Game matching method and device, electronic equipment and readable storage medium

Abstract

The invention provides a game matching method and device, electronic equipment and a readable storage medium. A matching mode is determined based on the number of participated players in a current game; when the matching mode is a first matching mode, distributing a preset number of to-be-matched game accounts to a target matching room based on the game information of each to-be-matched game account, and dividing a plurality of to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams to obtain a matching result; and when the matching mode is a second matching mode, dividing the to-be-matched game accounts into a second preset number of target matching teams based on the game information of each to-be-matched game account, and dividing target matching rooms according to the matching teams to obtain a matching result. Therefore, matching requirements of players can be met, different matching modes are adopted for different matching modes, matching calculation steps are reduced, and matching efficiency is improved.

Description

Technical Field

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

[0002] Before a game starts, the players participating in the game need to be matched, teammates need to be found for the players, and the enemy team that is in the same game and in a confrontation with the player's team needs to be determined to ensure the progress of the game.

[0003] In the related art, the matching waiting time and the balance of the game are usually used as optimization goals to determine the final matching result. However, the game matching method based on the above factors generally only considers the fairness in the game, but does not consider the matching needs of the players. At the same time, a unified matching method is used in the matching process, which may consume a lot of computing resources and affect the matching efficiency. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a game matching method, device, electronic device and readable storage medium. When matching game accounts to be matched, the game information of the game accounts to be matched is considered to meet the matching needs of the players. At the same time, different matching modes are determined according to the number of players participating in the game game, and different matching methods are used for different matching modes, which reduces the calculation steps in the matching process and helps to improve the matching efficiency.

[0005] In a first aspect, an embodiment of the present application provides a game matching method, the game matching method comprising:

[0006] Determine multiple unmatched game accounts to be matched and game information of each game account to be matched;

[0007] Determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold;

[0008] When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the match information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result;

[0009] When the matching mode is the second matching mode, based on the game information of each game account to be matched, the game accounts to be matched are divided into a second preset number of target matching teams, and the target matching rooms are divided according to the matching teams to obtain matching results.

[0010] In a second aspect, an embodiment of the present application further provides a game matching device, the game matching device comprising:

[0011] A game information determination module, used to determine a plurality of unmatched game accounts to be matched and game information of each game account to be matched;

[0012] A matching mode determination module, configured to determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold;

[0013] A first matching result acquisition module is used for, when the matching mode is the first matching mode, allocating a preset number of to-be-matched game accounts to a target matching room based on the game information of each to-be-matched game account, dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams, and obtaining a matching result;

[0014] The second matching result acquisition module is used to divide the game accounts to be matched into a second preset number of target matching teams based on the game information of each game account to be matched when the matching mode is the second matching mode, and then divide the target matching rooms according to the matching teams to obtain matching results.

[0015] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium through the bus, and the processor executes the machine-readable instructions to perform the game matching method as described in any one of the first aspects.

[0016] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the game matching method as described in any one of the first aspects is executed.

[0017] The game matching method, device, electronic device and readable storage medium provided by the embodiment of the present application determine multiple unmatched game accounts to be matched and the game information of each game account to be matched; determine the matching mode based on the number of players participating in the current game game; when the matching mode is the first matching mode, based on the game information of each game account to be matched, after allocating a preset number of game accounts to be matched to the target matching room, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result; when the matching mode is the second matching mode, based on the game information of each game account to be matched, after dividing the game account to be matched into a second preset number of target matching teams, the target matching room is divided according to the matching teams to obtain a matching result. In this way, when the game accounts to be matched are matched, the game information of the game accounts to be matched is considered to meet the matching needs of the players. At the same time, different matching modes are determined according to the number of players participating in the game game, and different matching methods are adopted for different matching modes, which reduces the calculation steps in the matching process and helps to improve the matching efficiency.

[0018] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A flowchart of a game matching method provided in an embodiment of the present application;

[0021] Figure 2 One of the structural schematic diagrams of the matching system provided in the embodiment of the present application;

[0022] Figure 3 The second structural diagram of the matching system provided in the embodiment of the present application;

[0023] Figure 4 This is one of the structural schematic diagrams of a game matching device provided in an embodiment of the present application;

[0024] Figure 5 The second structural diagram of a game matching device provided in an embodiment of the present application;

[0025] Figure 6A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0027] Before a game starts, the players participating in the game need to be matched, teammates need to be found for the players, and the enemy team that is in the same game and in a confrontation with the player's team needs to be determined to ensure the progress of the game.

[0028] The existing Player Versus Player (PVP) battle system matching mechanism usually takes the matching waiting time and game balance as optimization goals, mainly in the following two ways:

[0029] Solution 1: Matchmaking mechanism based on the Elo Rating System (ELO). This algorithm tries to put players of similar skill levels into the same game. It mainly considers the scores of kills, deaths, and assists (KDA) between the player and other players. It is highly universal. Each game usually makes fine adjustments to the ELO formula based on the situation of its own product.

[0030] Solution 2: A matching system based on Match Marking Rating (MMR), which calculates scores for matched players to reflect their actual combat level. MMR scores are calculated in different ways depending on the matching scenario, usually taking into account factors such as the player's win rate, the strength of the opponent, and personal performance.

[0031] In the related art, the matching waiting time and the balance of the game are usually used as optimization goals to determine the final matching result. However, the game matching method based on the above factors generally only considers the fairness in the game, but does not consider the needs of the game accounts to be matched. At the same time, a unified matching method is used in the matching process, which may consume a lot of computing resources and affect the matching efficiency.

[0032] Based on this, the embodiment of the present application provides a game matching method, which not only meets the matching needs of players, but also adopts different matching methods for different matching modes, reduces the calculation steps in the matching process, and helps to improve matching efficiency.

[0033] See also Figure 1 , Figure 1 This is a flow chart of a game matching method provided in an embodiment of the present application. Figure 1 As shown in , the game matching method provided in the embodiment of the present application includes:

[0034] S101: Determine a plurality of unmatched game accounts to be matched and game information of each of the game accounts to be matched.

[0035] S102. Determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold.

[0036] S103: When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the game information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result.

[0037] S104: When the matching mode is the second matching mode, the game accounts to be matched are divided into a second preset number of target matching teams based on the game information of each game account to be matched, and the target matching rooms are divided according to the matching teams to obtain matching results.

[0038] The game matching method provided in the embodiment of the present application considers the game information of the game accounts to be matched when matching the game accounts to meet the matching needs of the players. At the same time, different matching modes are determined according to the number of players participating in the game game, and different matching methods are used for different matching modes, which reduces the calculation steps in the matching process and helps to improve the matching efficiency.

[0039] The following is a description of the exemplary steps of the present application embodiment:

[0040] S101: Determine a plurality of unmatched game accounts to be matched and game information of each of the game accounts to be matched.

[0041] In an embodiment of the present application, before a game starts, it is necessary to match the players participating in the game, find teammates for the players, and determine the enemy team that is in the same game as the player's team and is in a confrontational relationship, so as to ensure the progress of the game. The matching of players in the game can be carried out through a matching system, and the matching system can be a system independent of the game server. Through the matching of the matching system, the game accounts corresponding to each player participating in the same game and the team division situation can be determined.

