Recommended methods, apparatus, devices, computer-readable media, and program products

Abstract

This disclosure presents embodiments of a method, apparatus, device, computer-readable medium, and program product for recommending items. One specific implementation of the method includes: in response to receiving a target item recommendation request, for each piece of business item information in a business item information set, determining the corresponding item sampling information based on the posterior distribution of the item sampling information corresponding to the business item information; allocating the business item information set to corresponding business information based on the obtained item sampling information set, obtaining business item information groups for each piece of business information; generating at least one recommended item information for the target item recommendation request based on the obtained business item information group set; and sending the at least one recommended item information to the user terminal corresponding to the target item recommendation request. This implementation is related to artificial intelligence and can accurately and efficiently generate at least one recommended item information.

Description

Technical Field

[0001] Embodiments of this disclosure relate to the field of computer technology, and more specifically to methods, apparatus, devices, computer-readable media, and program products for recommending items. Background Technology

[0002] Currently, in item search systems, item recall and ranking are classic objective optimization problems. The common approach to generating recommended item information is through exploration and exploitation (EE) algorithms.

[0003] However, the inventors discovered that when using the above method to generate recommended item information, the following technical problems often arise:

[0004] The failure to recommend item information based on the actual efficiency and uncertainty of the items resulted in at least one item recommendation being inaccurate.

[0005] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0007] Some embodiments of this disclosure provide methods, apparatus, devices, computer-readable media, and program products for recommending items to address the technical problems mentioned in the background section above.

[0008] In a first aspect, some embodiments of this disclosure provide an item recommendation method, comprising: in response to receiving a target item recommendation request, for each piece of business item information in a business item information set, determining the item sampling information corresponding to the business item information based on the posterior distribution of the item sampling information corresponding to the business item information, wherein the posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business items corresponding to the business item information; allocating the business item information set to corresponding business information according to the obtained item sampling information set to obtain a business item information group for each business item information; generating at least one recommended item information for the target item recommendation request based on the obtained set of business item information groups; and sending the at least one recommended item information to the user terminal corresponding to the target item recommendation request.

[0009] Optionally, the above-mentioned allocation of corresponding business information to the above-mentioned business item information set based on the obtained item sampling information set to obtain business item information groups for each business information includes: determining the business information set to be recommended; determining the business item recommendation quota information corresponding to each business information in the above-mentioned business information set; for each business information in the above-mentioned business information set, performing the following business item information filtering steps: filtering out business item information that has a business relationship with the above-mentioned business information from the above-mentioned business item information set as target business item information, obtaining a subset of target business item information; according to the business item recommendation quota information corresponding to the above-mentioned business information, filtering out target business item information whose corresponding item sampling information meets the preset sampling information conditions from the above-mentioned target business item information subset, obtaining a business item information group.

[0010] Optionally, the posterior distribution of the item sampling information corresponding to each business item in the above business item information set is determined by the following steps: determining the business item level corresponding to the above business item information at the current time as the target business item level; determining the prior distribution of the item sampling information corresponding to the target business item level; and determining the posterior distribution of the above item sampling information based on the prior distribution of the above item sampling information.

[0011] Optionally, determining the posterior distribution of the item sampling information based on the prior distribution of the item sampling information includes: obtaining the item click-through rate information and item exposure information of the business item information within a predetermined time period; substituting the item click-through rate information and the item exposure information into the item sampling information posterior distribution formula for the prior distribution of the item sampling information to generate the item sampling information posterior distribution.

[0012] Optionally, determining the business item level corresponding to the business item information at the current time includes: for each business item information in the business item information set, performing the following generation steps: determining the previous time period corresponding to the current time as the target historical time period; determining the click-through rate and conversion rate corresponding to the business item information within the target historical time period; generating item efficiency information for the business item information based on the click-through rate and conversion rate; sorting the obtained item efficiency information set to obtain an item efficiency information sequence; and determining the business item level corresponding to the business item information based on the position of the item efficiency information corresponding to the business item information in the item efficiency information sequence.

