CN113304485A - Operation data processing method of big data cloud game, server and storage medium - Google Patents
Operation data processing method of big data cloud game, server and storage medium Download PDFInfo
- Publication number
- CN113304485A CN113304485A CN202110597726.8A CN202110597726A CN113304485A CN 113304485 A CN113304485 A CN 113304485A CN 202110597726 A CN202110597726 A CN 202110597726A CN 113304485 A CN113304485 A CN 113304485A
- Authority
- CN
- China
- Prior art keywords
- operation data
- game operation
- game
- queue
- intention content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63F—CARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR
- A63F13/00—Video games, i.e. games using an electronically generated display having two or more dimensions
- A63F13/70—Game security or game management aspects
- A63F13/77—Game security or game management aspects involving data related to game devices or game servers, e.g. configuration data, software version or amount of memory
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Computer Security & Cryptography (AREA)
- General Business, Economics & Management (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The application relates to the technical field of big data and cloud games, in particular to an operation data processing method, a server and a storage medium of a big data cloud game, which can arrange a group of game operation data based on the matching degree between each game operation data and a target cloud game event and the operation behavior description of each game operation data, can automatically and intelligently generate a more complete and actual game behavior intention content queue based on an ordered game operation data queue, can help a game development end to quickly and accurately know the game requirements of game players, reduces the time consumed by the game development end for analyzing a large amount of game operation data, effectively improves the generation efficiency of behavior intention content, and can ensure the quality of the behavior intention content. In addition, the behavior intention contents generated by adopting the scheme have certain sequence, so that the game development end can quickly and accurately acquire the related contents.
Description
Technical Field
The embodiment of the application relates to the technical field of big data and cloud games, in particular to an operation data processing method, a server and a storage medium of a big data cloud game.
Background
The cloud game is a novel game mode formed based on cloud computing technology. Depending on the cloud game, most game logic and rendering can be operated at the cloud game server, and then compressed game content is transmitted to the client through the cloud game server, so that a client player does not need to configure a computer with higher performance.
Compared with the traditional game mode, the cloud game is free from the dependence on hardware. For the cloud game server, only the performance of the cloud game server needs to be improved without developing a new host, and for a client player, higher image quality can be obtained without purchasing a high-performance computer. Therefore, the application range of the cloud game is more and more extensive.
In the practical application process of the cloud game, in order to update and upgrade the game, the game requirements of a game player need to be analyzed. One of the links for analyzing the game requirement is to identify the game behavior intention content of the game player, however, the related art has a problem of low efficiency in identifying the game behavior intention content.
Disclosure of Invention
In view of this, the embodiment of the present application provides an operation data processing method of a big data cloud game, a server and a storage medium.
The embodiment of the application provides an operation data processing method of a big data cloud game, which is applied to a cloud game server and comprises the following steps:
obtaining a game operation data set aiming at a target cloud game event, wherein the game operation data set comprises at least two pieces of game operation data;
obtaining a matching degree between each piece of game operation data in the game operation data set and the target cloud game event;
according to the matching degree corresponding to each game operation data and the operation behavior description of each game operation data, sorting each game operation data to obtain a corresponding game operation data queue;
generating a target behavioral intention content queue for the target cloud game event based on the game operation data queue, the target behavioral intention content queue comprising at least two target behavioral intention content segments.
In an optional embodiment, the sorting the game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data to obtain a corresponding game operation data queue specifically includes:
dividing and processing each piece of game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data to obtain at least two game operation data subsets;
and sorting the game operation data subsets, and sorting the game operation data in the game operation data subsets to obtain the game operation data queue.
In an alternative embodiment,
the dividing and processing the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain at least two game operation data subsets specifically comprises:
respectively fusing the operation behavior descriptions of the game operation data according to the matching degrees corresponding to the game operation data to obtain the hotspot operation behavior description of the game operation data;
performing feature analysis on each piece of game operation data according to the hotspot operation behavior description of each piece of game operation data to obtain at least two game operation data subsets;
correspondingly, the sorting the game operation data subsets, and sorting the game operation data in the game operation data subsets, respectively, to obtain the game operation data queue specifically includes:
sorting the game operation data subsets according to the number of the game operation data contained in the game operation data subsets;
for each subset of game operation data, performing the following steps: according to the operation behavior description of each piece of game operation data in the game operation data subset and the relevant condition of the game operation data subset, sorting each piece of game operation data in the game operation data subset; generating the game operation data queue based on the arrangement result between the game operation data subsets and the arrangement result of the game operation data in the game operation data subsets.
