Competition evaluation processing method and device

Abstract

The embodiment of the invention provides a match evaluation processing method and device, and the method comprises the steps: carrying out the feature extraction from the attribute data of the members of a match team, obtaining the attribute features, and obtaining the member relation features based on a member graph network, and according to the attribute characteristics of the team members and the team member relationship characteristics, team characteristics of the game teams are constructed, and the team characteristics of the game teams are input into a game evaluation model for game evaluation processing to obtain an evaluation result, so that game evaluation on the game team level is realized.

Description

Technical Field

[0001] This document relates to the field of data processing technology, and in particular to a competition evaluation processing method and device. Background Art

[0002] With the continuous development and promotion of the Internet and artificial intelligence technologies, the application scope of artificial intelligence-related services provided on the Internet has become increasingly wide, and various forms of processing services based on artificial intelligence have emerged, making artificial intelligence services widely used in many fields. For example, in the field of sports competitions, artificial intelligence technology is used to analyze the performance of competitive teams or individuals, and game review is carried out based on artificial intelligence technology. However, as more and more service providers provide artificial intelligence services in various fields, the competitive pressure faced by service providers is also increasing. Summary of the invention

[0003] One or more embodiments of the present specification provide a method for game evaluation processing, including: extracting features from attribute data of players of a game team to obtain attribute features of each player. Obtaining player relationship features, wherein the player relationship features are obtained based on a player graph network. Constructing team features of each game team based on the attribute features and player relationship features of each player. Inputting the team features of each game team into a pre-trained game evaluation model to perform game evaluation processing and obtain an evaluation result.

[0004] One or more embodiments of the present specification provide a game evaluation processing device, including: an attribute feature extraction module, configured to extract features from the attribute data of the players of the game team to obtain the attribute features of each player. A relationship feature extraction module, configured to obtain the player relationship features. The player relationship features are obtained based on the player graph network. A feature construction module, configured to construct the team features of each game team according to the attribute features and player relationship features of each player. An evaluation processing module, configured to input the team features of each game team into a pre-trained game evaluation model to perform game evaluation processing and obtain an evaluation result.

[0005] One or more embodiments of the present specification provide a game evaluation processing device, including: a processor; and a memory configured to store computer executable instructions, wherein when the computer executable instructions are executed, the processor: extracts features from the attribute data of the players of the game team to obtain the attribute features of each player. Obtains the player relationship features, which are obtained based on the player graph network. Constructs the team features of each game team based on the attribute features and the player relationship features of each player. Inputs the team features of each game team into a pre-trained game evaluation model to perform game evaluation processing and obtain an evaluation result.

[0006] One or more embodiments of the present specification provide a computer-readable storage medium for storing computer-executable instructions, which implement the following process when executed: extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player. Obtaining the player relationship features, which are obtained based on the player graph network. Constructing the team features of each competition team based on the attribute features and the player relationship features of each player. Inputting the team features of each competition team into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate one or more embodiments of the present specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present specification. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor. Figure 1 A schematic diagram of a competition evaluation system provided for one or more embodiments of this specification; Figure 2 A processing flow chart of a competition evaluation processing method provided by one or more embodiments of this specification; Figure 3 A schematic diagram of a competition evaluation processing architecture provided for one or more embodiments of this specification; Figure 4 A flowchart of a match evaluation processing method applied to a football match scenario provided by one or more embodiments of this specification; Figure 5 A schematic diagram of an embodiment of a competition evaluation processing device provided in one or more embodiments of this specification; Figure 6 A schematic diagram of the structure of a competition evaluation processing device provided for one or more embodiments of this specification. DETAILED DESCRIPTION

[0008] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0009] The competition evaluation processing method provided in one or more embodiments of this specification can be applied to the implementation environment of the competition evaluation system. Figure 1 , the implementation environment at least includes: Feature processing module 101 and competition evaluation model 102; Among them, the feature processing module 101 is used to obtain the attribute characteristics and player relationship characteristics of each player of the competition team, construct the team characteristics of each competition team at the overall level based on the attribute characteristics and player relationship characteristics of each player, and input the constructed team characteristics into the competition evaluation model 102; the competition evaluation model 102 is used to evaluate the competition according to the input team characteristics to obtain the evaluation results.