[0042] In one possible implementation, see Figure 2 , Figure 2 This is one of the structural diagrams of the matching system provided in the embodiment of the present application. Figure 2 As shown, the matching system 200 includes an interaction processing module 210 and a matching processing module 220. The interaction processing module 210 can receive a matching request sent from a game terminal, and determine the game account information and the game information carried in the matching request according to the matching request sent by the game terminal; the game information of at least one to-be-matched game account included in the matching request is placed in the matching pool. For example, in order to reduce the pressure on the database, the game information of different to-be-matched game accounts can be placed in different databases, and then the game information of the to-be-matched game accounts is distributed according to the matching mode; the matching processing module 220 obtains multiple to-be-matched game accounts that have not been matched from the matching pool and each of the game accounts that have not been matched. The game information of the game accounts to be matched is filtered out through customized rules, and then different matching processes are performed according to different matching modes, and the matching results are pushed to the client; illustratively, the matching result under one matching mode can be determined through a matching prediction model (equivalent to a "prediction vector model"), a game balance model (equivalent to a "game score prediction model") and a post-game satisfaction model (equivalent to a "satisfaction score prediction model"); the matching result under another matching mode can be determined through a teammate matching model (equivalent to a "game score prediction model" combined with a "satisfaction score prediction model") and a matching degree prediction model (equivalent to a "prediction vector model").

[0043] In a possible implementation, in order to reduce the number of game accounts to be matched during the matching process, as well as the steps of dividing matching rooms and matching teams, and to improve the efficiency of the matching process, before matching the game accounts, the game accounts to be matched can be screened according to preset account screening rules.

[0044] Specifically, before the step of "determining a plurality of unmatched game accounts to be matched and the game information of each of the game accounts to be matched", the game matching method further includes:

[0045] a1: Obtain multiple candidate matching game accounts that have not been matched, and the game information of each candidate matching game account.

[0046] a2: Based on the game information of each candidate matching game account, detect whether the candidate matching game account meets the preset account screening rules.

[0047] a3: determining the matching game accounts that do not meet the preset account screening rule among the multiple candidate matching game accounts as the game accounts to be matched.

[0048] In a possible implementation, the candidate matching game accounts that have not been matched include the candidate matching game accounts that have not been successfully matched in the last matching process and the to-be-matched game accounts included in the matching request obtained in real time.

[0049] Specifically, when matching the game accounts of the players, candidate matching game accounts are obtained from the matching pool in the matching system. The candidate matching game accounts in the matching pool include candidate matching game accounts carried in real time through matching requests, and also include candidate matching game accounts that were not successfully matched in the previous matching process.

[0050] In a possible implementation, for candidate matching game accounts carried in real-time matching requests, players can issue matching requests, matching cancellation requests, and re-matching requests through mobile terminals. For different matching requests, the candidate matching game accounts carried in the requests can be added to the matching pool of the matching system (corresponding to matching requests), the candidate matching game accounts carried in the requests can be removed from the matching pool of the matching system (corresponding to matching cancellation requests), and the matching order of the candidate matching game accounts carried in the requests in the matching pool of the matching system can be adjusted (corresponding to re-matching requests).

[0051] Among them, the matching order of the candidate matching game accounts in the matching pool is determined according to the request time of the corresponding matching request. The earlier the request time of the matching request, the higher the matching order of the corresponding candidate matching game account. When the player initiates a re-matching request through the game terminal, the matching order of the corresponding candidate matching game account needs to be adjusted to the end of the matching order of the current multiple candidate matching game accounts.

[0052] Furthermore, for the candidate matching game accounts in the matching pool, it is possible to detect whether the candidate matching game accounts hit the preset account screening rules based on the game information of each candidate matching game account, and then screen out the candidate matching game accounts that hit the preset account screening rules.

[0053] Among them, the setting of the preset account screening rules is to ensure that the competitiveness between the matching teams divided after the matching is balanced as much as possible. Therefore, the preset account screening rules are set for players who affect the fairness of the game. For example, a penalty mechanism can be added for candidate matching game accounts that are idle (such game accounts are divided into the same room or the matching time of such game accounts is increased, etc.); at the same time, candidate matching game accounts for malicious team formation need to be isolated (during the game process, players may form alliances privately to achieve the purpose of improving rankings. This rule can prevent teams that have won the top few places in the same game from entering the same room again) to ensure the balance and fairness of the matching results as much as possible.

[0054] In one possible implementation, the judgment of whether a candidate matching game account hits a preset account screening rule is determined based on game information. For the above screening situation, if it is an idle penalty mechanism, it can be judged based on whether the player's operation information in the game information during the historical game process matches the target virtual character controlled by the player (for example, the player has not performed any control operations for a long time, but the target virtual character controlled by the player is still performing operations on its own). For the team isolation mechanism, it can be determined based on the interaction information between the player and other players during the game, as well as the game performance information, to determine whether it is necessary to group and isolate the teams or players that ranked top in the same game.

[0055] In a possible implementation, the preset account screening rules can be updated or added according to different player matching requirements. Specifically, the matching processing module of the matching system provides an add / delete interface for the preset account screening rules, and the preset account screening rules are added by adding a yml or xml configuration file.

[0056] Furthermore, after the candidate matching game accounts in the matching pool are screened according to the preset account screening rules, the remaining multiple game accounts to be matched in the matching pool and the game information of each game account to be matched are obtained to perform a subsequent matching process.

[0057] In a possible implementation, the game information of the game account to be matched includes at least one of the following:

[0058] Game performance information, in-game and out-game interaction information, and competitive level information.

[0059] Here, the game information of the game account to be matched represents the performance and competitive level of the player corresponding to the game account to be matched in the game from different aspects, so as to ensure the balance and accuracy of the matching result based on the game performance information and competitive level information, and at the same time combine the interaction information inside and outside the game to ensure that the matching result can meet the player's matching needs, and at the same time improve the player's satisfaction with the matching result.

[0060] Exemplarily, game performance information may include information such as kills, assists, damage, scores, and movement distance of the player in the game within a preset time period; in-game and out-of-game interaction information may include in-game cooperation, rescue, microphone, likes, reports, re-teaming, sending friend requests, blacklisting, deleting friends, and other interaction information; competitive level information may include information such as the player's win rate, rank, and score in a historical time period.

[0061] Furthermore, after obtaining multiple unmatched game accounts to be matched and the game information of each game account to be matched, it is necessary to determine the current matching mode. Different matching modes correspond to different matching rooms and the order of division of matching teams, thereby minimizing the matching calculation steps and computing power when matching the game accounts to be matched, thereby improving the player matching efficiency.

[0062] S102: Determine a matching mode based on the number of players participating in the current game.

[0063] In one possible implementation, the matching mode is determined based on the number of players participating in the game, and the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold.

[0064] For example, the first matching mode may be for multiplayer online tactical competitive games, and the second matching mode may be for battle royale games. For multiplayer online tactical competitive games, the number of players involved in a game is not large, and the matching result can be obtained by first determining a matching room, then dividing the game accounts to be matched in the matching room, and determining the matching team method; while for battle royale games, the number of players participating in a game at the same time is large, and the matching result can be obtained by first dividing the game accounts to be matched into multiple matching teams, and then determining the matching room according to the multiple matching teams.

[0065] S103: When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the game information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result.

[0066] In a possible implementation, when determining the target matching room, any to-be-matched game account in the matching pool may be first placed in an empty room (target matching room), and then all current to-be-matched game accounts in the matching pool may be traversed, and the to-be-matched game accounts to be assigned to the target matching room may be determined based on the degree of matching between the to-be-matched game accounts in the matching pool and the to-be-matched game accounts in the target matching room.

[0067] Specifically, the step of “when the matching mode is the first matching mode, allocating a preset number of to-be-matched game accounts to the target matching room based on the game information of each to-be-matched game account” includes:

[0068] b1: When the matching mode is the first matching mode, for each to-be-matched game account, based on the game information of the to-be-matched game account, a game feature vector of the to-be-matched game account is determined.

[0069] b2: Based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room.

[0070] In the embodiment of the present application, for each to-be-matched game account, when determining the game feature vector of the to-be-matched game account, it can be determined by a pre-trained prediction vector model.

