A method, server, medium and device for recommending a game live room

Abstract

The application provides a method, a server, a medium and equipment for recommending a game live room, and the method comprises the following steps: determining a reference live room in a live state and located in a target partition; acquiring anchor information and highlight moment pictures of each reference live room; screening the reference live room based on the highlight moment pictures and the anchor information to obtain a target live room; performing marking processing on the target live room through the anchor information to display corresponding anchor description information on a cover of the target live room; and the anchor description information is used for representing the game strength of the anchor; in this way, when the reference live room is screened by using the game data, the highlight moment pictures and the anchor information, the game strength of the anchor is used for screening, so that the target live room obtained is a live room with high anchor strength, the exposure rate of a new anchor with strength is improved; and a user can select a live room of interest according to the anchor description information to watch, and the effective watching amount of the user is improved.

Description

Technical Field

[0001] This application relates to the field of live streaming technology, and in particular to a method, server, medium and device for recommending live streaming rooms. Background Technology

[0002] Currently, when users select a game livestream on the homepage of a game section on a livestreaming platform, the room cover tags in the livestream list mostly consist of the room name, streamer name, and room popularity score, lacking game-related information. This makes users somewhat aimless when choosing a livestream, unable to directly select a streamer with strong skills. If a user randomly enters the livestream of a less skilled streamer, there may be ineffective views (short viewing time); at the same time, even talented new streamers cannot attract more viewers and attention.

[0003] Therefore, there is an urgent need for a live streaming recommendation method that can display the streamer's effective information to users in order to increase the number of effective viewers and the attention paid to promising new streamers. Summary of the Invention

[0004] To address the problems existing in the prior art, embodiments of the present invention provide a method, server, medium, and device for recommending game live streaming rooms, so as to solve or partially solve the technical problems in the prior art where platforms cannot display the streamer's strength information to users when recommending live streaming rooms, resulting in reduced effective viewership when users randomly select live streaming rooms to watch, and the inability to ensure the attention of promising new streamers.

[0005] A first aspect of the present invention provides a method for recommending game live streaming rooms, the method comprising:

[0006] Identify reference live stream rooms that are currently live and located in the target partition;

[0007] Obtain information about the streamers and highlights from each reference live stream;

[0008] Based on the aforementioned highlights and the streamer information, the reference live streams are filtered to obtain the target live stream;

[0009] The target live stream room is tagged using the streamer information so that the corresponding streamer description information is displayed on the cover of the target live stream room; the streamer description information is used to characterize the streamer's gaming skills.

[0010] In the above solution, obtaining the highlights of each reference live stream includes:

[0011] Acquire game video footage from each reference live stream;

[0012] Image recognition algorithms are used to identify the game video footage and pinpoint the highlights.

[0013] The method in the above scheme further includes:

[0014] When a confirmation message from the game server indicating that the streamer has started a live game broadcast is received, the streamer information and game video footage sent by the game server are received via the game data broadcast channel; or,

[0015] By using the Application Programming Interface (API) to reverse-engineer the game client's game data transmission channel, the program nodes that send and receive data can be obtained.

[0016] Hook the program node that sends data and the program node that receives data to obtain a data sending hook program module and a data receiving hook program module.

[0017] A second SDK is created based on the data sending hook module and the data receiving hook module, and the second SDK is sent to the broadcaster's client. The second SDK is injected into the game process when the broadcaster starts the game and broadcasts live. The second SDK is used to obtain the game video screen and broadcaster information published by the game client.

[0018] Receive the game video footage and streamer information sent by the streamer-side client.

[0019] In the above scheme, the step of filtering the reference live streams based on the highlights and the broadcast information to obtain the target live stream includes:

[0020] Determine the exposure level to which the target user belongs; the target user is the user who has an exposure request for the reference live broadcast room.

[0021] For any reference live streaming room, the total budget for the reference live streaming room on that day is determined based on the streamer information, and the budget allocation for the current time period is determined based on the total budget.

[0022] The budget allocation is adjusted based on the total actual consumption of the reference live streaming room before the current time period of the day to obtain the target allocation for the current time period;

[0023] Based on the target allocation amount for the current time period and the highlights, determine the probability of users at the current exposure level selecting the reference live stream during the current time period.

[0024] Based on the selection probability, the reference live streaming rooms are filtered to obtain the target live streaming room.

[0025] In the above scheme, the step of adjusting the budget allocation based on the total actual consumption of the reference live streaming room before the current time period of the day to obtain the target allocation for the current time period includes:

[0026] According to the formula The budget allocation is adjusted to obtain the target allocation C for the current time period. t ;in,

[0027] The B t The budget allocation for the reference live stream room in the current time period is defined as follows: B is the total budget for the reference live stream room for the day; i is any time period; t is the current time period; K is the number of time periods included in the day; and A is the budget allocation for the reference live stream room for the current time period. i To reference the actual consumption of the live stream in the i-th time period, the B represents the total actual consumption of the reference live stream room before the current time period on that day. i The budget allocation for the target live streaming room in the i-th time period.