In an alternative embodiment,
the obtaining of the matching degree between each piece of game operation data in the game operation data set and the target cloud game event specifically includes: respectively inputting the game operation data into a behavior intention content recognition network which is trained in advance, and detecting the matching degree of the game operation data based on a hot spot detection layer of an operation event layer in the behavior intention content recognition network which is trained in advance to obtain the matching degree corresponding to the game operation data output by the hot spot detection layer;
the sorting the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain a corresponding game operation data queue specifically includes: respectively inputting the game operation data and the matching degree corresponding to the game operation data into a feature analysis and arrangement layer in the action intention content recognition network which is trained in advance, performing feature analysis and arrangement on the game operation data based on the feature analysis and arrangement layer to obtain first synthetic operation data of a segment layer output by the feature analysis and arrangement layer, and combining game operation data segments in the first synthetic operation data to form the game operation data queue;
the generating a target behavior intention content queue for the target cloud game event based on the game operation data queue specifically includes: inputting the synthetic operation data into a behavior intention content layer in the behavior intention content recognition network which is trained in advance, and performing attention matching degree detection based on the behavior intention content layer to obtain the target behavior intention content queue output by the behavior intention content layer; the action intention content recognition network which is trained in advance is obtained by training according to a network model training set, sample training data in the network model training set comprises sample game operation data which already carries relevance content, and the relevance content represents whether the sample game operation data is relevant to a sample cloud game event or not.
In an optional embodiment, the respectively inputting the game operation data into a behavior intention content recognition network which is trained in advance, and based on a hotspot detection layer of an operation event layer in the behavior intention content recognition network which is trained in advance, obtaining a matching degree corresponding to each sample game operation data output by the hotspot detection layer specifically includes:
respectively inputting the game operation data into the hotspot detection layer, and mapping the game operation data to a set space based on an operation event mapping unit in the hotspot detection layer to obtain operation contents of the game operation data;
processing the operation content of each game operation data into corresponding operation habit expression content through continuous feature extraction;
based on the hot spot detection layer, respectively extracting attention features between the operation habit expression content of each piece of game operation data and the operation habit expression content of other game operation data except the piece of game operation data;
and obtaining the matching degree between each piece of game operation data and the target cloud game event based on the attention feature corresponding to each piece of game operation data.
In an optional embodiment, the performing, based on the feature analysis sorting layer, feature analysis and sorting on the pieces of game operation data to obtain first synthetic operation data of a segment level output by the feature analysis sorting layer specifically includes:
based on the behavior intention content recognition network which is trained in advance, a characteristic analysis and sorting layer is used for mapping each piece of game operation data to a set space to obtain a segment data queue corresponding to each piece of game operation data;
extracting operation tracks of the fragment data queues corresponding to the game operation data through linear data mining processing to obtain visual track data of the game operation data;
fusing the visual track data of the game operation data according to the matching degree corresponding to the game operation data to obtain the hot spot visual track data of the game operation data;
performing feature analysis based on the hotspot visual trajectory data of each piece of game operation data to obtain at least two game operation data subsets;
and arranging all game operation data subsets, arranging all game operation data in each game operation data subset, synthesizing hotspot visual track data of all game operation data, and performing segment level processing to obtain the first synthesized operation data.
In an optional embodiment, the inputting the first synthetic operation data into a behavioral intention content layer in the behavioral intention content recognition network that has been trained in advance, performing attention matching degree detection based on the behavioral intention content layer, and obtaining the target behavioral intention content queue output by the behavioral intention content layer specifically includes:
sequentially generating each behavior intention content segment in the target behavior intention content queue by adopting a repeated accumulation processing mode, wherein one behavior intention content in the target behavior intention content queue at least comprises one behavior intention content segment;
wherein, in a round of repeated accumulation processing, the following steps are implemented:
inputting the target behavior intention content segment output in the previous round into the behavior intention content layer, wherein the behavior intention content layer is input in the first round into an original reference segment configured in advance;
analyzing attention indexes of the target action intention content segments output in the previous round and the game operation data segments in the sample queue through an attention strategy, wherein the attention indexes represent attention degrees between the game operation data segments and the action intention content segments output in the previous round;
fusing the attention index and an operation habit expression content queue of the game operation data segment in the game operation data queue, and inputting the fused result into a long-term and short-term memory neural network to obtain target visual trajectory data of the game operation data queue output in the current round;
and generating the target behavior intention content segment output in the current round based on the target behavior intention content segment output in the previous round and the target visualization track data.
In an alternative embodiment,
before the analyzing the attention indexes of the target behavior intention content segment output in the previous round and the game operation data segments in the sample queue through the attention strategy, further comprising:
taking the target game operation data subsets selected in the current round and the associated subsets of the target game operation data subsets as dominant game operation data subsets, and taking other game operation data subsets as marginal game operation data subsets, wherein the target game operation data subsets selected each time are determined based on the sequence among the game operation data subsets;
adding a first attention description to a game operation data segment in an explicit game operation data subset in the game operation data queue, and adding a second attention description to a game operation data segment in an edge game operation data subset in the game operation data queue to obtain first attention operation habit expression contents corresponding to each game operation data segment in the sample queue;
adding the first attention description to the target behavior intention content segment output in the previous round to obtain corresponding second attention operation habit expression content;
the analyzing, by an attention strategy, the attention indexes of the target behavior intention content segment output in the previous round and the game operation data segments in the sample queue specifically includes:
and analyzing the attention indexes of the target behavior intention content segment output in the previous round and each game operation data segment in the sample queue based on an attention strategy by utilizing the first attention data corresponding to each game operation data segment in the sample queue and the second attention operation habit expression content corresponding to the target behavior intention content segment output in the previous round.
