Method, device and storage medium for determining to-be-recommended resource

Abstract

The present disclosure provides a method and device for determining recommended resources, and a storage medium, relates to the technical field of data processing, in particular to the technical field of big data, information flow, intelligent recommendation, and can be used in the information flow recommendation scene. The specific implementation scheme is as follows: obtaining a candidate resource to be recommended; determining a first resource gear where the candidate resource is located; comparing the resource quality of the candidate resource with the quality requirement corresponding to the second resource gear to obtain a comparison result, wherein the quality requirement corresponding to the second resource gear is higher than the quality requirement of the first resource gear; adjusting the recommendation weight of the candidate resource according to the comparison result to obtain a target recommendation weight; and determining a target resource to be recommended according to the target recommendation weight. Thus, by comparing with the quality requirements of different gears, the recommended resources are inclined to high-quality resources, and the resource recommendation quality is improved.

Description

Technical Field

[0001] This disclosure relates to the fields of big data, information flow, and intelligent recommendation in the field of data processing technology, and can be used in information flow recommendation scenarios. In particular, it relates to a method, apparatus, device, and storage medium for determining resources to be recommended. Background Technology

[0002] In information feed recommendation scenarios, users can act as both consumers and producers of resources. For example, when users write articles, shoot videos, and share them online, they are producers of resources; when users read articles by other users online, they are consumers of resources.

[0003] Given the varying quality of resources in news feed recommendation scenarios, and to prevent the over-distribution of low-quality resources, which could negatively impact the healthy development of the news feed recommendation system ecosystem, resource quality scoring can be implemented. Related technologies typically adjust the recommendation weight by multiplying the resource by its quality score and corresponding weight coefficient, thereby adjusting the distribution volume of the resource.

[0004] However, the aforementioned technologies cannot further tilt the information flow recommendation system towards high-quality resources, and the actual recommended resources still suffer from low quality. Summary of the Invention

[0005] This disclosure provides a method, apparatus, device, and storage medium for determining resources to be recommended in order to improve the quality of recommended resources.

[0006] According to a first aspect of this disclosure, a method for determining resources to be recommended is provided, comprising:

[0007] Obtain candidate resources to be recommended;

[0008] Determine the first resource tier of the candidate resource;

[0009] The quality of the candidate resource is compared with the quality requirements corresponding to the second resource tier to obtain a comparison result, wherein the quality requirements corresponding to the second resource tier are higher than the quality requirements of the first resource tier.

[0010] Based on the comparison results, the recommendation weights of the candidate resources are adjusted to obtain the target recommendation weights;

[0011] Based on the target recommendation weight, the target resources to be recommended are determined.

[0012] According to a second aspect of this disclosure, an apparatus for determining resources to be recommended is provided, comprising:

[0013] The resource acquisition unit is used to acquire candidate resources to be recommended.

[0014] A resource level determination unit is used to determine the first resource level in which the candidate resource is located.

[0015] A quality comparison unit is used to compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource tier to obtain a comparison result, wherein the quality requirements corresponding to the second resource tier are higher than the quality requirements of the first resource tier.

[0016] The weight adjustment unit is used to adjust the recommendation weight of the candidate resource according to the comparison result to obtain the target recommendation weight;

[0017] The resource determination unit is used to determine the target resources to be recommended based on the target recommendation weight.

[0018] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0019] At least one processor; and

[0020] A memory that is communicatively connected to the at least one processor;

[0021] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method for determining the recommended resources as described in the first aspect.

[0022] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the method for determining resources to be recommended as described in the first aspect.

[0023] According to a fifth aspect of this disclosure, a computer program product is provided, the computer program product comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, the at least one processor executing the computer program causing the electronic device to perform the method for determining resources to be recommended as described in the first aspect.

[0024] According to the technical solution provided in this disclosure, resources can be divided into multiple resource tiers. After determining the first resource tier to which a candidate resource to be recommended belongs, the candidate resource to be recommended is compared with the quality requirements corresponding to the second resource tier, where the quality requirements corresponding to the second resource tier are higher than those corresponding to the first resource tier. This enables cross-tier comparison of resources, especially comparison with tiers with higher quality requirements. The recommendation weight of the candidate resource is adjusted based on the comparison results to obtain the target recommendation weight. Based on the target recommendation weight, the target resource to be recommended is determined, so that the recommendation weight is tilted towards higher-quality candidate resources; that is, higher-quality candidate resources have higher recommendation weights and a higher probability of being recommended. Therefore, by classifying resources and performing cross-tier comparisons, the quality of recommended resources is improved, especially by promoting the information flow recommendation system to recommend higher-quality resources, thus ensuring the healthy development of the information flow recommendation system ecosystem.

[0025] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0026] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0027] Figure 1 This is a schematic diagram illustrating an application scenario to which this disclosure applies;

[0028] Figure 2 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 1 ;

[0029] Figure 3 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 2 ;

[0030] Figure 4 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 3 ;

[0031] Figure 5 A flowchart illustrating the method for determining resources to be recommended, as provided in this embodiment of the disclosure;

[0032] Figure 6 Schematic diagram of the structure of the apparatus for determining the resource to be recommended provided in the embodiments of this disclosure Figure 1 ;

[0033] Figure 7 Schematic diagram of the structure of the apparatus for determining the resource to be recommended provided in the embodiments of this disclosure Figure 2 ;

[0034] Figure 8 This is a schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure. Detailed Implementation

[0035] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0036] To prevent the over-distribution of low-quality resources and its negative impact on the healthy development of the information flow recommendation system ecosystem, resources can be scored for quality, and their distribution volume can be controlled based on these scores. Specifically, resource control in recommendation can be achieved through the following methods:

[0037] Method 1: Direct weight adjustment. Specifically, the recommendation weight of a resource can be adjusted by multiplying its quality score by a corresponding weight adjustment coefficient.

[0038] Method 2: Based on the amount of display or distribution of resources, categorize resources into different tiers, compare the quality scores of resources within the same tier, increase the recommendation weight of resources with higher quality scores, and decrease the recommendation weight of resources with lower quality scores.