[0071] Specifically, in order to take into account the personalized preferences and habits of each game account to be matched, and to take into account the satisfaction of the players in the game with the overall game, the behaviors of in-game interaction and post-game interaction can be used as fitting targets to train the model. Matches with positive behaviors such as high degree of in-game cooperation, post-game likes, adding friends, and continuing to form teams are selected as positive samples, and negative samples are selected from matching teams with no in-game interaction, reporting, blacklisting, etc., and the ratio of positive and negative samples can be selected based on the offline AUC (areaunder curve, an evaluation indicator for model classification tasks) indicator.

[0072] In a possible implementation, considering the high complexity of calculating the matching scores of all candidate matching units and players in the room, a game satisfaction index is used to train the EGES model to obtain a prediction vector model.

[0073] Furthermore, for each game account to be matched, based on the game information of the game account to be matched, the interaction information between the game account to be matched and other players inside and outside the game, such as rescuing, opening the microphone, liking, reporting, re-teaming, sending friend requests, blacklisting, and deleting friends, is determined, and the number of times the game account to be matched rescues, opens the microphone, likes, reports, re-teams, sends friend requests, blacklists, and deletes friends in the game is used as the input feature of the game account to be matched, and is input into the prediction vector model to obtain the game feature vector of the game account to be matched.

[0074] Exemplarily, the game feature vector may be an embedding vector, which is an N-dimensional real-valued vector that represents the input data as a point in a continuous numerical space.

[0075] Furthermore, after determining the game feature vectors of each matching game account, the similarity between the game feature vector of each to-be-matched game account and the game feature vectors of other to-be-matched game accounts is calculated, and then it is determined that the target matching room can be allocated.

[0076] In a possible implementation, when calculating the similarity between the game feature vector of each matching game account and the game feature vectors of other game accounts to be matched, any game account to be matched in the matching pool can be placed in the target matching room, and then the other game accounts to be matched in the matching pool are traversed to calculate the similarity between the game feature vectors of the two. Depending on the number of game accounts to be matched contained in the target matching room, there are certain differences in calculating the similarity between the game feature vectors and determining the game accounts to be matched assigned to the target matching room, which will be explained separately below.

[0077] Specifically, based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room, including:

[0078] c1: When the target matching room includes a target matching game account, the similarities between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account are calculated respectively, and the to-be-matched game account with the highest similarity with the target matching game account among the other to-be-matched game accounts is assigned to the target matching room.

[0079] c2: When the target matching room includes multiple target matching game accounts, for each target matching game account, the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account is calculated respectively, and the to-be-matched game account with the highest sum of similarities with each target matching game account among the other to-be-matched game accounts is allocated to the target matching room.

[0080] In a possible implementation, when the target matching room includes a target matching game account, the similarities between the game feature vectors of other to-be-matched game accounts in the matching pool and the game feature vector of the target matching game account are calculated respectively, and other to-be-matched game accounts with the highest similarity to the game feature vector of the target matching game account are screened out from the matching pool and assigned to the target matching room.

[0081] Among them, the calculation method for the similarity between the game feature vectors of other game accounts to be matched and the game feature vectors of the target matching game account can be calculated by cosine similarity, Euclidean distance, Manhattan distance, Minkowski distance, Chebyshev distance and other calculation methods, which are not specifically limited in this application.

[0082] In another possible implementation, if the target matching room includes multiple target matching game accounts, it is necessary to use the game feature vector of the target matching game account to calculate the similarity between the game feature vectors of other to-be-matched game accounts in the matching pool. For each other to-be-matched game account, after calculating the distance between the target matching game account in the target matching room, the calculated similarities are added up, and the to-be-matched game account with the highest sum of similarities is allocated to the target matching room. The above process is repeated until the number of to-be-matched game accounts included in the target matching room meets the requirement for the number of players in the matching room in the game game.

[0083] For example, if there are two target matching game accounts in the current target matching room, it is necessary to traverse each other to-be-matched game account in the matching pool, and calculate the similarity between the game feature vector of each other to-be-matched game account and the game feature vector of the target matching game account. When calculating the matching degree of each other to-be-matched game account with the target game account in the target matching room, the two similarities are added to obtain the result.

[0084] For example, if it is necessary to determine that 4 to-be-matched game accounts are to be added to the target matching room, at this time, there are 10 to-be-matched game accounts in the matching pool. First, any to-be-matched game account 1 is selected from the 10 to-be-matched game accounts to be added to the target matching room. It is necessary to calculate the similarities between the game feature vectors of the remaining 9 other to-be-matched game accounts and the game feature vector of the to-be-matched game account 1, and add the to-be-matched game account 2 with the highest similarity to the target matching room; further, calculate the similarities between the game feature vectors of the remaining 8 other to-be-matched game accounts and the game feature vector of the to-be-matched game account 1, and calculate the similarities between the game feature vectors of the remaining 8 other to-be-matched game accounts and the game feature vector of the to-be-matched game account 2, and add the to-be-matched game account with the highest sum of the similarities between the game feature vector of the to-be-matched game account 1 and the game feature vector of the to-be-matched game account 2. Matching game account 3 is added to the target matching room; further, the similarities between the game feature vectors of the remaining 7 other game accounts to be matched and the game feature vector of the game account 1 to be matched are calculated, the similarities between the game feature vectors of the remaining 7 other game accounts to be matched and the game feature vector of the game account 2 to be matched are calculated, and the similarities between the game feature vectors of the remaining 7 other game accounts to be matched and the game feature vector of the game account 3 to be matched are calculated. The game account 4 to be matched, which has the highest sum of the similarities with the game feature vector of the game account 1 to be matched, the similarities with the game feature vector of the game account 2 to be matched, and the similarities with the game feature vector of the game account 3 to be matched, is added to the target matching room. At this time, the number of game accounts to be matched in the current target matching room is 4, which meets the room number requirement, and the process of allocating the game accounts to be matched to the target matching room is completed.

[0085] Furthermore, after the process of allocating the to-be-matched game accounts to the target matching rooms is completed, it is necessary to perform a process of dividing the to-be-matched game accounts in the target matching game accounts into matching teams, and then obtain matching results.

[0086] Specifically, the step of “dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams to obtain matching results” includes:

[0087] d1: Determine a plurality of candidate matching teams based on the first preset number, the number of game accounts to be matched included in the target matching room, and the game account teaming information.

[0088] d2: For each candidate matching team, based on the game feature information of each to-be-matched game account included in the candidate matching team, determine the game score and interaction score of the candidate matching team.

[0089] d3: multiple candidate matching teams are matched according to the first preset number to determine multiple matching team groups.

[0090] d4: For each matching team group, the balance score and satisfaction score of the matching team group are calculated based on the game scores and interaction scores of each candidate matching team included in the matching team group.

[0091] d5: Get the matching result based on the matching team group with the highest sum of balance scores and satisfaction scores among multiple matching team groups.

[0092] In an embodiment of the present application, first, based on the first preset number of divided target matching teams, the number of game accounts to be matched included in the target matching room, and the game account teaming information, the number of game accounts to be matched included in each matching team that can be combined in the current target matching room and multiple candidate matching teams are determined.

[0093] Specifically, firstly, the number of game accounts to be matched included in each matching team is determined according to the first preset number of the target matching team and the number of game accounts to be matched included in the target matching room.

[0094] For example, if the number of game accounts to be matched included in the current target matching room is 10, and the first preset number of target matching teams to be divided is 2, then it can be determined that the number of game accounts to be matched included in each matching team is 5.

[0095] In a possible implementation manner, the first preset number is generally set to 2.

[0096] Furthermore, when listing the candidate matching teams that can be formed by the to-be-matched game accounts in the target matching room, the game account team information also needs to be considered. This game account team information can be related to the number of to-be-matched game accounts included in the matching request, that is, if multiple to-be-matched game accounts are included in the same matching request, it is necessary to determine the above multiple to-be-matched game accounts as a matching unit. When performing game account matching, the above multiple to-be-matched game accounts must be combined in a matching team. Based on the above matching factors, multiple candidate matching teams that better meet the players' team formation needs are determined.