[0028] In the above scheme, determining the probability of a user at the current exposure level selecting the reference live stream room during the current time period based on the target allocation amount for the current time period and the highlights includes:

[0029] If it is determined that the target allocation amount for the current time period is greater than the actual consumption amount for the previous time period, then according to the formula... Increase the probability that users at the current exposure level will select the reference live stream in the current time period. in,

[0030] The r l t-1 The probability that a user at the current exposure level selected the reference live stream in the previous time period is given by l, where l is the current exposure level, and C is the reference live stream. t The target allocation amount for the current time period, A t-1 The current time period represents the actual consumption in the previous time period, where j represents any exposure level, L represents the total number of exposure levels, and so on. Let e ​​be the probability that a user at the j-th exposure level selected the reference live stream in the previous time period of the current time period. The probability that a user at the current exposure level selected the reference live stream in the initial time period is a preset value. t-1 It is the number of highlights extracted from the previous time period of the reference live stream at the current moment, maxe t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

[0031] In the above scheme, determining the probability of a user at the current exposure level selecting the reference live stream room during the current time period based on the target allocation amount for the current time period, the game data, and the highlights includes:

[0032] If it is determined that the target allocation amount for the current time period is less than the actual consumption amount for the previous time period, then according to the formula... Reduce the probability that users at the current exposure level will select the reference live stream in the current time period. in,

[0033] The r l t-1 The probability that a user at the current exposure level selected the reference live stream in the previous time period is given by l, where l is the current exposure level, and C is the reference live stream. t The target allocation amount for the current time period, A t-1 The current time period represents the actual consumption in the previous time period, where j represents any exposure level, L represents the total number of exposure levels, and so on. Let e ​​be the probability that a user at the j-th exposure level selected the reference live stream in the previous time period of the current time period. The probability that a user at the current exposure level selected the reference live stream in the initial time period is a preset value. t-1 It is the number of highlights extracted from the previous time period of the reference live stream at the current moment, maxe t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

[0034] A second aspect of the present invention provides a server for recommending game live streaming rooms, the server comprising:

[0035] The determination unit is used to identify the reference live room that is in a live broadcast state and located in the target partition;

[0036] The acquisition unit is used to acquire the anchor information and highlights of each reference live stream room;

[0037] A filtering unit is used to filter the reference live streaming rooms based on the highlights and the anchor information to obtain the target live streaming room;

[0038] The tagging unit is used to tag the target live room based on the streamer information, so as to display the corresponding streamer description information on the cover of the target live room; the streamer description information is used to characterize the streamer's gaming skills.

[0039] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the method described in any of the first aspects.

[0040] A fourth aspect of the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method described in any of the first aspects.

[0041] This invention provides a method, server, medium, and device for recommending game live streaming rooms. The method includes: identifying reference live streaming rooms that are currently live and located in a target partition; acquiring the streamer information and highlight moments of each reference live streaming room; filtering the reference live streaming rooms based on the highlight moments and corresponding streamer information to obtain target live streaming rooms; tagging the target live streaming rooms to display corresponding streamer description information on the cover of the target live streaming room; the streamer description information is used to characterize the streamer's gaming skills; thus, when filtering reference live streaming rooms using game data, highlight moments, and streamer information, it is equivalent to filtering based on the streamer's gaming skills, thereby ensuring that the obtained target live streaming rooms are those with high-level streamers, increasing the exposure rate of new and skilled streamers; and by tagging the streamer description information on the cover of the corresponding live streaming room, users can select live streaming rooms of interest based on the streamer description information, increasing the effective viewership. Attached Figure Description

[0042] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0043] Figure 1 A schematic flowchart of a method for recommending game live streaming rooms according to an embodiment of the present invention is shown;

[0044] Figure 2 A schematic diagram of the server structure for a recommended game live streaming room according to an embodiment of the present invention is shown;

[0045] Figure 3 A schematic diagram of a computer-readable storage medium structure according to an embodiment of the present invention is shown;

[0046] Figure 4 A schematic diagram of a computer device structure according to an embodiment of the present invention is shown. Detailed Implementation

[0047] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0048] This invention provides a method for recommending game live streaming rooms, such as... Figure 1 As shown, the method includes:

[0049] S210, Identify the reference live room that is in live streaming status and located in the target partition;

[0050] The live streaming server records the live streaming status of each live streaming room and obtains the room IDs that are currently live streaming, storing these room IDs in the database.