[0039] Method 1 is simple to implement, but it lacks a clear distinction between high-quality and low-quality resources, making it difficult to further favor high-quality resources in resource recommendations. Furthermore, the weighting coefficient is difficult to determine, resulting in poor quality resources in the final recommendations. Method 2 compares resource quality within the same tier, but the comparison scope is limited. From a global perspective, it may not necessarily assign higher recommendation weights to higher-quality resources and lower recommendation weights to lower-quality resources. For example, a resource might be considered a low-quality resource at a low tier, and Method 2 would lower its recommendation weight; conversely, a resource might be considered a high-quality resource at a high tier, and Method 2 would increase its recommendation weight. Therefore, the recommendation weight of this resource does not follow a consistent trend. Thus, from a global perspective, Method 2 is unlikely to favor high-quality resources in resource recommendations, resulting in poor quality resources in the final recommendations.

[0040] To address the aforementioned shortcomings, this disclosure provides a method, apparatus, device, and storage medium for determining resources to be recommended, applicable to the fields of big data, information flow, and technology recommendation within the data processing technology field. In the method for determining resources to be recommended, the resource quality of candidate resources to be recommended is compared with higher-level quality requirements. Based on the comparison results, the recommendation weights of the candidate resources are adjusted to obtain target recommendation weights. The target resources to be recommended are then determined based on these target recommendation weights. Thus, by categorizing resources and comparing resource quality across different tiers, resource recommendations are tilted towards higher-quality resources, improving the overall quality of resource recommendations, especially encouraging information flow recommendation systems to prioritize recommending high-quality resources, thereby promoting the healthy development of the information flow recommendation system ecosystem.

[0041] Figure 1 This diagram illustrates an application scenario applicable to an embodiment of the present disclosure, which can be an information flow recommendation scenario. In this application scenario, the devices involved include a resource recommendation device 101 and a resource database 102. The resource recommendation device 101 can be a server or a terminal. Figure 1 Taking the resource recommendation device 101 as an example as a server; the resource database 102 stores multiple resources.

[0042] The resource recommendation device 101 can obtain candidate resources to be recommended from the resource database 102, adjust the weight of the candidate resources by means of resource classification and cross-class comparison, and determine the target resource to be recommended from the candidate resources with adjusted weight.

[0043] Optionally, the application scenario also includes terminal 103. Terminal 103 can send a resource recommendation request to resource recommendation device 101. Resource recommendation device 101 can further filter the target resource before sending it to terminal 103, or it can send the target resource directly to terminal 103 for resource recommendation.

[0044] The resource recommendation device 101 and the terminal 103 can communicate via wired or wireless means. Figure 1 Take the wireless method as an example.

[0045] As an example, when a user refreshes a video on the video display page shown on terminal 103, terminal 103 responds to the user's refresh operation by sending a video recommendation request to resource recommendation device 101, and resource recommendation device 101 responds to the video recommendation request by sending a recommended video to terminal 103.

[0046] The technical solutions of this disclosure and how they solve the aforementioned technical problems will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this disclosure will now be described with reference to the accompanying drawings.

[0047] For example, the execution subject of this disclosure embodiment can be an electronic device, which can be a server or a terminal. The server can be a centralized server, a distributed server, or a cloud server. The terminal can be a personal digital assistant (PDA) device, a handheld device with wireless communication capabilities (e.g., a smartphone, tablet), a computing device (e.g., a personal computer, PC), an in-vehicle device, a wearable device (e.g., a smartwatch, smart bracelet), and a smart home device (e.g., a smart speaker, smart display device), etc.

[0048] Figure 2 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 1 .like Figure 2 As shown, the methods for determining the resources to be recommended include:

[0049] S201, retrieve candidate resources to be recommended.

[0050] The resources can be media data. Media data can be data in a single media format, such as text, sound, or images; and / or, media data can be data formed by a combination of multimedia, such as video, articles with pictures and text, or music including melodies and lyrics.

[0051] The number of candidate resources can be one or more.

[0052] In this embodiment, candidate resources to be recommended can be obtained from a database. The database includes multiple resources. The candidate resources can be obtained from the database in various ways: sequentially according to their storage order, randomly, or based on the degree of matching between user attributes and resources, from highest to lowest. The matching process will not be described in detail here.

[0053] S202, determine the first resource tier of the candidate resource.

[0054] There can be multiple resource tiers, each with different requirements. These requirements can include both traffic and quality requirements; higher tiers generally have higher traffic and quality requirements. Resources can be tiered based on their traffic volume and the corresponding traffic requirements for each tier.

[0055] However, higher resource traffic does not necessarily equate to higher resource quality. For example, some resources may be frequently shared or commented on, but this does not guarantee their quality. Therefore, after categorizing resources based on their traffic, the recommendation weight can be further adjusted based on resource quality. This allows for the regulation of resource traffic in the future, reducing the likelihood of low-quality resources being excessively recommended to users.

[0056] The first resource tier refers to the resource tier that corresponds to the flow requirement satisfied by the resource flow of the candidate resource among multiple resource tiers.

[0057] In this embodiment, the first resource level of a candidate resource can be determined based on the resource flow of the candidate resource at a historical time or the resource flow of the candidate resource at the current time. Considering that the resource flow of the candidate resource will change over time, determining the first resource level of the candidate resource based on the resource flow of the candidate resource at the current time can improve the accuracy of the first resource level.

[0058] In determining the first resource tier of a candidate resource based on its current resource flow, the current resource flow of the candidate resource can be compared with the flow requirements corresponding to multiple resource tiers one by one to obtain the resource tier corresponding to the flow requirements satisfied by the candidate resource's resource flow, which is the first resource tier.

[0059] S203, compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource tier to obtain the comparison result, wherein the quality requirements corresponding to the second resource tier are higher than the quality requirements of the first resource tier.

[0060] In this embodiment, after determining the first resource level of the candidate resource, that is, after determining the resource level reached by the resource flow of the candidate resource, if no intervention is made to the candidate resource, the resource flow of the candidate resource will gradually increase over time, and the resource flow of the candidate resource will reach the flow requirements corresponding to a higher resource level, such as reaching the flow requirements corresponding to the second resource level. It is very likely that there will be a situation where the resource flow of low-quality resources is very large.