[0097] Furthermore, after determining multiple candidate matching teams that may be formed in the target matching room, the game score and interaction score of each candidate matching team are determined, and then different candidate matching teams are matched according to the first preset number, generally in pairs, to obtain multiple matching team groups, and the balance score and satisfaction score of each matching team group are calculated, and then the matching team group with the highest sum of the balance score and the satisfaction score is determined as the final matching team grouping to obtain the matching result.

[0098] Among them, calculating the match score and interaction score of each candidate matching team is the key to determining the final matching team group. The calculation process of the match score and interaction score will be explained below.

[0099] In a possible implementation, the match scores of the candidate matching teams are calculated:

[0100] Specifically, for each candidate matching team, the match score of the candidate matching team is determined by the following steps:

[0101] e1: For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team and the attribute information of each to-be-matched game account are input into a pre-trained game score prediction model to obtain the game score of the candidate matching team.

[0102] In a possible implementation, the attribute information of the game account to be matched may include player age, game time, online habits, and other portrait feature information of the player corresponding to the game account.

[0103] In one possible implementation, when training a game score prediction model, the score difference of game data excluding idle, immediate quit, and short game time is used as the fitting target, and the model features include attribute information such as player age, game time, and online habits; recent game performance information such as kills, assists, damage, scores, and movement distance; and competitive level information such as win rate, rank, and score. At the same time, the mean, variance, extreme value, and other aggregate features and difference features of the above features between the to-be-matched game accounts in the same matching team, as well as the mean, variance, extreme value, and other aggregate features and difference features of the above features between different matching teams must also be considered.

[0104] In a possible implementation, the game score prediction model may be a long short-term memory network (LSTM), and the LSTM time series model is used to predict the changing trend of the game score of the to-be-matched game account.

[0105] Furthermore, when calculating the game scores of the candidate matching teams, the player attribute characteristics, game performance characteristics, and competitive level characteristics of the players corresponding to each to-be-matched game account included in the candidate matching teams are input into the LSTM model, and the model predicts the game scores of each candidate matching team.

[0106] In another possible implementation, the interaction scores of the candidate matching teams are calculated:

[0107] Specifically, for each candidate matching team, the interaction score of the candidate matching team is determined by the following steps:

[0108] f1: For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team is input into the pre-trained satisfaction score prediction model to obtain the interaction score of the candidate matching team.

[0109] In one possible implementation, when training the satisfaction score prediction model, the likes, negative reviews, reports and other behaviors of the to-be-matched game account on the game and the cooperation and interaction behaviors within the game are quantified (generally the number of times the to-be-matched game account performs the above behaviors in the game) to obtain the satisfaction of the game, and the game satisfaction is used as the fitting target. The player's recent relevant data is used as the model feature, including the player's recent behavior data such as adding friends, liking, inviting teams, etc.; the distribution of the proportion of interactive behaviors such as joint debugging and microphone opening with non-friend players within the game; recent game performance, long-term game performance, rank, score and other competitive level information related to the instance, etc.

[0110] In a possible implementation, the satisfaction score prediction model may be a lightweight gradient boosting algorithm network (Light Gradient Boosting Machine, LightGBM).

[0111] Furthermore, when calculating the interaction scores of candidate matching teams, the in-game and out-of-game interaction features and competitive level features of each to-be-matched game account included in the candidate matching team are input into the LightGBM model, and the model predicts the interaction score of each candidate matching team.

[0112] Furthermore, in a possible implementation, after calculating the game scores of each matching candidate team, the matching candidate teams are combined in pairs to obtain a matching team group, and the balance score of the matching team group is determined by the game score of each matching candidate team in the matching team group. Specifically, the negative of the sum of the game score differences between the two matching candidate teams in the matching team group is used as the balance score of the matching team group.

[0113] In another possible implementation, after the interaction scores of the matching candidate teams are calculated, the matching candidate teams are combined in pairs to obtain a matching team group, and the satisfaction score of the matching team group is determined by the interaction score of each matching candidate team in the matching team group.

[0114] Furthermore, the matching team group with the highest sum of balance score and satisfaction score is determined as the final matching grouping result.

[0115] Exemplarily, candidate matching team 1, candidate matching team 2, candidate matching team 3 and candidate matching team 4 are determined in the target matching room; among them, the game score of candidate matching team 1 is 3, and the interaction score is 4; the game score of candidate matching team 2 is 5, and the interaction score is 1; the game score of candidate matching team 3 is 8, and the interaction score is 5; the game score of candidate matching team 4 is 4, and the interaction score is 7.

[0116] The above four candidate matching teams are combined in pairs to obtain matching team group 1: (candidate matching team 1, candidate matching team 2); matching team group 2: (candidate matching team 1, candidate matching team 3); matching team group 3: (candidate matching team 1, candidate matching team 4); matching team group 4: (candidate matching team 2, candidate matching team 3); matching team group 5: (candidate matching team 2, candidate matching team 4); matching team group 6: (candidate matching team 3, candidate matching team 4); After calculation, it can be seen that the balance score of matching team group 1 is -2, the satisfaction score is -3, and the sum of the balance score and the satisfaction score is -5; the balance score of matching team group 2 is -5, the satisfaction score is -1, and the sum of the balance score and the satisfaction score is -6; matching team The balance score of matching team group 3 is -1, and the satisfaction score is -3, and the sum of the balance score and the satisfaction score is -4; the balance score of matching team group 4 is -3, and the satisfaction score is -4, and the sum of the balance score and the satisfaction score is -7; the balance score of matching team group 5 is -1, and the satisfaction score is -6, and the sum of the balance score and the satisfaction score is -7; the balance score of matching team group 6 is -4, and the satisfaction score is -3, and the sum of the balance score and the satisfaction score is -7; the matching team group with the highest sum of balance scores and satisfaction scores among the multiple matching team groups is matching team group 3. At this time, the final matching result is determined to be candidate matching team 1, candidate matching team 4, and each to-be-matched game account included in candidate matching team 1 and candidate matching team 4.

[0117] S104: When the matching mode is the second matching mode, the game accounts to be matched are divided into a second preset number of target matching teams based on the game information of each game account to be matched, and the target matching rooms are divided according to the matching teams to obtain matching results.

[0118] In an embodiment of the present application, for the second matching mode in which the number of players participating in the game match is greater than a preset participation number threshold, if the matching room is determined first, when calculating the match room that can be assigned, a large amount of calculation is required when calculating the match feature vector of each game account to be matched, which consumes a lot of matching time, thereby affecting the matching efficiency. Based on this, in the second matching mode, matching calculations can be performed first between the game accounts to be matched to determine a limited number of target matching teams, and then the matching results are obtained by calculating the matching degree between each target matching team, which can reduce the number of calculations for the matching degree, reduce the matching time, and thus improve the matching efficiency.

[0119] Specifically, the step of “when the matching mode is the second matching mode, based on the game information of each game account to be matched, dividing the game accounts to be matched into a second preset number of target matching teams” includes:

[0120] g1: When the matching mode is the second matching mode, a plurality of matching teams are determined based on the second preset number and the number of game accounts of the game accounts to be matched.

[0121] g2: For each divided matching team, based on the game feature information of each to-be-matched game account included in the divided matching team, calculate the balance score and satisfaction score of the divided matching team.

[0122] g3: Based on the sum of the balance scores and satisfaction scores of the multiple divided matching teams, the game accounts to be matched are divided into a second preset number of target matching teams.

[0123] In the embodiment of the present application, the matching teams are divided in the same manner as in the first matching mode, and various behaviors of game interaction and post-game interaction are also used as fitting targets to train the model. Positive samples are selected from game account pairs of players with positive behaviors such as cooperation within the game, liking after the game, adding friends, and continuing team games. Negative samples are screened from teams with no interactive behaviors and with reporting, blacklisting, etc., and the ratio of positive and negative samples is selected according to the offline AUC indicator. At the same time, the game information such as the birth point, common behavior preference, average ranking, number of rescues, moving distance, and battle duration that are unique to the second matching mode in which the number of games is greater than the preset number threshold is combined to improve the accuracy of matching results.