The embodiment of the application also provides a cloud game server, which comprises a processor, a communication bus and a memory; the processor and the memory communicate via the communication bus, and the processor reads the computer program from the memory and runs the computer program to perform the method described above.
An embodiment of the present application further provides a computer storage medium, where a computer program is stored, and the computer program implements the method when running.
Compared with the prior art, the operation data processing method, the server and the storage medium of the big data cloud game provided by the embodiment of the application have the following technical effects: the method for generating the action intention content queue based on the game operation data can sort the game operation data based on the matching degree between the game operation data and the target cloud game event and the operation action description of the game operation data, and further automatically and intelligently generate a group of ordered action intention contents aiming at the target cloud game event based on the ordered game operation data queue. The method based on the embodiment of the application can automatically and intelligently generate a more complete and actual game behavior intention content queue, can help the game development end to quickly and accurately know the game requirements of game players, reduces the time consumed by the game development end for analyzing a large amount of game operation data, effectively improves the generation efficiency of behavior intention content, and can ensure the quality of the behavior intention content. In addition, the behavior intention contents generated by adopting the scheme have certain sequence, so that the game development end can quickly and accurately acquire the related contents.
In the description that follows, additional features will be set forth, in part, in the description. These features will be in part apparent to those skilled in the art upon examination of the following and the accompanying drawings, or may be learned by production or use. The features of the present application may be realized and attained by practice or use of various aspects of the methodologies, instrumentalities and combinations particularly pointed out in the detailed examples that follow.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a block diagram of a cloud game server according to an embodiment of the present disclosure.
Fig. 2 is a flowchart of an operation data processing method of a big data cloud game according to an embodiment of the present application.
Fig. 3 is a block diagram of an operation data processing apparatus of a big data cloud game according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
Fig. 1 is a block diagram illustrating a cloud game server 10 according to an embodiment of the present disclosure. In the embodiment of the present application, the cloud game server 10 may be a server with data storage, transmission, and processing functions, as shown in fig. 1, the cloud game server 10 includes: a memory 11, a processor 12, a communication bus 13, and an operation data processing device 20 of the big data cloud game.
The memory 11, processor 12 and communication bus 13 are electrically connected, directly or indirectly, to enable the transfer or interaction of data. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 11 stores an operation data processing device 20 of the big data cloud game, the operation data processing device 20 of the big data cloud game includes at least one software function module which can be stored in the memory 11 in a form of software or firmware (firmware), and the processor 12 executes various function applications and data processing by running software programs and modules stored in the memory 11, for example, the operation data processing device 20 of the big data cloud game in the embodiment of the present application, so as to implement the operation data processing method of the big data cloud game in the embodiment of the present application.
The Memory 11 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like. The memory 11 is used for storing a program, and the processor 12 executes the program after receiving an execution instruction.
The processor 12 may be an integrated circuit chip having data processing capabilities. The Processor 12 may be a general-purpose Processor including a Central Processing Unit (CPU), a Network Processor (NP), and the like. The various methods, steps and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The communication bus 13 is used for establishing communication connection between the cloud game server 10 and other communication terminal devices through a network, and implementing transceiving operation of network signals and data. The network signal may include a wireless signal or a wired signal.
It is to be understood that the configuration shown in fig. 1 is merely illustrative, and that cloud gaming server 10 may include more or fewer components than shown in fig. 1, or have a different configuration than shown in fig. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.
An embodiment of the present application further provides a computer storage medium, where a computer program is stored, and the computer program implements the method when running.
Fig. 2 shows a flowchart of operation data processing of a big data cloud game provided in an embodiment of the present application. The method steps defined by the flow related to the method are applied to the cloud game server 10 and can be implemented by the processor 12, and the method includes the following steps S21-S24.
And S21, acquiring the game operation data set aiming at the target cloud game event.
In an embodiment of the present application, the game play data set includes at least two pieces of game play data. The target cloud game event may be a game event corresponding to different cloud game types, such as a multiplayer strategy confrontation game event, a shooting game event, a formation game event, an online game event, a role playing game event, and the like. The at least two game operation data included in the game operation data set may be counted in chronological order.
And S22, obtaining the matching degree between each piece of game operation data in the game operation data set and the target cloud game event.
In the embodiment of the present application, the matching degree between each piece of game operation data in the game operation data set and the target cloud game event may be understood as a correlation degree, and for example, the matching degree may be described by using a correlation coefficient.
And S23, sorting the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain a corresponding game operation data queue.
In the embodiment of the present application, the operation behavior description may be an operation behavior feature, and may be expressed in the form of a feature vector or a feature map, for example. Sorting the pieces of game operation data may be understood as sorting the pieces of game operation data to obtain a game operation data queue (sequence).
In some possible embodiments, the sorting the game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data, which are described in S23, to obtain a corresponding game operation data queue may specifically include the following S231 and S232.
S231, dividing and processing the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain at least two game operation data subsets.