[0061] To avoid this situation and improve the quality of resources recommended to users, after determining the first resource tier, a second resource tier can be selected from multiple resource tiers based on the first resource tier. The resource quality of the candidate resources is then compared with the quality requirements corresponding to the second resource tier to obtain the comparison result. This comparison result is used to determine whether intervention is needed for the candidate resources.

[0062] S204. Based on the comparison results, adjust the recommendation weights of the candidate resources to obtain the target recommendation weights.

[0063] The recommendation weight of candidate resources can be determined based on the recommendation algorithm (e.g., matching candidate resources with user attributes to obtain the user's interest in the candidate resources, and determining the recommendation weight based on the degree of interest). There are no restrictions on the process of determining the recommendation weight of candidate resources.

[0064] The target recommendation weight refers to the adjusted recommendation weight of the candidate resources.

[0065] In this embodiment, based on the comparison results between the candidate resource and the second resource tier, it is determined whether intervention is needed for the candidate resource. If intervention is needed, the recommendation weight of the candidate resource can be increased or decreased to adjust the resource flow of the candidate resource in the future to obtain the target recommendation weight. Otherwise, the recommendation weight of the candidate resource can be kept unchanged.

[0066] S205, determine the target resources to be recommended based on the target recommendation weight.

[0067] In this embodiment, the target resource to be recommended can be selected from the candidate resources according to the target recommendation weight of the candidate resources. For example, one or more candidate resources can be selected in descending order of target recommendation weight, and the target resource can be determined as the selected one or more candidate resources.

[0068] In this embodiment, the recommendation weight of candidate resources is adjusted by classifying resources and comparing the quality of resources across different levels, thereby improving the accuracy and rationality of the recommendation weight adjustment. The target resources to be recommended are determined according to the adjusted recommendation weights, thereby controlling the future resource flow of the target resources. This increases the probability of recommending high-quality resources and decreases the probability of recommending low-quality resources, thus tilting resource recommendations towards high-quality resources and improving the quality of resource recommendations.

[0069] The following provides possible implementations of some steps in the above embodiments.

[0070] In some embodiments, the resource recommendation process may include multiple resource screening stages, such as recall, coarse ranking, and fine ranking. Based on this, one possible implementation of S201 includes: obtaining candidate resources to be recommended in one or more resource screening stages. Thus, by utilizing embodiments of this disclosure, the accuracy and rationality of the recommendation weights of resources in one or more resource screening stages are improved, the quality of resources screened in one or more resource screening stages is improved, and consequently, the quality of resource recommendations is improved.

[0071] In this implementation, during the resource filtering stage, candidate resources to be recommended can be obtained from multiple resources before or after filtering. For example, during the recall stage, multiple resources can be recalled from the database, and candidate resources to be recommended can be obtained from these multiple resources.

[0072] In some embodiments, the resource flow of a candidate resource includes the resource statistics of the candidate resource, which are statistics of the candidate resource in terms of resource flow. Based on this, a possible implementation of S202 includes: obtaining the resource statistics of the candidate resource; and determining a first resource level among multiple resource levels based on the resource statistics of the candidate resource and the statistical requirements corresponding to each of the multiple resource levels. Thus, utilizing resource statistics can more accurately and clearly reflect the characteristics of resource flow, improving the accuracy of determining the first resource level for the candidate resource.

[0073] In this implementation, the resource statistics of the candidate resource at the current moment can be compared one by one with the statistical requirements corresponding to multiple resource levels to obtain the resource level that the resource statistics of the candidate resource meet, namely the first resource level.

[0074] The statistical requirement can be either a threshold value or a range of values ​​for the statistical measure.

[0075] Specifically, when the statistical requirement is a statistical threshold, the resource statistics of the candidate resource are greater than the statistical threshold corresponding to the first resource level and less than the statistical threshold corresponding to the resource level one level higher than the first resource level; when the statistical requirement is a statistical value range, the resource statistics of the candidate resource are within the statistical value range corresponding to the first resource level.

[0076] Furthermore, resource statistics include the current display count of candidate resources and / or the current distribution count of candidate resources. Here, the display count refers to the number of times a resource is shown to users; a display means it is shown on a page, which can also be understood as the number of times a resource is recommended to users. The distribution count refers to the number of times a resource is clicked by users. The current display count of candidate resources refers to the number of times a candidate resource is displayed at the current moment, and the current distribution count of candidate resources refers to the number of times a candidate resource is distributed at the current moment.

[0077] Specifically, when the resource statistics include the current display volume of candidate resources, the statistics corresponding to the resource tier must include the display volume threshold or display volume range corresponding to the resource tier; when the resource statistics include the current distribution volume of candidate resources, the statistics corresponding to the resource tier must include the distribution volume threshold or distribution volume range corresponding to the resource tier.

[0078] As an example, in addition to the current number of impressions and the current number of distributions, the resource statistics of a candidate resource at the current moment can also include the current number of forwards and the current number of favorites.

[0079] In some embodiments, the resource quality of a candidate resource can be represented as a quality score, and the quality requirement corresponding to a resource tier can be represented as a quality score threshold or a quality score range corresponding to the resource tier. Based on this, a possible implementation of S203 includes: comparing the quality score of the candidate resource with the quality score threshold corresponding to the second resource tier, or comparing the quality score of the candidate resource with the quality score range corresponding to the second resource tier. Thus, by comparing the quality score, the quality score threshold, or the quality score range, the precision and accuracy of comparing the resource quality of the candidate resource with the quality requirement corresponding to the second resource tier are improved.

[0080] In addition to the quality score, the quality of a candidate resource can also be represented as its quality level, and the quality requirements corresponding to a resource tier can be represented as the quality level requirements corresponding to that tier. For example, the quality levels are tier one, tier two, and so on.

[0081] In some embodiments, the second resource tier can be an adjacent resource tier to the first resource tier; in other words, the second resource tier can be a resource tier one level higher than the first resource tier. Therefore, by comparing the resource quality of a candidate resource with the quality requirements corresponding to the next higher resource tier, the effectiveness and accuracy of weighting candidate resources are improved. This promotes the rapid attainment of the higher resource tier's traffic requirements by high-quality candidate resources after weight adjustment, while slowing down or preventing the higher resource tier's traffic requirements by low-quality candidate resources after weight adjustment.

