Resource recommendation method, device, equipment and storage medium

Abstract

The present disclosure provides a resource recommendation method, apparatus, device, and storage medium, which relate to technical fields such as big data, information flow, search technology, and recommendation technology in the field of data processing. The resource recommendation method includes: in response to a resource recommendation request, according to a first resource quantity threshold, filtering from a candidate resource set to obtain a first recommended resource and a recommendation score corresponding to the first recommended resource; obtaining a second recommended resource and a recommendation score corresponding to the second recommended resource, the second recommended resource and the recommendation score being obtained by filtering from the candidate resource set according to a second resource quantity threshold, the second resource quantity threshold being greater than the first resource quantity threshold; determining a target recommended resource from the first recommended resource and the second recommended resource based on the recommendation scores corresponding to the first recommended resource and the second recommended resource respectively; and recommending the target recommended resource. Thus, the first recommended resource is enriched with the second recommended resource filtered out on a larger scale, thereby reducing the resource duplication rate.

Description

Technical Field

[0001] The present disclosure relates to technical fields such as big data, information flow, search technology, and recommendation technology in the field of data processing, and in particular to a resource recommendation method, device, equipment, and storage medium. Background Art

[0002] As the amount of information gradually increases, filtering out resources suitable for users from massive resources and recommending them to users can effectively save users' time and improve user experience.

[0003] In related technologies, in response to a user refresh, resources in a resource set are recalled and sorted to obtain multiple resources, which are recommended to the user; thereafter, in response to the user's next refresh, resources in the resource set are recalled and sorted again to perform the next round of resource recommendations.

[0004] However, when the refresh interval is short, the resource recommendation repetition rate of the above method is high. Summary of the Invention

[0005] The present disclosure provides a resource recommendation method, apparatus, device, and storage medium for reducing the duplication rate of resource recommendations.

[0006] According to a first aspect of the present disclosure, a resource recommendation method is provided, comprising:

[0007] Receive resource recommendation requests from users;

[0008] In response to the resource recommendation request, screening a first recommended resource and a recommendation score corresponding to the first recommended resource from a candidate resource set according to a first resource quantity threshold;

[0009] Obtaining a second recommended resource and a recommendation score corresponding to the second recommended resource, where the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, where the second resource quantity threshold is greater than the first resource quantity threshold;

[0010] determining a target recommended resource from the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource;

[0011] Recommend the target recommended resource to the user.

[0012] According to a second aspect of the present disclosure, a resource recommendation device is provided, comprising:

[0013] A receiving unit, configured to receive a resource recommendation request from a user;

[0014] a first screening unit, configured to, in response to the resource recommendation request, screen a candidate resource set according to a first resource quantity threshold to obtain a first recommended resource and a recommendation score corresponding to the first recommended resource;

[0015] a resource acquisition unit, configured to acquire a second recommended resource and a recommendation score corresponding to the second recommended resource, wherein the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, wherein the second resource quantity threshold is greater than the first resource quantity threshold;

[0016] a resource determining unit, configured to determine a target recommended resource from among the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource;

[0017] A resource recommendation unit is configured to recommend the target recommended resource to the user.

[0018] According to a third aspect of the present disclosure, an electronic device is provided, comprising: 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, and the instructions are executed by the at least one processor to enable the at least one processor to execute the resource recommendation method described in the first aspect.

[0019] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the resource recommendation method described in the first aspect.

[0020] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising: a computer program, wherein the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and the at least one processor executes the computer program so that the electronic device executes the resource recommendation method described in the first aspect.

[0021] According to the technical solution provided by the present disclosure, when a user requests a resource recommendation, the first recommended resource and the recommendation score corresponding to the first recommended resource are screened from the candidate resource set according to the first resource quantity threshold. The candidate resource set changes less in a short period of time. If the target recommended resource is determined solely by the first recommended resource, the target recommended resource will have a high repetition rate with the historical recommended resource (such as the last recommended resource). To solve this problem, the target recommended resource is determined in the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource. Since the second recommended resource is screened from the candidate resource set based on the second resource quantity threshold, and the second resource quantity threshold is greater than the first resource quantity threshold, there are resources in the second recommended resource that are different from the first recommended resource. The second recommended resource can enrich the first recommended resource, thereby reducing the repetition rate between the target recommended resource and the last recommended resource.

[0022] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0024] Figure 1 This is an example of a resource recommendation. Figure 1 ;

[0025] Figure 2 is a schematic diagram of an application scenario applicable to the embodiments of the present disclosure;

[0026] Figure 3 is a schematic diagram according to a first embodiment of the present disclosure;

[0027] Figure 4 is a schematic diagram according to a second embodiment of the present disclosure;

[0028] Figure 5 This is an example of a resource recommendation. Figure 2 ;

[0029] Figure 6 is a schematic diagram of a third embodiment of the present disclosure;

[0030] Figure 7 FIG. 7 is a schematic block diagram of an example electronic device 700 that may be used to implement embodiments of the present disclosure. DETAILED DESCRIPTION

[0031] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0032] First, the terms used in this disclosure are explained:

[0033] Resource recall: The process of finding a small number of resources that a user might be interested in from a large number of candidate resources based on user characteristics. For example, this process involves filtering out tens of thousands of resources from tens of millions of candidate resources. Recall can be multi-channel, such as user-based collaborative filtering (UCF) and item-based collaborative filtering (ICF).

[0034] Sorting: This refers to the process of sorting resources and outputting the top K ranked resources. In resource recommendation systems, sorting can include coarse sorting and fine sorting. Coarse sorting focuses on sorting speed and cost, while fine sorting focuses on sorting accuracy.

[0035] Figure 1 This is an example of a resource recommendation. Figure 1 .like Figure 1 As shown, resource recommendation may include four stages: recall, rough sorting, fine sorting, and exposure. The corresponding magnitudes (i.e., the number of screened resources) can be tens of thousands, thousands, hundreds, and individual, respectively. The overall structure is like a funnel. The further down you go, the fewer resources are screened. Every time a user refreshes resources, the resource recommendation system can go through these stages and expose the resources to the user. If the time interval between two refreshes is short, then the calculated resource duplication rate will be relatively high in the recall, rough sorting, and fine sorting stages, and the resource duplication rate will also be high in the exposure stage.