[0013] Optionally, the above-mentioned method of generating at least one recommended item information for the target item recommendation request based on the obtained business item information set includes: obtaining a finely ranked item information set; inputting the finely ranked item information set and the business item information set into a user exploration score information generation model to generate a finely ranked exploration score information set for the finely ranked item information set and a business exploration score information set for the business item information set; determining the number of item recommendations for the target user; and, based on the finely ranked exploration score information set and the business exploration score information set, determining at least one item information from the finely ranked item information set and the business item information set as at least one item information to be recommended, wherein the number of item information corresponding to the at least one item information to be recommended is the same as the number of item recommendations.

[0014] Optionally, determining the number of items recommended for a target user includes: determining the user's item exploration intention information corresponding to the target user; and determining the number of items recommended based on the user's item exploration intention information.

[0015] Optionally, determining at least one item information from the refined ranking exploration score information set and the business exploration score information set as at least one item information to be recommended, based on the refined ranking exploration score information set and the business exploration score information set, includes: filtering refined ranking exploration score information with values ​​greater than the user's item exploration intention information from the refined ranking exploration score information set to obtain at least one refined ranking exploration score information; determining the difference between the number of refined ranking exploration score information included in the at least one refined ranking exploration score information and the number of recommended items; filtering business exploration score information sets whose corresponding score values ​​are among the top target number from the business exploration score information set, wherein the number of business exploration score information included in the business exploration score information set is the same as the difference; and combining the business item information set corresponding to the business exploration score information set and the at least one refined ranking item information corresponding to the at least one refined ranking exploration score information to generate at least one item information to be recommended.

[0016] Optionally, the above-mentioned business item information set is generated through the following steps: obtaining the offline dataset corresponding to the initial business item information set; performing a first item information filtering on the initial business item set based on the offline dataset to generate a filtered item information set; and removing item information whose corresponding exposure information meets preset exposure conditions from the filtered item information set based on the real-time dataset corresponding to the filtered item information set to obtain a removed item information set as the business item information set.

[0017] Optionally, the aforementioned initial business item information set includes: a self-driven mining business item information set and a business input item information set.

[0018] Secondly, some embodiments of this disclosure provide an item recommendation apparatus, comprising: a determining unit configured to, in response to receiving a target item recommendation request, determine, for each piece of business item information in a business item information set, item sampling information corresponding to the business item information based on the posterior distribution of item sampling information corresponding to the business item information, wherein the posterior distribution of item sampling information is determined based on the prior distribution of item sampling information associated with the hierarchical relationship of the business items corresponding to the business item information; an allocation unit configured to allocate corresponding business information to the business item information set according to the obtained item sampling information set, thereby obtaining a business item information group for each business item information; a generating unit configured to generate at least one recommended item information for the target item recommendation request based on the obtained business item information group set; and a sending unit configured to send the at least one recommended item information to the user terminal corresponding to the target item recommendation request.

[0019] Optionally, the allocation unit may be further configured to: determine a set of business information to be recommended; determine the business item recommendation quota information corresponding to each business information in the above business information set; for each business information in the above business information set, perform the following business item information filtering steps: filter out business item information that has a business relationship with the above business information from the above business item information set, as target business item information, to obtain a subset of target business item information; according to the business item recommendation quota information corresponding to the above business information, filter out target business item information whose corresponding item sampling information meets the preset sampling information conditions from the above target business item information subset, to obtain a group of business item information.

[0020] Optionally, the generation unit can be configured to: acquire a finely ranked item information set; input the finely ranked item information set and the business item information set into a user exploration score information generation model to generate a finely ranked exploration score information set for the finely ranked item information set and a business exploration score information set for the business item information set; determine the number of item recommendations for the target user; and, based on the finely ranked exploration score information set and the business exploration score information set, determine at least one item information from the finely ranked item information set and the business item information set as at least one item information to be recommended, wherein the number of item information corresponding to the at least one item information to be recommended is the same as the number of item recommendations.