Further, the dividing and controlling the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data, which are described in S231, to obtain at least two game operation data subsets may include the following embodiments: respectively fusing the operation behavior descriptions of the game operation data according to the matching degrees corresponding to the game operation data to obtain the hotspot operation behavior description of the game operation data; and performing characteristic analysis on each piece of game operation data according to the hotspot operation behavior description of each piece of game operation data to obtain at least two game operation data subsets. For example, the operation behavior descriptions of the game operation data can be weighted, so that the accuracy in feature analysis (clustering) is ensured, and further, the game operation data subsets are prevented from being missed.
S232, sorting the game operation data subsets, and sorting the game operation data in the game operation data subsets to obtain the game operation data queue.
In some possible embodiments, the sorting the game operation data subsets in S232, and sorting the game operation data in the game operation data subsets to obtain the game operation data queue may include the following technical solutions: sorting the game operation data subsets according to the number of the game operation data contained in the game operation data subsets; for each subset of game operation data, performing the following steps: arranging the game operation data in the game operation data subset according to the operation behavior description of each game operation data in the game operation data subset and the correlation condition (correlation degree) of the game operation data subset; generating the game operation data queue based on the arrangement result between the game operation data subsets and the arrangement result of the game operation data in the game operation data subsets. By the design, each game operation data subset can be independently analyzed, so that individual game operation data subsets are prevented from being missed, and the integrity and quality of a generated game operation data queue are ensured.
And S24, generating a target behavior intention content queue aiming at the target cloud game event based on the game operation data queue.
In an embodiment of the present application, the target behavioral intention content queue includes at least two target behavioral intention content segments. Correspondingly, the target behavior intention content segments comprise related intention requirements or tendency requirements of game players, and related game player requirement analysis can be carried out through different target behavior intention content segments, so that updating, upgrading and optimization of game products are realized.
In some other independently implementable embodiments, obtaining the matching degree between each piece of game operation data in the game operation data set and the target cloud game event as described in S22 may be implemented by the following embodiment a: and respectively inputting the game operation data into a behavior intention content recognition network which is trained in advance, and detecting the matching degree of the game operation data based on a hot spot detection layer of an operation event layer in the behavior intention content recognition network which is trained in advance to obtain the matching degree corresponding to the game operation data output by the hot spot detection layer.
In embodiment a, the behavioral intention content recognition network may be a convolutional neural network, a deep learning neural network, or a long-short term memory neural network. Further, the hot spot detection layer may be a functional network layer with a classification function, and is used for performing real-time and accurate matching degree detection.
Further, in the above embodiment a, the steps of inputting the game operation data into the action intention content recognition network trained in advance, obtaining the matching degree corresponding to each sample game operation data output by the hotspot detection layer based on the hotspot detection layer at the operation event level in the action intention content recognition network trained in advance may include the following steps a 1-a 4.
Step a1, inputting the game operation data into the hotspot detection layer, and mapping the game operation data to a set space based on an operation event mapping unit in the hotspot detection layer to obtain operation contents of the game operation data.
For example, the operation event mapping unit is used for performing normalization processing on each piece of game operation data, thereby ensuring the format consistency of the operation content.
And step A2, processing the operation content of each game operation data into corresponding operation habit expression content through continuous feature extraction.
It is understood that the operation habit expressions may be recorded in the form of text characters.
Step a3, based on the hotspot detection layer, extracting attention features between the operation habit expression content of each piece of game operation data and the operation habit expression content of other pieces of game operation data except the piece of game operation data.
It is understood that the feature of interest is an attention feature.
Step A4, obtaining the matching degree between each piece of game operation data and the target cloud game event based on the attention feature corresponding to each piece of game operation data.
It can be understood that, by implementing the steps a 1-a 4, the degree of matching between each piece of game operation data and the target cloud game event can be accurately determined by taking the attention feature corresponding to each piece of game operation data into consideration.
On the basis of the foregoing embodiment a, the step S23 of sorting the game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data to obtain a corresponding game operation data queue may be implemented by the following embodiment B: and respectively inputting the game operation data and the matching degree corresponding to the game operation data into a feature analysis and arrangement layer in the action intention content recognition network which is trained in advance, performing feature analysis and arrangement on the game operation data based on the feature analysis and arrangement layer to obtain first synthetic operation data of a segment layer output by the feature analysis and arrangement layer, and combining game operation data segments in the first synthetic operation data to form the game operation data queue.
In embodiment B, the feature analysis collation layer may be used to perform feature clustering and sorting, thereby ensuring the integrity and ordering of the first synthetic operation data at the output segment level.
In the above embodiment B, the performing feature analysis and sorting on the pieces of game operation data based on the feature analysis sorting layer to obtain the first synthetic operation data at the segment level output by the feature analysis sorting layer may include the following steps B1 to B5.
Step B1, recognizing a characteristic analysis and sorting layer in the network based on the behavior intention content which is trained in advance, and mapping each piece of game operation data to a set space to obtain a fragment data queue corresponding to each piece of game operation data;
and step B2, performing operation track extraction on the fragment data queue corresponding to each game operation data through linear data mining processing to obtain visual track data of each game operation data.
For example, the linear data mining process may be a graphical-level-based trajectory mining algorithm for ensuring smoothness and continuity of the visualized trajectory data.
And step B3, fusing the visual track data of the game operation data according to the matching degree corresponding to the game operation data respectively to obtain the hotspot visual track data of the game operation data.