[0036] For example: In the recall phase, there are 1 million candidate resources. When the user refreshes resources for the first time, 10,000 resources are recalled, and ultimately 6 resources are exposed to the user. When the user refreshes resources for the second time, assuming that all authors publish another 500 resources during the time interval between the two refreshes, since the exposed resources will no longer be recalled, the total number of candidate resources is "1 million + 500 - 6". Therefore, during the second resource refresh, the number of candidate resources in the recall phase is still around 1 million, most of which are duplicates of the previous 1 million. Therefore, most of the 10,000 resources recalled during the second resource refresh will also be duplicates of the 10,000 resources recalled during the first refresh. The same is true in the coarse and fine ranking stages. Most of the resources at the entrance are duplicates, and the recommendation scores of these resources in the two resource refreshes are basically the same. In the end, most of the resources exposed by the two resource refreshes are also duplicates.

[0037] It can be seen that when the candidate resources change little (for example, the time interval between two refreshes is short), the resource recommendation method mentioned above has a high resource duplication rate in different resource recommendations, and more resources cannot be explored, resulting in a poor user experience.

[0038] In order to solve the above problems, the present disclosure provides a resource recommendation method, apparatus, device and storage medium, which are applied to the field of data processing, specifically in the technical fields of big data, information flow, search technology, recommendation technology, etc. In the resource recommendation method, when a user requests a resource recommendation, according to a first resource quantity threshold, a first recommended resource and a recommendation score corresponding to the first recommended resource are screened from a candidate resource set, a second recommended resource and a recommendation score corresponding to the second recommended resource are obtained, and according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource, a target recommended resource is determined from the first recommended resource and the second recommended resource, and the target recommended resource is recommended to the user. The recommendation scores corresponding to the second recommended resource and the second recommended resource are obtained by screening from the candidate resource set based on a second resource quantity threshold. The second resource quantity threshold is greater than the first resource quantity threshold, so that a large number of resources different from the first recommended resource exist in the second recommended resource. Combining the first recommended resource and the second recommended resource can effectively reduce the repetition rate between the target recommended resource and the resource recommended last time.

[0039] Figure 2 2 is a schematic diagram of an application scenario applicable to the embodiment of the present disclosure. In the application scenario, the devices involved may include a resource recommendation device 201 and a terminal 202. The resource recommendation device 201 may be a terminal or a server. Figure 2 Take the resource recommendation device 201 as an example, which is a server.

[0040] A user can request resource recommendations on terminal 202. Terminal 202 sends the resource recommendation request to resource recommendation device 201. In response to the resource recommendation request, resource recommendation device 201 filters resources from the candidate set. After determining a recommended resource, resource recommendation device 201 sends the recommended resource to terminal 202, which then displays the recommended resource to the user.

[0041] Optionally, the above application scenario may further include a resource storage device 203 for storing the resource candidate set and for storing the second recommended resource and the recommendation score corresponding to the second recommended resource. The resource storage device 203 may be a storage device on the resource recommendation device 201 or an independent storage device. Figure 2 Taking the resource storage device 203 as a server as an example, the resource storage device 203 and the resource recommendation device 201 can communicate data.

[0042] Alternatively, the above application scenario can be a recommendation scenario for videos, news, or online products. For example, when a user watches a video on terminal 202, resource recommendation device 201 can recommend six videos to terminal 202 at a time. After the user scrolls to the last recommended video and then continues to scroll to the next video, a video recommendation request is triggered. Resource recommendation device 201 responds to the video recommendation request, selects a recommended video from the candidate set of videos, and sends the recommended video to terminal 202.

[0043] Optionally, the above application scenario may be a search scenario, and the resource recommendation request may include a search keyword. The resource recommendation device 201 may filter out recommended resources from candidate resources based on the search keyword and resource recommendation technology, and send the recommended resources to the terminal 202.

[0044] The embodiments of the present disclosure may be implemented on a terminal or server. The terminal may be a personal digital assistant (PDA), a handheld device with wireless communication capabilities (e.g., a smartphone or tablet), a computing device (e.g., a personal computer (PC)), a wearable device (e.g., a smartwatch or smart bracelet), or a smart home device (e.g., a smart speaker or smart display device). The server may be a standalone server or a server cluster, and may be a local server or a cloud server.

[0045] The following specific embodiments describe in detail the technical solutions of the present disclosure and how the technical solutions of the present disclosure solve the above-mentioned technical problems. The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The following embodiments of the present disclosure are described in conjunction with the accompanying drawings.

[0046] Figure 3 Schematic diagram of the first embodiment of the present disclosure. Figure 3 As shown, the resource recommendation method provided by the first embodiment of the present disclosure includes:

[0047] S301: Receive a resource recommendation request from a user.

[0048] The resource recommendation request is used to request recommended resources.

[0049] In different scenarios, the resource recommendation request may be different requests. For example, in a resource refresh scenario, the resource recommendation request may be a resource refresh request. In another example, in a resource search scenario, the resource recommendation request may be a resource search request.

[0050] In this embodiment, the user may trigger a resource recommendation request by performing interactive operations such as resource refresh and resource search on the terminal. Therefore, the current device may receive the resource recommendation request from the terminal.

[0051] S302 : In response to a resource recommendation request, obtain a first recommended resource and a recommendation score corresponding to the first recommended resource from a candidate resource set according to a first resource quantity threshold.

[0052] The first resource quantity threshold is used to constrain the quantity of resources screened from the candidate resource set.

[0053] There may be multiple first recommended resources, each of which has a corresponding recommendation score. The recommendation score corresponding to a first recommended resource reflects the degree to which the first recommended resource is suitable for recommendation to the user. For example, the higher the recommendation score corresponding to the first recommended resource, the more suitable it is for recommendation to the user.

[0054] In this embodiment, in response to a resource recommendation request, one or more stages of resource screening can be performed on the candidate resource set. During the resource screening process, the number of resources screened out is constrained by a first resource quantity threshold, and finally the first recommended resource and the recommendation score corresponding to the first recommended resource are obtained.

[0055] S303: Obtain a second recommended resource and a recommendation score corresponding to the second recommended resource, where the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, and the second resource quantity threshold is greater than the first resource quantity threshold.

[0056] There may be multiple second recommended resources, each with a corresponding recommendation score. The recommendation score corresponding to a second recommended resource reflects the degree to which the second recommended resource is suitable for recommendation to the user. For example, the higher the recommendation score corresponding to the second recommended resource, the more suitable it is for recommendation to the user.

