In-game original content recommendation method and apparatus, electronic device, and medium

Abstract

An in-game original content recommendation method and apparatus, an electronic device, and a medium. The method comprises: according to a first preset recommendation ratio, separately selecting recent original content from each first content pool of different levels, and recommending same to a user (S101); on the basis of feedback information from the user for target recent original content, determining recent inferior content, and removing the recent inferior content from multiple first content pools (S102); acquiring new recent original content and adding same to at least one target first content pool among the multiple first content pools, and, according to a recommendation rule, recommending the new recent original content to the user (S103). The new recent original content is recent original content that has not been put into the multiple first content pools. The foregoing method solves the problems in existing recommendation methods of cold starting, excessive recommendations, and narrow recommendations.

Description

A method, device, electronic device and medium for recommending original content in games

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to a Chinese patent application filed on October 7, 2023, with application number 202311284381.6, entitled “A method, device, electronic device and medium for recommending original content in a game”, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present disclosure relates to the field of computer technology, and more particularly to a method, device, electronic device, and medium for recommending original content in a game. Background Art

[0004] With the continuous development of the gaming market, UGC (User Generated Content) games are gradually entering the field of vision of users (players). Thanks to the evolution and advancement of gaming tools and engines, users can now more easily create and share their own game content. In UGC games, as the amount and variety of user-generated UGC content increases, a good UGC recommendation system can serve as a bridge between users and UGC content, increasing the level of interaction in UGC games. Therefore, how to recommend appropriate UGC content to different users has become a crucial aspect of the game.

[0005] Summary of the Invention

[0006] The present disclosure provides a method, device, electronic device, and medium for recommending original content in a game.

[0007] In a first aspect, an embodiment of the present disclosure provides a method for recommending original content in a game, comprising:

[0008] Selecting recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommending it to the user;

[0009] Determining recent low-quality content based on user feedback on target recent original content, and removing the recent low-quality content from the multiple first content pools;

[0010] New recent original content is obtained and added to at least one target first content pool among the multiple first content pools, and the new recent original content is recommended to users according to the recommendation rules. The new recent original content is recent original content that has not been placed in the multiple first content pools.

[0011] In a second aspect, an embodiment of the present disclosure further provides a device for recommending original content in a game, the device comprising:

[0012] A first recommendation module, configured to select recent original content from each first content pool of different levels and recommend it to the user according to a first preset recommendation ratio;

[0013] a content updating module, configured to determine recent low-quality content based on user feedback on target recent original content, and remove the recent low-quality content from the plurality of first content pools;

[0014] The second recommendation module is used to obtain new recent original content to supplement at least one target first content pool among the multiple first content pools, and recommend the new recent original content to users according to the recommendation rules. The new recent original content is recent original content that has not been placed in the multiple first content pools.

[0015] In a third aspect, an embodiment of the present disclosure further provides an electronic device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the original content recommendation method in the game as described above are performed.

[0016] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which executes the steps of the above-mentioned method for recommending original content in a game when the computer program is executed by a processor.

[0017] The embodiments of the present disclosure bring the following beneficial effects:

[0018] The embodiments of the present disclosure provide a method, device, electronic device, and medium for recommending original content in games. These methods can directly obtain new recent original content and recommend it to users, so that newly registered users can obtain new recent original content and display new recently released content, thus avoiding cold starts. At the same time, these methods can select recent original content in different user preference ranges from multiple content pools and recommend it to users, thus increasing the breadth of recommendations. Compared with the original content recommendation methods in games in the prior art, these methods solve the problems of cold starts, over-recommendations, and narrow recommendations in the existing recommendation methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] FIG1 shows a flowchart of a method for recommending original content in a game according to an embodiment of the present disclosure;

[0020] FIG2 is a schematic diagram showing a recent original content update process in a plurality of first content pools according to an embodiment of the present disclosure;

[0021] FIG3 shows a flow chart of one embodiment of the present disclosure for calculating a likeability score;

[0022] FIG4 shows a schematic structural diagram of an original content recommendation device in a game according to one embodiment of the present disclosure;

[0023] FIG5 shows a schematic structural diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. The components of the embodiments of the present disclosure generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure provided in the drawings is not intended to limit the scope of the disclosure for which protection is sought, but merely represents selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, every other embodiment obtained by those skilled in the art without making creative work falls within the scope of protection of the present disclosure.

[0025] Recommendation technology solutions are mainly divided into three types: recommendations based on user similarity, recommendations based on item similarity, and recommendations based on collaborative filtering. The above recommendation algorithms are all personalized recommendation methods, that is, by analyzing user behavior, finding content similar to the target user's interests, and then making recommendations.

[0026] However, these recommendation methods require extensive user data for model training, resulting in a severe cold-start problem. They are unable to effectively recommend content to newly registered users or newly published content. Furthermore, they are prone to over-recommendations and narrow recommendations when recommending content similar to user interests.

[0027] In view of the above problems, the present disclosure provides a method, device, electronic device and medium for recommending original content in games.

[0028] In one embodiment of the present disclosure, the method for recommending original content in a game can be run on a local terminal device or a server. When the method for recommending original content in a game is run on a server, the method can be implemented and executed based on a cloud interaction system, wherein the cloud interaction system includes a server and a client device.

[0029] In an optional embodiment, various cloud applications can be run under the cloud interaction system, such as cloud games. Taking cloud games as an example, cloud games refer to a gaming method based on cloud computing. In the cloud game operation mode, the operating body of the game program and the main body of the game screen presentation are separated. The storage and operation of the original content recommendation method in the game are completed on the cloud game server. The role of the client device is to receive and send data and present the game screen. For example, the client device can be a display device with data transmission function close to the user side, such as a mobile terminal, TV, computer, PDA, etc.; but the cloud game server in the cloud is responsible for information processing. When playing the game, the player operates the client device to send operation instructions to the cloud game server. The cloud game server runs the game according to the operation instructions, encodes and compresses the game screen and other data, and returns it to the client device through the network. Finally, the client device decodes and outputs the game screen.

[0030] In an optional embodiment, taking a game as an example, a local terminal device stores a game program and is used to present the game screen. The local terminal device is used to interact with the player through a graphical user interface, that is, conventionally downloading and installing the game program through an electronic device and running it. The local terminal device can provide the graphical user interface to the player in a variety of ways, for example, it can be rendered and displayed on the terminal's display screen, or provided to the player through holographic projection. For example, the local terminal device may include a display screen and a processor, the display screen is used to present the graphical user interface, the graphical user interface includes the game screen, and the processor is used to run the game, generate the graphical user interface, and control the display of the graphical user interface on the display screen.

[0031] In one possible implementation, an embodiment of the present disclosure provides a method for recommending original content in a game, providing a graphical user interface through a terminal device, wherein the terminal device can be the local terminal device mentioned above, or the client device in the cloud interaction system mentioned above.

[0032] Please refer to Figure 1, which is a flow chart of a method for recommending original content in a game provided by an embodiment of the present disclosure. As shown in Figure 1, the method for recommending original content in a game provided by an embodiment of the present disclosure includes:

[0033] Step S101 : selecting recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommending the content to the user.

[0034] In this step, the first preset recommendation ratio may refer to a recommendation ratio between recent original contents respectively selected from a plurality of first content pools, and the first preset recommendation ratio represents a recommendation weight of each first content pool.

[0035] The first content pool may refer to a resource pool for storing recent original content. The recent original content stored in each first content pool is different. A recent original content can only be stored in one of the multiple first content pools.

[0036] Exemplarily, the multiple first content pools include a first content pool D, a first content pool C, a first content pool B, a first content pool A, and a first content pool S.

[0037] The user preference interval may refer to a user preference range corresponding to each first content pool, and the user preference interval corresponding to each first content pool is determined by recent original content stored in the first content pool.

[0038] Assume that the first content pool C stores 100 recent original contents, the maximum user preference of these 100 recent original contents is 10,000, and the minimum user preference is 1,000. Then the user preference range corresponding to the first content pool C is 1,000 to 10,000.

[0039] Recent original content may refer to original content posted by users in a recent period of time, where the recent period of time is a set time range. For example, original content posted by users in the last 7 days is considered recent original content.

[0040] Original content can refer to content created by users in the game. In UGC games, users can share their own created content with other users for experience. These contents that can be shared and experienced are called original content.

[0041] For example, original content can be user-generated game levels, maps, and copies in UGC games, or user-generated pets and items.

[0042] In the disclosed embodiments, in order to promptly recommend recently released original content to users, a certain amount of recent original content can be randomly selected from the recent original content stored in each first content pool for recommendation to users. This ensures that the original content recommended to users is not only recently released content, but also that different recent original content falls within different user preference ranges. This increases the diversity of content recommendations and avoids the problem of narrow recommendations. Furthermore, even for newly registered users or new users who have just entered the game, recently released original content can be effectively recommended to them.

[0043] In one example, the content information of recent original content selected from multiple first content pools may be recommended to the user in response to the user's selection operation on the content recommendation interface, or the system may proactively recommend the content information of recent original content selected from multiple first content pools to the user when specific conditions are met.

[0044] Taking the above example, assuming that the first preset recommendation ratios of the first content pool D, the first content pool C, the first content pool B, the first content pool A, and the first content pool S are 10%, 20%, 35%, 25%, and 10% respectively, when selecting recent original content, a total of 100 recent original content need to be selected to recommend to users, then 100×10%=10 recent original content are selected from the first content pool D, 100×20%=20 recent original content are selected from the first content pool C, 100×35%=35 recent original content are selected from the first content pool B, 100×25%=25 recent original content are selected from the first content pool A, and 100×10%=10 recent original content are selected from the first content pool S.

[0045] It should be noted that the user preference interval corresponding to a first content pool is not fixed, and will change with changes in recent original content stored in the first content pool.