[0010] In this implementation environment, the feature processing module 101, on the one hand, extracts features from the attribute data of the players of the competition team to obtain the attribute characteristics of each player, and on the other hand, obtains the player relationship characteristics of each player based on the player graph network. On the basis of obtaining the attribute characteristics and player relationship characteristics of each player at the player level, the overall team characteristics of each competition team at the competition team level are constructed according to the attribute characteristics and player relationship characteristics of each player, and the constructed overall team characteristics of each competition team are input into the competition evaluation model 102. The competition evaluation model 102 performs competition evaluation processing according to the input team characteristics of each competition team to obtain an evaluation result, thereby realizing the evaluation of the competition or the competition team through the cooperation of the feature processing module 101 and the competition evaluation model 102.

[0011] One or more embodiments of a competition evaluation processing method provided in this specification are as follows: Reference Figure 2 The competition evaluation processing method provided in this embodiment specifically includes steps S202 to S208.

[0012] Step S202: extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player.

[0013] In this embodiment, the game can be a sports competition, such as a football match, a basketball match, a competition of other sports or a team match, etc.; in addition, it can also be an e-sports competition, or a simulation competition provided by a game or application, such as a football simulation competition provided in a game plot, or a team tactical simulation competition provided in a tactical simulation application.

[0014] The competing teams of a match refer to the various teams participating in the match. The competing teams may be two teams or more than one teams. The number of players in a competing team may be more than one or one. For example, the competing teams in a football match refer to the two competing teams participating in the football match: Team A and Team B. Team A and Team B each have 23 players.

[0015] The attribute data includes player's personal data, player's relevant data during the game, and may also include player's personal data or data between players that can directly or indirectly affect the game. Specifically, the attribute data includes player's court position data and / or player's personal data.

[0016] Among them, the player's field position data includes the player's role in the game and / or the player's actual field behavior data. For example, the first attribute data of the players of each team in a football game includes the player's position in the team lineup (forward, midfielder, defender, goalkeeper), the movement trajectory of the team members in the field, the movement hotspot position or the movement heat map, etc. The player's personal data includes data related to the player's personal attributes, such as the weight, speed, strength, kicking time, left and right foot usage of the players of each team in a football game. The third attribute data includes data that can directly or indirectly affect the game by the players themselves or between the players, such as the scrambling ability, short-distance passing ability, offensive intention, leadership ability, etc. of the players of each team in a football game.

[0017] In specific implementation, the attribute characteristics of each player are obtained by extracting the attribute data of the players of the game team. Optionally, the attribute characteristics include at least one of the following: player's court position characteristics, player's personal characteristics, and player's defensive coordination ability characteristics.

[0018] Among them, the player's field position feature is obtained by extracting the player's field position data, and the player's personal feature is obtained by extracting the player's personal feature. The player's coordination and defensive ability feature refers to the features related to the coordination and defensive ability between the players of the game team.

[0019] In an optional implementation manner provided by this embodiment, the player's defensive coordination ability characteristics are obtained in the following manner: the player's court position data and the player's personal data are input into a large language model to obtain the defensive coordination ability characteristics, thereby obtaining the player's defensive coordination ability characteristics.

[0020] Specifically, the big language model used to acquire the coordinated defensive capability characteristics may be a pre-trained dedicated big language model or a general big language model. In the process of acquiring the coordinated defensive capability characteristics, the big language model may perform feature retrieval in the knowledge base based on the players' court position data and the players' personal data to acquire the coordinated defensive capability characteristics.

[0021] For example, Figure 3 As shown, according to the player's field position data and player's personal data of each player of Team A and Team B in the football match, feature extraction is performed on the player's field position data of each player to obtain the player's field position feature, and the player's personal data of each player is extracted to obtain the player's personal feature; and, the player's field position data and player's personal data of each player are input into the large language model to obtain the coordinated defense ability feature, and the player's coordinated defense ability feature is obtained. The three features of each player's field position feature, the player's personal feature and the player's coordinated defense ability feature together constitute the attribute feature of each player.

[0022] Step S204, obtaining team member relationship characteristics.

[0023] The player relationship feature is used to characterize the characteristics of the relationship between the players within each team. The player relationship feature can be obtained based on the player graph network; the player graph network refers to a pre-built graph network used to characterize the player relationship. The players corresponding to the nodes in the graph network are not limited to the same game or the same region. It can be a graph network built based on all the players of each game that can be obtained, such as a player graph network built based on all football players in the football project.