[0057] The second resource quantity threshold is used to constrain the quantity of resources screened from the candidate resource set.

[0058] The second recommended resource and the corresponding recommendation score for the second recommended resource are obtained by filtering the candidate resource set according to the second resource quantity threshold before the first recommended resource and the corresponding recommendation score for the first recommended resource are obtained by filtering the candidate resource set according to the first resource quantity threshold. After the second recommended resource and the corresponding recommendation score for the second recommended resource are obtained by filtering, the second recommended resource and the corresponding recommendation score for the second recommended resource can be stored. Thus, after receiving a resource recommendation request from a user, the recommendation score for the second recommended resource and the corresponding recommendation score for the second recommended resource can be directly obtained, preventing the filtering of the second recommended resource from affecting the efficiency of resource recommendations.

[0059] Since the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to the second resource quantity threshold, and the first recommended resource and the recommendation score corresponding to the first recommended resource are not obtained by screening the candidate resource set according to the first resource quantity threshold at the same time, the candidate resource set may change. For example, the number of resources in the candidate resource set when screening the first recommended resource may be greater than the resource data in the candidate resource set when screening the second recommended resource. For another example, the number of resources has not changed, but the resources in the candidate resource set are different. This will, to a certain extent, make the second recommended resource different from the first recommended resource. Of course, if the time of screening the second recommended resource is relatively close to the time of screening the first recommended resource, then the candidate resource set may not change, but this does not affect the difference between the second recommended resource and the first recommended resource, because the second recommended resource is obtained by screening under the constraint of the second resource quantity threshold, and the second resource quantity threshold is greater than the first resource quantity threshold, resulting in the resources screened from the candidate resource set according to the second resource data threshold being different from the resources screened from the candidate resource set according to the first resource quantity threshold, that is, the second recommended resource is different from the first recommended resource.

[0060] The second recommended resource being different from the first recommended resource may include that the second recommended resource and the first recommended resource are different resources.

[0061] In this embodiment, the second recommended resource and the recommendation score corresponding to the second recommended resource may be obtained from a storage space (such as a database or a cache space).

[0062] S304: Determine a target recommended resource from the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource.

[0063] In this embodiment, since the number of resources ultimately recommended is limited, the target recommended resource may be selected from the first recommended resource and the second recommended resource based on the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource.

[0064] S305: Recommend target resources to the user.

[0065] In this embodiment, after the target recommended resource is determined, the target recommended resource may be sent to the terminal, and the terminal may display the target recommended resource, thereby recommending the target recommended resource to the user at the terminal.

[0066] In the disclosed embodiment, the first recommended resource and the second recommended resource are different due to the difference between the first resource quantity threshold and the second resource quantity threshold. When the current recommendation is close to the time of the historical recommendation, due to the small change in the candidate resource set, the first recommended resource may have a high duplication rate with the historically recommended resource. The second recommended resource, which is different from the first recommended resource, can improve the resource richness and explore more resources. Therefore, the combination of the first recommended resource and the second recommended resource can effectively reduce the duplication rate between the current recommended resource and the historically recommended resource while ensuring the accuracy of the resource recommendation.

[0067] In some embodiments, the process of obtaining a first recommended resource and a recommendation score corresponding to the first recommended resource from a candidate resource set, and the process of obtaining a second recommended resource and a recommendation score corresponding to the second recommended resource from a candidate resource set, can both include multiple screening stages. The first resource quantity threshold and the second resource quantity threshold can include a resource quantity threshold corresponding to at least one screening stage in the multiple screening stages. Thus, the first recommended resource and the second recommended resource are made different by having the resource quantity threshold corresponding to at least one screening stage in the screening process of the second recommended resource be greater than the resource quantity threshold corresponding to at least one screening stage in the screening process of the first recommended resource.

[0068] Optionally, the multiple screening stages include a recall stage and a ranking stage.

[0069] Based on multiple screening stages including recall stage and sorting stage, the following implementation methods can be provided:

[0070] In one possible implementation, the first resource quantity threshold includes a resource quantity threshold corresponding to the first recall and a resource quantity threshold corresponding to the first sorting, and the second resource quantity threshold includes a resource quantity threshold corresponding to the second recall and a resource quantity threshold corresponding to the second sorting. The second resource quantity threshold being greater than the first resource quantity threshold includes: the resource quantity threshold corresponding to the second recall is greater than the resource quantity threshold corresponding to the first recall, and / or the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting. Thus, a larger resource quantity threshold is set in the recall stage of screening the second recommended resource, and / or a larger resource quantity threshold is set in the sorting stage of screening the second recommended resource, so that the second recommended resource is different from the first recommended resource.

[0071] Among them, the first recall and the second recall are the recall stages in multiple screening stages, and the first sorting and the second sorting are the sorting stages in multiple screening stages.

[0072] In this implementation, in response to a resource recommendation request, a first recall can be performed on the candidate resource set according to the resource quantity threshold corresponding to the first recall, to obtain the first recalled resource and the recommendation score corresponding to the first recalled resource; a first sorting can be performed on the first recalled resource according to the recommendation score corresponding to the first recalled resource, to obtain the sorted first recalled resource; a first recommended resource can be selected from the sorted first recalled resource according to the resource quantity threshold corresponding to the first sorting, to obtain the first recommended resource and the recommendation score corresponding to the first recommended resource. In the process of screening and obtaining the second recommended resource and the second recommended resource, a second recall can be performed on the candidate resource set according to the resource quantity threshold corresponding to the second recall, to obtain the second recalled resource and the recommendation score corresponding to the second recalled resource; a second sorting can be performed on the second recalled resource according to the recommendation score corresponding to the second recalled resource, to obtain the sorted second recalled resource; a second recommended resource can be selected from the sorted second recalled resource according to the resource quantity threshold corresponding to the second sorting, to obtain the recommendation score corresponding to the second recommended resource and the second recommended resource.

[0073] It can be seen that if the resource quantity threshold corresponding to the second recall is greater than the resource quantity threshold corresponding to the first recall, then the number of second-recall resources is greater than the number of first-recall resources, and the second-recall resources contain resources different from the first-recall resources, and thus the second-recommended resources contain resources different from the first-recommended resources; if the resource quantity threshold corresponding to the second sort is greater than the resource quantity threshold corresponding to the first sort, the number of second-recommended resources is greater than the first-recommended resources, and the second-recommended resources contain resources different from the first-recommended resources. In other words, both can make the second-recommended resources different from the first-recommended resources.