For example, the visualized trajectory data of each piece of game operation data may be spliced to obtain the hotspot visualized trajectory data of each piece of game operation data.
Step B4, performing feature analysis based on the hotspot visualization track data of each game operation data to obtain at least two game operation data subsets;
and step B5, arranging all game operation data subsets, arranging all game operation data in each game operation data subset, synthesizing hotspot visual trajectory data of all game operation data, and performing segment level processing to obtain the first synthesized operation data.
By the design, based on the steps B1-B5, feature analysis and clustering can be performed from the visual track level, so that the game operation data subsets are sorted, and further, related game operation data are sorted, so that the data quality of the first synthetic operation data obtained by segment level processing cannot fluctuate to a large extent, and the usability of the first synthetic operation data can be ensured.
On the basis of the foregoing embodiment C, the generation of the target behavioral intention content queue for the target cloud game event based on the game operation data queue described in S24 may be implemented by the following embodiment C: and inputting the synthetic operation data into a behavior intention content layer in the behavior intention content recognition network which is trained in advance, and performing attention matching degree detection based on the behavior intention content layer to obtain the target behavior intention content queue output by the behavior intention content layer.
In embodiment C, the action intention content recognition network trained in advance is obtained by training according to a network model training set, where sample training data in the network model training set includes sample game operation data that already carries associated content, and the associated content indicates whether the sample game operation data is associated with a sample cloud game event. Therefore, the target behavior intention content segment can be accurately analyzed and identified through the behavior intention content layer, so that the integrity and the quality of the target behavior intention content queue are ensured, and the subsequent game player requirement analysis based on the target behavior intention content queue is facilitated.
In the above embodiment C, inputting the first synthesis operation data into a behavioral intention content layer in the behavioral intention content recognition network that has been trained in advance, performing attention matching degree detection based on the behavioral intention content layer, and obtaining the target behavioral intention content queue output by the behavioral intention content layer may include the following technical solutions: sequentially generating each behavior intention content segment in the target behavior intention content queue by adopting a repeated accumulation processing mode, wherein one behavior intention content in the target behavior intention content queue at least comprises one behavior intention content segment; wherein, in a round of repeated accumulation processing, the following steps are implemented: inputting the target behavior intention content segment output in the previous round into the behavior intention content layer, wherein the behavior intention content layer is input in the first round into an original reference segment configured in advance; analyzing attention indexes of the target action intention content segments output in the previous round and the game operation data segments in the sample queue through an attention strategy, wherein the attention indexes represent attention degrees between the game operation data segments and the action intention content segments output in the previous round; fusing the attention index and an operation habit expression content queue of the game operation data segment in the game operation data queue, and inputting the fused result into a long-term and short-term memory neural network to obtain target visual trajectory data of the game operation data queue output in the current round; and generating the target behavior intention content segment output in the current round based on the target behavior intention content segment output in the previous round and the target visualization track data.
For example, the iterative accumulation process may be understood as a loop iteration and the attention strategy may be an attention mechanism. By the design, different behavior intention content segments can be independently processed and analyzed, so that omission of the behavior intention content segments in the processing and analyzing process is avoided, and the integrity of the target behavior intention content queue is ensured.
In some optional embodiments, before analyzing the attention indexes of the target behavior intention content segment output in the previous round and the respective game operation data segments in the sample queue by the attention policy, the following embodiment C1 may be further included: taking the target game operation data subsets selected in the current round and the associated subsets of the target game operation data subsets as dominant game operation data subsets, and taking other game operation data subsets as marginal game operation data subsets, wherein the target game operation data subsets selected each time are determined based on the sequence among the game operation data subsets; adding a first attention description to a game operation data segment in an explicit game operation data subset in the game operation data queue, and adding a second attention description to a game operation data segment in an edge game operation data subset in the game operation data queue to obtain first attention operation habit expression contents corresponding to each game operation data segment in the sample queue; and adding the first attention description to the target behavior intention content segment output in the previous round to obtain corresponding second attention operation habit expression content.
On the basis of the above embodiment C1, analyzing the attention indexes of the target behavior intention content segment output in the previous round and the respective game operation data segments in the sample queue by the attention policy may include the following technical solutions: and analyzing the attention indexes of the target behavior intention content segment output in the previous round and each game operation data segment in the sample queue based on an attention strategy by utilizing the first attention data corresponding to each game operation data segment in the sample queue and the second attention operation habit expression content corresponding to the target behavior intention content segment output in the previous round. By the design, the accuracy and the timeliness of the attention index can be ensured.
In some alternative embodiments, the behavioral intent content recognition network described above may be trained by: obtaining the network model training set for at least one sample cloud game event; and according to sample training data in the network model training set, performing repeated accumulation processing training on the initial behavior intention content recognition network to obtain the behavior intention content recognition network which is trained in advance.
Accordingly, each round of the repeated cumulative processing training process comprises the following operations: selecting a group of sample training data aiming at the same sample cloud game event from the network model training set, respectively inputting the sample game operation data contained in each selected sample training data into a hot spot detection layer of an operation event layer in the initial behavior intention content identification network, and obtaining the matching degree corresponding to each sample game operation data output by the hot spot detection layer.