[0024] Among them, the attributes of each node in the player graph network include the attribute characteristics of the corresponding player; the attributes of the edge in the player graph network include the player relationship characteristics between the two players corresponding to the two nodes connected by the edge.

[0025] The player relationship features include the features common to the two players corresponding to the two nodes connected by the edge, and / or the relationship features that characterize the relationship between the two players. For example, in the pre-established graph network of the football project, there is an edge between the player nodes corresponding to the two players, indicating that the two players have a specific relationship. This relationship can be that the two players speak the same language, have the same nationality, or that the two players currently or have played for the same team in the past; accordingly, the player relationship features can be the common language and nationality of the two players, the length of time they have played for the same team, and the relevant attributes of the team that the two players play for (team ranking, game winning rate, etc.).

[0026] In the specific implementation process, in the process of obtaining the player relationship characteristics based on the player graph network, the nodes corresponding to the players of each team can be queried in the player graph network. After the nodes corresponding to the players of each team are queried, the player relationship characteristics of the edges between the queried nodes are read in the player graph network. Among them, in the process of querying the nodes corresponding to the players of each team in the player graph network, the nodes corresponding to each player can be queried in the player graph network according to the attribute characteristics of each player, and other data or characteristics other than the attribute characteristics of each player can also be queried, such as querying the nodes corresponding to each player in the player graph network according to the player identification of each player.

[0027] Step S206: constructing team characteristics of each competition team according to the attribute characteristics of each player and the relationship characteristics of the players.

[0028] In the specific implementation, on the basis of obtaining the attribute characteristics and player relationship characteristics of each player of each competition team, starting from the attribute characteristics and player relationship characteristics of each player at the player level, the overall team characteristics of each competition team at the team level are constructed. The overall team characteristics of each competition team are specifically obtained by constructing the attribute characteristics and player relationship characteristics of each player of the competition team.

[0029] Using the above example, Figure 3 As shown, after the player relationship characteristics of each player of Team A and Team B are obtained based on the player graph network of the football project, the team characteristics of Team A are constructed according to the attribute characteristics and player relationship characteristics of all the players of Team A. During the construction process, the attribute characteristics and player relationship characteristics of each player of Team A can be arranged in the form of a matrix, and the attribute characteristics and player relationship characteristics of each player are a matrix row. After the construction is completed, the team characteristics of Team A are obtained; similarly, the team characteristics of Team B are constructed.

[0030] Step S208, inputting the team characteristics of each competition team into a pre-trained competition evaluation model to perform competition evaluation processing to obtain an evaluation result.

[0031] The game evaluation model described in this embodiment refers to a model obtained by pre-training for game evaluation processing. The game evaluation processing can be an evaluation of the entire game, or an evaluation of the performance or evaluation of each team in the game, or an evaluation of the overall performance or overall evaluation of all teams in the game.

[0032] Optionally, the game evaluation process includes: evaluating the game score of each team, evaluating the overall score of the players of each team, evaluating the user evaluation of each team, or evaluating the game score. Among them, the overall score of the players refers to the score of the overall performance of the players of each team in the game; the user evaluation includes the viewing user evaluation, which refers to the evaluation of the game by the users who watched the game; the game score can be the score of the overall excitement of the game.

[0033] In addition to the four implementation methods of match evaluation processing provided above, the match evaluation processing performed by the match evaluation model can also be an evaluation of the match or the teams in the match from other angles, such as evaluating the match duration, so as to determine the live broadcast time period of the match based on the evaluated match duration.

[0034] In an optional implementation provided by this embodiment, the match evaluation model is trained in the following manner: the team characteristics of each team in the match sample are input into the model to be trained for match evaluation processing to obtain sample evaluation results; the parameters of the model to be trained are adjusted according to the sample evaluation results and the sample labels of the match sample; for example, the model to be trained is constructed using the Transformer framework, and the match evaluation model is obtained after the model training is completed. Among them, the team characteristics of each team in the match sample are obtained by constructing the attribute characteristics and the relationship characteristics of the players of each team in the match sample.