[0074] Optionally, the sorting stage may include one or more sub-sorting stages. The number of sub-sorting stages in the first sorting stage may be the same as or different from the number of sub-sorting stages in the second sorting stage.

[0075] Optionally, the first sorting includes at least one sub-sorting, the second sorting includes at least one sub-sorting, and the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting, including: the resource quantity threshold corresponding to at least one sub-sorting in the second sorting is greater than the resource quantity threshold corresponding to the corresponding sub-sorting in the first sorting. Thus, a larger resource quantity threshold is set during the at least one sub-sorting stage of screening the second recommended resources, so that the second recommended resources are different from the first recommended resources.

[0076] In the case where the first sorting and the second sorting include the same number of sub-sortings, the sub-sortings in the first sorting correspond one-to-one to the sub-sortings in the second sorting.

[0077] For example, the first sorting includes coarse sorting and fine sorting, and the second sorting also includes coarse sorting and fine sorting. The resource quantity threshold corresponding to the coarse sorting in the second sorting may be greater than the resource quantity threshold corresponding to the coarse sorting in the first sorting, and / or, the resource quantity threshold corresponding to the fine sorting in the second sorting may be greater than the resource quantity threshold corresponding to the coarse sorting in the second sorting.

[0078] In the case where the number of sub-sorts included in the first sorting is different from the number of sub-sorts included in the second sorting, the number of sub-sorts included in the first sorting may be less than the number of sub-sorts included in the second sorting to ensure the efficiency of the first sorting. In this case, the sub-sorts in the first sorting and the sub-sorts in the second sorting may correspond one to one in order.

[0079] Furthermore, the first sorting includes one sub-sorting, and the second sorting includes two sub-sortings. For ease of distinction, the two sub-sortings in the second sorting are referred to as the first sub-sorting and the second sub-sorting. Based on the fact that the first sorting includes one sub-sorting and the second sorting includes two sub-sortings, the following implementation is provided:

[0080] In one possible implementation, the resource quantity threshold corresponding to the second ranking being greater than the resource quantity threshold corresponding to the first ranking may include: the resource quantity threshold corresponding to the first sub-ranking being greater than the resource quantity threshold corresponding to the first ranking. Thus, a larger resource quantity threshold is set during the first sub-ranking stage of screening the second recommended resources, such that the second recommended resources are different from the first recommended resources.

[0081] For example, the first sorting includes a sub-sorting which is a coarse sorting, that is, the first sorting is a coarse sorting; the first sub-sorting is the coarse sorting in the second sorting, and the second sub-sorting is the fine sorting in the second sorting; the resource quantity threshold corresponding to the first sub-sorting is greater than the resource quantity threshold corresponding to the first sorting, that is, the resource quantity threshold corresponding to the coarse sorting stage in the process of screening the second recommended resources is greater than the resource quantity threshold corresponding to the coarse sorting stage in the process of screening the first recommended resources.

[0082] In this implementation, in response to a resource recommendation request, a first recall can be performed on the candidate resource set according to the resource quantity threshold corresponding to the first recall, and the first recalled resources and the recommendation scores corresponding to the first recalled resources can be obtained; the first recalled resources can be sorted according to the recommendation scores corresponding to the first recalled resources to obtain the sorted first recalled resources; and the first recommended resource can be selected from the sorted first recalled resources according to the resource quantity threshold corresponding to the first sorting to obtain the first recommended resource and the recommendation scores corresponding to the first recommended resource. This process can be understood as a rough sorting process.

[0083] In this implementation process, in the process of screening and obtaining the second recommended resource and the second recommended resource, the candidate resource set can be secondly recalled according to the resource quantity threshold corresponding to the second recall to obtain the recommendation score corresponding to the second recalled resource; the second recalled resource is first sub-sorted according to the recommendation score corresponding to the second recalled resource to obtain the sorted second recalled resource, and the resource quantity threshold corresponding to the first sub-sort is selected from the sorted second recalled resource to obtain the first selected resource. This process can be understood as a rough sorting process; the first selected resource is sorted according to the recommendation score corresponding to the first selected resource to obtain the sorted first selected resource, and the resource quantity threshold corresponding to the second sub-sort is selected from the sorted first selected resource to obtain the second selected resource, that is, the second recommended resource. This process can be understood as a fine sorting process.

[0084] Figure 4 Schematic diagram of the second embodiment of the present disclosure. Figure 4 As shown, the resource recommendation method provided by the second embodiment of the present disclosure includes:

[0085] S401, receiving a resource recommendation request from a user;

[0086] S402 : In response to a resource recommendation request, obtain a first recommended resource and a recommendation score corresponding to the first recommended resource from a candidate resource set according to a first resource quantity threshold.

[0087] S403: Obtain a second recommended resource and a recommendation score corresponding to the second recommended resource, where the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, and the second resource quantity threshold is greater than the first resource quantity threshold.

[0088] The implementation principles and technical effects of S401 to S403 may refer to the aforementioned embodiments and will not be described in detail.

[0089] S404: Filter out a third recommended resource from the first recommended resource according to the recommendation score corresponding to the first recommended resource.

[0090] In this embodiment, the third recommended resources may be screened out from the first recommended resources according to the order of recommendation scores from high to low and the number of selected third recommended resources.

[0091] S405 : Filter out a fourth recommended resource from the second recommended resources according to the recommendation score corresponding to the second recommended resource.

[0092] In this embodiment, the fourth recommended resource may be selected from the second recommended resources according to the order of recommendation scores from high to low and the number of selected fourth recommended resources.

[0093] S406: Obtain a target recommended resource based on the third recommended resource and the fourth recommended resource.

[0094] In this embodiment, the target recommended resource may be determined from the third recommended resource and the fourth recommended resource.

[0095] For example, it is determined that the target recommended resources include the third recommended resource and the fourth recommended resource.

[0096] For another example, a target recommended resource is randomly selected from the third recommended resource and the fourth recommended resource.

[0097] In a possible implementation of S406, the third recommended resource and the fourth recommended resource are sorted, and a target recommended resource is selected from the third recommended resource and the fourth recommended resource according to the resource sorting order. In this way, the rationality of selecting the target recommended resource is improved, and the accuracy of resource recommendation is improved.