[0021] Optionally, the generation unit can be configured to: determine the user item exploration intention information corresponding to the above target user; and determine the number of item recommendations based on the above user item exploration intention information.

[0022] Optionally, the generation unit can be configured to: filter out finely ranked exploration score information with values ​​greater than the user's item exploration intention information from the aforementioned finely ranked exploration score information set, to obtain at least one finely ranked exploration score information; determine the difference between the number of finely ranked exploration score information included in the aforementioned at least one finely ranked exploration score information and the number of recommended items; filter out a set of business exploration score information corresponding to the top target number of score sizes from the aforementioned business exploration score information set, wherein the number of business exploration score information included in the business exploration score information set is the same as the aforementioned difference; and combine the business item information set corresponding to the aforementioned business exploration score information set and at least one finely ranked item information corresponding to at least one finely ranked exploration score information to generate at least one item information to be recommended.

[0023] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0024] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0025] Fifthly, some embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0026] The above embodiments of this disclosure have the following beneficial effects: the item recommendation method of some embodiments of this disclosure can accurately and efficiently generate at least one recommended item information. Specifically, the reason why the related recommended item information is not accurate enough is that the item information recommendation fails to take into account the actual efficiency and uncertainty of the item, resulting in the inaccuracy of at least one recommended item information. Based on this, the item recommendation method of some embodiments of this disclosure firstly, in response to receiving a target item recommendation request, for each business item information in the business item information set, determines the item sampling information corresponding to the business item information according to the posterior distribution of the item sampling information corresponding to the business item information. The posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information. Here, determining the item sampling information corresponding to each business item information through the posterior distribution of the item sampling information fully considers the actual efficiency of the item and adds uncertainty, giving potentially high-quality items more exposure opportunities. Furthermore, the posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information. By dividing the business item information into business item hierarchical categories, the growth status of the items can be fully reflected. Therefore, by using the business item hierarchy, a precise prior distribution of item sampling information can be generated. Then, based on the obtained item sampling information set, the business item information set is precisely allocated to corresponding business information, resulting in business item information groups for each business item. Furthermore, based on the obtained business item information group set, at least one recommended item can be accurately generated for the target item recommendation request. Finally, the at least one recommended item is sent to the user terminal corresponding to the target item recommendation request. In summary, by using the posterior distribution of item sampling information corresponding to business item information, the precise generation of at least one recommended item can be achieved, enabling its transmission to the user terminal corresponding to the target item recommendation request. Attached Figure Description

[0027] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0028] Figure 1 This is a schematic diagram illustrating an application scenario of an item recommendation method according to some embodiments of the present disclosure;

[0029] Figure 2 This is a flowchart of some embodiments of the item recommendation method according to this disclosure;

[0030] Figure 3 This is a schematic diagram of the prior distribution of item sampling information associated with each business item level in some embodiments of the item recommendation method of this disclosure;

[0031] Figure 4 These are flowcharts of other embodiments of the item recommendation method according to this disclosure;

[0032] Figure 5 These are schematic diagrams illustrating the structure of some embodiments of the item recommendation device according to this disclosure;

[0033] Figure 6 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0034] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0035] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0036] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0037] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0038] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0039] Before performing any of the operations related to the collection, storage, and use of item information (such as business item information) involved in this disclosure, the relevant organizations or individuals shall fulfill their obligations, including conducting an item information security impact assessment, informing the item information subject, and obtaining the item information subject's prior authorization and consent.

[0040] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0041] Figure 1 This is a schematic diagram illustrating an application scenario of an item recommendation method according to some embodiments of the present disclosure.