[0046] Step S102 : determining recent low-quality content based on user feedback information on target recent original content, and removing the recent low-quality content from the plurality of first content pools.

[0047] In this step, feedback information may refer to information about feedback operations performed by users on the target's recent original content. Feedback information is used to represent the user's satisfaction with the target's recent original content. Feedback information includes but is not limited to likes, collections, shares, and plays.

[0048] The recent low-quality content may refer to a preset number of recent original contents with the lowest user preference rankings among all recent original contents stored in the plurality of first content pools.

[0049] In the embodiment of the present disclosure, while ensuring that recently released original content can be recommended to users, it is also necessary to ensure the user's acceptance of the recommended content. To this end, the target recent original content is evaluated for user preference based on the feedback operations of different users on the target recent original content, so as to determine recent low-quality content based on user preference, and remove recent low-quality content from the first content pool to avoid repeatedly recommending recent original content with low user preference to users.

[0050] Taking original content as game levels as an example, after a recently released game level is recommended to users, some users can view detailed information about the recommended game level on the game recommendation page, such as the level name, level screenshots, level description, number of level plays, and number of level likes. Users can independently choose whether to play the level based on their preferences, and after playing the level, they can express their love for the level by liking, sharing, collecting, and rating it. Based on the feedback actions of different users, the overall love of users for a particular recent original content can be comprehensively determined. Based on the overall love ranking of different recent original content, recent low-quality content can be selected from all recent original content stored in multiple first content pools to remove the recent low-quality content from the multiple first content pools.

[0051] Step S103: Acquire new recent original content and add it to at least one target first content pool among the multiple first content pools, and recommend the new recent original content to the user according to the recommendation rule.

[0052] In this step, the new recent original content is the recent original content that has not been put into multiple first content pools. The new recent original content is the recently released original content that has not been recommended to the user.

[0053] The recommendation rule may refer to a rule for recommending new recent original content, and the recommendation rule is used to ensure that new recent original content will definitely be recommended to users.

[0054] In the first example, the recommendation rule is a traffic decay rule. This rule sets a base traffic volume for new, recent original content. Each time a new, recent original content is exposed, the remaining traffic volume is equal to the base traffic volume minus one. When the remaining traffic volume reaches 0, the new, recent original content is no longer exposed. Traffic decay rules are more suitable for scenarios where users actively access recommended content.

[0055] In the second example, the recommendation rule is a push count rule. This rule recommends new, recent original content based on a set number of recommendations. Regardless of whether the new, recent original content has been exposed, the remaining number of recommendations decreases by one each time it is recommended. When the remaining number of recommendations reaches 0, it will no longer be recommended. This rule is more suitable for scenarios where users passively obtain recommended content.

[0056] In the disclosed embodiment, in order to timely update the recent original content in multiple first content pools and recommend users' newly released recent original content to users, the newly released recent original content can be stored in the first resource library in the order of release time. Then, the first content pool can be periodically checked. When new recent original content needs to be added, the earliest released recent original content is obtained from the first resource library, added to the first content pool, and removed from the first resource library, or marked as having been added to the first content pool, to avoid repeated addition of the earliest released recent original content to the first content pool in the future.

[0057] Here, when new recent original content is added to multiple first content pools, it may be added to only one of the multiple first content pools, or it may be added to multiple first content pools, depending on the user preference level of the added new recent original content.

[0058] In addition, if new recent original content is added to multiple first content pools and multiple first content pools are not updated, the user preference intervals for the recent original content stored in first content pool D may overlap with the user preference intervals for recent original content stored in other first content pools. However, after multiple first content pools are updated, the user preference intervals corresponding to different first content pools will not overlap. Multiple first content pool updates refer to updating the recent original content stored in the first content pools with the new recent original content added to the first content pools, thereby removing the recent low-quality content stored in the first content pools.

[0059] In the embodiment of the present disclosure, the multiple first content pools include a first active content pool D and multiple first passive content pools, and the multiple first passive content pools are respectively a first passive content pool C, a first passive content pool B, a first passive content pool A, and a first passive content pool S. After removing recent low-quality content from the first content pool, new recent original content will be added to the first content pool. The newly added recent original content first enters the first active content pool and then enters the multiple first passive content pools. The following is an introduction to the update process of recent original content in the multiple first content pools in conjunction with Figure 2. Among them, the active content pool refers to a content pool that actively obtains original content. The passive content pool refers to a content pool that passively obtains original content.

[0060] FIG2 is a schematic diagram showing a recent original content update process in multiple first content pools provided by an embodiment of the present disclosure.

[0061] As shown in Figure 2, in this example, at least one first target content pool includes a first active content pool D, and obtaining new recently released original content to supplement at least one target first content pool among multiple first content pools includes: when the number of recent original content in the first active content pool D is less than the content pool quantity threshold corresponding to the first active content pool D, obtaining new recent original content and adding it to the first active content pool D, so that the number of recent original content in the first active content pool D is equal to the content pool quantity threshold corresponding to the first active content pool D.

[0062] In order to replenish the first active content pool D in a timely manner, the amount of recent original content in the first active content pool D is checked every 30 seconds to see if it is less than the content pool quantity threshold of 1600 for the first active content pool D. If the amount of recent original content in the first active content pool D is less than 1600, the first resource library is scanned to obtain new recent original content. The amount of new recent original content obtained is equal to the difference between the content pool quantity threshold of 1600 and the current amount of recent original content stored in the first active content pool, so that the amount of recent original content in the first active content pool D is equal to 1600. The time for checking the first active content pool D and the content pool quantity threshold of the first active content pool D are set values. Those skilled in the art can select the specific values ​​of the detection time and the content pool quantity threshold according to actual conditions, and this disclosure does not limit them here.

[0063] In an optional embodiment, after acquiring new recent original content and adding it to the first active content pool D, the following steps are included: step a1 to step a3.

[0064] Step a1: Set basic traffic for new recent original content and set the new recent original content to an exposure state.

[0065] In order to enable the new recent original content in the first active content pool D to be effectively recommended to users, a basic flow is set for each new recent original content added to the first active content pool D. For example, the basic flow set for each new recent original content is 500. After setting the basic flow, each new recent original content is set to an exposure state, and the original content in the exposure state is recommended to users.

[0066] Step a2: When new recent original content is browsed by a user, the basic traffic of the new recent original content in the exposure state is reduced by one.

[0067] Taking the new recent original content A as an example, after the recent original content A is recommended to the user, if a user browses the recent original content on the recommendation page for a time greater than or equal to 1 second, it is determined to be exposed once. At this time, the basic traffic of the recent original content A in the exposed state will be reduced by one.

[0068] Step a3: When the basic traffic of the new recent original content decreases to 0, the status of the new recent original content is changed from the exposure status to the waiting status. The recent original content in the waiting status will not be recommended to the user.

[0069] When recent original content A is viewed 500 times, the basic traffic of recent original content A is reduced to 0. At this time, the status of the recent original content A is changed from exposure status to waiting status. Recent original content A in the waiting status will not be recommended to users.

[0070] The waiting period is set because different game levels have different play times. Some levels may take longer to play, and feedback is only uploaded after the game is completed. Therefore, a waiting period is required to collect as complete player feedback as possible. In one example, the waiting period is one hour, meaning that the number of plays, likes, favorites, and shares for recent original content A within that hour will be accumulated and recorded on the recent original content A.

[0071] In an optional embodiment, step S102 includes: steps b1 to b4.

[0072] Step b1: Determine the number of recent original contents in a ranked state in the first active content pool.

[0073] When original content changes from exposure status to waiting status, the cumulative time that recent original content has been in waiting status is calculated. When the cumulative time reaches the waiting time threshold, the waiting status ends and the new recent original content is changed from waiting status to ranking status.

[0074] Whenever an original piece of content moves from the waiting state to the ranking state, the number of recently ranked original content in the first active content pool is counted. This number is used to determine whether to begin scoring the popularity of the ranked original content. To ensure fairness, feedback information will no longer be collected once the original content enters the ranking state.

[0075] Step b2: When the number of recent original contents in the ranking state reaches a first rating number threshold, determine the likeability ratings of the recent original contents in the ranking state.

[0076] When the number of ranked original content in the first active content pool D reaches a first rating threshold of 100, a likeability score calculation is performed for each recently ranked original content. The likeability score is determined by user feedback on the original content. For example, the more shares and likes a piece receives, the higher the likeability score; the fewer shares and likes a piece receives, the lower the likeability score.

[0077] Step b3: taking a first set number of recent original contents with high popularity scores in the first active content pool as recent high-quality contents, and transferring the recent high-quality contents to a plurality of first passive content pools.

[0078] Upgrade original content with higher likeability scores to a more advanced first content pool, that is, use the top M original content with higher likeability scores as recent high-quality content, put the M recent high-quality content into the first passive content pool C, and remove the M recent high-quality content from the first active content pool D.

[0079] The multiple first content pools are ranked, from lowest to highest, as first active content pool D, first passive content pool C, first passive content pool B, first passive content pool A, and first passive content pool S. That is, first active content pool D is the lowest-ranked first content pool, and the first passive content pools are all ranked higher than the first active content pool. The rank of a first content pool corresponds to the likeability score of the original content within that content pool. The higher the likeability score of the original content, the higher the rank of the first content pool; the lower the likeability score of the original content, the lower the rank of the first content pool.

[0080] Step b4: utilizing recent high-quality content to remove recent low-quality content from the plurality of first content pools.

[0081] After placing M recent high-quality content into the first passive content pool and removing M recent high-quality content from the first active content pool, the remaining original content in the first active content pool D is deleted, completing a round of updates to the first active content pool D. Simultaneously, recent low-quality content in multiple first passive content pools is identified and removed from the multiple first passive content pools.