[0035] In specific implementation, in order to improve the accuracy of the game evaluation process, an attention mechanism can be introduced in the game evaluation process to perform corresponding game evaluation processing. In this case, a corresponding module for performing attention calculation can be set in the game evaluation model, so that in the process of the game evaluation model performing game evaluation processing, the corresponding attention calculation can be performed through the set module, thereby improving the accuracy of the game evaluation processing on the basis of the attention calculation.

[0036] Optionally, the game evaluation model includes: an attention calculation module and a feedforward neural network. The attention calculation module is used to use the attention mechanism to calculate the positive and negative impacts of team features, and input the obtained positive and negative impact features of each team into the feedforward neural network; the feedforward neural network is used to perform game evaluation processing based on the input positive and negative impact features and output the evaluation results. For example, Figure 3 As shown, the positive influence features and negative influence features of team A and team B in the football match are input into a feedforward neural network, and the feedforward neural network performs team score evaluation on the overall team scores of team A and team B according to the input positive influence features and negative influence features of team A and team B, and outputs the overall team scores of team A and team B in the game.

[0037] Among them, the attention calculation module uses the attention mechanism to calculate the positive impact of team characteristics. The purpose is to use the attention mechanism to calculate the degree of cooperation between players within each team in the game. This degree of cooperation helps the team improve its performance in the game, which is called positive impact calculation. The attention calculation module uses the attention mechanism to calculate the negative impact of team characteristics. The purpose is to use the attention mechanism to calculate the degree of confrontation between the teams in the game, that is, to use the attention mechanism to calculate the degree of confrontation between each team in the game relative to other teams. This degree of cooperation may cause a negative impact on the performance of the team in the game, which is called negative impact calculation.

[0038] Specifically, in the process of using the attention mechanism to calculate the positive impact of team characteristics, the multi-head attention mechanism can be used to calculate the positive impact of the team characteristics of each competition team, that is, the multi-head attention mechanism is used to calculate the degree of cooperation between the players within each competition team in the game, so that the calculated positive impact characteristics can more accurately represent the degree of cooperation between the players within the competition team in the game.

[0039] In an optional implementation provided by this embodiment, an attention mechanism is used to calculate the positive impact of team characteristics, including: Performing feature transformation on the team features of each competition team according to the attention parameter to obtain transformed features; A multi-head attention mechanism is used to calculate the transformation features and obtain the positive impact features of each team.

[0040] Among them, transforming the features of the team characteristics of each competing team according to the attention parameters means that according to the three parameters Query, Key and Value required for the multi-head attention calculation, the team characteristics of each competing team are respectively transformed into query features, key features and value features to obtain query features, key features and value features, and further, the query features, key features and value features are input into the multi-head attention algorithm for calculation, and the features obtained by calculation are the positive influence features.

[0041] Furthermore, in the process of calculating the transformed features using a multi-head attention mechanism, a scaled dot product attention algorithm can be used within the multi-head attention mechanism for calculation. The scaled dot product attention algorithm can be used to measure the degree of mutual cooperation between all players within the team, so that the positive influence features obtained on this basis are more accurate. Specifically, in an optional implementation provided by this embodiment, each head of the multi-head attention mechanism performs the following coordination calculation: the query features, key features, and value features contained in the transformed features are input into the scaled dot product attention algorithm for calculation to obtain the attention sub-features of each head. In this case, the positive influence features can be obtained by splicing and converting the attention sub-features of each head.

[0042] Continuing with the above example, after obtaining the team characteristics of team A and team B, according to the three parameters Query, Key and Value required for the multi-head attention calculation, the query feature transformation, key feature transformation and value feature transformation are performed on the team characteristics of team A and team B respectively to obtain the query features, key features and value features of team A and team B respectively. Further, the query features, key features and value features of team A are input into the multi-head attention algorithm for calculation, and the positive influence features of team A are output; similarly, the query features, key features and value features of team B are input into the multi-head attention algorithm for calculation, and the positive influence features of team B are output.

[0043] In addition, in the process of using the attention mechanism to calculate the negative impact of team characteristics, a multi-head attention mechanism can be used to calculate the negative impact of the team characteristics of each team. That is, the multi-head attention mechanism is used to calculate the degree of confrontation of each team in the game relative to other teams, so that the calculated negative impact characteristics can more comprehensively express the degree of confrontation between the teams in the game.