[0042] exist Figure 1In this application scenario, firstly, in response to receiving a target item recommendation request, for each piece of business item information in the business item information set 102, the electronic device 101 can determine the item sampling information corresponding to the aforementioned business item information based on the posterior distribution of the item sampling information corresponding to the aforementioned business item information. The posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the aforementioned business item information. In this application scenario, the business item information set 102 may include: first business item information 1021, second business item information 1022, third business item information 1023, and fourth business item information 1024. First business item information 1021 may be a "washing machine." Second business item information 1022 may be "laundry detergent." Third business item information 1023 may be a "pen." Fourth business item information 1024 may be a "backpack." The posterior distribution of the item sampling information corresponding to the first business item information 1021 may be the first item sampling information posterior distribution 103. The posterior distribution of the item sampling information corresponding to the second business item information 1022 may be the second item sampling information posterior distribution 104. The posterior distribution of the item sampling information corresponding to the third business item information 1023 can be the posterior distribution 105 of the third item sampling information. The posterior distribution of the item sampling information corresponding to the fourth business item information 1024 can be the posterior distribution 106 of the fourth item sampling information. The item sampling information 1071 corresponding to the first business item information 1021 can be "76". The item sampling information 1072 corresponding to the second business item information 1022 can be "56". The item sampling information 1073 corresponding to the third business item information 1023 can be "95". The item sampling information 1074 corresponding to the fourth business item information 1024 can be "56". Then, the electronic device 101 can allocate corresponding business information to the above business item information set 102 according to the obtained item sampling information set 107, to obtain business item information groups for each business item information. In this application scenario, for the business information being laundry business information 111, the corresponding business item information group 109 can include: the first business item information 1021 and the second business item information 1022. For the business information 110 (laundry business information), the corresponding business item information group 110 may include: third business item information 1023 and fourth business item information 1024. Furthermore, the electronic device 101 can generate at least one recommended item information 113 for the aforementioned target item recommendation request based on the obtained business item information group set 108. In this application scenario, at least one recommended item information 113 may include: first business item information 1021 and third business item information 1022. Finally, the electronic device 101 can send the at least one recommended item information 113 to the user terminal 114 corresponding to the aforementioned target item recommendation request.

[0043] It should be noted that the aforementioned electronic device 101 can be either hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the electronic device is software, it can be installed in the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0044] It should be understood that Figure 1 The number of electronic devices shown is merely illustrative. Any number of electronic devices can be used depending on the implementation requirements.

[0045] Continue to refer to Figure 2 The diagram illustrates a flow 200 of some embodiments of an item recommendation method according to the present disclosure. This item recommendation method includes the following steps:

[0046] Step 201: In response to receiving the target item recommendation request, for each piece of business item information in the business item information set, determine the item sampling information corresponding to the business item information based on the posterior distribution of the item sampling information corresponding to the business item information.

[0047] In some embodiments, in response to receiving a target item recommendation request, for each business item in the business item information set, the execution entity of the above item recommendation method (e.g., Figure 1The electronic device 101 shown can determine the item sampling information corresponding to the above-mentioned business item information based on the posterior distribution of the item sampling information corresponding to the above-mentioned business item information. The target item recommendation request can be a recommendation request indicating recommended item information. Specifically, for a shopping scenario, the corresponding target item recommendation request is a recommendation request recommending a target product. The business item information set can be a pre-selected set of item information. Business item information can be the item information of a business item. Specifically, business item information can include, but is not limited to, at least one of the following: business item identifier, business item price. Business items can be items related to a business domain. Specifically, items related to a business domain can be items sold in the business domain. In practice, a business domain can be, but is not limited to, at least one of the following: home appliance business domain, fresh food business domain, electronic product business domain. The posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the business item hierarchy corresponding to the above-mentioned business item information. The posterior distribution of the item sampling information can characterize the distribution of the sampling scores of the business item information within the time period corresponding to the receiving time of the target item recommendation request. Specifically, the business item hierarchy can be the classification criteria for business items. For example, the business item hierarchy can be one of the following: first business item hierarchy, second business item hierarchy, and third business item hierarchy. The first business item hierarchy has a corresponding first business sales range. The second business item hierarchy has a corresponding second business sales range. The third business item hierarchy has a corresponding third business sales range. The values ​​in the first business sales range are less than the values ​​in the second business sales range. The values ​​in the second business sales range are less than the values ​​in the third business sales range. Therefore, by hierarchically dividing the items in the business item information set according to their sales volume, subsets of business item information at each hierarchy can be obtained. The prior distribution of item sampling information can characterize the normal sampling score distribution of business item information. It should be noted that each business item information has a corresponding business item hierarchy, and each business item has a corresponding prior distribution of item sampling information. Item sampling information can characterize the probability information of the item corresponding to the business item information being sampled. Specifically, item sampling information can be in fractional form or probabilistic form. Further details will not be elaborated. The prior distribution of item sampling information can be a Beta distribution. In practice, the prior distribution of item sampling information can be “X~Be(α,β)”. The expected value of the prior distribution of the item sampling information can be α / (α+β). The variance of the prior distribution of the item sampling information can be α*β / [(α+β)]. 2 (α+β+1). The prior parameters α and β determine the expected value and uncertainty of the distribution.