[0082] In an optional embodiment, multiple first passive content pools have corresponding levels from high to low according to the recommendation degree of their respective stored recent original content, and recent low-quality content is removed from the multiple first content pools by using recent high-quality content, including: steps c1 to c3.

[0083] Here, since the popularity score of the new recent original content added to the first passive content pool C may be higher than the popularity score of the recent original content in other first passive content pools, a progressive upgrade method is used to determine which first passive content pool the newly added recent original content should be stored in.

[0084] Because recent original content undergoes a progressive upgrade, the popularity rating range within each first content pool is not fixed. To facilitate the description of the level of a first content pool, the level of the first content pool can be determined based on the recommendation level of the recent original content stored in the first content pool. A higher recommendation level indicates a higher level. A higher overall popularity rating indicates a higher recommendation level; a lower overall popularity rating indicates a lower recommendation level.

[0085] Step c1: Add a first set number of recent high-quality content to a first passive content pool of the lowest level.

[0086] Taking the above example, if M recent high-quality contents are placed into multiple first passive content pools, the first set number is M. When placing M recent high-quality contents into multiple first passive content pools, these M recent high-quality contents are first placed into the first passive content pool C, that is, the lowest-level first passive content pool, because the lowest-level first passive content pool stores the recent original content with the lowest popularity rating among all first passive content pools.

[0087] Step c2 starts with adding a first set number of recent high-quality contents to the lowest-level first passive content pool, and progressively upgrades the recent high-quality contents in each first passive content pool to a higher-level first passive content pool, until the recent high-quality contents in the second-highest-level first passive content pool are upgraded to the highest-level first passive content pool.

[0088] Compare M recent high-quality contents with the recent original contents stored in the first passive content pool C to determine which recent high-quality contents in the first passive content pool C can be upgraded to the first passive content pool B. The recent high-quality contents selected from the first passive content pool C that can be upgraded to the first passive content pool B are called the first recent high-quality contents.

[0089] Then, the first recent high-quality content upgraded to the first passive content pool B is compared with the recent original content stored in the first passive content pool B to determine which recent high-quality content in the first passive content pool B can be upgraded to the first passive content pool A. The recent high-quality content selected from the first passive content pool B that can be upgraded to the first passive content pool A is called the second recent high-quality content.

[0090] The same process is repeated until the third recent high-quality content in the first passive content pool A is upgraded to the first passive content pool S.

[0091] Step c3, starting from the highest-level first passive content pool to which recent high-quality content is added, progressively downgrade the recent low-quality content in each first passive content pool to a first passive content pool of a lower level, until a first set number of recent low-quality content is removed from the lowest-level first passive content pool.

[0092] Determine the first recent low-quality content in the first passive content pool S, transfer the first recent low-quality content to the first passive content pool A, determine which recent low-quality content in the first passive content pool A can be downgraded to the first passive content pool B, and call the recent low-quality content selected from the first passive content pool A that can be downgraded to the first passive content pool B the second recent low-quality content.

[0093] Then, it is determined which recent low-quality content in the first passive content pool B can be downgraded to the first passive content pool C. The recent low-quality content selected from the first passive content pool B that can be downgraded to the first passive content pool C is called the third recent high-quality content.

[0094] This process is repeated in this way until the fourth recent low-quality content in the first passive content pool C is removed from the first passive content pool C. During a round of promotion and demotion, a single recent high-quality content can only be promoted once, and a single recent low-quality content can only be demoted once, to avoid situations where a single recent high-quality content is continuously promoted or demoted within a round.

[0095] In an optional embodiment, step c2 includes: steps c21 to c24.

[0096] Step c21: Select the lowest-level first passive content pool as the first upgraded passive content pool.

[0097] When performing recent high-quality content upgrade, the first passive content pool C is first used as the first upgraded passive content pool.

[0098] Step c22: Determine whether the first upgraded passive content pool is the highest-level first passive content pool.

[0099] Step c23: If the first upgraded passive content pool is not the highest-level first passive content pool, select recent high-quality content from the recent original content in the first upgraded passive content pool and transfer the recent high-quality content to a first passive content pool that is one level higher than the first upgraded passive content pool.

[0100] Since the first passive content pool C is not the highest-level first passive content pool, when the conditions are met, the first recent candidate high-quality content is selected from the recent original content of the first passive content pool C, the first recent high-quality content is selected from the first recent candidate high-quality content, and then the first recent high-quality content is transferred from the first passive content pool C to the first passive content pool B.

[0101] If the first passive content pool is the highest-level first passive content pool, the recent high-quality content upgrade process ends, and the recent low-quality content downgrade process begins.

[0102] Step c24: Use the higher-level first passive content pool as a new first upgraded passive content pool, and return to the step of selecting recent high-quality content from the recent original content in the first upgraded passive content pool.

[0103] After the first recent high-quality content is transferred to the first passive content pool B, the first passive content pool B is used as a new first upgraded passive content pool, and the second recent high-quality content is continuously selected from the first passive content pool B.

[0104] After transferring the first recent high-quality content to the first passive content pool B, when the conditions are met, select the second recent candidate high-quality content from the recent original content of the first passive content pool B, select the second recent high-quality content from the second recent candidate high-quality content, and transfer the second recent high-quality content from the first passive content pool B to the first passive content pool A.

[0105] Then, the first passive content pool A is used as the new first upgraded passive content pool, and when the conditions are met, the third recent candidate high-quality content is selected from the recent original content of the first passive content pool A, and the third recent high-quality content is selected from the third recent candidate high-quality content, and the third recent high-quality content is transferred from the first passive content pool A to the first passive content pool S.

[0106] Since the first passive content pool S is the highest-level first passive content pool, the recent high-quality content upgrade process ends, and the recent low-quality content downgrade process begins.

[0107] In an optional embodiment, step c3 includes: steps c31 to c34.

[0108] Step c31: Select the first passive content pool with the highest level as the first degraded passive content pool.

[0109] When performing recent low-quality content degradation, the first passive content pool S is first used as the first degraded passive content pool.

[0110] Step c32: Determine whether the first degraded passive content pool is the lowest-level first passive content pool.

[0111] Step c33: If the first degraded passive content pool is not the lowest-level first passive content pool, select recent low-quality content from the recent original content in the first degraded passive content pool and transfer the recent low-quality content to a first passive content pool one level lower than the first degraded passive content pool.

[0112] The first passive content pool S accepts the third recent high-quality content upgraded from the first passive content pool A. Since the first passive content pool S is not the lowest-level first passive content pool, when the conditions are met, the first recent candidate low-quality content is selected from the recent original content of the first passive content pool S, and the first recent low-quality content is selected from the first recent candidate low-quality content, and the first recent low-quality content is downgraded from the first passive content pool S to the first passive content pool A.

[0113] Step c34: Use the first passive content pool at a lower level as a new first degraded passive content pool, and return to the step of selecting recent low-quality content from the recent original content in the first degraded passive content pool.

[0114] The first passive content pool A is used as the new first downgraded passive content pool. When the conditions are met, the second recent candidate low-quality content is selected from the recent original content of the first passive content pool A, and the second recent low-quality content is selected from the second recent candidate low-quality content. The second recent low-quality content is transferred from the first passive content pool A to the first passive content pool B.

[0115] The first passive content pool B is used as the new first degraded passive content pool. When the conditions are met, the third recent candidate low-quality content is selected from the recent original content of the first passive content pool B, and the third recent low-quality content is selected from the third recent candidate low-quality content. The third recent low-quality content is transferred from the first passive content pool B to the first passive content pool C.

[0116] The first passive content pool C is used as the new first downgraded passive content pool. Since the first passive content pool C is the lowest-level first passive content pool, when the conditions are met, the fourth recent candidate low-quality content is selected from the recent original content of the first passive content pool C, and the fourth recent low-quality content is selected from the fourth recent candidate low-quality content. The fourth recent low-quality content is removed from the first passive content pool C, completing a round of recent low-quality content removal process.

[0117] In an optional embodiment, step c23 selects recent high-quality content from the recent original content in the first upgraded passive content pool, including: steps c231 to c233.

[0118] Step c231: When the evaluation conditions are met, recent original content is sorted according to the time sequence of entering the first upgraded passive content pool, and a fixed proportion of recent original content with high rankings is selected as recent candidate high-quality content.

[0119] The evaluation condition may refer to a condition for scoring the likeability. In the embodiment of the present disclosure, the evaluation condition is that the recent original content in the first upgraded passive content pool reaches a corresponding content pool quantity threshold.

[0120] Taking the first upgraded passive content pool, first passive content pool C, as an example, the time when new recent original content enters first passive content pool C is recorded. When the recent original content in first passive content pool C reaches the content pool quantity threshold of first passive content pool C, all recent original content stored in first passive content pool C is sorted according to the time of entry. The top 50% of recent original content is selected as the first candidate recent high-quality content.

[0121] Taking the first upgraded passive content pool as the first passive content pool B as an example, after the first recent high-quality content is transferred to the first passive content pool B, the time when the first recent high-quality content enters the first passive content pool B is recorded. When the recent original content in the first passive content pool B reaches the content pool quantity threshold of the first passive content pool B, all recent original content stored in the first passive content pool B are sorted according to the time of entering the content pool, and the top 50% of recent original content are selected as the second recent candidate high-quality content.

[0122] Step c232: Determine the current likeability score of the recent candidate high-quality content.

[0123] Obtain the latest feedback information of recent candidate high-quality content, and calculate the current likeability score of the recent candidate high-quality content based on the latest feedback information and the likeability score formula.

[0124] Step c233: sort the recent high-quality content candidates in descending order of current likeability scores, and select a preset upgraded number of recent high-quality content candidates corresponding to the first upgraded passive content pool from a first set number of top rankings as the final recent high-quality content.