[0044] The following takes the competition teams including the first competition team and the second competition team as an example, and explains the calculation process of using the multi-head attention mechanism to calculate the negative impact characteristics of the first competition team relative to the second competition team in the competition, and the calculation process of the negative impact characteristics of the second competition team relative to the first competition team in the competition.

[0045] (1) In the process of using a multi-head attention mechanism to calculate the negative impact features of the first team relative to the second team in the game, the multi-head attention calculation is performed using the team features of the first team as a query and the team features of the second team as a key and a value. Specifically, in an optional implementation provided by this embodiment, the attention mechanism is used to calculate the negative impact of team features, including: Performing query feature transformation on the team feature of the first competition team to obtain query features, and performing key feature transformation and value feature transformation on the team feature of the second competition team to obtain key features and value features; A multi-head attention mechanism is used to calculate the query features, key features, and value features to obtain the negative impact features of the first-place team.

[0046] Among them, in the process of using the multi-head attention mechanism to calculate the negative impact characteristics of the first competition team relative to the second competition team in the game, in order to conduct a more fine-grained analysis of the degree of confrontation between the first competition team and the second competition team at the level of the first competition team and the second competition team, the scaled dot product attention algorithm can be used within the multi-head attention mechanism to calculate the degree of confrontation between all players within the competition team, that is, the scaled dot product attention algorithm is used to more accurately measure the degree of confrontation between the players of the first competition team and the second competition team.

[0047] Specifically, in an optional implementation provided by this embodiment, each head of the multi-head attention mechanism performs the following adversarial calculation: the query feature, key feature, and value feature are input into the scaled dot product attention algorithm for calculation to obtain the attention sub-feature of each head. In this case, the negative impact feature can be obtained by concatenating and transforming the attention sub-features calculated by each head.

[0048] Still taking the above football game as an example, after obtaining the overall team characteristics of team A and team B, in the process of calculating the negative impact of team A relative to team B in the game, the team characteristics of team A are transformed into query features to obtain query features, and the team characteristics of team B are transformed into key features and value features to obtain key features and value features. Furthermore, the obtained query features, key features and value features are input into the multi-head attention algorithm for calculation, and the negative impact features representing the degree of confrontation between team A and team B in the game are output.

[0049] (2) In the process of using the multi-head attention mechanism to calculate the negative impact features of the second team relative to the first team in the game, the multi-head attention calculation is performed using the team features of the second team as a query and the team features of the first team as a key and a value. Specifically, in an optional implementation provided by this embodiment, the attention mechanism is used to calculate the negative impact of team features, including: Performing query feature transformation on the team features of the second competition team to obtain query features, and performing key feature transformation and value feature transformation on the team features of the first competition team to obtain key features and value features; A multi-head attention mechanism is used to calculate the query features, key features, and value features to obtain the negative impact features of the second-place team.

[0050] Similar to the above-mentioned process of using the multi-head attention mechanism to calculate the negative impact characteristics of the first competition team relative to the second competition team, in the process of using the multi-head attention mechanism to calculate the negative impact characteristics of the second competition team relative to the first competition team, the scaled dot product attention algorithm can also be used within the multi-head attention mechanism to calculate the degree of confrontation between all players within the competition team.

[0051] It should be noted that the above-mentioned positive and negative influence calculations on team characteristics using the attention mechanism can also be performed on only one of the positive and negative influence calculations as needed in actual scenarios, that is, the implementation method of using the attention mechanism to perform positive and negative influence calculations on team characteristics can be replaced by using the attention mechanism to perform positive influence calculations on team characteristics to obtain positive influence characteristics, or, it can also be replaced by using the attention mechanism to perform negative influence calculations on team characteristics to obtain negative influence characteristics; on the basis of obtaining the positive influence characteristics or negative influence characteristics of each competition team, the positive influence characteristics or negative influence characteristics of each competition team can be input into the feedforward neural network.

[0052] In actual applications, on the basis of obtaining evaluation results through game evaluation processing, the evaluation results can be further processed according to the needs of actual application scenarios. For example, in a game commentary scenario, a game commentary label can be generated based on the overall player score, viewing user evaluation and / or game score. Alternatively, a game commentary text can be generated based on the overall player score, user evaluation and / or game score. Alternatively, a commentary label or commentary text can be generated based on the evaluation results (at least one of the overall player score and user evaluation) and game information.