[0082] Figure 3 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 2 .like Figure 3 As shown, the methods for determining the resources to be recommended include:

[0083] S301, retrieve candidate resources to be recommended.

[0084] S302, determine the first resource tier of the candidate resource.

[0085] S303, compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource tier to obtain a comparison result, wherein the quality requirements corresponding to the second resource tier are higher than the quality requirements of the first resource tier.

[0086] The implementation principles and technical effects of S301 to S303 can be referred to the relevant descriptions in the foregoing embodiments, and will not be repeated here.

[0087] S304. Based on the comparison results, adjust the recommendation weights of the candidate resources to obtain the target recommendation weights.

[0088] In this embodiment, the resource quality of the candidate resource is compared with the quality requirements corresponding to the second resource tier to obtain a comparison result. The comparison result can be that the resource quality of the candidate resource meets the quality requirements corresponding to the second resource tier, the resource quality of the candidate resource is higher than the quality requirements corresponding to the second resource tier, or the resource quality of the candidate resource is lower than the quality requirements corresponding to the second resource tier. Based on different comparison results, different adjustment methods can be adopted for the recommendation weight to improve the rationality and accuracy of the recommendation weight adjustment, that is, to improve the rationality and accuracy of the target recommendation weight.

[0089] Among them, such as Figure 3 As shown, S304 includes S3041 to S3043:

[0090] S3041, if the comparison result shows that the resource quality of the candidate resource meets the quality requirements corresponding to the second resource level, then the target recommendation weight is determined as the first weight, where the first weight is the current recommendation weight of the candidate resource.

[0091] In this embodiment, the resource quality of the candidate resource meets the quality requirements corresponding to the second resource tier, indicating that the resource traffic of the candidate resource is suitable for meeting the traffic requirements corresponding to the second resource tier. Without intervention, the resource traffic of the candidate resource will increase over time to reach the traffic requirements corresponding to the second resource tier. Therefore, no intervention is needed for the candidate resource, and the target recommendation weight can be determined as the first weight, i.e., the current recommendation weight of the candidate resource remains unchanged.

[0092] S3042, if the comparison result shows that the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource level, then the target recommendation weight is determined as the second weight based on the current recommendation weight of the candidate resource, and the second weight is greater than the current recommendation weight of the candidate resource.

[0093] In this embodiment, the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource tier. This indicates that the resource quality of the candidate resource not only meets the quality requirement corresponding to the second resource tier but also exceeds it, reaching the quality requirement of the next higher resource tier. Therefore, this means that the resource traffic of the candidate resource can reach a higher resource tier than the second resource tier. In this case, without intervention, the resource traffic of the candidate resource may take a long time to reach the traffic requirement corresponding to the higher resource tier. To promote a faster increase in the resource traffic of high-quality candidate resources—that is, to promote the recommendation of high-quality resources to users more frequently and with a higher probability—the recommendation weight of the candidate resource can be increased. In the process of increasing the recommendation weight of the candidate resource, a target recommendation weight can be determined based on the current recommendation weight of the candidate resource. In this case, the target recommendation weight is a second weight that is greater than the current recommendation weight.

[0094] S3043, if the comparison result shows that the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource level, then the target recommendation weight is determined as the third weight based on the current recommendation weight of the candidate resource, and the third weight is less than the current recommendation weight of the candidate resource.

[0095] In this embodiment, the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource tier, indicating that the resource traffic of the candidate resource is not suitable for meeting the traffic requirement corresponding to the second resource tier. However, without intervention, the resource traffic of the candidate resource will naturally reach the traffic requirement corresponding to the second resource tier over time. Therefore, it is necessary to intervene in the candidate resource to slow down or even prevent the resource traffic of the candidate resource from reaching the traffic requirement corresponding to the second resource tier. To achieve this, a target recommendation weight can be determined based on the current recommendation weight of the candidate resource. In this case, the target recommendation weight is a third weight that is less than the current recommendation weight.

[0096] In one possible implementation, where the resource quality of candidate resources is represented by quality scores, the meanings of different comparison results are as follows:

[0097] The quality of the candidate resource meets the quality requirements corresponding to the second resource tier: the quality score of the candidate resource is within the range of the quality score value corresponding to the second resource tier, or the quality score of the candidate resource is greater than or equal to the quality score threshold corresponding to the second resource tier and the quality score of the candidate resource is less than or equal to the quality score threshold corresponding to the resource tier one tier higher than the second resource tier.

[0098] The quality of the candidate resource is higher than the quality requirement of the second resource tier: the quality score of the candidate resource is greater than the maximum value within the range of the quality score corresponding to the second resource tier, or the quality score of the candidate resource is greater than the quality score threshold corresponding to the resource tier one level higher than the second resource tier.

[0099] The quality of the candidate resource is lower than the quality requirement corresponding to the second resource tier: the quality score of the candidate resource is less than the minimum value within the range of the quality score corresponding to the second resource tier, or the quality score of the candidate resource is less than the quality score threshold corresponding to the second resource tier.

[0100] In one possible implementation, when the second resource tier is an adjacent resource tier to the first resource tier, if the quality score of a candidate resource is less than the quality score threshold corresponding to the second resource tier, then the resource quality of the candidate resource is determined to be lower than the quality requirement corresponding to the second resource tier, and the recommendation weight of the candidate resource needs to be reduced. If the quality score of the candidate resource is greater than the quality score threshold corresponding to the second resource tier, then it can be determined whether the quality score of the candidate resource is greater than or equal to the quality score threshold corresponding to the third resource tier. If so, then the resource quality of the candidate resource is determined to be higher than the quality requirement corresponding to the second resource tier, and the recommendation weight of the candidate resource needs to be increased; otherwise, the resource quality of the candidate resource is determined to meet the quality requirement corresponding to the second resource tier.

[0101] Among them, the third resource tier is the adjacent resource tier of the second resource tier, and the quality requirements corresponding to the third resource tier are higher than those corresponding to the second resource tier.