[0098] The process of selecting the first recommended resource includes a first recall and a first sorting. The first sorting can be understood as the rough sorting stage of the resource recommendation process, while sorting the third and fourth recommended resources can be understood as the fine sorting stage of the resource recommendation process. In the fine sorting stage, the third recommended resource from the first recommended resource and the fourth recommended resource from the second recommended resource are screened to obtain the target recommended resource, thus achieving the fusion of the first and second recommended resources.

[0099] In this implementation method, the recommendation score corresponding to the third recommended resource can be obtained from the recommendation score corresponding to the first recommended resource; the recommendation score corresponding to the fourth recommended resource can be obtained from the recommendation score corresponding to the second recommended resource; according to the recommendation score corresponding to the third recommended resource and the recommendation score corresponding to the fourth recommended resource, the third recommended resource and the fourth recommended resource are sorted in order of recommendation scores from high to low; then, according to the selected number of target recommended resources and the resource sorting order, the target recommended resource is selected from the sorted third recommended resource and the sorted fourth recommended resource.

[0100] Optionally, when the second sorting includes a first sub-sorting and a second sub-sorting, the resource quantity threshold corresponding to the second sub-sorting is greater than the selected quantity of target recommended resources, wherein the second sub-sorting can be understood as the resource quantity threshold corresponding to the fine sorting stage in the resource recommendation process of screening the second recommended resources, and the selected quantity of target recommended resources can be understood as the resource quantity threshold corresponding to the fine sorting stage in the resource recommendation process of screening the target recommended resources. Thus, a larger resource quantity threshold is set for the resource recommendation process of screening the second recommended resources in the fine sorting stage, so that the second recommended resources can provide more abundant resources for the selection of the target recommended resources, thereby reducing the resource duplication rate.

[0101] Optionally, in the process of obtaining the target recommended resource based on the third recommended resource and the fourth recommended resource, if there are duplicate recommended resources between the third recommended resource and the fourth recommended resource, the duplicate recommended resources are deleted from the third recommended resource or the fourth recommended resource, thereby avoiding duplicate recommended resources in the target recommended resource and improving user experience.

[0102] Furthermore, after deleting duplicate recommended resources from the third recommended resources or the fourth recommended resources, the next batch of recommended resources can be screened out from the second recommended resources. The next batch of recommended resources is used to supplement the third recommended resources and the fourth recommended resources to provide the target recommended resources with sufficiently rich resources for selection, thereby reducing the duplication rate between the target recommended resources and historical recommended resources.

[0103] Furthermore, if N duplicate recommended resources are deleted from the third or fourth recommended resources, N recommended resources can be selected from the remaining resources in the second recommended resources excluding the fourth recommended resource, i.e., the next batch of recommended resources will be N recommended resources, where N is greater than or equal to 1. This will make up for the number of resources lost due to deduplication in the third and fourth recommended resources.

[0104] As an example, Figure 5 This is an example of a resource recommendation. Figure 2 .like Figure 5As shown, in normal refresh, the stages of resource recommendation include recall, rough sorting, fine sorting, and exposure. The corresponding orders of magnitude for these stages are tens of thousands, thousands, hundreds, and individuals respectively; in large refresh, the stages of resource recommendation include recall, rough sorting, fine sorting, and caching. The corresponding orders of magnitude for these stages are hundreds of thousands, tens of thousands, thousands, and thousands respectively. It can be seen that in recall, rough sorting, and fine sorting, large refresh can obtain more and richer resources than normal refresh; in large refresh, the resources obtained after fine sorting are cached, and when the user requests a resource refresh, a normal refresh can be triggered. Based on the resources obtained by fine sorting in the normal refresh and the resources obtained by fine sorting in the large refresh in the cache, the resources to be exposed are determined, and finally the resources are exposed. It can be seen that the use of large refresh provides more resources for normal refresh and reduces the duplication rate of resource exposure.

[0105] in, Figure 5 The big refresh in the process is equivalent to the resource recommendation process of screening out the second candidate resource from the candidate resource set and the recommendation score corresponding to the second candidate resource. For example, the resource quantity threshold corresponding to the second recall is 100,000, the resource quantity threshold corresponding to the first sub-sort is 10,000, and the resource quantity threshold corresponding to the second sub-sort is 1,000; Figure 5 The recall and rough ranking of small and medium refreshes are equivalent to the process of selecting the first candidate resource and the recommendation score corresponding to the first candidate resource from the candidate resource set. For example, the resource quantity threshold corresponding to the first recall is 10,000, and the resource quantity threshold corresponding to the first ranking is 1,000. Figure 5 The precise sorting of small and medium refreshes is equivalent to the process of sorting and screening the third and fourth recommended resources in the aforementioned embodiment. The magnitude of the precise sorting of small refreshes is "pieces", which can be understood as the number of selected target recommended resources can be 10, 6, and other single digits.

[0106] by Figure 5 Taking the resource refresh shown as an example, the 50 resources with the highest scores can be obtained from the cached results of the large refresh (equivalent to selecting the 50 resources with the highest recommendation scores from the second recommended resources as the fourth recommended resources) and sent to the fine sorting layer of the ordinary refresh. The ordinary refresh will also obtain the 50 resources with the highest scores from the coarse sorting layer (equivalent to selecting the 50 resources with the highest recommendation scores from the first recommended resources as the third recommended resources). In this way, 100 resources are gathered together, and these 100 resources are finely sorted in the ordinary refresh to obtain the K exposures with the highest scores (K is 6, for example, which is equivalent to taking K target recommended resources).

[0107] If there are 10 duplicate resources between the 50 resources selected from the cached results of the major refresh and the 50 resources selected from the normal refresh, you can delete the 10 duplicate resources from the 50 resources selected from the cached results of the major refresh or the 50 resources selected from the normal refresh, and then select 10 more resources from the cached results of the major refresh to make up 100 resources again.

[0108] S407: Recommend target resources to the user.

[0109] The implementation principle and technical effects of S407 may refer to the aforementioned embodiments and will not be described in detail.

[0110] In the disclosed embodiment, by differentiating the first resource quantity threshold from the second resource quantity threshold, the first recommended resource and the second recommended resource are different. Combining the first and second recommended resources effectively reduces the duplication rate between the currently recommended resource and the previously recommended resources while ensuring the accuracy of the resource recommendation. Furthermore, in the process of combining the first and second recommended resources, the accuracy of the resource recommendation is improved through resource screening and resource sorting.