[0124] In a possible implementation, when determining the balance score and satisfaction score for dividing the matching teams, the game scores and interaction scores of the game accounts corresponding to the players in the team are calculated in combination with a pre-trained model. The above calculation process is consistent with the calculation process of the balance score and satisfaction score for the target matching team in the first matching mode, and will not be repeated here.

[0125] Furthermore, after calculating the sum of the balance scores and satisfaction scores of multiple divided matching teams, the sum of the balance scores and the satisfaction scores are used to sort the divided matching teams from high to low, and the divided matching teams sorted before the second preset number are determined as target matching teams.

[0126] Exemplarily, after calculating the similarity of the game accounts to be matched in the current matching pool, the divided matching team 1, the divided matching team 2, the divided matching team 3 and the divided matching team 4 are determined; according to the four divided matching teams currently determined, after calculation, the sum of the balance score and the satisfaction score of the divided matching team 1 is -3; the sum of the balance score and the satisfaction score of the divided matching team 2 is -5; the sum of the balance score and the satisfaction score of the divided matching team 3 is -1; the sum of the balance score and the satisfaction score of the divided matching team 4 is -7; if the second preset number is 2, then at this time, the divided matching team 1 and the divided matching team 3, which are ranked in the top 2 in terms of the sum of the balance score and the satisfaction score, are determined as the target matching teams.

[0127] Furthermore, after determining multiple target matching teams, you can continue to divide the target matching rooms according to the determined multiple target matching teams. When dividing the target matching, you can first place any target matching team in the matching pool into an empty room (target matching room), and then traverse all the current target matching teams in the matching pool, and determine the target matching team to be divided into the target matching room based on the matching degree between the target matching team in the matching pool and the target matching team in the target matching room.

[0128] Specifically, the step of "dividing the target matching room according to the matching team and obtaining the matching result" includes:

[0129] h1: For each matching team, determine the team feature vector of the matching team based on the game information of each to-be-matched game account included in the matching team.

[0130] h2: Based on the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams, the target matching room is divided to obtain the matching results.

[0131] In an embodiment of the present application, for each matching team, when determining the team feature vector of the matching team, it can also be determined by a pre-trained prediction vector model. Similarly, the game satisfaction index can be used to train the EGES model to obtain a prediction vector model.

[0132] Furthermore, for each matching team, based on the game information of each to-be-matched game account in the matching team, determine the interaction information between each to-be-matched game account and other players inside and outside the game, such as rescuing, turning on the microphone, liking, reporting, re-teaming, sending friend requests, blacklisting, and deleting friends in the game, and take the number of times each to-be-matched game account rescues, turns on the microphone, likes, reports, re-teaming, sending friend requests, blacklisting, and deleting friends in the game as the input features of each to-be-matched game account, and input them into the prediction vector model to obtain the team feature vector of the matching team.

[0133] Further, after determining the team feature vectors of each matching team, the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams is calculated, and then the target matching teams included in the target matching room are obtained to obtain the matching results.

[0134] In a possible implementation, the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams is calculated through the following steps to divide the target matching room:

[0135] j1: When the target matching room includes a target matching team, the similarities between the team feature vectors of other matching teams and the team feature vector of the target matching team are calculated respectively, and the matching team with the highest similarity with the target matching team among the other matching teams is assigned to the target matching room.

[0136] j2: When the target matching room includes multiple target matching teams, for each target matching team, the similarity between the team feature vectors of other matching teams and the team feature vector of the target matching team is calculated respectively, and the matching team with the highest sum of similarities with each target matching team among other matching teams is assigned to the target matching room.

[0137] In a possible implementation, the similarity calculation between the team feature vector of each matching team and the team feature vectors of other matching teams is consistent with the similarity calculation between the game feature vector of each matching game account and the game feature vector of other to-be-matched game accounts in the first matching mode, which is not repeated here.

[0138] Furthermore, in order to ensure the efficient and normal operation of the matching system, after determining the matching result, the matching process is detected according to the preset matching detection indicators to detect whether there are any abnormal situations in the matching process. If there are any abnormal situations, it is necessary to locate the abnormal matching information in the abnormal situation and adjust the matching system to ensure the efficient and accurate operation of the matching system.

[0139] Specifically, after the step of "dividing the target matching room according to the matching teams to obtain the matching results", the game matching method further includes:

[0140] k1: Based on the matching results and the matching configuration information during the matching process, combined with the preset matching detection indicators, determine whether there are abnormal situations in the matching process and the abnormal matching information under abnormal situations.

[0141] In an embodiment of the present application, when the matching results are pushed to the client, they will be recorded in the log of the matching system. The ELK technology is used to collect the matching push results in real time, and the Flink real-time jobs and scheduled data tasks are constructed to aggregate and classify them, and the matching-related preset matching detection indicators are generated.

[0142] In one possible implementation, the preset matching detection indicators may include at least one of the following: a matching time consumption indicator, a matching request volume indicator, and a matching model prediction sub-indicator; the above-mentioned preset matching detection indicators can be used to reflect the operating status and matching effect of the matching system, and then the corresponding matching rules can be adjusted to improve the matching system.

[0143] Specifically, the matching time consumption index includes the average time consumption of the entire matching request, the maximum request time consumption, the number of timeouts, the timeout ratio, and the average time consumption of each matching step. The time consumption and timeout indicators can be used to analyze the response status of the matching system. The increase in time consumption and the increase in the number of timeouts often indicate that the system load is too high. The average time consumption of each matching step can accurately find the steps with high computational complexity in the entire matching link, which is convenient for precise optimization of the system. Through the matching request volume indicator analysis, the number of matching requests can be monitored in real time. In response to request peaks and troughs, different adjustments can be made to the matching resources. The matching model prediction score indicator can give the prediction and scoring results of each model in the matching system. By analyzing the model prediction score, the matching model effect can be monitored in time. For example, when the model calculates a generally lower balance score, it means that the model effect has deteriorated and the model needs to be adjusted in time.

[0144] For example, see Figure 3 , Figure 3 The second structural diagram of the matching system provided in the embodiment of the present application; Figure 3As shown, the matching system 200 also includes a guarantee detection module 230. After the matching processing module 220 obtains the matching result based on the game information obtained from the interaction processing module 210 and pushes the matching result, the guarantee detection module 230 collects the pushed result, completes the indicator analysis and rule modification through the preset matching detection indicators generated in real time, and adjusts the customized rules in the matching processing module 220 by modifying the rules.

[0145] The game matching method provided in the embodiment of the present application determines multiple unmatched game accounts to be matched and the game information of each game account to be matched; determines the matching mode based on the number of players participating in the current game game; when the matching mode is the first matching mode, based on the game information of each game account to be matched, after allocating a preset number of game accounts to be matched to the target matching room, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result; when the matching mode is the second matching mode, based on the game information of each game account to be matched, after dividing the game account to be matched into a second preset number of target matching teams, the target matching room is divided according to the matching teams to obtain a matching result. In this way, when the game accounts to be matched are matched, the game information of the game accounts to be matched is taken into account to meet the matching needs of the players. At the same time, different matching modes are determined according to the number of players participating in the game game, and different matching methods are adopted for different matching modes, thereby reducing the calculation steps in the matching process and helping to improve the matching efficiency.