[0051] When starting a live stream, the streamer can manually select a live stream section, and the live stream server can set a corresponding live stream section identifier for that live stream based on the selected section. However, in practice, there may be a small number of cases where the live stream section selected by the streamer is inconsistent with the live stream content. Therefore, during the live stream, there will be manual review, and the live stream section identifier will be modified according to the live stream content.

[0052] In this way, the live streaming server can determine the reference live streaming room that is in a live streaming state and located in the target partition based on the live streaming status of each live streaming room and the live streaming partition identifier of each live streaming room.

[0053] For example, if the target partition is game partition A, then the live streaming server can identify the live streaming room that is currently streaming and located in game partition A as the reference live streaming room.

[0054] S211, obtain the anchor information and highlights of each reference live stream;

[0055] The streamer's information can include: the game character the streamer chooses, the character's skills, equipment, and other information, as well as information about various events in the game, such as the streamer's rank and performance (number of kills, assists, and deaths, etc.) when the streamer is playing the game.

[0056] When a streamer starts broadcasting a game, the streaming server can obtain game video footage from each reference stream and use image recognition algorithms to identify highlight moments. Highlight moments are those featuring continuous special effects (such as consecutive kills).

[0057] In this embodiment, the methods by which the live streaming server acquires game video footage and streamer information include the following, and in one implementation of acquiring game data for each reference live streaming room, it includes:

[0058] When the game server receives confirmation from the streamer that the streamer has started live streaming the game, the game data broadcast channel is used to receive the streamer information and game video footage sent by the game server.

[0059] Specifically, the live streaming platform server can make an agreement with the third-party game server in advance. The streamer can fill in the name of the live streaming platform and the game account in the registration portal provided by the third-party game server. When the streamer starts the game live stream, the third-party game server can determine the live streaming platform to which the streamer belongs based on the name of the live streaming platform. At this time, the third-party game server will send a confirmation message to the live streaming platform server to confirm the start of the game live stream. Then, the live streaming platform server establishes a communication connection with the game data broadcast channel provided by the third-party game provider and receives game data sent by the game server through the game data broadcast channel.

[0060] Furthermore, when a streamer changes equipment or skills, the live streaming platform server can also receive game data sent by the game server through the game data broadcast channel.

[0061] Another method for obtaining real-time game data for each reference live stream includes:

[0062] Receive game data sent by the broadcaster-side client; wherein the game data is obtained by the broadcaster-side client from the game client using a first software development kit (SDK) provided by a third-party game developer.

[0063] Specifically, if a third-party game developer provides the first SDK, then when a streamer is live streaming a game, the game client will push the game data to the first SDK through inter-process communication, and the first SDK will then obtain the game data.

[0064] The streamer's client can communicate with the first SDK and obtain game data through the first SDK. Then, the streamer's client will send the game data to the live streaming server, and the live streaming server will then obtain the game data.

[0065] Furthermore, if the third-party game developer does not provide the first SDK, the live streaming server in this embodiment can perform reverse engineering analysis on the game client to create a second SDK, and then obtain game data through the second SDK. That is, the implementation method for obtaining game data includes:

[0066] By using the Application Programming Interface (API), the game client's game data transmission channel is reverse-engineered to obtain the program nodes that send and receive data.

[0067] Hook the program nodes that send data and the program nodes that receive data to obtain the data sending hook program module and the data receiving hook program module.

[0068] A second SDK is created based on the data sending hook module and the data receiving hook module, and the second SDK is sent to the broadcaster's client. The second SDK is injected into the game process when the broadcaster starts the game and broadcasts live. The second SDK is used to obtain the game video screen and broadcaster information published by the game client.

[0069] Receive game video footage and streamer information sent by the streamer-side client.

[0070] Here, the game client can be located on the same host as the streamer's client, meaning the streamer is broadcasting the game they are currently controlling; or the game client can be located on a different host than the streamer's client, meaning the streamer is broadcasting the game screen controlled by other players.

[0071] Specifically, since most game data is transmitted over the network, this embodiment first performs reverse analysis on the game client's game data transmission channel to find the program nodes used to receive data and the program nodes used to send data. Then, after hooking the program nodes used to receive data and the program nodes used to send data, the data sent and received by the game client can be intercepted. After creating a second SDK based on the sending data hook program module and the receiving data hook program module, the second SDK can also intercept the game video screen and streamer information sent by the game client.

[0072] In this way, the above methods can be used to obtain the host information and highlights of each reference live stream room.