[0048] As an example, the posterior distribution of item sampling information can be generated through the following steps:

[0049] The first step is to obtain sales performance information for business items within the target time period. The target time period can be a pre-set time frame.

[0050] The second step involves inputting the prior distribution of the sales performance information and the item sampling information corresponding to the above business item information into a pre-trained convolutional neural network (CNN) model to output the posterior distribution of the item sampling information corresponding to the business item information.

[0051] As an example, the aforementioned execution entity can determine the item sampling information corresponding to the aforementioned business item information by using random numbers in the form of the posterior distribution of the item sampling information corresponding to the aforementioned business item information.

[0052] In some optional implementations of certain embodiments, the posterior distribution of the item sampling information corresponding to each business item information in the above-mentioned business item information set is determined through the following steps:

[0053] The first step is to determine the business item level corresponding to the above business item information at the current time, which will be used as the target business item level. The current time can be the time when the target item recommendation request was received.

[0054] As an example, the aforementioned executing entity can determine the sales volume information of the items corresponding to the aforementioned business item information within the previous 10 minutes corresponding to the current time. In response to determining that the sales volume information is between 0 and 100 items, the business item level corresponding to the business item information is set to the first business item level. In response to determining that the sales volume information is between 100 and 1000 items, the business item level corresponding to the business item information is set to the second business item level. In response to determining that the sales volume information is between 1000 and 10000 items, the business item level corresponding to the business item information is set to the third business item level. In response to determining that the sales volume information is greater than 10000 items, the business item level corresponding to the business item information is set to the fourth business item level.

[0055] The second step is to determine the prior distribution of item sampling information corresponding to the aforementioned target business item levels. Each business item level has a corresponding prior distribution of item sampling information.

[0056] As an example, firstly, the aforementioned executing entity can use relevant statistical methods based on historical datasets to generate a prior distribution of item sampling information corresponding to the target business item level.

[0057] The third step is to determine the posterior distribution of the above-mentioned item sampling information based on the prior distribution of the above-mentioned item sampling information.

[0058] As an example, firstly, the aforementioned executing entity can determine the sales volume of the target business item information in the historical data of the previous 10 minutes at the current time. Then, the sales volume information is input into the formula for the posterior distribution of the item sampling information based on the prior distribution of the aforementioned item sampling information to generate the posterior distribution of the aforementioned item sampling information.

[0059] Optionally, determining the posterior distribution of the item sampling information based on the prior distribution of the item sampling information may include the following steps:

[0060] The first step is to obtain the click-through rate (CTR) and exposure information of the aforementioned business items within a predetermined time period. The predetermined time can be a pre-set time period corresponding to the current time. For example, for a 24-hour day, the 24 hours can be divided into four phases: Phase 1, Phase 2, Phase 3, and Phase 4. Phase 1 corresponds to the time period "0:00-6:00". Phase 2 corresponds to the time period "6:00-12:00". Phase 3 corresponds to the time period "12:00-18:00". Phase 4 corresponds to the time period "18:00-24:00". Given that the current time is 14:00, the predetermined time is Phase 3. The click-through rate information can be the click-through rate (CTR) of the business item. The exposure information can be the conversion rate (CVR) of the business item information.