[0146] Based on the same inventive concept, a game matching device corresponding to the game matching method is also provided in the embodiment of the present application. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned game matching method in the embodiment of the present disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0147] See also Figure 4 , Figure 5 , Figure 4 This is one of the structural schematic diagrams of a game matching device provided in an embodiment of the present application. Figure 5 This is a second structural diagram of a game matching device provided in an embodiment of the present application. Figure 4 As shown in , the game matching device 400 includes:

[0148] A game information determination module 410 is used to determine a plurality of unmatched game accounts to be matched and game information of each of the game accounts to be matched;

[0149] A matching mode determination module 420, configured to determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold;

[0150] The first matching result acquisition module 430 is used for, when the matching mode is the first matching mode, allocating a preset number of to-be-matched game accounts to a target matching room based on the game information of each to-be-matched game account, dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams, and obtaining a matching result;

[0151] The second matching result acquisition module 440 is used to divide the game accounts to be matched into a second preset number of target matching teams based on the game information of each game account to be matched, and then divide the target matching rooms according to the matching teams to obtain matching results when the matching mode is the second matching mode.

[0152] In a possible implementation, when the matching mode is the first matching mode, the first matching result acquisition module 430 is used to allocate a preset number of to-be-matched game accounts to a target matching room based on the game information of each to-be-matched game account, and the first matching result acquisition module 430 is used to:

[0153] When the matching mode is the first matching mode, for each to-be-matched game account, based on the game information of the to-be-matched game account, determining a game feature vector of the to-be-matched game account;

[0154] Based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room.

[0155] In a possible implementation manner, when the first matching result acquisition module 430 is used to allocate a preset number of to-be-matched game accounts to a target matching room based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, the first matching result acquisition module 430 is used to:

[0156] When the target matching room includes a target matching game account, respectively calculate the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account, and assign the to-be-matched game account with the highest similarity to the target matching game account among the other to-be-matched game accounts to the target matching room;

[0157] When the target matching room includes multiple target matching game accounts, for each target matching game account, the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account is calculated respectively, and the to-be-matched game account with the highest sum of similarities with each target matching game account among the other to-be-matched game accounts is allocated to the target matching room.

[0158] In a possible implementation manner, when the first matching result acquisition module 430 is used to divide a plurality of to-be-matched game accounts contained in a target matching room into a first preset number of target matching teams to obtain a matching result, the first matching result acquisition module 430 is used to:

[0159] Determine a plurality of candidate matching teams based on the first preset number, the number of to-be-matched game accounts included in the target matching room, and the game account teaming information;

[0160] For each candidate matching team, based on the game feature information of each to-be-matched game account included in the candidate matching team, determine the game score and interaction score of the candidate matching team;

[0161] Matching multiple candidate matching teams according to the first preset number to determine multiple matching team groups;

[0162] For each matching team group, based on the match scores and interaction scores of each candidate matching team included in the matching team group, calculate the balance score and satisfaction score of the matching team group;

[0163] The matching result is obtained based on the matching team group with the highest sum of balance scores and satisfaction scores among multiple matching team groups.

[0164] In a possible implementation manner, for each candidate matching team, the first matching result acquisition module 430 is used to determine the match score of the candidate matching team through the following steps:

[0165] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team and the attribute information of each to-be-matched game account are input into a pre-trained game score prediction model to obtain the game score of the candidate matching team.

[0166] In a possible implementation, for each candidate matching team, the first matching result acquisition module 430 is used to determine the interaction score of the candidate matching team through the following steps:

[0167] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team is input into a pre-trained satisfaction score prediction model to obtain the interaction score of the candidate matching team.

[0168] In a possible implementation manner, when the matching mode is the second matching mode, the second matching result obtaining module 440 is used to divide the to-be-matched game accounts into a second preset number of target matching teams based on the game information of each to-be-matched game account, the second matching result obtaining module 440 is used to:

[0169] When the matching mode is the second matching mode, determining a plurality of matching teams based on the second preset number and the number of game accounts of the game accounts to be matched;

[0170] For each divided matching team, based on the game feature information of each to-be-matched game account included in the divided matching team, calculate the balance score and satisfaction score of the divided matching team;

[0171] Based on the sum of the balance scores and the satisfaction scores in the multiple divided matching teams, the game accounts to be matched are divided into a second preset number of target matching teams.

[0172] In a possible implementation manner, when the second matching result acquisition module 440 is used to divide the target matching room according to the matching team to obtain the matching result, the second matching result acquisition module 440 is used to:

[0173] For each matching team, based on the game information of each to-be-matched game account included in the matching team, determine the team feature vector of the matching team;

[0174] Based on the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams, the target matching room is divided to obtain the matching result.

[0175] In a possible implementation, Figure 5 As shown, the game matching device 400 further includes a screening module 450, and the screening module 450 is used to:

[0176] Obtain multiple candidate matching game accounts that have not been matched, and game information of each candidate matching game account;

[0177] Based on the game information of each candidate matching game account, detecting whether the candidate matching game account meets the preset account screening rules;

[0178] The matching game accounts that do not meet the preset account screening rule among the multiple candidate matching game accounts are determined as the game accounts to be matched.

[0179] In a possible implementation manner, the unmatched candidate matching game accounts include candidate matching game accounts that were not successfully matched in the last matching process and the to-be-matched game accounts included in the matching request obtained in real time.

[0180] In a possible implementation, Figure 5 As shown, the game matching device 400 further includes an abnormality detection module 460, and the abnormality detection module 460 is used to:

[0181] Based on the matching results and the matching configuration information during the matching process, combined with the preset matching detection indicators, determine whether there are abnormal situations in the matching process and the abnormal matching information under abnormal situations.

[0182] In a possible implementation manner, the game information of the game account to be matched includes at least one of the following:

[0183] Game performance information, in-game and out-game interaction information, and competitive level information.

[0184] The game matching device provided in the embodiment of the present application determines multiple unmatched game accounts to be matched and the game information of each game account to be matched; determines the matching mode based on the number of players participating in the current game game; when the matching mode is the first matching mode, based on the game information of each game account to be matched, after allocating a preset number of game accounts to be matched to the target matching room, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result; when the matching mode is the second matching mode, based on the game information of each game account to be matched, after dividing the game account to be matched into a second preset number of target matching teams, the target matching room is divided according to the matching teams to obtain a matching result. In this way, when the game accounts to be matched are matched, the game information of the game accounts to be matched is taken into consideration to meet the matching needs of the players. At the same time, different matching modes are determined according to the number of players participating in the game game, and different matching methods are adopted for different matching modes, thereby reducing the calculation steps in the matching process and helping to improve the matching efficiency.

[0185] See also Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 6 As shown in , the electronic device 600 includes a processor 610 , a memory 620 and a bus 630 .

[0186] The memory 620 stores machine-readable instructions executable by the processor 610. When the electronic device 600 is running, the processor 610 communicates with the memory 620 via the bus 630, so that the processor 610 executes the following instructions when running:

[0187] Determine multiple unmatched game accounts to be matched and game information of each game account to be matched;

[0188] Determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold;

[0189] When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the match information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result;

[0190] When the matching mode is the second matching mode, based on the game information of each game account to be matched, the game accounts to be matched are divided into a second preset number of target matching teams, and the target matching rooms are divided according to the matching teams to obtain matching results.

[0191] In a possible implementation, in the instructions executed by the processor 610, when the matching mode is the first matching mode, based on the game information of each to-be-matched game account, allocating a preset number of to-be-matched game accounts to the target matching room includes:

[0192] When the matching mode is the first matching mode, for each to-be-matched game account, based on the game information of the to-be-matched game account, determining a game feature vector of the to-be-matched game account;

[0193] Based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room.

[0194] In a possible implementation, in the instructions executed by the processor 610, allocating a preset number of to-be-matched game accounts to a target matching room based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts includes:

[0195] When the target matching room includes a target matching game account, respectively calculate the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account, and assign the to-be-matched game account with the highest similarity to the target matching game account among the other to-be-matched game accounts to the target matching room;

[0196] When the target matching room includes multiple target matching game accounts, for each target matching game account, the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account is calculated respectively, and the to-be-matched game account with the highest sum of similarities with each target matching game account among the other to-be-matched game accounts is allocated to the target matching room.