[0102] As an example, Figure 4 This is an example diagram illustrating a resource comparison on a per-level basis according to embodiments of this disclosure. For example... Figure 4 As shown, resource tiers include Tier 1, Tier 2, Tier 3, ..., Tier n. Resources within each tier can be categorized as high-quality, intermediate, or low-quality. Taking Tier 1 as an example, the quality of resources in Tier 1 can be compared to the quality requirements of Tier 2. If the resource's quality score is lower than the Tier 2 quality score threshold, it is classified as low-quality, and its recommendation weight can be reduced to suppress or filter it. If the resource's quality score is higher than the Tier 2 quality score threshold, it can be compared to the Tier 3 quality score threshold. If the resource's quality score is higher than the Tier 3 quality score threshold, it is classified as high-quality, and its recommendation weight can be increased to enhance its quality. If the resource's quality score is greater than or equal to the Tier 2 quality score threshold but less than the Tier 3 quality score threshold, it is classified as intermediate, and its recommendation weight remains unchanged.

[0103] S305, determine the target resources to be recommended based on the target recommendation weight.

[0104] The implementation principle and technical effects of S305 can be referred to the relevant descriptions in the foregoing embodiments, and will not be repeated here.

[0105] In this embodiment, the recommendation weight of candidate resources is adjusted by classifying resources and comparing resource quality across different levels. During the adjustment process, targeted adjustment schemes are provided based on different comparison results to increase the recommendation weight of high-quality candidate resources and decrease the recommendation weight of low-quality candidate resources. The target resources to be recommended are determined according to the adjusted recommendation weights, thereby controlling the future resource traffic of the target resources. This increases the probability of high-quality resources being recommended and decreases the probability of low-quality resources being recommended, thus tilting resource recommendations towards high-quality resources and improving the quality of resource recommendations.

[0106] exist Figure 3 In the illustrated embodiment, if the comparison result shows that the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource tier, or the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource tier, then the target recommendation weight needs to be determined based on the current recommendation weight of the candidate resource. The following will explain... Figure 5 The illustrated embodiment provides a possible implementation for determining the target recommendation weight based on the current recommendation weight of the candidate resources.

[0107] Figure 5 This is a flowchart illustrating the method for determining resources to be recommended according to embodiments of this disclosure. Figure 3 .like Figure 5 As shown, the methods for determining the resources to be recommended include:

[0108] S501, retrieve candidate resources to be recommended.

[0109] S502, determine the first resource tier of the candidate resource.

[0110] S503, compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource tier to obtain a comparison result, wherein the quality requirements corresponding to the second resource tier are higher than the quality requirements of the first resource tier.

[0111] The implementation principles and technical effects of S501 to S503 can be referred to the relevant descriptions in the foregoing embodiments, and will not be repeated here.

[0112] S504, determine the comparison result as either the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource tier, or the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource tier.

[0113] S505, based on the resource quality of the candidate resources, determines the target distribution amount of the candidate resources at future times.

[0114] The future time can be preset or the prediction duration can be preset, and the future time can be determined based on the prediction duration and the current time.

[0115] In this embodiment, if the comparison result shows that the resource quality of a candidate resource is higher than or lower than the quality requirement corresponding to the second resource tier, the recommendation weight of the candidate resource needs to be adjusted. That is, the target recommendation weight needs to be determined based on the current recommendation weight of the candidate resource. In this process, the target distribution amount of the candidate resource at future times can be determined first based on its resource quality.

[0116] In one possible implementation, the traffic requirement corresponding to the resource tier includes the distribution volume threshold corresponding to the resource tier. Based on this, S505 may include: predicting the distribution volume of the candidate resource to obtain the predicted distribution volume of the candidate resource at a future time; determining the target tier whose resource quality meets the requirements of the candidate resource among multiple resource tiers; and determining the target distribution volume of the candidate resource at a future time based on the distribution volume threshold corresponding to the target tier and the predicted distribution volume.

[0117] In this implementation, a distribution volume prediction model can be used to predict the distribution volume of candidate resources, obtaining a relatively accurate predicted distribution volume that conforms to the distribution rules of candidate resources. Since low-quality resources may have a large distribution volume, relying solely on distribution volume prediction might result in a large predicted distribution volume for low-quality resources. To avoid this unreasonable situation, the predicted distribution volume can be constrained by the resource quality of the candidate resources to obtain a more reasonable target distribution volume: First, the resource quality of the candidate resources can be compared with the quality requirements corresponding to multiple resource tiers to determine the resource tier corresponding to the quality requirements met by the candidate resources' resource quality, i.e., to determine the target tier; considering that the resource quality of the candidate resources meets the target tier, the predicted distribution volume can be adjusted based on the distribution volume threshold corresponding to the target tier to obtain the target distribution volume of the candidate resources at future times.

[0118] Therefore, by using a distribution volume prediction method, a relatively accurate predicted distribution volume is obtained. Based on the target level satisfied by the resource quality of the candidate resources, the predicted distribution volume is constrained to obtain the target distribution volume of the candidate resources at future times, thus improving the rationality of the target distribution volume. In summary, this approach improves both the accuracy and rationality of the target distribution volume.

[0119] Furthermore, the distribution volume prediction model can be a time series model. Predicting the distribution volume of candidate resources to obtain the predicted distribution volume of candidate resources at future times can include: predicting the distribution volume of candidate resources at future times based on historical distribution volumes and the time series model. Thus, utilizing historical distribution volumes and the time series model improves the accuracy of distribution volume prediction.

[0120] The time series model can be a neural network model or a fitted model obtained through mathematical fitting. When the time series model is a neural network model, it needs to be pre-trained, and then the trained time series model is used to predict the predicted distribution amount of candidate resources at future times.

[0121] The historical time points can be multiple. Based on the historical distribution volume and time series model of the candidate resources at each historical time point, the distribution volume of the candidate resources can be predicted. The historical distribution volume of the candidate resources at each historical time point can reflect the development pattern of the distribution volume of the candidate resources, so that the predicted distribution volume of the candidate resources at future time points also conforms to the development pattern of the distribution volume, resulting in high accuracy.