[0061] In practice, the click-through rate (CTR) of items within a predetermined time period can be the CTR from the initial time corresponding to the predetermined time period to the current time period. Similarly, the item exposure information within a predetermined time period can be the item exposure from the initial time corresponding to the predetermined time period to the current time period. The initial time corresponding to the predetermined time period can be the earliest time point within that time segment.

[0062] The second step involves substituting the aforementioned item click-through rate information and item exposure information into the posterior distribution formula for the item sampling information prior to the item sampling information, to generate the posterior distribution of the item sampling information. The posterior distribution formula for the item sampling information can be {Beta(α + item click-through rate information, β + item exposure information - item click-through rate information)}.

[0063] Optionally, determining the business item level corresponding to the aforementioned business item information at the current time may include the following steps:

[0064] The first step is to perform the following generation steps for each business item in the above business item information set:

[0065] Sub-step 1: Determine the previous time period corresponding to the current time as the target historical time period.

[0066] For example, a 24-hour day can be pre-divided into four periods: the first time period, the second time period, the third time period, and the fourth time period. The first time period corresponds to the time frame "0:00-6:00". The second time period corresponds to the time frame "6:00-12:00". The third time period corresponds to the time frame "12:00-18:00". The fourth time period corresponds to the time frame "18:00-24:00". If the current time is 14:00, then the target historical time period is the second time period.

[0067] Sub-step 2: Determine the click-through rate and conversion rate of the aforementioned business item information within the target historical time period.

[0068] As an example, the click-through rate and conversion rate of the aforementioned business items within the target historical time period are determined by querying click-through rate and conversion rate.

[0069] Sub-step 3: Based on the click-through rate and conversion rate mentioned above, generate item efficiency information for the aforementioned business item information. This item efficiency information can be item exploration efficiency information. Item efficiency information can characterize the effectiveness of exploring the corresponding item. Specifically, item efficiency information can be an item exploration efficiency score. The higher the item exploration efficiency score, the higher the growth in the item's attention.

[0070] As an example, firstly, the aforementioned execution entity can add the click-through rate to the first value to obtain a first summed value. Then, the aforementioned execution entity can add the conversion rate to the second value to obtain a second summed value. Next, the first target value and the second target value are added together to obtain the item efficiency information for the aforementioned business item information. Here, the first target value is a value with the first summed value as the base and the first search ranking refinement fusion parameter as the exponent. The second target value is a value with the second summed value as the base and the second search ranking refinement fusion parameter as the exponent. The first and second search ranking refinement fusion parameters are pre-set parameters.

[0071] The second step is to sort the obtained set of item efficiency information to obtain the item efficiency information sequence.

[0072] As an example, the obtained set of item efficiency information is sorted according to the order of item efficiency information from largest to smallest to obtain the item efficiency information sequence.

[0073] The third step is to determine the business item level corresponding to the above business item information based on the position of the item efficiency information corresponding to the above business item information in the above item efficiency information sequence.

[0074] As an example, firstly, determine the quantile position of the item efficiency information corresponding to the above business item information within the item efficiency information sequence. Then, when the quantile is between 0% and 33%, the corresponding business item level can be the first business item level. When the quantile is between 33% and 66%, the corresponding business item level can be the second business item level. When the quantile is between 66% and 100%, the corresponding business item level can be the third business item level.

[0075] It should be noted that, based on the different item efficiency information, the item information set within each time period can be divided into item information groups for each business item level. For the same time period, the more item information corresponding to a business item level, the lower the corresponding business item level, indicating that the probability of that item information being recommended is lower. Each business item level is divided based on the top 30% and top 60% cutoff points of item efficiency information.

