategy, and the label of the deduced battle report data is annotated based on the intention description information, including:

[0085] For each deduction battle report data, extract keywords in the deduction battle report data to construct a keyword group of the deduction battle report data;

[0086] Perform cluster analysis based on each keyword group to obtain multiple cluster centers, and set a label for each cluster center;

[0087] Based on the keyword groups contained in the cluster centers, a keyword set of tags is constructed.

[0088] In this embodiment, as an optional embodiment, the method further includes:

[0089] Save the intent tag.

[0090] In this embodiment, the extracted intent tags are saved in a tag database.

[0091] S102, collecting real-time war game simulation data and voice data, extracting description information from the voice data based on the label recognition strategy, and labeling the voice data based on the description information and the label recognition strategy;

[0092] In this embodiment, as an optional embodiment, the game data includes the player's voice data and the real-time game data sent by the war game engine. Among them, the voice data refers to the voice description data of the player's intention to participate in the game recorded by the recorder during the game. If the start and end time and related operators are not specified in the voice, it is assumed that the player is playing the game when the voice is described. It should be noted that the player does not need to describe all the game intentions. If the description will affect the player's own operation, you can choose not to make a voice description.

[0093] S103, for each label annotated with the voice data, obtain the start time and end time of the voice data corresponding to the label, obtain the start frame and end frame in the real-time deduction data corresponding to the voice data based on the voice data between the start time and the end time, and mark the real-time deduction data between the start frame and the end frame as the label;

[0094] In this embodiment, the initial automatic labeling is performed.

[0095] In this embodiment, the first annotation module first processes the received voice data, matches the intention description contained in the voice data with the keyword set in the tag database, and automatically annotates the real-time deduction data corresponding to the voice data according to the annotation framework of 4W1H (When, Who, Where, What and How) if the match is successful. Among them, When refers to the start and end frames of the tag, Who refers to the relevant operator of the tag. If it is not specified in the voice data, it is searched in the deduction data according to the time of recording the voice; Where refers to the operator trajectory, and the coordinate sequence record of the relevant operator is automatically extracted by the first annotation module; What refers to the intention, that is, the intention tag; How refers to the situation, which is automatically extracted from the real-time deduction data by the first annotation module.

[0096] In this embodiment, as an optional embodiment, the pre-labeled data of the label is saved. For example, the pre-labeled data is saved in a pre-labeled database.

[0097] In this embodiment, as an optional embodiment, labels of real-time deduction data are annotated based on the 4W1H annotation framework.

[0098] S104, replaying the annotated real-time deduction data, receiving label information input according to the annotated real-time deduction data replayed and the label identification strategy displayed on the playback interface, and re-labeling the annotated real-time deduction data.

[0099] In this embodiment, after the war game, the game data (annotated real-time deduction data) is visualized. As an optional embodiment, the game data is uploaded to the visualization module and the game process is replayed in the user interface.

[0100] In this embodiment, as an optional embodiment, receiving the label information input according to the annotated real-time deduction data played back and the label identification strategy displayed on the playback interface, and re-labeling the annotated real-time deduction data includes:

[0101] Determine the starting frame, target operator, and ending frame for intention labeling;

[0102] Obtaining the trajectory and situation of the target operator from the real-time deduction data starting from the start frame to the end frame;

[0103] Determine the keywords in the real-time deduction data of the target operator from the start frame to the end frame, determine the intention label based on the keywords and the label recognition strategy displayed in the playback interface, and construct a mapping relationship between the intention label and the start frame, target operator, end frame, trajectory, situation, and keyword.