[0125] The preset upgrade quantity corresponding to each first passive content pool is different. Taking the first set quantity as M as an example, the preset upgrade quantity corresponding to the first passive content pool C is M / 2, the preset upgrade quantity corresponding to the first passive content pool B is M / 4, and the preset upgrade quantity corresponding to the first passive content pool A is M / 8.

[0126] Taking the first upgraded passive content pool as the first passive content pool C as an example, if a total of 16 new recent original content are added to the first passive content pool C, 8 first recent high-quality content will be selected from the first passive content pool C and transferred to the first passive content pool B. These 8 first recent high-quality content are the top-ranked recent candidate high-quality content corresponding to the preset upgraded number of the first passive content pool C, and are also the final recent high-quality content selected from the first passive content pool C.

[0127] In an optional embodiment, step c33 selects recent low-quality content from recent original content in the first degraded passive content pool, including steps c331 to c333.

[0128] Step c331 : When the evaluation conditions are met, recent original content is sorted according to the time sequence of entering the first degraded passive content pool, and a fixed proportion of recent original content with high rankings is selected as recent candidate low-quality content.

[0129] The evaluation condition may refer to a condition for scoring the likeability. In the embodiment of the present disclosure, the evaluation condition is that the recent original content in the first degraded passive content pool reaches a corresponding content pool quantity threshold.

[0130] Taking the first downgraded passive content pool as first passive content pool A, for example, first passive content pool A accepts M / 8 first recent low-quality content downgraded from first passive content pool S and records the time when the first recent low-quality content enters first passive content pool A. When the number of recent original content in first passive content pool A reaches the content pool quantity threshold for first passive content pool A, all recent original content stored in first passive content pool A is sorted by the time it entered the content pool. The top 50% of recent original content is selected as the second recent candidate low-quality content.

[0131] Taking the first downgraded passive content pool as the first passive content pool B as an example, the first passive content pool B accepts M / 4 second recent low-quality content downgraded from the first passive content pool A, records the time when the second recent high-quality content enters the first passive content pool B, and when the recent original content in the first passive content pool B reaches the content pool quantity threshold of the first passive content pool B, all recent original content stored in the first passive content pool B are sorted according to the time of entry into the content pool, and the top 50% of recent original content are selected as the second recent candidate low-quality content.

[0132] Step c332: Determine the current likeability score of the recent candidate low-quality content.

[0133] Obtain the latest feedback information of the recent candidate low-quality content, and calculate the current likeability score of the recent candidate low-quality content based on the latest feedback information and the likeability score formula.

[0134] Step c333 : sort the recent low-quality content candidates according to the current popularity score from high to low, and select the preset number of recent low-quality content candidates corresponding to the first downgraded passive content pool from the first set number of low-ranked recent low-quality content candidates as the final recent low-quality content.

[0135] The preset downgrade number corresponding to each first passive content pool is different. Taking the first set number as M as an example, the preset downgrade number corresponding to the first passive content pool S is M / 8, the preset downgrade number corresponding to the first passive content pool A is M / 4, the preset downgrade number corresponding to the first passive content pool B is M / 2, and the preset downgrade number corresponding to the first passive content pool C is M.

[0136] Taking the first degraded passive content pool as the first passive content pool C as an example, if a total of 16 new recent original contents are added to the first passive content pool C, 16 fourth recent low-quality contents will be selected and deleted from the first passive content pool C.

[0137] It can be seen that if all the first passive content pools are regarded as a whole, then in the first passive content pool combination, every time M recent original content is input, M recent original content will be eliminated, thereby ensuring the stability of the number of recent original content in the first passive content pool combination.

[0138] In an optional embodiment, the numerical values ​​of the preset upgrade quantities corresponding to the plurality of first passive content pools gradually decrease as the level of the first passive content pool increases.

[0139] Taking the above example, if the first set number is M, then the preset upgrade number corresponding to the first passive content pool C is M / 2, the preset upgrade number corresponding to the first passive content pool B is M / 4, and the preset upgrade number corresponding to the first passive content pool A is M / 8. It can be seen that each time the level of the first passive content pool increases by one level, the value of the preset upgrade number is reduced by half, that is, it gradually decreases as the level of the first passive content pool increases. It should be noted that the reduction rate of the preset upgrade number is not fixed to half. Those skilled in the art can select the reduction rate of the preset upgrade number based on actual circumstances, and this disclosure does not limit it.

[0140] In an optional embodiment, the numerical values ​​of the preset downgrade numbers corresponding to the plurality of first passive content pools gradually increase as the level of the first passive content pool decreases.

[0141] Taking the above example, if the first set number is M, then the preset downgrade number corresponding to the first passive content pool S is M / 8, the preset downgrade number corresponding to the first passive content pool A is M / 4, the preset downgrade number corresponding to the first passive content pool B is M / 2, and the preset downgrade number corresponding to the first passive content pool C is M. It can be seen that each time the level of the first passive content pool decreases by one level, the value of the preset downgrade number will double, that is, it will gradually increase as the level of the first passive content pool decreases. It should be noted that the increase in the preset downgrade number is not fixed to one, and those skilled in the art can select the increase in the preset downgrade number according to actual circumstances, and this disclosure does not limit it here.

[0142] It should be noted that the preset upgrade number and the preset downgrade number of the same first passive content pool are equal, and the number of recent original content in the first passive content pool is stable. When the number of recent original content released by users is small and the number of exposures is large, the pyramid content pool will not be consumed to emptiness by the exposure operation, and a stable recommendation function can be maintained.

[0143] FIG3 shows a flowchart of calculating a likeability score provided by an embodiment of the present disclosure.

[0144] Determine the popularity score of recent original content through steps S201, S202, S203, and S204:

[0145] Step S201: Determine a static quality score based on the static data of recent original content.

[0146] For each piece of recent original content created by users, a standard needs to be set to evaluate its popularity. The more popular recent original content is, the more likely it is to be promoted to a higher-level content pool to gain greater exposure. Based on this, this paper adopts a three-in-one evaluation method that combines static quality scoring, dynamic quality scoring, and a time decay coefficient. This method comprehensively considers the creator, reviewer, and time, ultimately resulting in a comprehensive method for calculating popularity scores.

[0147] The static quality score mainly considers the static data carried by the recent original content when it is first released. It is a value evaluated from the perspective of the effort put into the creation of the recent original content. The more effort is put into the creation of the recent original content, the more likely the recent original content will have a higher quality. Specifically, the following 10 categories of static features are considered, as shown in Table 1 below. These static features all represent the cost of creating recent original content. Taking into account the different importance of these static features, the present disclosure uses machine learning methods to learn the relationship between these static features and the future dynamic quality of recent original content from a large amount of data on recently released original content, and uses deep neural networks to fit these static features together, and finally obtains the static quality score corresponding to the recent original content.

[0148] Table 1: Static characteristics data table.

[0149] As shown in Table 1, static features include creation time, component usage, capacity, description length, total number of components, number of component types, whether the title is original, whether music is included, genre, and the number of components of each type. Genre refers to the type of recent original content. For example, if recent original content is a game level, the genre could be racing, puzzle, or other types. Components refer to the components used to create recent original content.

[0150] A deep neural network model is used to evaluate ten static features of recent original content to determine the static quality score of the recent original content.

[0151] Step S202: Determine a dynamic quality score based on user feedback on recent original content.

[0152] Dynamic quality scores are used to represent the quality of user feedback on recent original content after a certain number of exposures. User feedback includes, but is not limited to, play, like, favorite, and share. Play refers to whether a user plays the recent original content after viewing it; like refers to whether a user likes the content after playing it; favorite refers to whether a user adds the content to their favorites after playing it; and share refers to whether a user shares the content with friends or other users after playing it.

[0153] Here, the dynamic quality score of a recent original content is determined based on the total number of plays, likes, favorites, and shares of the content by different users. Because different content pools have different exposures, these dynamic features need to be normalized. The ratio of total plays to total exposures is used as the play rate, the ratio of total likes to total plays is used as the like rate, the ratio of total favorites to total plays is used as the favorite rate, and the ratio of total shares to total plays is used as the share rate. The play rate, like rate, favorite rate, and share rate are the four dynamic quality evaluation indicators for measuring dynamic quality.

[0154] The importance of these four dynamic quality evaluation indicators is different. The importance, from high to low, is sharing rate, like rate, collection rate, and play rate. Therefore, a corresponding weight is set for each indicator according to its importance, and the sum of the weights of the four dynamic quality evaluation indicators is used as the dynamic quality score.

[0155] For example, the dynamic quality score = 1×play rate + 18×like rate + 10×collection rate + 20×share rate.

[0156] Step S203: Determine a time decay coefficient based on the recent publishing duration of the original content.

[0157] Time decay refers to the fact that after a recent original content is released, its popularity score will gradually decrease over time, until it reaches zero. This is to prevent classic recent original content from dominating the premium content pool for a long time, causing user aesthetic fatigue. The time decay coefficient can be calculated using the following formula: Time decay coefficient = max(1 - number of days the recent original content was released × 0.02, 0), where 0.02 is the coefficient in the formula. Those skilled in the art can select the specific value of this coefficient based on actual circumstances.

[0158] Step S204 : determining the likeability score of the recent original content based on the static quality score, the dynamic quality score, and the time decay coefficient.

[0159] Here, the popularity score of recent original content can be determined based on the static quality score, dynamic quality score and time decay coefficient. For example, the product of the static quality score, dynamic quality score and time decay coefficient can be used as the popularity score of recent original content, or the product of the dynamic quality score and the time decay coefficient can be used as the popularity score of recent original content. Alternatively, the sum of the static quality score and the dynamic quality score can be calculated first, and then the sum of the two and the product of the time decay coefficient can be used as the popularity score of recent original content.