[0073] S212, Based on the highlights and the anchor information, filter the reference live stream rooms to obtain the target live stream room;

[0074] Understandably, highlight reels and streamer information can indicate a streamer's gaming skill. For example, a higher streamer level and rank indicate stronger gaming ability. More highlight reels also suggest stronger gaming skill. Therefore, filtering reference streams based on highlight reels and corresponding streamer information is essentially filtering streams based on the streamer's gaming skill. This means that regardless of whether a streamer is new or experienced, as long as they are skilled enough, they can be selected, thereby increasing the visibility of talented new streamers.

[0075] In one implementation, the reference live streams are filtered based on the highlights and corresponding streamer information to obtain the target live stream, including:

[0076] Determine the exposure level of the target users; the target users are those who have an exposure request for the reference live stream.

[0077] For any reference live streaming room, determine the total budget for the reference live streaming room for the day based on the streamer information, and determine the budget allocation for the current time period based on the total budget.

[0078] The budget allocation is adjusted based on the total actual consumption of the reference live streaming room before the current time period of the day to obtain the target allocation for the current time period;

[0079] The probability of users at the current exposure level selecting the target live stream is determined based on the target allocation volume and highlights of the current time period.

[0080] The target live stream is obtained by filtering the reference live streams based on the selection probability.

[0081] Specifically, the streamer's information includes: the game character the streamer chooses, the character's skills, equipment, and a series of other information; the streamer's rank and performance in the game (number of kills, assists, and deaths, etc.); and information on various events in the game.

[0082] If the streamer's rank is higher and their game rank and performance are better, then a larger total budget will be allocated to that streamer's reference live stream room that day.

[0083] There are various types of users on live streaming platforms, such as ordinary users, premium users, and other users. Premium users can be understood as users who watch for a longer period of time in the section where the target live stream is located (e.g., more than 1 hour of viewing time per day, or this can be determined based on the actual situation of the platform) or have a higher click rate. Ordinary users can be understood as users who watch for a shorter period of time in the section where the target live stream is located (e.g., less than 20 minutes of viewing time per day, or this can be determined based on the actual situation of the platform) or have a lower click rate.

[0084] If the live stream can be exposed to high-quality users, whose viewing time is longer, then the conversion rate and return on investment of the live stream can be improved.

[0085] Therefore, in order to increase the likelihood of exposing the live stream to high-quality users, it is necessary to segment the users on the platform in this embodiment.

[0086] In one implementation, determining the exposure level to which the target user belongs includes:

[0087] Get the total viewing time of each user in the live streaming section of the reference live streaming room within a preset historical time period;

[0088] Sort the total viewing time to obtain the set W of viewing times;

[0089] Based on the number of users and the preset number of levels, the viewing time set is divided into viewing time intervals of the corresponding number of levels;

[0090] Obtain the target viewing time of the target user in the live streaming section of the reference live streaming room, and determine the target viewing time range into which the target viewing time falls;

[0091] The level corresponding to the target viewing time interval is taken as the exposure level to which the target user belongs.

[0092] The historical time period can be one month, or it can be determined based on the actual situation of the platform, without restriction. The target user is any user who has an exposure request for the reference live stream; the reference live stream is any live stream on the live streaming platform.

[0093] In this embodiment, when dividing the viewing time set into viewing time intervals corresponding to the number of levels based on the number of users and the preset number of levels, the division is performed using quantiles.

[0094] For example, if the number of tiers is L and the number of users is N, then the viewing duration range corresponding to the current exposure tier l is: Where W is the set of viewing durations, For W Quantiles. The initial quantile is +∞.

[0095] For ease of understanding, let's take the viewing time of 10 users as an example (in real-world applications, there will be a massive number of users). The viewing times of these 10 users, arranged from longest to shortest, are 100 minutes, 90 minutes, 80 minutes, 70 minutes, 60 minutes, 50 minutes, 40 minutes, 30 minutes, 20 minutes, and 10 minutes. If we need to divide them into 5 levels, then the viewing time range corresponding to the first exposure level is [90, 100], the viewing time range corresponding to the second exposure level is [70, 90), and the higher the level, the shorter the viewing time. In this way, by determining the exposure level to which the target user belongs, users with longer viewing times can be given a higher probability of being selected for the initial time slot, making it more likely that high-quality users will see the reference live stream. Since high-quality users are more likely to click, this stratification can improve the conversion rate and ROI of the live stream.

[0096] In this embodiment, the anchor information mainly includes the total budget (price) of the reference live room corresponding to the anchor on that day. In this embodiment, the day is divided into multiple time periods, and the total budget is divided according to the time period. Each time period has a corresponding budget allocation.

[0097] For example, if the total budget is B, and the day is divided into K time periods, then:

[0098] B = (B 1 ,...B t ,...,B K )

[0099] Among them, B t K represents the budget allocation for the target live streaming room during the current time period. K is a constant, and can typically be 24, 48, 96, etc.