[0104] Figure 3 Schematic diagram of a user interface provided by an embodiment of the present invention. Figure 3 As shown, the middle of the interface is the playback area for data playback, which plays back the wargame process frame by frame, including displaying the coordinates of each hexagonal grid on the chessboard, the position of the capture point, the position of the operator, the operator number, the operator health and other information. The top of the playback box displays the current frame number, and the bottom of the playback box is the progress bar and the speed selection button. In the upper right corner are the data upload button, the visibility measurement button, the playback perspective switch button and the annotated data save button. Click the game data upload button to select the game data to be annotated twice and upload it to the visualization module; click the visibility measurement button and then click to select any two hexagonal grids on the chessboard to measure the visibility between the two hexagonal grids to assist in analyzing the operator's behavior intention; click the playback perspective switch button to select the red side or blue side participating in the deduction or the God's perspective to present the deduction process. When the God's perspective is selected, the playback area displays all operators of both the red and blue sides; click the annotated data save button to send the annotated data to the cloud for storage. On the left is the force formation display area, which displays all existing red and blue operators in the current frame of playback. Click an operator to select the icon of the operator in the playback area. On the right is the intention label selection area, which displays the intention labels in the label database. You can also add, delete labels, and select labels for intention labeling. Below is the labeling area, which displays all labeled data (including pre-labeled data) and data being labeled. You can add, delete, and save labeled data.

[0105] In this embodiment, after uploading the game data, the game process is replayed by clicking the play button to check whether the pre-annotated data (annotated real-time deduction data) needs to be modified and the unannotated real-time deduction data is labeled to perform secondary intention labeling. As an optional embodiment, the annotator can modify the existing annotation entry by clicking the new button, add an intention label annotation, and delete the annotation by clicking the delete button. After starting the annotation, select the frame and operator, enter the start and end frames (When) and the operator (Who); the trajectory (Where) is automatically extracted by the second annotation module. The coordinate sequence of the operator is entered; select the intention label and enter the behavior intention (What); the situation (How) is automatically extracted and entered from the real-time game data by the system; finally, click the Save button to save.

[0106] In this embodiment, as an optional embodiment, the method further includes:

[0107] Save the mapping relationship between real-time deduction data and intent labels.

[0108] In this embodiment, as an optional embodiment, the labeled data is sent to the cloud for centralized storage.

[0109] In this embodiment, as another optional embodiment, the method further includes:

[0110] Based on the annotated battle report data and the re-annotated real-time game data, a war game simulation model is constructed using a pre-set neural network algorithm.

[0111] In this embodiment, the annotated war game report data and the re-annotated real-time war game data are used as sample data and trained using a neural network algorithm or an AI algorithm to obtain a war game model.

[0112] The semi-automatic computer war game data annotation system proposed in this embodiment, by designing a user interface suitable for real-time war game data annotation, presents the war game simulation process more concisely and intuitively, and can make comprehensive judgments based on information such as operator health, weapon status, terrain, and visibility. At the same time, in computer war game simulation, the same action sequence of the operator corresponds to different labels under different situations. Even experienced annotators often find it difficult to judge the true intention of the operator, but the player who controls the operator can correctly describe the label corresponding to the issued action sequence. Therefore, by adding modules such as the simulation battle report collection module and the first annotation module in which players participate, the label extraction and automatic labeling functions are realized through interaction with the player's voice data. In this way, pre-labeling is realized through the assistance of voice data and interaction with players, which not only effectively improves the labeling efficiency and labeling accuracy, but also reduces the time cost and labor cost of labeling.

[0113] Based on the same inventive concept, an embodiment of the present invention further provides a storage medium on which a computer program is stored, and when the program is executed by a processor, the steps of the method for marking war game data in any possible implementation manner described above are implemented.

[0114] Alternatively, the storage medium may be a non-transitory computer-readable storage medium, for example, the non-transitory computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0115] Based on the same inventive concept, see Figure 4 The embodiment of the present invention further provides an electronic device, including a memory 101 (such as a non-volatile memory), a processor 102, and a computer program stored in the memory 101 and executable on the processor 102. When the processor 102 executes the program, the steps of the method for marking war game data in any possible implementation manner described above can be equivalent to the marking system for war game data as described above. Of course, the processor can also be used to process other data or operations. The electronic device can be a PC, a server, a terminal, and other devices.

[0116] like Figure 4As shown, the electronic device may also generally include: a memory 103, a network interface 104, and an internal bus 105. In addition to these components, other hardware may also be included, which will not be described in detail.