[0111] It can be seen from the above embodiments that the screening of the first recommended resource from the candidate resource set is triggered by a resource recommendation request, while the screening of the second recommended resource from the candidate resource set is not triggered by a resource recommendation request. The frequency of screening the second recommended resource from the candidate resource set can be lower than the frequency of screening the first recommended resource from the candidate resource set to avoid the screening of the second recommended resource occupying too many resources.

[0112] The following provides an implementation method for triggering the selection of the second recommended resource from the candidate resource set:

[0113] In one possible implementation, the second recommended resource and the recommendation score corresponding to the second recommended resource can be obtained by screening the candidate resource set according to a set period and the second resource quantity threshold. Thus, the second recommended resource and the recommendation score corresponding to the second recommended resource can be periodically updated.

[0114] In one possible implementation, in response to a user's resource duplication feedback message, a second recommended resource and its corresponding recommendation score can be obtained by screening from the candidate resource set according to a second resource quantity threshold. The resource duplication feedback message indicates that adjacent recommended resources are duplicated. Thus, based on the user's resource duplication feedback, the second recommended resource can be screened in a timely manner, thereby promptly providing an alternative second recommended resource for the target recommended resource in addition to the first recommended resource, thereby effectively and timely reducing the duplication rate of recommended resources.

[0115] In one possible implementation, in response to the user characteristics of the user satisfying the screening trigger condition of the second recommended resource, according to the second resource quantity threshold, the second recommended resource and the recommendation score corresponding to the second recommended resource are screened from the candidate resource set according to the second resource quantity threshold.

[0116] Among them, the user's user characteristics meet the screening trigger conditions for the second recommended resources, indicating that the user is prone to resource duplication when obtaining recommended resources. Screening the second recommended resources for such users can effectively reduce the duplication rate between the current recommended resources and the historical recommended resources.

[0117] In this implementation method, the user characteristics of the user can be obtained and compared with the screening trigger conditions of the second recommended resource. If the user characteristics of the user meet the screening trigger conditions of the second recommended resource, the second recommended resource and the recommendation score corresponding to the second recommended resource are screened from the candidate resource set according to the second resource quantity threshold.

[0118] Furthermore, the screening trigger conditions for the second recommended resource include at least one of the following: the cumulative number of times the user requests resource refresh is greater than the number threshold, the cumulative time the user browses the resource is greater than the time threshold, and the user's user tag includes the trigger tag of the second recommended resource.

[0119] In this implementation, if the cumulative number of times a user requests a resource refresh is greater than the number threshold, it indicates, to a certain extent, that the user frequently refreshes resources, and it is likely that the current recommended resource will be repeated with a historically recommended resource, thus triggering the screening of the second recommended resource. If the cumulative duration of the user's resource browsing is greater than the duration threshold, and the longer the cumulative time, the more times the user refreshes resources, and if the cumulative duration of the user's resource browsing is greater than the duration threshold, it can be considered that the user frequently refreshes resources, and it is likely that the current recommended resource will be repeated with a historically recommended resource, thus triggering the screening of the second recommended resource. If the user's user tag includes the trigger tag of the second recommended resource, it indicates that the user frequently refreshes resources or has the habit of browsing resources for a long time, and it is likely that the current recommended resource will be repeated with a historically recommended resource, thus triggering the screening of the second recommended resource. Thus, based on these screening trigger conditions, the rationality of triggering the screening of the second recommended resource can be improved, and the duplication rate between the current recommended resource and the historically recommended resource can be effectively reduced.

[0120] Furthermore, after the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening from the candidate resource set according to the second resource quantity threshold, the cumulative number of times the user requests a resource refresh, and / or the cumulative duration of the user's resource browsing can be cleared. Afterwards, the number of times the user requests a resource refresh and / or the duration of the user's resource browsing can continue to be accumulated. When the cumulative number of times the user requests a resource refresh is greater than the number threshold again, or the cumulative duration of the user's resource browsing is greater than the duration threshold again, the second recommended resource and the recommendation score corresponding to the second recommended resource can be screened from the candidate resource set again according to the second resource quantity threshold. In this way, whenever the user frequently refreshes the resource, the screening trigger condition for the second recommended resource is triggered, thereby reducing the repetition rate between the current recommended resource and the historical recommended resource when the user frequently refreshes the resource.

[0121] In some embodiments, if the user's resource browsing behavior in the past meets the conditions of the trigger tag of the second recommended resource, a trigger tag is added for the user. Thus, the trigger tag is added based on the user's resource browsing behavior in the past, improving the accuracy of adding trigger tags for the user.

[0122] For example, if the resource refresh duration of a user in the past is greater than a duration threshold, a trigger tag is added to the user; for another example, if the number of resource refreshes by a user in the past is greater than a number threshold, a trigger tag is added to the user; for another example, if a user browses resources in a fixed time period within T consecutive days, a trigger tag is added to the user, where T is a constant.

[0123] Figure 6 Schematic diagram of the third embodiment of the present disclosure. Figure 6 As shown, the resource recommendation device 600 provided in the third embodiment of the present disclosure includes:

[0124] Receiving unit 601, configured to receive a resource recommendation request from a user;

[0125] A first screening unit 602 is configured to, in response to a resource recommendation request, screen a candidate resource set according to a first resource quantity threshold to obtain a first recommended resource and a recommendation score corresponding to the first recommended resource;

[0126] A resource acquisition unit 603 is configured to acquire a second recommended resource and a recommendation score corresponding to the second recommended resource, where the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, where the second resource quantity threshold is greater than the first resource quantity threshold;

[0127] A resource determining unit 604 is configured to determine a target recommended resource from the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource;

[0128] The resource recommendation unit 605 is configured to recommend target resources to the user.

[0129] In some embodiments, the first resource quantity threshold includes the resource quantity threshold corresponding to the first recall and the resource quantity threshold corresponding to the first sort, the second resource quantity threshold includes the resource quantity threshold corresponding to the second recall and the resource quantity threshold corresponding to the second sort, and the second resource quantity threshold is greater than the first resource quantity threshold, including: the resource quantity threshold corresponding to the second recall is greater than the resource quantity threshold corresponding to the first recall, and / or, the resource quantity threshold corresponding to the second sort is greater than the resource quantity threshold corresponding to the first sort.