[0160] In an optional embodiment, determining the likeability score of recent original content based on the static quality score, the dynamic quality score, and the time decay coefficient includes steps e1 to e3.

[0161] Step e1: Determine the type and level of the content pool where the recent original content currently resides.

[0162] Specifically, different likeability score calculation formulas are used according to the characteristics of different usage scenarios.

[0163] In the low-level content pool, since the preset recommendation ratio corresponding to recent original content is low, it will result in fewer exposures and the dynamic quality score has a lower confidence level. Therefore, it is necessary to introduce static quality scores to balance it, so that recent original content that creators have spent more effort on has a greater probability of entering the high-level content pool.

[0164] In the premium content pool, recent original content has a high preset recommendation ratio, which results in more exposures and sufficient user feedback data. This makes dynamic quality ratings more reliable, and there is no need to consider static quality ratings.

[0165] To sum up, multiple first content pools are divided into low-level content pools and high-level content pools. Taking the above example, the first active content pool D and the first passive content pool C can be divided into low-level content pools, and the first passive content pool B, the first passive content pool A and the first passive content pool S can be divided into high-level content pools.

[0166] Step e2: If the current content pool is the first active content pool or the low-level first passive content pool, select the first likeability calculation formula, substitute the static quality score, dynamic quality score and time decay coefficient of the recent original content into the first likeability calculation formula to determine the likeability score of the recent original content.

[0167] For the first active content pool D and the first passive content pool C, the product of the static quality score, the dynamic quality score and the time decay coefficient of the recent original content is used as the likeability score of the recent original content.

[0168] Step e3: If the current content pool is a high-level first passive content pool, select the second likeability calculation formula, substitute the dynamic quality score and time decay coefficient of the recent original content into the second likeability calculation formula, and determine the likeability score of the recent original content.

[0169] For the first passive content pool B, the first passive content pool A, and the first passive content pool S, the product of the dynamic quality score and the time decay coefficient of the recent original content is used as the likeability score of the recent original content.

[0170] In an optional embodiment, the threshold value of the number of content pools of a first passive content pool with a higher level among the multiple first passive content pools is lower.

[0171] In one example, the content pool quantity threshold for the first active content pool D is 1600, the content pool quantity threshold for the first passive content pool C is 800, the content pool quantity threshold for the first passive content pool B is 400, the content pool quantity threshold for the first passive content pool A is 200, and the content pool quantity threshold for the first passive content pool S is 100. As the level of the content pool increases, the content pool quantity threshold gradually decreases, and the multiple first content pools form a pyramid shape as the level increases.

[0172] In an optional embodiment, the method also includes: selecting historical popular content in different user preference ranges from each second content pool according to a second preset recommendation ratio and recommending it to the user; determining low-quality historical popular content based on user feedback information on the target historical popular content, and removing the low-quality historical popular content from multiple second content pools; obtaining new historical popular content to supplement at least one target second content pool among the multiple second content pools, and recommending the new historical popular content to the user according to the recommendation rules.

[0173] According to statistics, 40% of users play original content through the recommendation system, which means that another 60% of the maps are spread through other systems in the game, including but not limited to the search system, sharing system, room system and task system.

[0174] To further leverage feedback from all systems globally, this publication also utilizes a leaderboard system for global feedback collection. Feedback from every game session is recorded to form a unified server-wide popularity ranking. Unlike the pyramid content pool recommendation method, the ranking system's feedback data isn't limited to the recommendation system itself; it encompasses all feedback data from the entire server.

[0175] While adopting a progressive graph pool of recent original content, the present disclosure also creates an additional progressive graph pool of historical popular content, namely, a second content pool. The main difference between the second content pool and the first content pool is that the server-wide popularity ranking list is used as the source of scanning data, that is, the historical popular content in the server-wide popularity ranking list is placed in the second resource library, and the historical popular content in the second resource library is placed in the second active content pool in the second content pool. The historical popular content in the second content pool is updated according to the same progressive method as the first content pool to remove low-quality historical popular content from the second content pool.

[0176] It should be noted that when obtaining historical popular content from the second resource library, randomness is added to the selection process. First, a preset number of historical popular content with a high ranking is selected from the second resource library as a candidate historical popular content set. Then, the historical popular content in the candidate historical popular content set is randomly shuffled, and the required number of historical popular content is selected from the beginning to the end of the randomly shuffled candidate historical popular content set and added to multiple second content pools. At a set time point every day, the candidate historical popular content set is reset, returned to the head position, and historical popular content is reselected and added to multiple second content pools, and the same historical popular content will only be scanned into the second content pool once on the same day.

[0177] Compared to the first content pool, the initial original content in the second content pool is of higher quality, which can offset the impact of low-quality recommendations caused by recent original content during its initial exposure. It also allows users to easily review classic original content, ensuring that original content that truly meets the needs of all users on the server can appear in users' recommendation lists for a long time.

[0178] Since the update process for historical popular content in multiple second content pools is the same as the update process for recent original content in multiple first content pools, it will not be further described here. It should be noted that multiple first content pools and multiple second content pools can exist simultaneously. When a user opens the recommendation page, a first preset number of recent original content will be selected from the multiple first content pools, and a second preset number of historical popular content will be selected from the multiple second content pools. The first preset number of recent original content and the second preset number of historical popular content will then be mixed and recommended to the user.

[0179] Compared with the original content recommendation method in the game in the prior art, the present invention can directly obtain new recent original content and recommend the new recent original content to users, so that newly registered users can obtain new recent original content, and can also display new recently released content, avoiding cold start. At the same time, it can select recent original content in different user preference ranges from multiple content pools and recommend it to users, increasing the breadth of recommendation and solving the problems of cold start, over-recommendation and narrow recommendation in the existing recommendation methods.

[0180] Based on the same inventive concept, the embodiments of the present disclosure also provide an original content recommendation device in the game corresponding to the original content recommendation method in the game. Since the principle of solving the problem by the device in the embodiments of the present disclosure is similar to the original content recommendation method in the game mentioned above in the embodiments 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.

[0181] Please refer to FIG4 , which is a schematic diagram of the structure of an original content recommendation device in a game provided by an embodiment of the present disclosure. As shown in FIG4 , the original content recommendation device 300 in a game includes:

[0182] The first recommendation module 301 is configured to select recent original content from each first content pool of different levels and recommend it to the user according to a first preset recommendation ratio;

[0183] A content updating module 302 is configured to determine recent low-quality content based on user feedback on target recent original content, and remove the recent low-quality content from the plurality of first content pools;

[0184] The second recommendation module 303 is used to obtain new recent original content to supplement at least one target first content pool among the multiple first content pools, and recommend the new recent original content to the user according to the recommendation rules. The new recent original content is recent original content that has not been placed in the multiple first content pools.

[0185] In a feasible implementation scheme, the method also includes: selecting historical popular content from each second content pool of different levels and recommending it to the user according to a second preset recommendation ratio; determining low-quality historical popular content based on user feedback information on the target historical popular content, and removing the low-quality historical popular content from multiple second content pools; obtaining new historical popular content to supplement at least one target second content pool among the multiple second content pools, and recommending the new historical popular content to the user according to the recommendation rules.

[0186] In a feasible implementation scheme, at least one first target content pool includes a first active content pool, and obtaining new recent original content to supplement at least one target first content pool among multiple first content pools includes: when the amount of recent original content in the first active content pool is less than the content pool quantity threshold corresponding to the first active content pool, obtaining new recent original content and adding it to the first active content pool so that the amount of recent original content in the first active content pool is equal to the content pool quantity threshold corresponding to the first active content pool.

[0187] In a feasible implementation scheme, after acquiring new recent original content and adding it to the first active content pool, it includes: setting a basic flow for the new recent original content, and setting the new recent original content to an exposure state; when the new recent original content is browsed by a user, reducing the basic flow of the new recent original content in the exposure state by one; when the basic flow of the new recent original content is reduced to 0, changing the state of the new recent original content from the exposure state to the waiting state, and the recent original content in the waiting state will not be recommended to the user; calculating the cumulative time that the new recent original content enters the waiting state, and when the cumulative time reaches the waiting time threshold, changing the new recent original content from the waiting state to the ranking state.

[0188] In a feasible implementation scheme, the multiple first content pools include multiple first passive content pools, and recent low-quality content is determined based on user feedback information on target recent original content, and the recent low-quality content is removed from the multiple first content pools, including: determining the number of recent original content in a ranked state in the first active content pool; when the number of recent original content in a ranked state reaches a first score quantity threshold, determining the likeability score of the recent original content in a ranked state; taking a first set number of recent original content with high likeability rankings in the first active content pool as recent high-quality content, and transferring the recent high-quality content to the multiple first passive content pools; using the recent high-quality content to remove the recent low-quality content from the multiple first content pools.

[0189] In a feasible implementation scheme, multiple first passive content pools have corresponding levels from high to low according to the recommendation degree of recent original content stored in each of them, and recent low-quality content is removed from the multiple first content pools by using recent high-quality content, including: adding a first set number of recent high-quality content to the first passive content pool of the lowest level; starting from the first passive content pool of the lowest level to which the first set number of recent high-quality content is added, progressively upgrading the recent high-quality content in each first passive content pool to a first passive content pool of a higher level, until the recent high-quality content in the first passive content pool of the second highest level is upgraded to the first passive content pool of the highest level; starting from the first passive content pool of the highest level to which the recent high-quality content is added, progressively downgrading the recent low-quality content in each first passive content pool to a first passive content pool of a lower level, until the first set number of recent low-quality content is removed from the first passive content pool of the lowest level.