[0117] It should be pointed out that the above-mentioned wargame data annotation system can be implemented through software. As a device in a logical sense, it is formed by the processor 102 of the electronic device in which it is located reading the computer program instructions stored in the non-volatile memory into the memory 103 for execution.

[0118] The embodiments of the subject matter and functional operations described in this specification may be implemented in the following: digital electronic circuits, tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or a combination of one or more of them. The embodiments of the subject matter described in this specification may be implemented as one or more computer programs, i.e., one or more modules in computer program instructions encoded on a tangible non-temporary program carrier to be executed by a data processing device or to control the operation of the data processing device. Alternatively or additionally, the program instructions may be encoded on an artificially generated propagation signal, such as a machine-generated electrical, optical or electromagnetic signal, which is generated to encode information and transmit it to a suitable receiver device for execution by a data processing device. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.

[0119] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform the corresponding functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuits, such as FPGAs (field programmable gate arrays) or ASICs (application specific integrated circuits), and the apparatus can also be implemented as special purpose logic circuits.

[0120] Computers suitable for executing computer programs include, for example, general and / or special microprocessors, or any other type of central processing unit. Typically, the central processing unit will receive instructions and data from a read-only memory and / or a random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, the computer will also include one or more large-capacity storage devices for storing data, such as magnetic disks, magneto-optical disks, or optical disks, or the computer will be operably coupled to this large-capacity storage device to receive data from it or to transmit data to it, or both. However, the computer does not necessarily have such a device. In addition, the computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.

[0121] Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including, for example, semiconductor memory devices (e.g., EPROM, EEPROM and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. The processor and memory may be supplemented by, or incorporated in, special purpose logic circuitry.

[0122] Although this specification includes many specific implementation details, these should not be interpreted as limiting the scope of any invention or the scope of protection claimed, but are mainly used to describe the features of the specific embodiments of specific inventions. Certain features described in multiple embodiments in this specification may also be implemented in combination in a single embodiment. On the other hand, the various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. In addition, although the features may work as above in certain combinations and even initially claim protection, one or more features from the claimed combination may be removed from the combination in some cases, and the claimed combination may point to a sub-combination or a variation of a sub-combination.

[0123] Similarly, although operations are depicted in a particular order in the accompanying drawings, this should not be understood as requiring that these operations be performed in the particular order shown or performed sequentially, or requiring that all illustrated operations be performed to achieve the desired results. In some cases, multitasking and parallel processing may be advantageous. In addition, the separation of various system modules and components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product, or packaged into multiple software products.

[0124] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired results. In some implementations, multitasking and parallel processing may be advantageous.

[0125] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0126] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features applied herein.

[0127] This embodiment takes voice-based wargame data annotation as an example. It should be noted that the system and method are also applicable to the annotation of real-time strategy game data.

Claims

1. A wargame data annotation system, characterized in that: include: The war game simulation client is used to parse and display real-time simulation data to players, send the action instructions issued by players and / or agents based on the real-time simulation data to the war game simulation engine module, and send the pre-acquired players' simulation battle report data to the simulation battle report collection module; The war game report collection module is used to receive the war game report data sent by the war game client and send it to the label extraction module; A label extraction module is used to obtain the intention description information recorded in the deduction battle report data based on a preset label recognition strategy, label the deduction battle report data based on the intention description information, and send the label and the corresponding deduction battle report data to the label database; The label database is used to store the mapping relationship between the labels sent by the label extraction module and the corresponding deduction battle report data; The war game simulation engine is used to receive action instructions from players and / or agents, update the environmental status and battlefield situation of the simulated environment of the war game according to the action instructions, generate real-time simulation data based on the updated simulation environment, and send it to the war game simulation client and the data acquisition module respectively; A recorder is used to collect the voice data of players during the war game and send it to the data collection module; A data collection module is used to extract description information from the voice data, label the voice data based on the description information and the label recognition strategy, and send the labeled voice data and the real-time deduction data sent by the war game deduction engine to the first labeling module; A first labeling module is used to extract a label from the labeled voice data using the labeled voice data, and for each label, obtain a start time and an end time of the voice data corresponding to the label; based on the voice data between the start time and the end time, obtain a start frame and an end frame in the real-time deduction data corresponding to the voice data, label the real-time deduction data between the start frame and the end frame as the label, and send the labeled real-time deduction data to a pre-labeling database; Pre-annotated database, used to store annotated real-time deduction data; A visualization module is used to receive the real-time deduction data of the annotations sent from the pre-annotation database and display them visually on the annotation terminal; The annotation terminal is used to visualize the annotated real-time deduction data, receive the annotated real-time deduction data based on the visual playback, make a comprehensive judgment based on the operator health, weapon status, terrain information presented in the playback, and then perform secondary annotation of the intention label, wherein the secondary annotation includes: modifying the annotation, adding the annotation, and deleting the annotation, and sending it to the second annotation module; The second labeling module is used to label the real-time deduction data of the visual display label based on the received label information if it is determined that the received label information is one of the labels stored in the label database.