[0122] In determining the target distribution amount of candidate resources at future times based on the distribution amount threshold corresponding to the target tier and the predicted distribution amount, one possible implementation is to determine the target distribution amount as the distribution amount threshold corresponding to the target tier if the predicted distribution amount is greater than or equal to the distribution amount threshold corresponding to the target tier; otherwise, the target distribution amount is determined as the predicted distribution amount. This limits the target distribution amount of candidate resources at future times to within the distribution amount threshold corresponding to the resource tier that the resource quality of the candidate resources satisfies, improving the rationality of the target distribution amount and preventing candidate resources from reaching distribution amounts inconsistent with their quality.

[0123] As an example, among multiple resource tiers, the distribution threshold for the first tier is 10,000, for the second tier it's 50,000, for the third tier it's 100,000, and for the fourth tier it's 500,000. Comparing the resource quality of the candidate resource with the quality requirements of each resource tier, we find that the candidate resource's resource quality reaches the highest level, the third tier. Since the predicted distribution volume of the candidate resource at a future time may exceed the third tier's threshold of 100,000, the predicted distribution volume needs to be limited, resulting in a target distribution volume of 100,000.

[0124] Besides the methods mentioned above, another approach is to increase or decrease the predicted distribution volume of candidate resources in the future based on the distribution volume threshold corresponding to the target tier, thus obtaining the target distribution volume of the candidate resources in the future. However, this method may still result in the target distribution volume exceeding the distribution volume threshold corresponding to the target tier, making it less effective than the methods described above. Alternatively, one can directly determine the target distribution volume of the candidate resources in the future as the distribution volume threshold corresponding to the target tier without using a distribution volume prediction model. However, this method is relatively crude and may broaden the user scope of distribution, resulting in many invalid distributions. For example, if the candidate resource's quality reaches the fourth tier, without prediction, the target distribution volume of the candidate resource in the future would be determined according to the distribution volume threshold of 500,000 corresponding to the fourth tier. With prediction, the predicted distribution volume of the candidate resource in the future might be 200,000, and the target distribution volume would be determined to be 200,000.

[0125] S506, determine the resource distribution speed of the candidate resource based on the difference between the target distribution amount and the current distribution amount of the candidate resource.

[0126] Resource distribution speed refers to the speed at which candidate resources are distributed to users, or the growth rate of the distribution volume of candidate resources. The faster the resource distribution speed of candidate resources, the more frequently and with a higher probability the candidate resources should be recommended to users.

[0127] In this embodiment, the target distribution amount of the candidate resource at a future time can be compared with the current distribution amount of the candidate resource to obtain the difference between the target distribution amount and the current distribution amount. Then, the resource distribution speed of the candidate resource can be determined based on the difference between the target distribution amount and the current distribution amount, the current time, and the future time. For example, the resource distribution speed of the candidate resource can be obtained by dividing the difference by the time difference between the current time and the future time.

[0128] In one possible implementation, S506 includes: determining the difference between the target distribution amount of the candidate resource at a future time and the current distribution amount of the candidate resource; and determining the resource distribution rate of the candidate resource using a proportional-integral-derivative (PID) controller based on the difference. Thus, the accuracy of the resource distribution rate is improved by utilizing the PID controller.

[0129] The principle formula of the PID controller can be expressed as:

[0130]

[0131] err(t) = rin(t) - rout(t)

[0132] Discretizing U(t) yields:

[0133] U(t) = K p err(t)+K i ∑err(t)+K d (err(t)-err(t-1))

[0134] U(t) represents the resource distribution rate of the candidate resource, t represents a future time, rin(t) represents the target distribution amount of the candidate resource at the future time, rout(t) identifies the current distribution amount of the candidate resource, err(t) represents the difference between the target distribution amount and the current distribution amount, T1 represents the integration time constant, and T... D K represents the differential time constant. p It is the proportionality coefficient, K i It is the integral coefficient, K d These are differential coefficients, all of which are hyperparameters. Therefore, the only remaining calculation in the formula is err(t), which can be determined by the target distribution amount and the distribution amount at the current moment. Thus, through the PID algorithm, precise control of resource flow can be achieved for specific resources based on resource quality and the resource tier they occupy.

[0135] S507. Determine the target recommendation weight based on the current recommendation weight of the candidate resource and the resource distribution speed of the candidate resource.

[0136] In this embodiment, the current recommendation weight of the candidate resource can be increased or decreased based on the resource distribution speed to obtain the target recommendation weight.

[0137] In one possible implementation, the resource distribution speed can be multiplied by the current recommendation weight of the candidate resource to obtain the target recommendation weight. When the resource distribution speed is greater than 1, the recommendation weight of the candidate resource is increased; when the resource distribution speed is less than 1, the recommendation weight of the candidate resource is decreased. Thus, the recommendation weight of the candidate resource is precisely adjusted through the resource distribution speed, improving the rationality of the recommendation weight.

[0138] S508: Determine the target resources to be recommended based on the target recommendation weight.

[0139] The implementation principle and technical effects of S508 can be referred to the corresponding descriptions in the foregoing embodiments, and will not be repeated here.

[0140] In this embodiment, the recommendation weight of candidate resources is adjusted by classifying resources and comparing the quality of resources across different levels. During the adjustment process, the resource distribution speed of candidate resources is determined in combination with the resource quality of the candidate resources. The recommendation weight of candidate resources is adjusted based on the resource distribution speed, which improves the accuracy and rationality of the recommendation weight adjustment. This enables precise control of the future resource traffic of target resources, increases the probability of recommending high-quality resources and reduces the probability of recommending low-quality resources, thus tilting resource recommendations towards high-quality resources and improving the quality of resource recommendations.

[0141] Figure 6 Schematic diagram of the structure of the apparatus for determining the resource to be recommended provided in the embodiments of this disclosure Figure 1 .like Figure 6 As shown, the device 600 for determining the resources to be recommended includes:

[0142] Resource acquisition unit 601 is used to acquire candidate resources to be recommended;

[0143] The grade determination unit 602 is used to determine the first resource grade of the candidate resource;

[0144] The quality comparison unit 603 is used to compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource level to obtain a comparison result, wherein the quality requirements corresponding to the second resource level are higher than the quality requirements of the first resource level.