Further, a first network performance index is generated based on the difference between the matching degree corresponding to each sample game operation data and the corresponding relevance content; respectively inputting the sample game operation data in each selected sample training data and the matching degree corresponding to each sample game operation data into a characteristic analysis and sorting layer in the initial behavior intention content recognition network, and performing characteristic analysis on each sample game operation data based on the characteristic analysis and sorting layer to obtain at least two game operation data subsets; sorting each game operation data subset based on the characteristic analysis sorting layer to obtain second synthetic operation data of a segment layer output by the characteristic analysis sorting layer; inputting the second synthetic operation data into a behavior intention content layer in the initial behavior intention content identification network, performing attention matching degree detection based on the behavior intention content layer, and obtaining a group of test behavior intention content queues output by the behavior intention content layer, wherein the test behavior intention content queues comprise at least two test behavior intention content segments; generating a second network performance index based on the global difference condition of the test behavior intention content segment in the test behavior intention content queue and the actual behavior intention content segment in the actual behavior intention content queue; and generating a third network performance index based on the attention of the game operation data segments in each of the game operation data subsets.
And finally, performing parameter improvement on the initial behavior intention content identification network according to the first network performance index, the second network performance index and the third network performance index.
It can be understood that the first network performance index reflects a difference between a matching degree corresponding to each piece of the sample game operation data and a corresponding correlation content, the second network performance index reflects a global difference between a test behavior intention content segment in the test behavior intention content queue and an actual behavior intention content segment in the actual behavior intention content queue, the third network performance index reflects a degree of attention of the game operation data segment in each game operation data subset, and different network performance indexes can be understood as different loss functions. In this way, loss analysis of the behavioral intention content recognition network can be realized from different angles, so that losses from different angles are taken into consideration as much as possible when model parameters are adjusted and optimized, and the behavioral intention content recognition network can be applied to different cloud game scenes as much as possible.
In some alternative embodiments, the generating the second network performance indicator based on the global difference between the content segment of the test behavioral intention in the content queue of the test behavioral intention and the content segment of the actual behavioral intention in the content queue of the actual behavioral intention described in the above steps may include: for any one test behavior intention content segment, determining the global difference condition between the test behavior intention content segment in the test behavior intention content queue and the actual behavior intention content segment in the actual behavior intention content queue based on the global possibility description of the test behavior intention content segment in a preset behavior intention content segment set and the global possibility description of the test behavior intention content segment in the game operation data set; generating the second network performance indicator based on the determined global difference condition. In the embodiment of the present application, the global difference condition may be understood as a distribution error. In this way, the second network performance indicator can be generated as accurately as possible by taking into account the global likelihood description (distribution probability).
In summary, the method for generating the action intention content queue based on the group of game operation data is provided in the embodiment of the present application, so that the group of game operation data can be sorted based on the matching degree between each piece of game operation data and the target cloud game event and the operation behavior description of each piece of game operation data, and then an ordered group of action intention contents for the target cloud game event can be automatically and intelligently generated based on the ordered game operation data queue. The method based on the embodiment of the application can automatically and intelligently generate a more complete and actual game behavior intention content queue, can help the game development end to quickly and accurately know the game requirements of game players, reduces the time consumed by the game development end for analyzing a large amount of game operation data, effectively improves the generation efficiency of behavior intention content, and can ensure the quality of the behavior intention content. In addition, the behavior intention contents generated by adopting the scheme have certain sequence, so that the game development end can quickly and accurately acquire the related contents.
Based on the same inventive concept, there is also provided an operation data processing device 20 for a big data cloud game, applied to a cloud game server 10, the device including:
a data obtaining module 21, configured to obtain a game operation data set for a target cloud game event, where the game operation data set includes at least two pieces of game operation data;
a matching analysis module 22, configured to obtain a matching degree between each piece of game operation data in the game operation data set and the target cloud game event;
the data sorting module 23 is configured to sort the game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data, so as to obtain a corresponding game operation data queue;
an intention analysis module 24, configured to generate a target behavior intention content queue for the target cloud game event based on the game operation data queue, where the target behavior intention content queue includes at least two target behavior intention content segments.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a cloud game server 10, or a network device) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (10)
1. An operation data processing method of a big data cloud game is applied to a cloud game server, and comprises the following steps:
obtaining a game operation data set aiming at a target cloud game event, wherein the game operation data set comprises at least two pieces of game operation data;
obtaining a matching degree between each piece of game operation data in the game operation data set and the target cloud game event;
according to the matching degree corresponding to each game operation data and the operation behavior description of each game operation data, sorting each game operation data to obtain a corresponding game operation data queue;
generating a target behavioral intention content queue for the target cloud game event based on the game operation data queue, the target behavioral intention content queue comprising at least two target behavioral intention content segments.
2. The method according to claim 1, wherein the sorting the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain a corresponding game operation data queue specifically includes:
dividing and processing each piece of game operation data according to the matching degree corresponding to each piece of game operation data and the operation behavior description of each piece of game operation data to obtain at least two game operation data subsets;
and sorting the game operation data subsets, and sorting the game operation data in the game operation data subsets to obtain the game operation data queue.