[0076] For example, consider time periods such as "[0:00-6:00], [6:00-12:00], [18:00-24:00]". For the time period "[0:00-6:00]", the corresponding business item levels can include: a first business item level, a second business item level, and a third business item level. The item efficiency value of each item in the item information group corresponding to the first business item level is less than the item efficiency value of each item in the item information group corresponding to the second business item level. The item efficiency value of each item in the item information group corresponding to the second business item level is less than the item efficiency value of each item in the item information group corresponding to the third business item level. The number of items in the item information group corresponding to the first business item level is greater than the number of items in the item information group corresponding to the second business item level. The number of items in the item information group corresponding to the second business item level is greater than the number of items in the item information group corresponding to the third business item level. The item information group corresponding to the first business item level includes: first item information, second item information, third item information, and fourth item information. The item information group corresponding to the second business item level includes: first item information, second item information, third item information, and fourth item information.

[0077] Depending on the time period, the business item level corresponding to each item will change based on the item efficiency information of the previous time period. This change in the business item level reflects the item's growth potential. The newly deployed business item level is the target business item level. For item A, the business item level during [0:00-6:00] is the first business item level, and during [6:00-12:00] it is the second business item level, indicating increased popularity. Similarly, for item B, the business item level during [0:00-6:00] is the third business item level, and during [6:00-12:00] it is the second business item level, indicating decreased popularity. For initial item information (i.e., newly deployed item information), the initial business item level can be the second business item level.

[0078] To ensure that each item at each business item level has a certain probability of being recommended, a prior distribution of item sampling information is set for each business item level. The higher the business item level, the higher the expected value and variance of the prior distribution of the corresponding item sampling information.

[0079] See details Figure 3 The prior distribution of the item sampling information corresponding to the first business item level corresponds to distribution curve L1. The prior distribution of the item sampling information corresponding to the second business item level corresponds to distribution curve L2. The prior distribution of the item sampling information corresponding to the third business item level corresponds to distribution curve L3.

[0080] In some optional implementations of certain embodiments, the above-mentioned business item information set is generated through the following steps:

[0081] The first step is to obtain the offline dataset corresponding to the initial business item information set.

[0082] The initial business item information set can be the complete set of business item information. There is a one-to-one correspondence between the offline data in the offline dataset and the initial business item information in the initial business item information set. The offline data can be a set of offline item feature information representing the corresponding initial business item information.

[0083] The second step is to filter the initial business item set based on the offline dataset to generate a filtered item information set.

[0084] As an example, the aforementioned execution entity can control the offline product selection module to perform a first item information filter on the initial business item set based on the aforementioned offline dataset, thereby generating a filtered item information set. The offline product selection module can be a module that selects item information offline.

[0085] In practice, the offline product selection module can filter product information from all business product information sets based on various product selection strategies or related selection methods to generate an initial business product information set.

[0086] The third step involves removing item information that meets preset exposure conditions from the aforementioned filtered item information set, based on the corresponding real-time dataset. This results in a removed item information set, which serves as the business item information set. There is a one-to-one correspondence between the filtered item information in the filtered item information set and the real-time data in the aforementioned real-time dataset. The real-time data can be the current data of the filtered item information. The preset exposure conditions can include sub-conditions indicating that the exposure ratio of item information within the past 7 days is less than a predetermined ratio, and sub-conditions indicating that the cumulative exposure of item information is less than a predetermined value while the conversion efficiency meets a predetermined conversion condition. For example, the predetermined ratio can be a pre-defined percentage. The exposure ratio of item information can be the ratio between the exposure of the item information and the exposure of each item in the same category. Conversion efficiency can include item click-through rate and item conversion rate. The predetermined conversion condition can be that the item click-through rate is less than the target pass rate and the item conversion rate is less than the target conversion rate.

[0087] Specifically, the preset exposure conditions can be that the item information has "less than 10,000 exposures in the past 15 days, a click-through rate of less than 0.01 in the past 15 days, a conversion rate of less than 0.01 in the past 15 days, more than 1,000 exposures in the past day, a click-through rate of less than 0.001 in the past day, and a conversion rate of less than 0.008 in the past day".

[0088] Compared to the original exit method, this approach improves the liquidity and richness of products in the overall search domain without reducing the efficiency of EE-triggered traffic or affecting the efficiency of the search domain, effectively mitigating the Matthew effect.