[0145] The weight adjustment unit 604 is used to adjust the recommendation weight of the candidate resources according to the comparison results to obtain the target recommendation weight;

[0146] Resource determination unit 605 is used to determine the target resources to be recommended based on the target recommendation weight.

[0147] Figure 7 Schematic diagram of the structure of the apparatus for determining the resource to be recommended provided in the embodiments of this disclosure Figure 2 .like Figure 7 As shown, the device 700 for determining the resources to be recommended includes:

[0148] Resource acquisition unit 701 is used to acquire candidate resources to be recommended;

[0149] The grade determination unit 702 is used to determine the first resource grade of the candidate resource;

[0150] The quality comparison unit 703 is used to compare the resource quality of the candidate resource with the quality requirements corresponding to the second resource level and obtain a comparison result, wherein the quality requirements corresponding to the second resource level are higher than the quality requirements of the first resource level.

[0151] The weight adjustment unit 704 is used to adjust the recommendation weight of the candidate resources according to the comparison results to obtain the target recommendation weight;

[0152] Resource determination unit 705 is used to determine the target resources to be recommended based on the target recommendation weight.

[0153] In some embodiments, the weight adjustment unit 704 includes: a first weight determination module 7041, configured to determine a target recommendation weight as a first weight, where the first weight is the current recommendation weight of the candidate resource, if the comparison result indicates that the resource quality of the candidate resource meets the quality requirements corresponding to the second resource tier; a second weight determination module 7042, configured to determine a target recommendation weight as a second weight, where the second weight is greater than the current recommendation weight, if the comparison result indicates that the resource quality of the candidate resource is higher than the quality requirements corresponding to the second resource tier; and a third weight determination module 7043, configured to determine a target recommendation weight as a third weight, where the third weight is less than the current recommendation weight, if the comparison result indicates that the resource quality of the candidate resource is lower than the quality requirements corresponding to the second resource tier.

[0154] In some embodiments, the second weight determination module 7042 and / or the third weight determination module 7043 include: a distribution volume prediction submodule (not shown in the figure), used to determine the target distribution volume of the candidate resource at a future time based on the resource quality of the candidate resource; a distribution speed determination submodule (not shown in the figure), used to determine the resource distribution speed of the candidate resource based on the difference between the target distribution volume and the current distribution volume of the candidate resource; and a weight determination submodule (not shown in the figure), used to determine the target recommendation weight based on the current recommendation weight and the resource distribution speed.

[0155] In some embodiments, the distribution volume prediction submodule is specifically used for: predicting the distribution volume of candidate resources to obtain the predicted distribution volume of candidate resources at future times; determining the target level that the resource quality of candidate resources meets among multiple resource levels; and determining the target distribution volume based on the distribution volume threshold corresponding to the target level and the predicted distribution volume.

[0156] In some embodiments, in the process of predicting the distribution volume of candidate resources to obtain the predicted distribution volume of candidate resources at future times, the distribution volume prediction submodule is specifically used to: predict the distribution volume of candidate resources based on the historical distribution volume and time series model of candidate resources at historical times to obtain the predicted distribution volume.

[0157] In some embodiments, during the process of determining the target distribution amount based on the distribution amount threshold corresponding to the target tier and the predicted distribution amount, the distribution amount prediction submodule is specifically used to: if the predicted distribution amount is greater than or equal to the distribution amount threshold, then determine the target distribution amount as the distribution amount threshold; if the predicted distribution amount is less than the distribution amount threshold, then determine the target distribution amount as the predicted distribution amount.

[0158] In some embodiments, the distribution speed determination submodule is specifically used to: determine the difference between the target distribution amount and the current distribution amount; and determine the resource distribution speed through a PID controller based on the difference.

[0159] In some embodiments, the tier determination unit 702 includes: a resource statistics acquisition module 7021, used to acquire resource statistics of candidate resources, the resource statistics including the current display volume of candidate resources and / or the current distribution volume of candidate resources; and a tier determination module 7022, used to determine a first resource tier among multiple resource tiers based on the resource statistics and the statistics requirements corresponding to multiple resource tiers respectively.

[0160] Figures 6-7 The device for determining the resources to be recommended can execute the corresponding method embodiments described above. Its implementation principle and technical effect are similar, and will not be repeated here.

[0161] According to embodiments of this disclosure, this disclosure also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the scheme provided in any of the above embodiments.

[0162] According to embodiments of this disclosure, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the scheme provided in any of the above embodiments.

[0163] According to embodiments of this disclosure, this disclosure also provides a computer program product comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, and the at least one processor executing the computer program causing the electronic device to perform the scheme provided in any of the above embodiments.

[0164] Figure 8This is a schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0165] like Figure 8 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) (e.g., ROM 802) or a computer program loaded from a storage unit 808 into a random access memory (RAM) (e.g., RAM 803). The RAM 803 may also store various programs and data required for the operation of the electronic device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface (e.g., I / O interface 805) is also connected to the bus 804.

[0166] Multiple components in electronic device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of displays, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows electronic device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0167] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the method for determining resources to be recommended. For example, in some embodiments, the method for determining resources to be recommended may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the method for determining resources to be recommended described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform a method for determining the resources to be recommended by any other suitable means (e.g., by means of firmware).

[0168] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard parts (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0169] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0170] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0171] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0172] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0173] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem, addressing the shortcomings of traditional physical hosts and VPS (Virtual Private Server, or simply "VPS") services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0174] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this disclosure can be achieved, and this is not limited herein.

[0175] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for determining resources to be recommended, comprising: Obtain candidate resources to be recommended; Among multiple resource tiers, determine the first resource tier in which the candidate resource is located; The multiple resource tiers are divided according to the resource flow rate and the flow rate requirements corresponding to the resource tier; The quality of the candidate resource is compared with the quality requirement corresponding to the second resource level among the multiple resource levels to obtain a comparison result, wherein the quality requirement corresponding to the second resource level is higher than the quality requirement of the first resource level. If the comparison result indicates that the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource level, then based on the current recommendation weight, the target recommendation weight is determined as the second weight, and the second weight is greater than the current recommendation weight. If the comparison result indicates that the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource tier, then based on the current recommendation weight, the target recommendation weight is determined to be the third weight, and the third weight is less than the current recommendation weight. Based on the target recommendation weight, determine the target resources to be recommended; Determining the target recommendation weight based on the current recommendation weight includes: The distribution volume of the candidate resources is predicted to obtain the predicted distribution volume of the candidate resources at future times; Among multiple resource tiers, determine the target tier that the resource quality of the candidate resource meets; Based on the distribution volume threshold corresponding to the target tier and the predicted distribution volume, the target distribution volume of the candidate resource at a future time is determined; The resource distribution speed of the candidate resource is determined based on the difference between the target distribution amount and the current distribution amount of the candidate resource. The target recommendation weight is determined based on the current recommendation weight and the resource distribution speed.