[0100] Therefore, the budget allocation for the reference live stream room in the current time period is determined based on the streamer information, including:

[0101] According to the formula Determine the budget allocation B for the reference live streaming room in the current time period. t ;

[0102] Where t represents the current time period, B represents the total daily budget of the reference live streaming room, K represents the number of time periods included in the day, and E represents the total daily budget. t This represents the actual consumption amount corresponding to the current time period t of the previous day.

[0103] The principle behind the above formula is: based on the proportion of the actual consumption of the current day in time period t of the previous day to the total consumption of the previous day. To estimate the budget allocation for the current time period t of the day.

[0104] It's important to note that the budget can be expressed in terms of either traffic or monetary value; that is, the unit of budget can be either traffic or monetary value (e.g., yuan). For example, suppose the target livestream's total budget for the day is 10,000 yuan, the actual spending in the first session of the previous day was 10,000 yuan, and the total actual spending across all sessions of the previous day was 500,000 yuan. Then, the budget allocation for the first session of the day would be: Yuan.

[0105] However, in practical applications, actual consumption is related to the competition at the time. If no winner is won, even with a budget allocation, the live stream may not get exposure. That is, it's impossible to guarantee that the actual consumption and budget allocation for each time period will be exactly equal. To improve the accuracy of subsequent selection probability determination, the budget allocation for the current time period t needs to be adjusted based on the total actual consumption before the current time period t, including:

[0106] According to the formula Adjust the budget allocation for the current period to obtain the target allocation C for the current period. t ;in,

[0107] B t To provide a reference for the budget allocation of the live stream in the current time period, B represents the total budget of the reference live stream for the day, i represents any time period, t represents the current time period, K represents the number of time periods included in the day, and A represents the budget allocation of the reference live stream in the current time period. i To reference the actual consumption of the live stream in the i-th time period, To reference the total actual consumption of the live stream room before the current time period on that day, B i This is the budget allocation for the reference live streaming room in the i-th time period.

[0108] The principle behind the above formula is:

[0109] The total actual consumption of the reference live stream up to time period t-1 is: Represents the remaining amount of the total budget for the day; This represents the remaining budget allocation from the current time period t to the end of the day. If the remaining amount of the total budget is greater than the remaining budget allocation, the excess budget needs to be evenly distributed across the current time period t and the subsequent Kt time periods to determine the average distribution across the current time period t and the subsequent Kt time periods. Then add the average allocation to the budget allocation corresponding to the current time period t to get the target allocation for the current time period t.

[0110] This approach divides the budget by time period and adjusts the budget allocation for the current period based on actual consumption. This prevents the live stream from quickly exhausting its budget and avoids users watching the same live stream consecutively during certain time periods, thereby increasing user interest in watching the platform's live streams.

[0111] Furthermore, users at different exposure levels should be assigned different selection probabilities. At the current time t, the selection probability vector is r. t , Where, r l t l represents the probability of a user at the current exposure level selecting a reference live stream during the current time period, where l is the current exposure level.

[0112] Therefore, it is also necessary to determine the probability of users at the current exposure level choosing the reference live stream room during the current time period based on the target allocation volume and highlights of the current time period.

[0113] Specifically, if the target allocation for the current time period is determined to be greater than the actual consumption in the previous time period, then it is necessary to apply the formula... Increase the probability of users at the current exposure level selecting the reference live stream in the current time period. l t This is done to increase the probability that the reference live stream will become the target live stream.

[0114] The r l t-1 Let l represent the probability that a user at the current exposure level selected the reference live stream in the previous time period, and let C represent the current exposure level. t Assign A the target amount for the current time period. t-1 Here, j represents the actual consumption in the previous time period, j represents any exposure level, and L represents the total number of exposure levels. Let e ​​be the probability that a user at the j-th exposure level selected the reference live stream in the previous time period, while the probability that a user at the current exposure level selected the reference live stream in the initial time period is a preset value. t-1 It refers to the number of highlight moments extracted from the previous time slot of the live stream at the current moment, maxe t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

[0115] The principle behind the above formula: If C t >A t-1 This indicates that, based on the previous time period, it is necessary to increase the probability of users at the current exposure level selecting the reference live stream during the current time period; C t -A t-1This is the amount of budget that needs to be consumed, and this budget needs to be allocated to users at exposure levels 1 to L, specifically:

[0116] Determine the percentage of the budget that needs to be consumed. This percentage can be understood as the proportion of the budget that needs to be consumed in the current time period t compared to the time period t-1. For example, suppose C... t For 120, A t-1 If the value is 100, then 20% more budget needs to be consumed in the current time period t.