[0190] In one feasible implementation, starting with a first passive content pool of the lowest level to which a first set number of recent high-quality contents are added, the recent high-quality contents in each first passive content pool are progressively upgraded to a first passive content pool of a higher level, including: selecting the first passive content pool of the lowest level as the first upgraded passive content pool; determining whether the first upgraded passive content pool is the first passive content pool of the highest level; if the first upgraded passive content pool is not the first passive content pool of the highest level, selecting recent high-quality contents from the recent original contents in the first upgraded passive content pool, and transferring the recent high-quality contents to a first passive content pool of a higher level than the first upgraded passive content pool; using the first passive content pool of the higher level as the new first upgraded passive content pool, and returning to the step of selecting recent high-quality contents from the recent original contents in the first upgraded passive content pool.

[0191] In one feasible implementation, starting with the highest-level first passive content pool to which recent high-quality content is added, the recent low-quality content in each first passive content pool is progressively downgraded to a first passive content pool of a lower level, including: selecting the highest-level first passive content pool as the first downgraded passive content pool; determining whether the first downgraded passive content pool is the lowest-level first passive content pool; if the first downgraded passive content pool is not the lowest-level first passive content pool, selecting recent low-quality content from the recent original content in the first downgraded passive content pool, and transferring the recent low-quality content to a first passive content pool of a lower level than the first downgraded passive content pool; using the lower-level first passive content pool as the new first downgraded passive content pool, and returning to the step of selecting recent low-quality content from the recent original content in the first downgraded passive content pool.

[0192] In a feasible implementation scheme, recent high-quality content is selected from the recent original content in the first upgraded passive content pool, including: when the evaluation conditions are met, sorting the recent original content in the chronological order of their entry into the first upgraded passive content pool, and selecting a fixed proportion of the top-ranked recent original content as the recent candidate high-quality content; determining the current likeability ratings of the recent candidate high-quality content; sorting the recent candidate high-quality content in descending order according to the current likeability ratings, and selecting a preset upgraded number of the top-ranked recent candidate high-quality content corresponding to the first upgraded passive content pool as the final recent high-quality content.

[0193] In a feasible implementation scheme, recent low-quality content is selected from the recent original content in the first degraded passive content pool, including: when the evaluation conditions are met, sorting the recent original content in the chronological order of their entry into the first degraded passive content pool, and selecting a fixed proportion of the top-ranked recent original content as the recent candidate low-quality content; determining the current likeability ratings of the recent candidate low-quality content; sorting the recent candidate low-quality content in descending order according to the current likeability ratings, and selecting a preset number of the recent candidate low-quality content from a first set number of the bottom-ranked recent candidate low-quality content corresponding to the first degraded passive content pool as the final recent low-quality content.

[0194] In a feasible implementation manner, the numerical values ​​of the preset upgrade quantities corresponding to the plurality of first passive content pools gradually decrease as the level of the first passive content pool increases.

[0195] In a feasible implementation manner, the numerical values ​​of the preset downgrade numbers corresponding to the plurality of first passive content pools gradually increase as the level of the first passive content pool decreases.

[0196] In a feasible implementation scheme, the likeability score of recent original content is determined by the following processing: determining a static quality score based on the static data of the recent original content; determining a dynamic quality score based on user feedback information on the recent original content; determining a time decay coefficient based on the published length of the recent original content; and determining a likeability score of the recent original content based on the static quality score, the dynamic quality score, and the time decay coefficient.

[0197] In a feasible implementation scheme, the popularity score of recent original content is determined based on the static quality score, dynamic quality score and time decay coefficient, including: determining the type and level of the content pool in which the recent original content is currently located; if the current content pool is the first active content pool or the low-level first passive content pool, then selecting the first popularity calculation formula, substituting the static quality score, dynamic quality score and time decay coefficient of the recent original content into the first popularity calculation formula, and determining the popularity score of the recent original content; if the current content pool is the high-level first passive content pool, then selecting the second popularity calculation formula, substituting the dynamic quality score and time decay coefficient of the recent original content into the second popularity calculation formula, and determining the popularity score of the recent original content.

[0198] In one feasible implementation manner, the threshold value of the number of content pools of a first passive content pool with a higher level among the plurality of first passive content pools is lower.

[0199] Please refer to FIG5 , which is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. As shown in FIG5 , the electronic device 400 includes a processor 410 , a memory 420 , and a bus 430 .

[0200] The memory 420 stores machine-readable instructions executable by the processor 410. When the electronic device executes a method for recommending original content in a game as described in the embodiment, the processor 410 communicates with the memory 420 via the bus 430. The processor 410 executes the machine-readable instructions, including the preamble of the method item, to perform the following steps:

[0201] Selecting recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommending it to the user;

[0202] Determining recent low-quality content based on user feedback on target recent original content, and removing the recent low-quality content from the multiple first content pools;

[0203] New recent original content is obtained and added to at least one target first content pool among the multiple first content pools, and the new recent original content is recommended to users according to the recommendation rules. The new recent original content is recent original content that has not been placed in the multiple first content pools.

[0204] In a feasible implementation scheme, the processor 410 is also used to: select historical popular content from each second content pool of different levels and recommend it to the user according to a second preset recommendation ratio; determine low-quality historical popular content based on user feedback information on the target historical popular content, and remove the low-quality historical popular content from multiple second content pools; obtain new historical popular content and add it to at least one target second content pool among the multiple second content pools, and recommend the new historical popular content to the user according to the recommendation rules.

[0205] In a feasible embodiment, at least one first target content pool includes a first active content pool. When the processor 410 executes the acquisition of new recent original content to supplement at least one target first content pool among multiple first content pools, it is specifically used to: when the number of recent original content in the first active content pool is less than the content pool quantity threshold corresponding to the first active content pool, acquire new recent original content and add it to the first active content pool, so that the number of recent original content in the first active content pool is equal to the content pool quantity threshold corresponding to the first active content pool.

[0206] In a feasible implementation scheme, after executing the acquisition of new recent original content and adding it to the first active content pool, the processor 410 is specifically used to: set a basic flow for the new recent original content, and set the new recent original content to an exposure state; when the new recent original content is browsed by a user, the basic flow of the new recent original content in the exposure state is reduced by one; when the basic flow of the new recent original content is reduced to 0, the state of the new recent original content is changed from an exposure state to a waiting state, and the recent original content in the waiting state will not be recommended to the user; calculate the cumulative time that the new recent original content enters the waiting state, and when the cumulative time reaches the waiting time threshold, the new recent original content is changed from a waiting state to a ranking state.

[0207] In a feasible implementation scheme, the multiple first content pools include multiple first passive content pools. When the processor 410 determines the recent low-quality content based on the user's feedback information on the target recent original content and removes the recent low-quality content from the multiple first content pools, it is specifically used to: determine the number of recent original content in a ranked state in the first active content pool; when the number of recent original content in a ranked state reaches a first score quantity threshold, determine the likeability score of the recent original content in a ranked state; use a first set number of recent original content with a high likeability score in the first active content pool as recent high-quality content, and transfer the recent high-quality content to the multiple first passive content pools; use the recent high-quality content to remove the recent low-quality content from the multiple first content pools.

[0208] In a feasible embodiment, multiple first passive content pools have corresponding levels from high to low according to the recommendation degree of the recent original content stored in each of them. When executing the process of removing recent low-quality content from the multiple first content pools by utilizing recent high-quality content, the processor 410 is specifically configured to: add a first set number of recent high-quality content to the first passive content pool of the lowest level; starting from the first passive content pool of the lowest level to which the first set number of recent high-quality content is added, progressively upgrade the recent high-quality content in each first passive content pool to a first passive content pool of a higher level, until the recent high-quality content in the first passive content pool of the second highest level is upgraded to the first passive content pool of the highest level; starting from the first passive content pool of the highest level to which the recent high-quality content is added, progressively downgrade the recent low-quality content in each first passive content pool to a first passive content pool of a lower level, until the first set number of recent low-quality content is removed from the first passive content pool of the lowest level.

[0209] In a feasible embodiment, when executing the process of progressively upgrading the recent high-quality content in each first passive content pool to a first passive content pool of a higher level starting from the lowest-level first passive content pool to which a first set number of recent high-quality content is added, the processor 410 is specifically configured to: select the lowest-level first passive content pool as the first upgraded passive content pool; determine whether the first upgraded passive content pool is the highest-level first passive content pool; if the first upgraded passive content pool is not the highest-level first passive content pool, select recent high-quality content from the recent original content in the first upgraded passive content pool and transfer the recent high-quality content to a first passive content pool of a higher level than the first upgraded passive content pool; use the higher-level first passive content pool as the new first upgraded passive content pool and return to the step of selecting recent high-quality content from the recent original content in the first upgraded passive content pool.

[0210] In a feasible embodiment, when executing the process of progressively downgrading the recent low-quality content in each first passive content pool to a first passive content pool of a lower level starting from the first passive content pool of the highest level in which recent high-quality content is added, the processor 410 is specifically configured to: select the first passive content pool of the highest level as the first downgraded passive content pool; determine whether the first downgraded passive content pool is the first passive content pool of the lowest level; if the first downgraded passive content pool is not the first passive content pool of the lowest level, select recent low-quality content from the recent original content in the first downgraded passive content pool, and transfer the recent low-quality content to the first passive content pool of the lower level of the first downgraded passive content pool; use the first passive content pool of the lower level as the new first downgraded passive content pool, and return to the step of selecting recent low-quality content from the recent original content in the first downgraded passive content pool.

[0211] In a feasible implementation scheme, when the processor 410 selects recent high-quality content from the recent original content in the first upgraded passive content pool, it is specifically used to: when the evaluation conditions are met, sort the recent original content in the chronological order of their entry into the first upgraded passive content pool, and select a fixed proportion of the top-ranked recent original content as the recent candidate high-quality content; determine the current likeability rating of the recent candidate high-quality content; sort the recent candidate high-quality content in descending order according to the current likeability rating, and select a preset upgraded number of recent candidate high-quality content from the first set number of top-ranked content corresponding to the first upgraded passive content pool as the final recent high-quality content.