2. The determination method according to claim 1, wherein, If the comparison result indicates that the resource quality of the candidate resource meets the quality requirements corresponding to the second resource tier, then the target recommendation weight is determined to be the first weight, and the first weight is the current recommendation weight of the candidate resource.

3. The determination method according to claim 1, wherein, The step of predicting the distribution volume of the candidate resources to obtain the predicted distribution volume of the candidate resources at future times includes: Based on the historical distribution volume and time series model of the candidate resources at historical moments, the distribution volume of the candidate resources is predicted to obtain the predicted distribution volume.

4. The determination method according to claim 1, wherein, Determining the target distribution volume of the candidate resource at a future time based on the distribution volume threshold corresponding to the target tier and the predicted distribution volume includes: If the predicted distribution amount is greater than or equal to the distribution amount threshold, then the target distribution amount is determined to be the distribution amount threshold; If the predicted distribution amount is less than the distribution amount threshold, then the target distribution amount is determined to be the predicted distribution amount.

5. The determination method according to claim 1, wherein, Determining the resource distribution speed of the candidate resource based on the difference between the target distribution amount and the current distribution amount of the candidate resource includes: Determine the difference between the target distribution amount and the current distribution amount; Based on the difference, the resource distribution speed is determined by a PID controller.

6. The determining method according to any one of claims 1-5, characterized in that, in, Determining the first resource tier of the candidate resource includes: Obtain resource statistics for candidate resources, including the current display volume and / or the current distribution volume of the candidate resources; Based on the resource statistics and the statistical requirements corresponding to the multiple resource tiers, the first resource tier is determined among the multiple resource tiers.

7. A device for determining resources to be recommended, comprising: The resource acquisition unit is used to acquire candidate resources to be recommended. A resource level determination unit is used to determine the first resource level in which the candidate resource is located among multiple resource levels. The multiple resource tiers are divided according to the resource flow rate and the flow rate requirements corresponding to the resource tier; A quality comparison unit is used to compare the resource quality of the candidate resource with the quality requirement corresponding to the second resource level among the plurality of resource levels, and obtain a comparison result, wherein the quality requirement corresponding to the second resource level is higher than the quality requirement of the first resource level. The weight adjustment unit is used to adjust the recommendation weight of the candidate resource according to the comparison result to obtain the target recommendation weight; The resource determination unit is used to determine the target resource to be recommended based on the target recommendation weight. The weight adjustment unit includes: a second weight determination module, used to determine a target recommendation weight as a second weight based on the current recommendation weight if the comparison result is that the resource quality of the candidate resource is higher than the quality requirement corresponding to the second resource level; the second weight is greater than the current recommendation weight. The third weight determination module is used to determine the target recommendation weight as the third weight based on the current recommendation weight if the comparison result shows that the resource quality of the candidate resource is lower than the quality requirement corresponding to the second resource level. The third weight is less than the current recommendation weight. The second weight determination module and / or the third weight determination module include: The distribution volume prediction submodule is used to determine the target distribution volume of the candidate resource at a future time based on the resource quality of the candidate resource; The distribution speed determination submodule is used to determine the resource distribution speed of the candidate resource based on the difference between the target distribution amount and the current distribution amount of the candidate resource; The weight determination submodule is used to determine the target recommendation weight based on the current recommendation weight and the resource distribution speed; The distribution volume prediction submodule is specifically used for: The distribution volume of the candidate resources is predicted to obtain the predicted distribution volume of the candidate resources at future times; Among multiple resource tiers, determine the target tier that the resource quality of the candidate resource meets; The target distribution volume is determined based on the distribution volume threshold corresponding to the target tier and the predicted distribution volume.

8. The determining device according to claim 7, wherein, The weight adjustment unit further includes: The first weight determination module is used to determine the target recommendation weight as the first weight if the comparison result shows that the resource quality of the candidate resource meets the quality requirements corresponding to the second resource level. The first weight is the current recommendation weight of the candidate resource.

9. The determining device according to claim 7, wherein, In the process of predicting the distribution volume of the candidate resources to obtain the predicted distribution volume of the candidate resources at future times, the distribution volume prediction submodule is specifically used for: Based on the historical distribution volume and time series model of the candidate resources at historical moments, the distribution volume of the candidate resources is predicted to obtain the predicted distribution volume.

10. The determining device according to claim 7, wherein, In the process of determining the target distribution volume based on the distribution volume threshold corresponding to the target tier and the predicted distribution volume, the distribution volume prediction submodule is specifically used for: If the predicted distribution amount is greater than or equal to the distribution amount threshold, then the target distribution amount is determined to be the distribution amount threshold; If the predicted distribution amount is less than the distribution amount threshold, then the target distribution amount is determined to be the predicted distribution amount.

11. The determining device according to claim 7, wherein, The distribution speed determination submodule is specifically used for: Determine the difference between the target distribution amount and the current distribution amount; Based on the difference, the resource distribution speed is determined by a PID controller.

12. The determining device according to any one of claims 7-11, characterized in that, in, The gear selection unit includes: The resource statistics acquisition module is used to acquire the resource statistics of the candidate resources, wherein the resource statistics include the current display volume and / or the current distribution volume of the candidate resources; The tier determination module is used to determine the first resource tier among multiple resource tiers based on the resource statistics and the statistical requirements corresponding to the multiple resource tiers respectively.

13. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method for determining the resource to be recommended according to any one of claims 1-6.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method for determining the resource to be recommended according to any one of claims 1-6.

15. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the method for determining the resource to be recommended according to any one of claims 1-6.

Images