[0117] In this embodiment, it is desirable to give users with higher exposure levels a higher probability of selection, thereby allocating a larger proportion of the budget to be consumed. Therefore, for users at the current exposure level l, the normalized selection probability from the previous time period is used. The excess budget amount is allocated using a weighted average method.

[0118] Additionally, in this example, we aim to give more exposure to live streams with more exciting moments; therefore, we use... To further increase the selection probability of live streams with exciting content in the t-1 time period, thus increasing their exposure in the current t time period, the number of exciting moments in the t-1 time period is normalized and used as a further weighting factor to increase the selection probability.

[0119] Similarly, if the target allocation for the current time period is determined to be less than the actual consumption in the previous time period, then according to the formula... Reduce the probability r of users at the current exposure level selecting the reference live stream in the current time period. l t ;in,

[0120] r l t-1 Let l represent the probability that a user at the current exposure level selected the reference live stream in the previous time period, and let C represent the current exposure level. t Assign A the target amount for the current time period. t-1 Here, j represents the actual consumption in the previous time period, j represents any exposure level, and L represents the total number of exposure levels. e represents the probability that a user at the j-th exposure level would select the reference live stream in the previous time period. The probability that a user at the current exposure level would select the reference live stream in the initial time period is a preset value. For example, the preset value can be between 0.3 and 0.5, with smaller exposure levels receiving a higher probability of selection in the initial time period. t-1 It refers to the number of highlight moments extracted from the previous time period of the live stream at the current moment, max e t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

[0121] The principle behind the above formula: If C t <=A t-1 This indicates that, based on the previous time period, the probability of users at the current exposure level selecting the reference live stream during the current time period needs to be reduced; A t-1 -C t This is the amount of budget that needs to be reduced. This portion of the budget needs to be allocated to users at the 1-L exposure level, specifically:

[0122] Determine the percentage of the budget that needs to be reduced. This percentage can be understood as the proportion by which the budget needs to be reduced in the current time period t compared to the time period t-1. For example, suppose C... t For 100, A t-1 If the value is 120, then the budget needs to be reduced by 20% in the current period t.

[0123] In this embodiment, it is desirable to give users with higher exposure levels a higher probability of selection, thereby allocating a larger proportion of the budget that needs to be reduced. Therefore, for users at the current exposure level l, the normalized selection probability from the previous time period is used. The proportion of the budget reduction is weighted and allocated.

[0124] Furthermore, in this example, we aim to give more exposure to live streams with more engaging content. Therefore, for live streams with engaging content in time slot t-1, we want to reduce the weight of decreased selection probability in the current time slot t, so that they receive more exposure than live streams without engaging content. Thus, we normalize the number of engaging content in live streams during time slot t-1 using... This has mitigated the decline in the probability of selecting live streams with exciting visuals.

[0125] Since the selection probability for the current time period t needs to be reduced based on the selection probability for time period t-1, it is necessary to subtract the weighted budget percentage from 1. Ensure that the probability of selection in the current time period t is necessarily less than the probability of selection in the time period t-1.

[0126] In this way, for each reference live stream, the reference live stream can be filtered by selecting a probability to obtain the target live stream.

[0127] In one implementation, the target live stream is obtained by filtering reference live streams based on selected probabilities, including:

[0128] For the reference live stream, a reference probability is randomly generated using a random function;

[0129] If the probability of selection is greater than the reference probability, then the reference live stream will be selected as the target live stream.

[0130] For example, suppose there are three live streaming rooms D, E and F; the probability of a user in the l-th layer choosing live streaming room D is 0.7, the probability of a user in the l-th layer choosing live streaming room E is 0.5, and the probability of a user in the l-th layer choosing live streaming room F is 0.3.

[0131] Assuming the randomly generated reference probability is 0.4, the selection probabilities of live stream room D and live stream room E are both greater than 0.4. Therefore, live stream rooms D and E can be identified as target live stream rooms. Since the selection probability of live stream room F is less than the reference probability, live stream room F will not be identified as a target live stream room.

[0132] This embodiment can use the rand() function to randomly generate the reference probability, or it can be generated using other methods; no restrictions are imposed here.

[0133] In this way, when filtering reference live streams using game data, highlight moments, and streamer information, it is equivalent to using the streamer's gaming skills for filtering, thereby ensuring that the target live streams obtained are those with high-level streamers, and increasing the exposure of talented new streamers.

[0134] S213, the target live stream room is tagged using the streamer information so that the corresponding streamer description information is displayed on the cover of the target live stream room; the streamer description information is used to characterize the streamer's gaming skills.