[0053] In summary, the match evaluation processing method provided in this embodiment, on the one hand, extracts features from the attribute data of the players of the match teams to obtain the attribute features of each player, and on the other hand, obtains the player relationship features of each player based on the player graph network. On the basis of obtaining the attribute features and player relationship features of each player at the player level, the overall team features of each match team at the match team level are constructed according to the attribute features and player relationship features of each player, that is, the team features at the match team level are constructed from the attribute features and player relationship features at the player level, and the constructed team features of each match team are input into the match evaluation model. The match evaluation model performs match evaluation processing according to the input team features of each match team to obtain an evaluation result, thereby combining the data of the players of the match teams in multiple dimensions to perform match evaluation, which helps to improve the accuracy of match evaluation. Furthermore, in the process of match evaluation processing, the match evaluation model can use the attention mechanism to calculate the positive and negative influence of team characteristics, obtain the positive and negative influence characteristics of each team, and input the positive and negative influence characteristics of each team into the feedforward neural network match evaluation processing, so as to perform match evaluation processing on the basis of considering the positive and negative influences between the players of the teams in the game, thereby further improving the accuracy of match evaluation.

[0054] The following uses the application of a match evaluation processing method provided by this embodiment in a football match scenario as an example to further illustrate the match evaluation processing method provided by this embodiment. Figure 4 , a match evaluation processing method applied to a football match scenario specifically includes the following steps.

[0055] Step S402: extracting features from the player's field position data of each player of the team in the football match to obtain player's field position features.

[0056] Step S404: extracting features from the individual data of each player to obtain individual features of the player.

[0057] Step S406, inputting the player's court position data and the player's personal data into the large language model to obtain the coordinated defense ability characteristics, thereby obtaining the player's coordinated defense ability characteristics.

[0058] Step S408, obtaining the player relationship characteristics of each player based on the player graph network.

[0059] Step S410, constructing team characteristics of each team based on the player's field position characteristics, player's individual characteristics, player's defensive coordination ability characteristics and player's relationship characteristics.

[0060] Step S412, inputting the team characteristics of each team into the game evaluation model to perform team score evaluation and obtain the overall team score of each team.

[0061] It should be noted that any one of steps S402 to S412 or any combination of multiple steps can be combined with any one of steps S202 to S208 to form a new implementation method according to the needs of implementation deployment; in addition, according to the needs of actual deployment, any one or multiple technical features can be selected from steps S402 to S412 and combined with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method; or, any one or multiple technical features in steps S402 to S412 can be replaced with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method according to the needs of actual deployment, which will not be repeated here.

[0062] An embodiment of a competition evaluation processing device provided in this specification is as follows: In the above-mentioned embodiment, a competition evaluation processing method is provided, and correspondingly, a competition evaluation processing device is also provided, which will be described below with reference to the accompanying drawings.

[0063] Reference Figure 5 , which shows a schematic diagram of an embodiment of a competition evaluation processing device provided in this embodiment.

[0064] Since the device embodiment corresponds to the method embodiment, the description is relatively simple, and the relevant parts can refer to the corresponding description of the method embodiment provided above. The device embodiment described below is only illustrative.

[0065] This embodiment provides a competition evaluation processing device, the device comprising: The attribute feature extraction module 502 is configured to extract features from the attribute data of the players of the competition team to obtain the attribute features of each player; The relationship feature extraction module 504 is configured to obtain the team member relationship features, which are obtained based on the team member graph network. A feature construction module 506 is configured to construct team features of each competition team according to the attribute features of each player and the player relationship features; The evaluation processing module 508 is configured to input the team characteristics of each competition team into a pre-trained competition evaluation model to perform competition evaluation processing and obtain an evaluation result.

[0066] An embodiment of a game evaluation processing device provided in this specification is as follows: Corresponding to the above-described competition evaluation processing method, based on the same technical concept, one or more embodiments of this specification further provide a competition evaluation processing device, which is used to execute the above-described competition evaluation processing method. Figure 6 A schematic diagram of the structure of a competition evaluation processing device provided for one or more embodiments of this specification.