[0197] In a possible implementation manner, in the instructions executed by the processor 610, dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams to obtain a matching result includes:

[0198] Determine a plurality of candidate matching teams based on the first preset number, the number of to-be-matched game accounts included in the target matching room, and the game account teaming information;

[0199] For each candidate matching team, based on the game feature information of each to-be-matched game account included in the candidate matching team, determine the game score and interaction score of the candidate matching team;

[0200] Matching multiple candidate matching teams according to the first preset number to determine multiple matching team groups;

[0201] For each matching team group, based on the match scores and interaction scores of each candidate matching team included in the matching team group, calculate the balance score and satisfaction score of the matching team group;

[0202] The matching result is obtained based on the matching team group with the highest sum of balance scores and satisfaction scores among multiple matching team groups.

[0203] In a possible implementation manner, in the instructions executed by the processor 610, for each candidate matching team, the game score of the candidate matching team is determined by the following steps:

[0204] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team and the attribute information of each to-be-matched game account are input into a pre-trained game score prediction model to obtain the game score of the candidate matching team.

[0205] In a possible implementation manner, in the instructions executed by the processor 610, for each candidate matching team, the interaction score of the candidate matching team is determined by the following steps:

[0206] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team is input into a pre-trained satisfaction score prediction model to obtain the interaction score of the candidate matching team.

[0207] In a possible implementation manner, in the instructions executed by the processor 610, when the matching mode is the second matching mode, the game accounts to be matched are divided into a second preset number of target matching teams based on the game information of each game account to be matched, including:

[0208] When the matching mode is the second matching mode, determining a plurality of matching teams based on the second preset number and the number of game accounts of the game accounts to be matched;

[0209] For each divided matching team, based on the game feature information of each to-be-matched game account included in the divided matching team, calculate the balance score and satisfaction score of the divided matching team;

[0210] Based on the sum of the balance scores and the satisfaction scores in the multiple divided matching teams, the game accounts to be matched are divided into a second preset number of target matching teams.

[0211] In a possible implementation manner, in the instructions executed by the processor 610, dividing the target matching room according to the matching teams to obtain the matching results includes:

[0212] For each matching team, based on the game information of each to-be-matched game account included in the matching team, determine the team feature vector of the matching team;

[0213] Based on the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams, the target matching room is divided to obtain the matching result.

[0214] In a possible implementation manner, the instructions executed by the processor 610 further include:

[0215] Obtain multiple candidate matching game accounts that have not been matched, and game information of each candidate matching game account;

[0216] Based on the game information of each candidate matching game account, detecting whether the candidate matching game account meets the preset account screening rules;

[0217] The matching game accounts that do not meet the preset account screening rule among the multiple candidate matching game accounts are determined as the game accounts to be matched.

[0218] In a possible implementation manner, the unmatched candidate matching game accounts include candidate matching game accounts that were not successfully matched in the last matching process and the to-be-matched game accounts included in the matching request obtained in real time.

[0219] In a possible implementation manner, the instructions executed by the processor 610 further include:

[0220] Based on the matching results and the matching configuration information during the matching process, combined with the preset matching detection indicators, determine whether there are abnormal situations in the matching process and the abnormal matching information under abnormal situations.

[0221] In a possible implementation manner, the game information of the game account to be matched includes at least one of the following:

[0222] Game performance information, in-game and out-game interaction information, and competitive level information.

[0223] Through the above method, when matching game accounts, the game information of the game accounts to be matched is considered to meet the matching needs of players. At the same time, different matching modes are determined according to the number of players participating in the game, and different matching methods are used for different matching modes to reduce the calculation steps in the matching process, which helps to improve the matching efficiency. At the same time, after determining the matching result, the matching process is detected according to the preset matching detection indicators, and the matching process is detected to detect whether there are abnormal situations in the matching process. If there are abnormal situations, it is necessary to locate the abnormal matching information in the abnormal situation and adjust the matching system to ensure the efficient and accurate operation of the matching system.

[0224] The present application also provides a computer-readable storage medium, on which a computer program is stored, so that when the computer program is executed by a processor, the following instructions are executed:

[0225] Determine multiple unmatched game accounts to be matched and game information of each game account to be matched;

[0226] Determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold;

[0227] When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the match information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result;

[0228] When the matching mode is the second matching mode, based on the game information of each game account to be matched, the game accounts to be matched are divided into a second preset number of target matching teams, and the target matching rooms are divided according to the matching teams to obtain matching results.

[0229] In a possible implementation, in the instructions executed by a computer-readable storage medium, when the matching mode is the first matching mode, based on the game information of each to-be-matched game account, a preset number of to-be-matched game accounts are allocated to a target matching room, including:

[0230] When the matching mode is the first matching mode, for each to-be-matched game account, based on the game information of the to-be-matched game account, determining a game feature vector of the to-be-matched game account;

[0231] Based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room.

[0232] In a possible implementation manner, in the instructions executed by a computer-readable storage medium, allocating a preset number of to-be-matched game accounts to a target matching room based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts includes:

[0233] When the target matching room includes a target matching game account, respectively calculate the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account, and assign the to-be-matched game account with the highest similarity to the target matching game account among the other to-be-matched game accounts to the target matching room;

[0234] When the target matching room includes multiple target matching game accounts, for each target matching game account, the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account is calculated respectively, and the to-be-matched game account with the highest sum of similarities with each target matching game account among the other to-be-matched game accounts is allocated to the target matching room.

[0235] In a possible implementation manner, in the instructions executed by the computer-readable storage medium, dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams to obtain the matching results includes:

[0236] Determine a plurality of candidate matching teams based on the first preset number, the number of to-be-matched game accounts included in the target matching room, and the game account teaming information;

[0237] For each candidate matching team, based on the game feature information of each to-be-matched game account included in the candidate matching team, determine the game score and interaction score of the candidate matching team;

[0238] Matching multiple candidate matching teams according to the first preset number to determine multiple matching team groups;

[0239] For each matching team group, based on the match scores and interaction scores of each candidate matching team included in the matching team group, calculate the balance score and satisfaction score of the matching team group;

[0240] The matching result is obtained based on the matching team group with the highest sum of balance scores and satisfaction scores among multiple matching team groups.

[0241] In a possible implementation manner, in the instructions executed by the computer-readable storage medium, for each candidate matching team, the game score of the candidate matching team is determined by the following steps:

[0242] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team and the attribute information of each to-be-matched game account are input into a pre-trained game score prediction model to obtain the game score of the candidate matching team.

[0243] In a possible implementation manner, in the instructions executed by the computer-readable storage medium, for each candidate matching team, the interaction score of the candidate matching team is determined by the following steps:

[0244] For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team is input into a pre-trained satisfaction score prediction model to obtain the interaction score of the candidate matching team.

[0245] In a possible implementation manner, in the instructions executed by the computer-readable storage medium, when the matching mode is the second matching mode, the game accounts to be matched are divided into a second preset number of target matching teams based on the game information of each game account to be matched, including:

[0246] When the matching mode is the second matching mode, determining a plurality of matching teams based on the second preset number and the number of game accounts of the game accounts to be matched;

[0247] For each divided matching team, based on the game feature information of each to-be-matched game account included in the divided matching team, calculate the balance score and satisfaction score of the divided matching team;

[0248] Based on the sum of the balance scores and the satisfaction scores in the multiple divided matching teams, the game accounts to be matched are divided into a second preset number of target matching teams.

[0249] In a possible implementation manner, in the instructions executed by a computer-readable storage medium, dividing the target matching room according to the matching teams to obtain the matching results includes:

[0250] For each matching team, based on the game information of each to-be-matched game account included in the matching team, determine the team feature vector of the matching team;

[0251] Based on the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams, the target matching room is divided to obtain the matching result.