[0130] In some embodiments, the first sorting includes at least one sub-sorting, the second sorting includes at least one sub-sorting, and the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting, including: the resource quantity threshold corresponding to at least one sub-sorting in the second sorting is greater than the resource quantity threshold corresponding to the corresponding sub-sorting in the first sorting.

[0131] In some embodiments, the resource determination unit 604 includes: a first screening module (not shown in the figure), used to screen out a third recommended resource from the first recommended resource according to the recommendation score corresponding to the first recommended resource; a second screening module (not shown in the figure), used to screen out a fourth recommended resource from the second recommended resource according to the recommendation score corresponding to the second recommended resource; and a resource determination module (not shown in the figure), used to obtain a target recommended resource based on the third recommended resource and the fourth recommended resource.

[0132] In some embodiments, the resource determination module includes: a sorting submodule (not shown in the figure), which is used to sort the third recommended resource and the fourth recommended resource; and a selection submodule (not shown in the figure), which is used to select the target recommended resource from the third recommended resource and the fourth recommended resource according to the resource sorting order.

[0133] In some embodiments, the resource determination unit 604 further includes: a resource deduplication module (not shown in the figure), which is used to delete the duplicate recommended resources from the third recommended resource or the fourth recommended resource if there are duplicate recommended resources between the third recommended resource and the fourth recommended resource.

[0134] In some embodiments, the resource determination unit 604 also includes: a third screening module (not shown in the figure), which is used to screen out the next batch of recommended resources from the second recommended resources based on the number of deleted duplicate recommended resources and the recommendation score corresponding to the second recommended resource, and the next batch of recommended resources are used to supplement the third recommended resources and the fourth recommended resources.

[0135] In some embodiments, the resource recommendation device 600 further includes: a second screening unit (not shown in the figure), which is used to respond to the user characteristics of the user satisfying the screening trigger condition of the second recommended resource, and to screen the second recommended resource and the recommendation score corresponding to the second recommended resource from the candidate resource set according to the second resource quantity threshold.

[0136] In some embodiments, the screening trigger conditions for the second recommended resource include at least one of the following: the cumulative number of times the user requests resource refresh is greater than a number threshold; the cumulative time the user browses the resource is greater than a time threshold; the user's user tag includes a trigger tag for the second recommended resource.

[0137] In some embodiments, the resource recommendation device 600 further includes a clearing unit (not shown in the figure) configured to clear the accumulated times and / or the accumulated duration.

[0138] In some embodiments, the resource recommendation apparatus 600 further includes: a tag adding unit (not shown in the figure) configured to add a trigger tag to the user if the user's resource browsing behavior in the past meets the conditions for adding a trigger tag.

[0139] Figure 6 The resource recommendation device provided can execute the above-mentioned corresponding method embodiments, and its implementation principles and technical effects are similar, which will not be repeated here.

[0140] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the solution provided by any of the above embodiments.

[0141] According to an embodiment of the present disclosure, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the solution provided by any of the above embodiments.

[0142] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0143] Figure 77 is a schematic block diagram of an example electronic device 700 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 can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0144] like Figure 7 As shown, the electronic device 700 includes a computing unit 701, which can store data in a read-only memory (ROM) ( Figure 7 A computer program in ROM 702 is loaded from storage unit 708 into random access memory (RAM) ( Figure 7 The computer programs in the RAM 703 are used as an example to perform various appropriate actions and processes. The RAM 703 may also store various programs and data required for the operation of the electronic device 700. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. The input / output (I / O) interface ( Figure 7 The I / O interface 705 (for example) is also connected to the bus 704 .

[0145] Multiple components in the electronic device 700 are connected to the I / O interface 705, including an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a magnetic disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the electronic device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0146] The computing unit 701 can be various general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above, such as the resource recommendation method. For example, in some embodiments, the resource recommendation method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the resource recommendation method described above can be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to execute the resource recommendation method in any other appropriate manner (for example, by means of firmware).

[0147] Various embodiments of the systems and techniques described above 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), system on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0148] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0149] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0150] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the 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 acoustic input, voice input, or tactile input).

[0151] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0152] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. This client-server relationship is established by computer programs running on the respective computers, establishing a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosts and VPS services ("Virtual Private Servers" or simply "VPS"). The server may also be a server in a distributed system or a server integrated with blockchain.

[0153] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0154] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A resource recommendation method, comprising: Receive resource recommendation requests from users; In response to the resource recommendation request, screening a first recommended resource and a recommendation score corresponding to the first recommended resource from a candidate resource set according to a first resource quantity threshold; Obtaining a second recommended resource and a recommendation score corresponding to the second recommended resource, where the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, where the second resource quantity threshold is greater than the first resource quantity threshold; determining a target recommended resource from the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource; recommending the target recommended resource to the user; The method further comprises: In response to the user characteristics of the user satisfying the screening trigger conditions of the second recommended resource, the second recommended resource and the recommendation score corresponding to the second recommended resource are screened from the candidate resource set according to the second resource quantity threshold, wherein the screening trigger conditions of the second recommended resource include at least one of the following: the cumulative number of times the user requests resource refresh is greater than the number threshold, the cumulative time the user browses resources is greater than the time threshold, and the user tag of the user includes the trigger tag of the second recommended resource, wherein the user tag including the trigger tag of the second recommended resource indicates that the user is the type that frequently refreshes resources or has the habit of browsing resources for a long time.

2. The resource recommendation method according to claim 1, wherein: The first resource quantity threshold includes a resource quantity threshold corresponding to a first recall and a resource quantity threshold corresponding to a first sort; the second resource quantity threshold includes a resource quantity threshold corresponding to a second recall and a resource quantity threshold corresponding to a second sort; and the second resource quantity threshold being greater than the first resource quantity threshold includes: The resource quantity threshold corresponding to the second recall is greater than the resource quantity threshold corresponding to the first recall, and / or the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting.

3. The resource recommendation method according to claim 2, wherein: The first sorting includes at least one sub-sorting, the second sorting includes at least one sub-sorting, and the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting, including: A resource quantity threshold corresponding to at least one sub-sequence in the second sequence is greater than a resource quantity threshold corresponding to a corresponding sub-sequence in the first sequence.