2. The wargame data annotation system according to claim 1, characterized in that: The system further comprises: The user interface is used to provide an interface for human-computer interaction with players in the annotation terminal, so that the annotator can annotate according to the interface.

3. The wargame data annotation system according to claim 1, characterized in that: The system further comprises: The annotation database is used to store the mapping relationship between the intention label sent by the second annotation module for annotating the annotated real-time deduction data and the corresponding real-time deduction data.

4. The wargame data annotation system according to claim 3, characterized in that: The system further comprises: The cloud service engine is used for the mapping relationships stored in the cloud storage annotation database, the mapping relationships stored in the tag database, and the mapping relationships stored in the pre-annotated database.

5. A method for marking war game data, characterized in that: include: Obtain the war game's battle report data, obtain the intention description information recorded in the war game's battle report data based on a pre-set label recognition strategy, and label the war game's battle report data based on the intention description information; Collecting real-time wargame data and voice data, extracting description information of the wargame in the voice data based on the label recognition strategy, and labeling the voice data based on the description information of the wargame and the label recognition strategy; For each label annotated with the voice data, obtain the start time and end time of the voice data corresponding to the label, obtain the start frame and end frame in the real-time deduction data corresponding to the voice data based on the voice data between the start time and the end time, and annotate the real-time deduction data between the start frame and the end frame as the label; The real-time deduction data of the annotations is displayed visually. The annotator makes a comprehensive judgment based on the real-time deduction data of the annotations in the visual playback, according to the operator health, weapon status, terrain information presented in the playback, and then performs secondary annotation of the intention label. The secondary annotation includes: modifying the annotation, adding the annotation, and deleting the annotation.

6. The method for marking war game data according to claim 5, characterized in that: The method of acquiring the intention description information recorded in the deduced battle report data based on the preset label recognition strategy, and labeling the deduced battle report data based on the intention description information, includes: For each deduction battle report data, extract keywords in the deduction battle report data to construct a keyword group of the deduction battle report data; Perform cluster analysis based on each keyword group to obtain multiple cluster centers, and set a label for each cluster center; Based on the keyword groups contained in the cluster centers, a keyword set of tags is constructed.

7. The method for marking wargame data according to claim 5, characterized in that: The method of acquiring the intention description information recorded in the deduced battle report data based on the preset label recognition strategy, and labeling the deduced battle report data based on the intention description information, includes: Splitting the deduction battle report data into unit data according to the paragraphs or chapters contained in the deduction battle report data; For each unit of data, extract keywords from the unit of data; According to the preset mapping relationship between each tag and the keyword group, the similarity between the keywords extracted from the unit data and the keyword group of each tag is calculated, the tag corresponding to the highest similarity is selected, and the unit data is marked with the corresponding tag.

8. The method for marking wargame data according to claim 7, characterized in that: The method further comprises: Based on the annotated battle report data and the re-annotated real-time game data, a war game simulation model is constructed using a pre-set neural network algorithm.

9. A storage medium, characterized in that: The storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the method for marking war game data as described in any one of claims 5 to 8 are implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for marking war game data described in any one of claims 5 to 8 are implemented.

Images