[0067] This embodiment provides a competition evaluation processing device, including: like Figure 6 As shown, the game evaluation processing device may have relatively large differences due to different configurations or performances, and may include one or more processors 601 and memory 602, and the memory 602 may store one or more storage applications or data. Among them, the memory 602 may be a short-term storage or a persistent storage. The application stored in the memory 602 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the game evaluation processing device. Furthermore, the processor 601 may be configured to communicate with the memory 602, and execute a series of computer executable instructions in the memory 602 on the game evaluation processing device. The game evaluation processing device may also include one or more power supplies 603, one or more wired or wireless network interfaces 604, one or more input / output interfaces 605, one or more keyboards 606, etc.

[0068] In a specific embodiment, the game evaluation processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer executable instructions in the game evaluation processing device, and the one or more programs are configured to be executed by one or more processors, including computer executable instructions for performing the following: Extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player; Acquire a team member relationship feature, wherein the team member relationship feature is obtained based on a team member graph network; According to the attribute characteristics and relationship characteristics of each player, the team characteristics of each competition team are constructed; the team characteristics of each competition team are input into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result.

[0069] An embodiment of a computer-readable storage medium provided in this specification is as follows: Corresponding to the above-described competition evaluation processing method, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0070] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and the computer-executable instructions implement the following process when executed: Extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player; Acquire a team member relationship feature, wherein the team member relationship feature is obtained based on a team member graph network; According to the attribute characteristics and relationship characteristics of each player, the team characteristics of each competition team are constructed; the team characteristics of each competition team are input into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result.

[0071] It should be noted that an embodiment of a computer-readable storage medium in this specification and an embodiment of a competition evaluation processing method in this specification are based on the same inventive concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned corresponding method, and the repeated parts will not be repeated.

[0072] An embodiment of a computer program product provided in this specification is as follows: Corresponding to the above-described competition evaluation processing method, based on the same technical concept, one or more embodiments of this specification also provide a computer program product.

[0073] A computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the following steps: Extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player; Acquire a team member relationship feature, wherein the team member relationship feature is obtained based on a team member graph network; According to the attribute characteristics and relationship characteristics of each player, the team characteristics of each competition team are constructed; the team characteristics of each competition team are input into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result.

[0074] It should be noted that an embodiment of a computer program product in this specification and an embodiment of a competition evaluation processing method in this specification are based on the same inventive concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned corresponding method, and the repeated parts will not be repeated.

[0075] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referenced to each other. Each embodiment focuses on the differences from other embodiments. For example, the device embodiment, equipment embodiment, and computer-readable storage medium embodiment are similar to the method embodiment, so the description is relatively simple. To read the relevant content in the device embodiment, equipment embodiment, and computer-readable storage medium embodiment, please refer to the partial description of the method embodiment.

[0076] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0077] In the 1930s, improvements to a technology could be clearly distinguished as hardware improvements (for example, improvements to circuit structures such as diodes, transistors, switches, etc.) or software improvements (improvements to method flows). However, with the development of technology, many of today's method flow improvements can be regarded as direct improvements to hardware circuit structures. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that an improvement in a method flow cannot be implemented using hardware entity modules. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to ask chip manufacturers to design and produce dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one kind of HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0078] The controller may be implemented in any suitable manner, for example, the controller may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) executable by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller may also be implemented as part of the control logic of the memory. It is also known to those skilled in the art that, in addition to implementing the controller in a purely computer-readable program code manner, the controller may be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, such a controller may be considered as a hardware component, and the devices for implementing various functions included therein may also be considered as structures within the hardware component. Or even, the devices for implementing various functions may be considered as both software modules for implementing the method and structures within the hardware component.

[0079] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0080] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing the embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0081] It should be understood by those skilled in the art that one or more embodiments of the present specification may be provided as a method, system or computer program product. Therefore, one or more embodiments of the present specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present specification may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0082] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0086] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0087] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0088] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of further restrictions, the elements defined by the sentence "includes at least one ..." do not exclude the presence of other identical elements in the process, method, commodity or device including the elements.

[0089] One or more embodiments of the present specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. One or more embodiments of the present specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0090] The above description is only an embodiment of this document and is not intended to limit this document. For those skilled in the art, this document may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this document should be included in the scope of the claims of this document.