[0252] In a possible implementation manner, the instructions executed by the computer-readable storage medium further include:

[0253] Obtain multiple candidate matching game accounts that have not been matched, and game information of each candidate matching game account;

[0254] Based on the game information of each candidate matching game account, detecting whether the candidate matching game account meets the preset account screening rules;

[0255] The matching game accounts that do not meet the preset account screening rule among the multiple candidate matching game accounts are determined as the game accounts to be matched.

[0256] In a possible implementation manner, the unmatched candidate matching game accounts include candidate matching game accounts that were not successfully matched in the last matching process and the to-be-matched game accounts included in the matching request obtained in real time.

[0257] In a possible implementation manner, the instructions executed by the computer-readable storage medium further include:

[0258] Based on the matching results and the matching configuration information during the matching process, combined with the preset matching detection indicators, determine whether there are abnormal situations in the matching process and the abnormal matching information under abnormal situations.

[0259] In a possible implementation, in the instructions executed by the computer-readable storage medium, the game information of the game account to be matched includes at least one of the following:

[0260] Game performance information, in-game and out-game interaction information, and competitive level information.

[0261] Through the above method, when matching game accounts, the game information of the game accounts to be matched is considered to meet the matching needs of players. At the same time, different matching modes are determined according to the number of players participating in the game, and different matching methods are used for different matching modes to reduce the calculation steps in the matching process, which helps to improve the matching efficiency. At the same time, after determining the matching result, the matching process is detected according to the preset matching detection indicators, and the matching process is detected to detect whether there are abnormal situations in the matching process. If there are abnormal situations, it is necessary to locate the abnormal matching information in the abnormal situation and adjust the matching system to ensure the efficient and accurate operation of the matching system.

[0262] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0263] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0264] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0265] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0266] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0267] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A game matching method, characterized in that: The game matching method comprises: Determine multiple unmatched game accounts to be matched and game information of each game account to be matched; Determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold; When the matching mode is the first matching mode, after allocating a preset number of game accounts to be matched to a target matching room based on the match information of each game account to be matched, the multiple game accounts to be matched contained in the target matching room are divided into a first preset number of target matching teams to obtain a matching result; When the matching mode is the second matching mode, based on the game information of each game account to be matched, the game accounts to be matched are divided into a second preset number of target matching teams, and the target matching rooms are divided according to the matching teams to obtain matching results.

2. The game matching method according to claim 1, characterized in that: When the matching mode is the first matching mode, based on the game information of each to-be-matched game account, a preset number of to-be-matched game accounts are allocated to the target matching room, including: When the matching mode is the first matching mode, for each to-be-matched game account, based on the game information of the to-be-matched game account, determining a game feature vector of the to-be-matched game account; Based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts, a preset number of to-be-matched game accounts are allocated to the target matching room.

3. The game matching method according to claim 2, characterized in that: The method of allocating a preset number of to-be-matched game accounts to the target matching room based on the similarity between the game feature vector of each matching game account and the game feature vectors of other to-be-matched game accounts includes: When the target matching room includes a target matching game account, respectively calculate the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account, and assign the to-be-matched game account with the highest similarity to the target matching game account among the other to-be-matched game accounts to the target matching room; When the target matching room includes multiple target matching game accounts, for each target matching game account, the similarity between the game feature vectors of other to-be-matched game accounts and the game feature vector of the target matching game account is calculated respectively, and the to-be-matched game account with the highest sum of similarities with each target matching game account among the other to-be-matched game accounts is allocated to the target matching room.

4. The game matching method according to claim 1, characterized in that: The step of dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams to obtain a matching result includes: Determine a plurality of candidate matching teams based on the first preset number, the number of to-be-matched game accounts included in the target matching room, and the game account teaming information; For each candidate matching team, based on the game feature information of each to-be-matched game account included in the candidate matching team, determine the game score and interaction score of the candidate matching team; Matching multiple candidate matching teams according to the first preset number to determine multiple matching team groups; For each matching team group, based on the match scores and interaction scores of each candidate matching team included in the matching team group, calculate the balance score and satisfaction score of the matching team group; The matching result is obtained based on the matching team group with the highest sum of balance scores and satisfaction scores among multiple matching team groups.

5. The game matching method according to claim 4, characterized in that: For each candidate matching team, the match score of the candidate matching team is determined through the following steps: For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team and the attribute information of each to-be-matched game account are input into a pre-trained game score prediction model to obtain the game score of the candidate matching team.

6. The game matching method according to claim 4, characterized in that: For each candidate matching team, the interaction score of the candidate matching team is determined by the following steps: For each candidate matching team, the game information of each to-be-matched game account in the candidate matching team is input into a pre-trained satisfaction score prediction model to obtain the interaction score of the candidate matching team.

7. The game matching method according to claim 1, characterized in that: When the matching mode is the second matching mode, based on the game information of each to-be-matched game account, the to-be-matched game accounts are divided into a second preset number of target matching teams, including: When the matching mode is the second matching mode, determining a plurality of matching teams based on the second preset number and the number of game accounts of the game accounts to be matched; For each divided matching team, based on the game feature information of each to-be-matched game account included in the divided matching team, calculate the balance score and satisfaction score of the divided matching team; Based on the sum of the balance scores and the satisfaction scores in the multiple divided matching teams, the game accounts to be matched are divided into a second preset number of target matching teams.

8. The game matching method according to claim 1, characterized in that: The step of dividing the target matching room according to the matching team and obtaining the matching result includes: For each matching team, based on the game information of each to-be-matched game account included in the matching team, determine the team feature vector of the matching team; Based on the similarity between the team feature vector of each matching team and the team feature vectors of other matching teams, the target matching room is divided to obtain the matching result.

9. The game matching method according to claim 1, characterized in that: Before determining the multiple unmatched game accounts to be matched and the game information of each of the game accounts to be matched, the game matching method further includes: Obtain multiple candidate matching game accounts that have not been matched, and game information of each candidate matching game account; Based on the game information of each candidate matching game account, detecting whether the candidate matching game account meets the preset account screening rules; The matching game accounts that do not meet the preset account screening rule among the multiple candidate matching game accounts are determined as the game accounts to be matched.

10. The game matching method according to claim 9, characterized in that: The unmatched candidate matching game accounts include candidate matching game accounts that were not successfully matched in the last matching process and the to-be-matched game accounts included in the matching request obtained in real time.

11. The game matching method according to claim 1, characterized in that: After dividing the target matching room according to the matching teams and obtaining the matching results, the game matching method further includes: Based on the matching results and the matching configuration information during the matching process, combined with the preset matching detection indicators, determine whether there are abnormal situations in the matching process and the abnormal matching information under abnormal situations.

12. The game matching method according to claim 1, characterized in that: The game information of the game account to be matched includes at least one of the following: Game performance information, in-game and out-game interaction information, and competitive level information.

13. A game matching device, characterized in that: The game matching device comprises: A game information determination module, used to determine a plurality of unmatched game accounts to be matched and game information of each game account to be matched; A matching mode determination module, configured to determine a matching mode based on the number of players participating in the current game; the matching mode includes a first matching mode in which the number of players participating in the current game is less than a preset participation number threshold, and a second matching mode in which the number of players participating in the current game is greater than the preset participation number threshold; A first matching result acquisition module is used for, when the matching mode is the first matching mode, allocating a preset number of to-be-matched game accounts to a target matching room based on the game information of each to-be-matched game account, dividing the multiple to-be-matched game accounts contained in the target matching room into a first preset number of target matching teams, and obtaining a matching result; The second matching result acquisition module is used to divide the game accounts to be matched into a second preset number of target matching teams based on the game information of each game account to be matched when the matching mode is the second matching mode, and then divide the target matching rooms according to the matching teams to obtain matching results.

14. An electronic device, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the game matching method as described in any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the game matching method according to any one of claims 1 to 12 are executed.

Images