[0212] In a feasible implementation scheme, when the processor 410 selects recent low-quality content from the recent original content in the first degraded passive content pool, it is specifically used to: when the evaluation conditions are met, sort the recent original content in the chronological order of their entry into the first degraded passive content pool, and select a fixed proportion of the top-ranked recent original content as the recent candidate low-quality content; determine the current likeability score of the recent candidate low-quality content; sort the recent candidate low-quality content in descending order according to the current likeability score, and select a preset number of recent candidate low-quality content from the first set number of the bottom-ranked content corresponding to the first degraded passive content pool as the final recent low-quality content.

[0213] In a feasible implementation manner, the numerical values ​​of the preset upgrade quantities corresponding to the plurality of first passive content pools gradually decrease as the level of the first passive content pool increases.

[0214] In a feasible implementation manner, the numerical values ​​of the preset downgrade numbers corresponding to the plurality of first passive content pools gradually increase as the level of the first passive content pool decreases.

[0215] In a feasible embodiment, the processor 410 determines the likeability score of recent original content through the following processing: determining a static quality score based on the static data of the recent original content; determining a dynamic quality score based on user feedback information on the recent original content; determining a time decay coefficient based on the published length of the recent original content; and determining a likeability score of the recent original content based on the static quality score, the dynamic quality score and the time decay coefficient.

[0216] In a feasible implementation scheme, when the processor 410 determines the likeability score of recent original content based on the static quality score, the dynamic quality score and the time decay coefficient, it is specifically used to: determine the type and level of the content pool in which the recent original content is currently located; if the current content pool is the first active content pool or the low-level first passive content pool, then select the first likeability calculation formula, substitute the static quality score, the dynamic quality score and the time decay coefficient of the recent original content into the first likeability calculation formula, and determine the likeability score of the recent original content; if the current content pool is the high-level first passive content pool, then select the second likeability calculation formula, substitute the dynamic quality score and the time decay coefficient of the recent original content into the second likeability calculation formula, and determine the likeability score of the recent original content.

[0217] In one feasible implementation manner, the threshold value of the number of content pools of a first passive content pool with a higher level among the plurality of first passive content pools is lower.

[0218] Through the above method, new recent original content can be directly obtained and recommended to users, so that newly registered users can obtain new recent original content, and new recently released content can be displayed, avoiding cold start. At the same time, recent original content in different user preference ranges can be selected from multiple content pools and recommended to users, increasing the breadth of recommendations and solving the problems of cold start, over-recommendation and narrow recommendation in existing recommendation methods.

[0219] The present disclosure also provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium. The computer program is executed by a processor when the processor is running, and the processor performs the following steps:

[0220] Selecting recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommending it to the user;

[0221] Determining recent low-quality content based on user feedback on target recent original content, and removing the recent low-quality content from the multiple first content pools;

[0222] New recent original content is obtained and added to at least one target first content pool among the multiple first content pools, and the new recent original content is recommended to users according to the recommendation rules. The new recent original content is recent original content that has not been placed in the multiple first content pools.

[0223] In a feasible implementation scheme, the processor is also used to: select historical popular content from each second content pool of different levels and recommend it to the user according to a second preset recommendation ratio; determine low-quality historical popular content based on user feedback information on the target historical popular content, and remove the low-quality historical popular content from multiple second content pools; obtain new historical popular content and supplement it to at least one target second content pool among the multiple second content pools, and recommend the new historical popular content to the user according to the recommendation rules.

[0224] In a feasible embodiment, at least one first target content pool includes a first active content pool. When the processor executes the function of acquiring new recent original content to supplement at least one target first content pool among multiple first content pools, it is specifically used to: when the number of recent original content in the first active content pool is less than the content pool quantity threshold corresponding to the first active content pool, acquire new recent original content and add it to the first active content pool so that the number of recent original content in the first active content pool is equal to the content pool quantity threshold corresponding to the first active content pool.

[0225] In a feasible implementation scheme, after executing the acquisition of new recent original content and adding it to the first active content pool, the processor is specifically used to: set a basic flow for the new recent original content, and set the new recent original content to an exposure state; when the new recent original content is browsed by a user, reduce the basic flow of the new recent original content in the exposure state by one; when the basic flow of the new recent original content is reduced to 0, change the state of the new recent original content from the exposure state to the waiting state, and the recent original content in the waiting state will not be recommended to the user; calculate the cumulative time that the new recent original content enters the waiting state, and when the cumulative time reaches the waiting time threshold, change the new recent original content from the waiting state to the ranking state.

[0226] In a feasible embodiment, the multiple first content pools include multiple first passive content pools. When the processor determines recent low-quality content based on user feedback information on target recent original content and removes the recent low-quality content from the multiple first content pools, it is specifically used to: determine the number of recent original content in a ranked state in the first active content pool; when the number of recent original content in a ranked state reaches a first score quantity threshold, determine the likeability score of the recent original content in the ranked state; use a first set number of recent original content with high likeability rankings in the first active content pool as recent high-quality content, and transfer the recent high-quality content to the multiple first passive content pools; use the recent high-quality content to remove the recent low-quality content from the multiple first content pools.

[0227] In a feasible embodiment, multiple first passive content pools have corresponding levels from high to low according to the recommendation degree of recent original content stored in each of them. When executing the process of removing recent low-quality content from the multiple first content pools by utilizing recent high-quality content, the processor is specifically configured to: add a first set number of recent high-quality content to the first passive content pool of the lowest level; starting from the first passive content pool of the lowest level to which the first set number of recent high-quality content is added, progressively upgrade the recent high-quality content in each first passive content pool to a first passive content pool of a higher level, until the recent high-quality content in the first passive content pool of the second highest level is upgraded to the first passive content pool of the highest level; starting from the first passive content pool of the highest level to which the recent high-quality content is added, progressively downgrade the recent low-quality content in each first passive content pool to a first passive content pool of a lower level, until the first set number of recent low-quality content is removed from the first passive content pool of the lowest level.

[0228] In one feasible embodiment, when executing the process of progressively upgrading the recent high-quality content in each first passive content pool to a first passive content pool of a higher level, starting from the lowest-level first passive content pool to which a first set number of recent high-quality content is added, the processor is specifically configured to: select the lowest-level first passive content pool as the first upgraded passive content pool; determine whether the first upgraded passive content pool is the highest-level first passive content pool; if the first upgraded passive content pool is not the highest-level first passive content pool, select recent high-quality content from the recent original content in the first upgraded passive content pool and transfer the recent high-quality content to a first passive content pool of a higher level than the first upgraded passive content pool; use the higher-level first passive content pool as the new first upgraded passive content pool, and return to the step of selecting recent high-quality content from the recent original content in the first upgraded passive content pool.

[0229] In one feasible embodiment, when executing the process of progressively downgrading recent low-quality content in each first passive content pool to a first passive content pool of a lower level, starting from the highest-level first passive content pool in which recent high-quality content is added, the processor is specifically configured to: select the highest-level first passive content pool as the first downgraded passive content pool; determine whether the first downgraded passive content pool is the lowest-level first passive content pool; if the first downgraded passive content pool is not the lowest-level first passive content pool, select recent low-quality content from the recent original content in the first downgraded passive content pool, and transfer the recent low-quality content to a first passive content pool of a lower level than the first downgraded passive content pool; use the lower-level first passive content pool as the new first downgraded passive content pool, and return to the step of selecting recent low-quality content from the recent original content in the first downgraded passive content pool.

[0230] In a feasible implementation scheme, when the processor selects recent high-quality content from the recent original content in the first upgraded passive content pool, it is specifically used to: when the evaluation conditions are met, sort the recent original content in the chronological order of their entry into the first upgraded passive content pool, and select a fixed proportion of the top-ranked recent original content as the recent candidate high-quality content; determine the current likeability score of the recent candidate high-quality content; sort the recent candidate high-quality content in descending order according to the current likeability score, and select a preset upgraded number of recent candidate high-quality content from the first set number of top-ranked content corresponding to the first upgraded passive content pool as the final recent high-quality content.

[0231] In a feasible implementation scheme, when the processor executes the selection of recent low-quality content from the recent original content in the first degraded passive content pool, it is specifically used to: when the evaluation conditions are met, sort the recent original content in the chronological order of their entry into the first degraded passive content pool, and select a fixed proportion of the top-ranked recent original content as the recent candidate low-quality content; determine the current likeability score of the recent candidate low-quality content; sort the recent candidate low-quality content in descending order according to the current likeability score, and select a preset downgraded number of recent candidate low-quality content from the first set number of the bottom-ranked content corresponding to the first degraded passive content pool as the final recent low-quality content.

[0232] In a feasible implementation manner, the numerical values ​​of the preset upgrade quantities corresponding to the plurality of first passive content pools gradually decrease as the level of the first passive content pool increases.

[0233] In a feasible implementation manner, the numerical values ​​of the preset downgrade numbers corresponding to the plurality of first passive content pools gradually increase as the level of the first passive content pool decreases.

[0234] In a feasible embodiment, the processor determines the likeability score of recent original content by the following processing: determining a static quality score based on static data of the recent original content; determining a dynamic quality score based on user feedback information on the recent original content; determining a time decay coefficient based on the published length of the recent original content; and determining a likeability score of the recent original content based on the static quality score, the dynamic quality score, and the time decay coefficient.