Claims

1. A competition evaluation processing method, comprising: Extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player; Acquire a team member relationship feature, wherein the team member relationship feature is obtained based on a team member graph network; Constructing team characteristics of each competition team according to the attribute characteristics of each player and the characteristics of the relationship between the players; The team characteristics of each competition team are input into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result.

2. According to the match evaluation processing method of claim 1, the attribute characteristics include at least one of the following: player's court position characteristics, player's personal characteristics, and player's defensive coordination ability characteristics.

3. The match evaluation processing method according to claim 2, wherein the attribute data includes player court position data and player personal data; The player's defensive coordination capability characteristics are obtained in the following manner: The player's court position data and the player's personal data are input into a large language model to obtain coordinated defensive ability characteristics, thereby obtaining the player's coordinated defensive ability characteristics.

4. The competition evaluation processing method according to claim 1, wherein the competition evaluation processing comprises: Evaluate the game score of each team, evaluate the overall score of the players of each team, evaluate the user evaluation of each team, or evaluate the game score.

5. According to the competition evaluation processing method of claim 1, the competition evaluation model is trained in the following manner: Input the team characteristics of each team in the competition sample into the to-be-trained model for competition evaluation processing to obtain the sample evaluation result; Adjusting parameters of the model to be trained according to the sample evaluation results and the sample labels of the competition samples; in, The team characteristics of each competition team of the competition sample are obtained by constructing the attribute characteristics and the relationship characteristics of the players of each competition team of the competition sample.

6. The competition evaluation processing method according to claim 1, wherein the competition evaluation model comprises: Attention calculation module and feedforward neural network; The attention calculation module is used to use the attention mechanism to calculate the positive influence and negative influence of the team characteristics, and input the obtained positive influence characteristics and negative influence characteristics of each competition team into the feedforward neural network; The feedforward neural network is used to perform competition evaluation processing according to the input positive impact features and negative impact features and output the evaluation results.

7. The game evaluation processing method according to claim 6, wherein the step of using an attention mechanism to perform positive impact calculation on the team characteristics comprises: Performing feature transformation on the team features of each of the competition teams according to the attention parameters, Obtain transformation features; A multi-head attention mechanism is used to calculate the transformation features to obtain the positive impact features of each competing team.

8. According to the competition evaluation processing method of claim 7, each head of the multi-head attention mechanism performs the following calculation: Input the query feature, key feature and value feature included in the transformed feature into the scaled dot product attention algorithm for calculation to obtain the attention sub-feature of each head; in, The positive influence feature is obtained by concatenating and transforming the attention sub-features of each head.

9. The game evaluation processing method according to claim 6, wherein the step of using an attention mechanism to calculate the negative impact of the team characteristics comprises: Performing query feature transformation on the team feature of the first competition team to obtain query features, and performing key feature transformation and value feature transformation on the team feature of the second competition team to obtain key features and value features; A multi-head attention mechanism is used to calculate the query feature, the key feature and the value feature to obtain the negative impact feature of the first competition team.

10. The competition evaluation processing method according to claim 9, wherein each head of the multi-head attention mechanism performs the following calculation: Input the query feature, the key feature, and the value feature into a scaled dot product attention algorithm for calculation to obtain an attention sub-feature of each head; in, The negative impact feature is obtained by concatenating and transforming the attention sub-features obtained by calculating each head.

11. A competition evaluation processing device, comprising: The attribute feature extraction module is configured to extract features from the attribute data of the players of the competition team to obtain the attribute features of each player; A relationship feature extraction module is configured to obtain team member relationship features, wherein the team member relationship features are obtained based on the team member graph network. A feature construction module is configured to construct team features of each competition team according to the attribute features of each player and the player relationship features; The evaluation processing module is configured to input the team characteristics of each competition team into a pre-trained competition evaluation model to perform competition evaluation processing and obtain an evaluation result.

12. A competition evaluation processing device, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Extracting features from the attribute data of the players of the competition team to obtain the attribute features of each player; Acquire a team member relationship feature, wherein the team member relationship feature is obtained based on a team member graph network; Constructing team characteristics of each competition team according to the attribute characteristics of each player and the characteristics of the relationship between the players; The team characteristics of each competition team are input into a pre-trained competition evaluation model for competition evaluation processing to obtain an evaluation result.

13. A computer-readable storage medium for storing computer-executable instructions, wherein the computer-executable instructions implement the steps of the method of claim 1 when executed.

Images