3. The method of claim 2,
the dividing and processing the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain at least two game operation data subsets specifically comprises:
respectively fusing the operation behavior descriptions of the game operation data according to the matching degrees corresponding to the game operation data to obtain the hotspot operation behavior description of the game operation data;
performing feature analysis on each piece of game operation data according to the hotspot operation behavior description of each piece of game operation data to obtain at least two game operation data subsets;
correspondingly, the sorting the game operation data subsets, and sorting the game operation data in the game operation data subsets, respectively, to obtain the game operation data queue specifically includes:
sorting the game operation data subsets according to the number of the game operation data contained in the game operation data subsets;
for each subset of game operation data, performing the following steps: according to the operation behavior description of each piece of game operation data in the game operation data subset and the relevant condition of the game operation data subset, sorting each piece of game operation data in the game operation data subset; generating the game operation data queue based on the arrangement result between the game operation data subsets and the arrangement result of the game operation data in the game operation data subsets.
4. The method of claim 1,
the obtaining of the matching degree between each piece of game operation data in the game operation data set and the target cloud game event specifically includes: respectively inputting the game operation data into a behavior intention content recognition network which is trained in advance, and detecting the matching degree of the game operation data based on a hot spot detection layer of an operation event layer in the behavior intention content recognition network which is trained in advance to obtain the matching degree corresponding to the game operation data output by the hot spot detection layer;
the sorting the game operation data according to the matching degree corresponding to the game operation data and the operation behavior description of the game operation data to obtain a corresponding game operation data queue specifically includes: respectively inputting the game operation data and the matching degree corresponding to the game operation data into a feature analysis and arrangement layer in the action intention content recognition network which is trained in advance, performing feature analysis and arrangement on the game operation data based on the feature analysis and arrangement layer to obtain first synthetic operation data of a segment layer output by the feature analysis and arrangement layer, and combining game operation data segments in the first synthetic operation data to form the game operation data queue;
the generating a target behavior intention content queue for the target cloud game event based on the game operation data queue specifically includes: inputting the synthetic operation data into a behavior intention content layer in the behavior intention content recognition network which is trained in advance, and performing attention matching degree detection based on the behavior intention content layer to obtain the target behavior intention content queue output by the behavior intention content layer; the action intention content recognition network which is trained in advance is obtained by training according to a network model training set, sample training data in the network model training set comprises sample game operation data which already carries relevance content, and the relevance content represents whether the sample game operation data is relevant to a sample cloud game event or not.
5. The method according to claim 4, wherein the respectively inputting the pieces of game operation data into a behavior intention content recognition network that is trained in advance, and obtaining the matching degree corresponding to the pieces of sample game operation data output by a hotspot detection layer based on a hotspot detection layer at an operation event level in the behavior intention content recognition network that is trained in advance, specifically comprises:
respectively inputting the game operation data into the hotspot detection layer, and mapping the game operation data to a set space based on an operation event mapping unit in the hotspot detection layer to obtain operation contents of the game operation data;
processing the operation content of each game operation data into corresponding operation habit expression content through continuous feature extraction;
based on the hot spot detection layer, respectively extracting attention features between the operation habit expression content of each piece of game operation data and the operation habit expression content of other game operation data except the piece of game operation data;
and obtaining the matching degree between each piece of game operation data and the target cloud game event based on the attention feature corresponding to each piece of game operation data.
6. The method according to claim 4, wherein the performing feature analysis and sorting on the pieces of game operation data based on the feature analysis sorting layer to obtain first synthetic operation data at a segment level output by the feature analysis sorting layer specifically includes:
based on the behavior intention content recognition network which is trained in advance, a characteristic analysis and sorting layer is used for mapping each piece of game operation data to a set space to obtain a segment data queue corresponding to each piece of game operation data;
extracting operation tracks of the fragment data queues corresponding to the game operation data through linear data mining processing to obtain visual track data of the game operation data;
fusing the visual track data of the game operation data according to the matching degree corresponding to the game operation data to obtain the hot spot visual track data of the game operation data;
performing feature analysis based on the hotspot visual trajectory data of each piece of game operation data to obtain at least two game operation data subsets;
and arranging all game operation data subsets, arranging all game operation data in each game operation data subset, synthesizing hotspot visual track data of all game operation data, and performing segment level processing to obtain the first synthesized operation data.
7. The method according to claim 4, wherein the inputting the first synthetic operation data into a behavioral intention content layer in the behavioral intention content recognition network that has been trained in advance, performing attention matching degree detection based on the behavioral intention content layer, and obtaining the target behavioral intention content queue output by the behavioral intention content layer specifically includes:
sequentially generating each behavior intention content segment in the target behavior intention content queue by adopting a repeated accumulation processing mode, wherein one behavior intention content in the target behavior intention content queue at least comprises one behavior intention content segment;
wherein, in a round of repeated accumulation processing, the following steps are implemented:
inputting the target behavior intention content segment output in the previous round into the behavior intention content layer, wherein the behavior intention content layer is input in the first round into an original reference segment configured in advance;
analyzing attention indexes of the target action intention content segments output in the previous round and the game operation data segments in the sample queue through an attention strategy, wherein the attention indexes represent attention degrees between the game operation data segments and the action intention content segments output in the previous round;
fusing the attention index and an operation habit expression content queue of the game operation data segment in the game operation data queue, and inputting the fused result into a long-term and short-term memory neural network to obtain target visual trajectory data of the game operation data queue output in the current round;
and generating the target behavior intention content segment output in the current round based on the target behavior intention content segment output in the previous round and the target visualization track data.