4. The resource recommendation method according to any one of claims 1 to 3, wherein: The determining a target recommended resource from the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource includes: Filtering a third recommended resource from the first recommended resources according to the recommendation score corresponding to the first recommended resource; Filtering a fourth recommended resource from the second recommended resources according to the recommendation score corresponding to the second recommended resource; The target recommended resource is obtained according to the third recommended resource and the fourth recommended resource.

5. The resource recommendation method according to claim 4, wherein: Obtaining the target recommended resource according to the third recommended resource and the fourth recommended resource includes: sorting the third recommended resource and the fourth recommended resource; The target recommended resource is selected from the third recommended resource and the fourth recommended resource according to the resource sorting order.

6. The resource recommendation method according to claim 4, before obtaining the target recommended resource based on the third recommended resource and the fourth recommended resource, further comprising: If there is a duplicate recommended resource between the third recommended resource and the fourth recommended resource, the duplicate recommended resource is deleted from the third recommended resource or the fourth recommended resource.

7. The resource recommendation method according to claim 6, further comprising: after deleting the duplicate recommended resource from the third recommended resource or the fourth recommended resource; According to the number of deleted duplicate recommended resources and the recommendation score corresponding to the second recommended resource, the next batch of recommended resources is screened out from the second recommended resources, and the next batch of recommended resources is used to supplement the third recommended resource and the fourth recommended resource.

8. The resource recommendation method according to claim 1, after screening the candidate resource set according to the second resource quantity threshold to obtain the second recommended resource and the recommendation score corresponding to the second recommended resource, further comprising: When the screening trigger condition includes that the cumulative number of times the user requests resource refresh is greater than a number threshold, the cumulative number is cleared; When the cumulative duration of resource browsing of the user included in the screening trigger condition is greater than a duration threshold, the cumulative duration is cleared.

9. The resource recommendation method according to claim 1, further comprising: If the resource browsing behavior of the user in the past meets the conditions for adding the trigger tag, the trigger tag is added for the user.

10. A resource recommendation device, comprising: A receiving unit, configured to receive a resource recommendation request from a user; a first screening unit, configured to, in response to the resource recommendation request, screen a candidate resource set according to a first resource quantity threshold to obtain a first recommended resource and a recommendation score corresponding to the first recommended resource; a resource acquisition unit, configured to acquire a second recommended resource and a recommendation score corresponding to the second recommended resource, wherein the second recommended resource and the recommendation score corresponding to the second recommended resource are obtained by screening the candidate resource set according to a second resource quantity threshold, wherein the second resource quantity threshold is greater than the first resource quantity threshold; a resource determining unit, configured to determine a target recommended resource from among the first recommended resource and the second recommended resource according to the recommendation score corresponding to the first recommended resource and the recommendation score corresponding to the second recommended resource; A resource recommendation unit, configured to recommend the target recommended resource to the user; Also includes: a second screening unit configured to, in response to the user characteristic of the user satisfying a screening trigger condition for the second recommended resource, screen the candidate resource set according to a second resource quantity threshold to obtain the second recommended resource and a recommendation score corresponding to the second recommended resource; the screening trigger condition for the second recommended resource comprising at least one of the following: The cumulative number of times the user requests resource refresh is greater than a threshold number; The cumulative duration of resource browsing by the user is greater than the duration threshold; The user tag of the user includes a trigger tag of the second recommended resource; The fact that the user tag of the user includes the trigger tag of the second recommended resource indicates that the user is a type that frequently refreshes resources or has the habit of browsing resources for a long time.

11. The resource recommendation device according to claim 10, wherein: The first resource quantity threshold includes a resource quantity threshold corresponding to a first recall and a resource quantity threshold corresponding to a first sort; the second resource quantity threshold includes a resource quantity threshold corresponding to a second recall and a resource quantity threshold corresponding to a second sort; and the second resource quantity threshold being greater than the first resource quantity threshold includes: The resource quantity threshold corresponding to the second recall is greater than the resource quantity threshold corresponding to the first recall, and / or the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting.

12. The resource recommendation device according to claim 11, wherein: The first sorting includes at least one sub-sorting, the second sorting includes at least one sub-sorting, and the resource quantity threshold corresponding to the second sorting is greater than the resource quantity threshold corresponding to the first sorting, including: A resource quantity threshold corresponding to at least one sub-sequence in the second sequence is greater than a resource quantity threshold corresponding to a corresponding sub-sequence in the first sequence.

13. The resource recommendation device according to any one of claims 10 to 12, wherein: The resource determination unit includes: a first screening module, configured to screen out a third recommended resource from the first recommended resources according to a recommendation score corresponding to the first recommended resource; a second screening module, configured to screen out a fourth recommended resource from the second recommended resources according to the recommendation score corresponding to the second recommended resource; The resource determination module is configured to obtain the target recommended resource based on the third recommended resource and the fourth recommended resource.

14. The resource recommendation device according to claim 13, wherein: The resource determination module includes: a sorting submodule, configured to sort the third recommended resource and the fourth recommended resource; The selection submodule is configured to select the target recommended resource from the third recommended resource and the fourth recommended resource according to a resource sorting order.

15. The resource recommendation device according to claim 13, wherein the resource determination unit further comprises: The resource deduplication module is configured to delete the duplicate recommended resource from the third recommended resource or the fourth recommended resource if there is a duplicate recommended resource between the third recommended resource and the fourth recommended resource.

16. The resource recommendation device according to claim 15, wherein the resource determination unit further comprises: The third screening module is used to screen out the next batch of recommended resources from the second recommended resources based on the number of deleted duplicate recommended resources and the recommendation score corresponding to the second recommended resource, and the next batch of recommended resources are used to supplement the third recommended resources and the fourth recommended resources.

17. The resource recommendation device according to claim 10, further comprising: A clearing unit is used to clear the accumulated number of times when the filtering trigger condition includes that the accumulated number of times the user requests resource refresh is greater than a number threshold, and to clear the accumulated time when the filtering trigger condition includes that the accumulated time of the user's resource browsing is greater than a time threshold.

18. The resource recommendation device according to claim 10, further comprising: The tag adding unit is configured to add the trigger tag to the user if the resource browsing behavior of the user in the past meets the adding condition of the trigger tag.

19. 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 that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the resource recommendation method according to any one of claims 1 to 9.

20. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable the computer to execute the resource recommendation method according to any one of claims 1 to 9.

21. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the resource recommendation method according to any one of claims 1 to 9 are implemented.

Images