[0135] Once the target live stream is identified, it is tagged so that the corresponding streamer description information is displayed on the target live stream's cover. The streamer description information is used to represent the streamer's gaming skills. For example, the streamer information can include the streamer's rank, record, and highlights in the current game.

[0136] Specifically, this embodiment mainly performs tagging by setting an enumeration object, the format of which is as follows:

[0137] Public enum CornerTag{}

[0138] The characteristic of an enumeration object is that its enumeration value is unique. Therefore, in this embodiment, the enumeration value and its index are mapped one-to-one. Thus, the constructor function for the enumeration object is:

[0139] CornerTag(String info, intres)

[0140] Here, `info` contains descriptive information for the index, and `res` is an integer variable used to mark the resource ID value. For example, if the user selects game character A, then `res` will be A.

[0141] This establishes the basic data structure information of the enumeration object. Next, the subscript information to be displayed needs to be defined in the enumeration class CornerTag, as follows:

[0142] GUESS (“Platinum”, R.drawable.guess_icon)

[0143] Here, GUESS represents an enumeration value in the CornerTag enumeration class, "Platinum" represents the description information of the badge (i.e., the player's current rank in a certain game), and R.drawable.guess_icon represents the icon resource file information of the badge.

[0144] Similarly, all badge information (such as character badges, stat badges, and highlight badges) can be set in the CornerTag class using the same method, and then displayed on the cover of the reference live stream. The CornerTag class can be understood as a badge data pool for all badge information that needs to be displayed.

[0145] Based on the same inventive concept as in the foregoing embodiments, this embodiment also provides a server for recommending game live streaming rooms, such as... Figure 2 As shown, the server includes:

[0146] Determining unit 41 is used to determine the reference live room that is in live streaming status and located in the target partition;

[0147] Acquisition unit 42 is used to acquire the anchor information and highlights of each reference live broadcast room;

[0148] The filtering unit 43 is used to filter the reference live streaming rooms based on the anchor information to obtain the target live streaming room;

[0149] The tagging unit 44 is used to tag the target live room using the streamer information, so as to display the corresponding streamer description information on the cover of the target live room; the streamer description information is used to characterize the streamer's gaming skills.

[0150] Since the server described in this embodiment of the invention is a server for the recommended game live streaming room for implementing this embodiment, those skilled in the art can understand the specific structure and variations of this server based on the method described in this embodiment, and therefore will not be repeated here. All servers used in the methods of this embodiment of the invention fall within the scope of protection of this invention.

[0151] Based on the same inventive concept, this embodiment provides a computer device 500, such as... Figure 3As shown, it includes a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 511, it implements any step of the method described above.

[0152] Based on the same inventive concept, this embodiment provides a computer-readable storage medium 600, such as... Figure 3 As shown, a computer program 611 is stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0153] Through one or more embodiments of the present invention, the present invention has the following beneficial effects or advantages:

[0154] This invention provides a method, server, medium, and device for recommending game live streaming rooms. The method includes: identifying reference live streaming rooms that are currently live and located in a target partition; acquiring the streamer information and highlight moments of each reference live streaming room; filtering the reference live streaming rooms based on the highlight moments and streamer information to obtain target live streaming rooms; and tagging the target live streaming rooms using the streamer information to display corresponding streamer description information on the cover of the target live streaming room. The streamer description information is used to characterize the streamer's gaming skills. Thus, when filtering reference live streaming rooms using game data, highlight moments, and streamer information, it is equivalent to filtering using the streamer's gaming skills, thereby ensuring that the obtained target live streaming rooms are those with high-level streamers, increasing the exposure rate of new and skilled streamers. Furthermore, by tagging the streamer description information on the cover of the corresponding live streaming room, users can select live streaming rooms of interest based on the streamer description information, increasing the effective viewership.

[0155] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0156] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0157] Similarly, it should be understood that, in order to simplify this disclosure and aid in understanding one or more of the various aspects of the invention, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, this method of disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into this detailed description, wherein each claim itself is a separate embodiment of the invention.