[0235] In a feasible implementation scheme, when the processor determines the likeability score of recent original content based on the static quality score, dynamic quality score and time decay coefficient, it is specifically used to: determine the type and level of the content pool in which the recent original content is currently located; if the current content pool is the first active content pool or the low-level first passive content pool, then select the first likeability calculation formula, substitute the static quality score, dynamic quality score and time decay coefficient of the recent original content into the first likeability calculation formula, and determine the likeability score of the recent original content; if the current content pool is the high-level first passive content pool, then select the second likeability calculation formula, substitute the dynamic quality score and time decay coefficient of the recent original content into the second likeability calculation formula, and determine the likeability score of the recent original content.

[0236] In one feasible implementation manner, the threshold value of the number of content pools of a first passive content pool with a higher level among the plurality of first passive content pools is lower.

[0237] Through the above method, new recent original content can be directly obtained and recommended to users, so that newly registered users can obtain new recent original content, and new recently released content can be displayed, avoiding cold start. At the same time, recent original content in different user preference ranges can be selected from multiple content pools and recommended to users, increasing the breadth of recommendations and solving the problems of cold start, over-recommendation and narrow recommendation in existing recommendation methods.

[0238] Those skilled in the art will 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.

[0239] In the several embodiments provided in the present disclosure, 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 merely a logical function division. There may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0240] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0241] In addition, each functional unit in each embodiment of the present disclosure 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.

[0242] 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 disclosure, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present disclosure. 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.

[0243] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present disclosure, which are used to illustrate the technical solutions of the present disclosure, rather than to limit them. The scope of protection of the present disclosure is not limited thereto. Although the present disclosure has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-mentioned embodiments within the technical scope disclosed in the present disclosure, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in the scope of protection of the present disclosure. Therefore, the scope of protection of the present disclosure should be based on the scope of protection of the claims.

Claims

1. A method for recommending original content in a game, comprising: Selecting recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommending it to the user; Determine recent low-quality content according to user feedback information on target recent original content, and remove the recent low-quality content from the plurality of first content pools; Acquire new recent original content to supplement at least one target first content pool among the multiple first content pools, and recommend the new recent original content to the user according to the recommendation rule, wherein the new recent original content is recent original content that has not been placed in the multiple first content pools.

2. The method according to claim 1, wherein: The method further comprises: Selecting historical popular content from each second content pool of different levels according to a second preset recommendation ratio and recommending it to the user; Determine low-quality historical hot content according to user feedback information on target historical hot content, and remove the low-quality historical hot content from the plurality of second content pools; New historical hot content is acquired and added to at least one target second content pool among the plurality of second content pools, and the new historical hot content is recommended to the user according to the recommendation rule.

3. The method according to claim 1, wherein: The at least one first target content pool includes a first active content pool, and the acquiring of new recent original content to supplement at least one target first content pool among the plurality of first content pools includes: When the number of recent original contents in the first active content pool is less than the content pool quantity threshold corresponding to the first active content pool, new recent original contents are acquired and added to the first active content pool so that the number of recent original contents in the first active content pool is equal to the content pool quantity threshold corresponding to the first active content pool.

4. The method according to claim 3, wherein: After acquiring new recent original content and adding it to the first active content pool, the method includes: Setting a basic flow rate for the new recent original content, and setting the new recent original content to an exposure state; When the new recent original content is browsed by the user, the basic flow of the new recent original content in the exposure state is reduced by one; When the basic flow of the new recent original content decreases to 0, the state of the new recent original content is changed from the exposure state to the waiting state, and the recent original content in the waiting state will not be recommended to the user; The accumulated time length of the waiting state of the new recent original content is calculated, and when the accumulated time length reaches a waiting time length threshold, the new recent original content is changed from the waiting state to the ranking state.

5. The method according to claim 1, wherein: The multiple first content pools include multiple first passive content pools, and determining recent low-quality content according to user feedback information on target recent original content, and removing the recent low-quality content from the multiple first content pools includes: Determine the number of recent original content in a ranked state in the first active content pool; When the number of recent original contents in a ranking state reaches a first rating number threshold, determining a popularity rating of the recent original contents in a ranking state; taking a first set number of recent original contents with high popularity scores in the first active content pool as recent high-quality contents, and transferring the recent high-quality contents to the plurality of first passive content pools; The recent high-quality content is used to remove recent low-quality content from multiple first content pools.

6. The method according to claim 5, wherein: The multiple first passive content pools have corresponding levels from high to low according to the recommendation degree of the recent original content stored in each of them, and the recent low-quality content is removed from the multiple first content pools by using the recent high-quality content, including: adding a first set number of recent high-quality content to a first passive content pool of the lowest level; Starting from the lowest level first passive content pool to which a first set number of recent high-quality contents are added, the recent high-quality contents in each first passive content pool are progressively upgraded to a first passive content pool of a higher level, until the recent high-quality contents in the second highest level first passive content pool are upgraded to the highest level first passive content pool; Starting from the highest level first passive content pool to which recent high-quality content is added, recent low-quality content in each first passive content pool is progressively downgraded to a first passive content pool of a lower level until a first set number of recent low-quality content is removed from the lowest level first passive content pool.

7. The method according to claim 6, wherein: The method starts with adding a first set number of recent high-quality contents to a first passive content pool of the lowest level, and progressively upgrades the recent high-quality contents in each first passive content pool to a first passive content pool of a higher level, including: Selecting the lowest level first passive content pool as the first upgraded passive content pool; determining whether the first upgraded passive content pool is the highest-level first passive content pool; If the first upgraded passive content pool is not the highest level first passive content pool, the recent content in the first upgraded passive content pool is Selecting recent high-quality content from the original content, and transferring the recent high-quality content to a first passive content pool that is one level higher than the first upgraded passive content pool; The higher-level first passive content pool is used as a new first upgraded passive content pool, and the step of selecting recent high-quality content from the recent original content in the first upgraded passive content pool is returned to be executed.

8. The method according to claim 6, wherein: The method starts with adding the highest level first passive content pool with recent high-quality content, and progressively downgrading the recent low-quality content in each first passive content pool to a first passive content pool of a lower level, including: Selecting the highest-level first passive content pool as the first downgraded passive content pool; determining whether the first degraded passive content pool is a first passive content pool of the lowest level; If the first degraded passive content pool is not the first passive content pool of the lowest level, selecting recent low-quality content from recent original content in the first degraded passive content pool, and transferring the recent low-quality content to a first passive content pool of a lower level than the first degraded passive content pool; The first passive content pool at a lower level is used as a new first degraded passive content pool, and the step of selecting recent low-quality content from the recent original content in the first degraded passive content pool is returned to be executed.

9. The method according to claim 7, wherein: The selecting of recent high-quality content from the recent original content in the first upgraded passive content pool includes: When the evaluation conditions are met, recent original content is sorted in the order of time when it enters the first upgraded passive content pool, and a fixed proportion of recent original content with a high ranking is selected as recent candidate high-quality content; Determine the current likeability score of the recent candidate high-quality content; The recent candidate high-quality contents are sorted in descending order according to the current likeability score, and the recent candidate high-quality contents of the preset upgraded number corresponding to the first upgraded passive content pool in the first set number with the top ranking are selected as the final recent high-quality contents.

10. The method according to claim 8, wherein: The selecting of recent low-quality content from recent original content in the first downgraded passive content pool includes: When the evaluation conditions are met, recent original content is sorted in the order of time when it enters the first downgraded passive content pool, and a fixed proportion of recent original content with a high ranking is selected as recent candidate low-quality content; Determining a current likeability score of the recent candidate low-quality content; The recent candidate low-quality contents are sorted in descending order according to the current likeability score, and a preset number of recent candidate low-quality contents of the first set number with lower rankings and corresponding to the first degraded passive content pool are selected as the final recent low-quality contents.

11. The method according to claim 9, wherein: The numerical values ​​of the preset upgrade quantities corresponding to the plurality of first passive content pools gradually decrease as the levels of the first passive content pools increase.

12. The method according to claim 10, wherein: The numerical values ​​of the preset downgrade quantities corresponding to the plurality of first passive content pools gradually increase as the level of the first passive content pool decreases.

13. The method according to claim 1, wherein: The likeability score of recent original content is determined by the following process: Determining a static quality score based on the static data of the recent original content; Determining a dynamic quality score based on user feedback on the recent original content; Determine a time decay coefficient according to the published duration of the recent original content; A likeability score of the recent original content is determined according to the static quality score, the dynamic quality score, and the time decay coefficient.

14. The method according to claim 13, wherein: The determining the likeability score of the recent original content according to the static quality score, the dynamic quality score and the time decay coefficient includes: Determine the type and level of the content pool where the recent original content currently resides; If the current content pool is the first active content pool or the low-level first passive content pool, the first likeability calculation formula is selected, and the static quality score, the dynamic quality score and the time decay coefficient of the recent original content are substituted into the first likeability calculation formula to determine the likeability score of the recent original content; If the current content pool is a high-level first passive content pool, the second likeability calculation formula is selected, and the dynamic quality score and time decay coefficient of the recent original content are substituted into the second likeability calculation formula to determine the likeability score of the recent original content.

15. The method according to claim 6, wherein: A first passive content pool with a higher level among the plurality of first passive content pools has a lower content pool quantity threshold.

16. A device for recommending original content in a game, comprising: A first recommendation module is configured to select recent original content from each first content pool of different levels according to a first preset recommendation ratio and recommend it to the user; A content updating module, configured to determine recent low-quality content according to user feedback information on target recent original content, and remove the recent low-quality content from the plurality of first content pools; The second recommendation module is configured to acquire new recent original content to supplement at least one target first content pool among the multiple first content pools, and recommend the new recent original content to the user according to the recommendation rules, wherein the new recent original content is recent original content that has not been placed in the multiple first content pools.

17. 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 and the storage medium communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of the original content recommendation method in the game as described in any one of claims 1 to 15.

18. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, executes the steps of the original content recommendation method in a game as claimed in any one of claims 1 to 15.

Images