8. The method of claim 7,
before the analyzing the attention indexes of the target behavior intention content segment output in the previous round and the game operation data segments in the sample queue through the attention strategy, further comprising:
taking the target game operation data subsets selected in the current round and the associated subsets of the target game operation data subsets as dominant game operation data subsets, and taking other game operation data subsets as marginal game operation data subsets, wherein the target game operation data subsets selected each time are determined based on the sequence among the game operation data subsets;
adding a first attention description to a game operation data segment in an explicit game operation data subset in the game operation data queue, and adding a second attention description to a game operation data segment in an edge game operation data subset in the game operation data queue to obtain first attention operation habit expression contents corresponding to each game operation data segment in the sample queue;
adding the first attention description to the target behavior intention content segment output in the previous round to obtain corresponding second attention operation habit expression content;
the analyzing, by an attention strategy, the attention indexes of the target behavior intention content segment output in the previous round and the game operation data segments in the sample queue specifically includes:
and analyzing the attention indexes of the target behavior intention content segment output in the previous round and each game operation data segment in the sample queue based on an attention strategy by utilizing the first attention data corresponding to each game operation data segment in the sample queue and the second attention operation habit expression content corresponding to the target behavior intention content segment output in the previous round.
9. A cloud game server is characterized by comprising a processor, a communication bus and a memory; the processor and the memory communicate via the communication bus, the processor reading a computer program from the memory and operating to perform the method of any of claims 1-8.
10. A computer storage medium, characterized in that it stores a computer program which, when executed, implements the method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597726.8A CN113304485A (en) | 2021-05-31 | 2021-05-31 | Operation data processing method of big data cloud game, server and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597726.8A CN113304485A (en) | 2021-05-31 | 2021-05-31 | Operation data processing method of big data cloud game, server and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113304485A true CN113304485A (en) | 2021-08-27 |
Family
ID=77376201
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110597726.8A Withdrawn CN113304485A (en) | 2021-05-31 | 2021-05-31 | Operation data processing method of big data cloud game, server and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113304485A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114445542A (en) * | 2022-01-17 | 2022-05-06 | 上海光追网络科技有限公司 | Game role model map processing method and system based on big data |
-
2021
- 2021-05-31 CN CN202110597726.8A patent/CN113304485A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114445542A (en) * | 2022-01-17 | 2022-05-06 | 上海光追网络科技有限公司 | Game role model map processing method and system based on big data |
| CN114445542B (en) * | 2022-01-17 | 2024-05-28 | 上海光追网络科技有限公司 | Game role model mapping processing method and system based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113508378B (en) | Training method, recommendation method, device and computer readable medium for recommendation model | |
| Jiang et al. | MULocDeep: a deep-learning framework for protein subcellular and suborganellar localization prediction with residue-level interpretation | |
| JP6629678B2 (en) | Machine learning device | |
| Richert | Building machine learning systems with Python | |
| US11436617B2 (en) | Behavior pattern search system and behavior pattern search method | |
| CN108304373B (en) | Semantic dictionary construction method and device, storage medium and electronic device | |
| CN108846097B (en) | User interest tag representation method, article recommendation device and equipment | |
| JP5454827B1 (en) | Document evaluation apparatus, document evaluation method, and program | |
| CN114117213A (en) | Recommendation model training and recommendation method, device, medium and equipment | |
| CN110322254B (en) | Online fraud identification method, device, medium and electronic equipment | |
| CN111447574B (en) | Short message classification method, device, system and storage medium | |
| Saleh | Machine Learning Fundamentals: Use Python and scikit-learn to get up and running with the hottest developments in machine learning | |
| CN111881671A (en) | Attribute word extraction method | |
| KR20200107389A (en) | Rating augmentation and item recommendation method and system based on generative adversarial networks | |
| CN108959550B (en) | User focus mining method, device, equipment and computer readable medium | |
| CN112529024A (en) | Sample data generation method and device and computer readable storage medium | |
| CN116648698A (en) | Dynamic facet ordering | |
| CN112633976A (en) | Data processing method based on big data and cloud service server | |
| CN113304485A (en) | Operation data processing method of big data cloud game, server and storage medium | |
| CN113157871B (en) | News public opinion text processing method, server and medium applying artificial intelligence | |
| US20230253076A1 (en) | Local steps in latent space and descriptors-based molecules filtering for conditional molecular generation | |
| CN116579803A (en) | Multi-class joint demand prediction method and device based on substitution and association effects | |
| CN112017634B (en) | Data processing method, device, equipment and storage medium | |
| CN108875060A (en) | A kind of website identification method and identifying system | |
| CN115769194A (en) | Automatic data linking across datasets |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210827 |