[0158] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0159] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0160] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some or all of the components of the gateway, proxy server, or system according to embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing some or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0161] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0162] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0163] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for recommending game live streaming rooms, characterized in that, The method includes: Identify reference live stream rooms that are currently live and located in the target partition; Obtain information about the streamers and highlights from each reference live stream; Based on the aforementioned highlights and the streamer information, the reference live streams are filtered to obtain the target live stream; The target live stream is tagged using the streamer information so that the corresponding streamer description information is displayed on the cover of the target live stream; the streamer description information is used to characterize the streamer's gaming skills. The step of filtering the reference live streams based on the highlight moments and the streamer information to obtain target live streams includes: determining the exposure level of the target user, where the target user is a user who has an exposure request for the reference live streams; for any reference live stream, determining the total budget for the reference live stream for the day based on the streamer information, and determining the budget allocation for the current time period based on the total budget; adjusting the budget allocation based on the total actual consumption of the reference live stream before the current time period on the day to obtain the target allocation for the current time period; determining the probability of a user at the current exposure level selecting the reference live stream for the current time period based on the target allocation for the current time period and the highlight moments; and filtering the reference live streams based on the selection probability to obtain target live streams. The process of determining the exposure level of a target user includes: obtaining the total viewing time of each user in the live streaming section of the reference live streaming room within a preset historical time period; sorting the total viewing time to obtain a viewing time set W; dividing the viewing time set into viewing time intervals corresponding to the number of levels based on the number of users and the preset number of levels; obtaining the target viewing time of the target user in the live streaming section of the reference live streaming room and determining the target viewing time interval into which the target viewing time falls; and taking the level corresponding to the target viewing time interval as the exposure level of the target user.

2. The method as described in claim 1, characterized in that, The process of acquiring highlights from each reference live stream includes: Acquire game video footage from each reference live stream; The game video footage is identified using an image recognition algorithm to determine the highlight moments.

3. The method as described in claim 2, characterized in that, The method further includes: When a confirmation message from the game server indicating that the streamer has started a live game broadcast is received, the streamer information and the game video feed sent by the game server are received via the game data broadcast channel. or, By using the Application Programming Interface (API) to reverse-engineer the game client's game data transmission channel, the program nodes that send and receive data can be obtained. Hook the program node that sends data and the program node that receives data to obtain a data sending hook program module and a data receiving hook program module. A second SDK is created based on the data sending hook module and the data receiving hook module, and the second SDK is sent to the broadcaster's client. The second SDK is injected into the game process when the broadcaster starts the game and broadcasts live. The second SDK is used to obtain the game video screen and broadcaster information published by the game client. Receive the game video footage and the streamer information sent by the streamer-side client.

4. The method as described in claim 1, characterized in that, The step of adjusting the budget allocation based on the total actual consumption of the reference live stream room before the current time period of the day to obtain the target allocation for the current time period includes: According to the formula The budget allocation is adjusted to obtain the target allocation C for the current time period. t ;in, The B t The budget allocation for the reference live stream room in the current time period is defined as follows: B is the total budget for the reference live stream room for the day; i is any time period; t is the current time period; K is the number of time periods included in the day; and A is the budget allocation for the reference live stream room for the current time period. i To reference the actual consumption of the live stream in the i-th time period, the B represents the total actual consumption of the reference live stream room before the current time period on that day. i The budget allocation for the target live streaming room in the i-th time period.

5. The method as described in claim 4, characterized in that, The step of determining the probability of a user at the current exposure level selecting the reference live stream room during the current time period based on the target allocation amount for the current time period and the highlights includes: If it is determined that the target allocation amount for the current time period is greater than the actual consumption amount for the previous time period, then according to the formula... Increase the probability that users at the current exposure level will select the reference live stream in the current time period. in, The The probability that a user at the current exposure level selected the reference live stream in the previous time period is given by l, where l is the current exposure level, and C is the reference live stream. t The target allocation amount for the current time period, A t-1 The current time period represents the actual consumption in the previous time period, where j represents any exposure level, L represents the total number of exposure levels, and so on. Let e ​​be the probability that a user at the j-th exposure level selected the reference live stream in the previous time period of the current time period. The probability that a user at the current exposure level selected the reference live stream in the initial time period is a preset value. t-1 It is the number of highlights extracted from the previous time period of the reference live stream at the current moment, maxe t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

6. The method as described in claim 4, characterized in that, The step of determining the probability of a user at the current exposure level selecting the reference live stream during the current time period based on the target allocation amount, game data, and the highlights includes: If it is determined that the target allocation amount for the current time period is less than the actual consumption amount for the previous time period, then according to the formula... Reduce the probability that users at the current exposure level will select the reference live stream in the current time period. in, The The probability that a user at the current exposure level selected the reference live stream in the previous time period is given by l, where l is the current exposure level, and C is the reference live stream. t The target allocation amount for the current time period, A t-1 The current time period represents the actual consumption in the previous time period, where j represents any exposure level, L represents the total number of exposure levels, and so on. Let e ​​be the probability that a user at the j-th exposure level selected the reference live stream in the previous time period of the current time period. The probability that a user at the current exposure level selected the reference live stream in the initial time period is a preset value. t-1 It is the number of highlights extracted from the previous time period of the reference live stream at the current moment, maxe t-1 It represents the maximum number of highlights extracted from the previous time period across all live streams at the current moment.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-6.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1-6.

Images