[0089] Optionally, the aforementioned initial business item information set includes: a self-driven business item information set and a business input item information set. The self-driven business item information set can be a set of item information automatically mined by the system through various operational signals. The business input item information set can be a set of item information input by the relevant business side.

[0090] Step 202: Based on the obtained item sampling information set, allocate the corresponding business information to the above business item information set to obtain business item information groups for each business information.

[0091] In some embodiments, the aforementioned executing entity can allocate corresponding business information to the aforementioned business item information set based on the obtained item sampling information set, thereby obtaining business item information groups for each business item. The business information can be domain information corresponding to a business domain. Specifically, the domain information can include, but is not limited to, at least one of the following: a business domain identifier, and business content involved in the business domain. The business information corresponding to each business item in the business item information group (i.e., the domain information corresponding to the business domain) is the same.

[0092] As an example, firstly, the aforementioned execution entity can remove item sampling information whose corresponding values ​​are less than the target value from the aforementioned item sampling information set, obtaining at least one item sampling information. Then, the at least one item sampling information is matched with various business information to obtain a business item information group for each business information.

[0093] In some optional implementations of certain embodiments, the above-mentioned allocation of corresponding business information to the business item information set based on the obtained item sampling information set to obtain business item information groups for each business information may include the following steps:

[0094] The first step is to determine the set of business information to be recommended. This set can contain information about items (e.g., goods) to be recommended within the corresponding business domain.

[0095] The second step is to determine the recommended product quota information for each business information item in the aforementioned business information set. This recommended product quota information represents the number of recommended products available for each business information item. For example, under the business information item for the home appliance business, the recommended product quota is 10. Under the business information item for the fresh produce business, the recommended product quota is also 10.

[0096] Third, for each piece of business information in the above business information set, perform the following business item information filtering steps:

[0097] Sub-step 1 involves filtering out business item information that has a business relationship with the aforementioned business information from the above business item information set, and using this as the target business item information subset. Each target business item in the target business item information subset corresponds to the same business information.

[0098] Sub-step 2 involves selecting target business item information that meets preset sampling conditions from the aforementioned subset of target business item information, based on the business item recommendation quota information corresponding to the above business information. The preset sampling conditions can be to select target business item information whose corresponding numerical ranking is among the top target values. The target value corresponds to the same size as the business item recommendation quota information.

[0099] Step 203: Based on the obtained set of business item information, generate at least one recommended item information for the above-mentioned target item recommendation request.

[0100] In some embodiments, the executing entity may generate at least one recommended item information for the target item recommendation request based on the obtained set of business item information. The recommended item information in the at least one recommended item information may be business item information to be recommended.

[0101] As an example, the aforementioned executing entity can randomly select a predetermined number of business item information from the business item information set to obtain at least one business item information. This at least one business item information is then combined with the aforementioned refined item information set to obtain an item information set, which serves as at least one recommended item information.

[0102] Step 204: Send at least one of the above-mentioned recommended item information to the user terminal corresponding to the above-mentioned target item recommendation request.

[0103] In some embodiments, the executing entity may send at least one recommended item information to the user terminal corresponding to the target item recommendation request. The user terminal may be the user-associated terminal that sent the target item recommendation request. Specifically, the user terminal may be a mobile phone or a tablet.

[0104] The above embodiments of this disclosure have the following beneficial effects: the item recommendation method of some embodiments of this disclosure can accurately and efficiently generate at least one recommended item information. Specifically, the reason why the related recommended item information is not accurate enough is that the item information recommendation fails to take into account the actual efficiency and uncertainty of the item, resulting in the inaccuracy of at least one recommended item information. Based on this, the item recommendation method of some embodiments of this disclosure firstly, in response to receiving a target item recommendation request, for each business item information in the business item information set, determines the item sampling information corresponding to the business item information according to the posterior distribution of the item sampling information corresponding to the business item information. The posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information. Here, determining the item sampling information corresponding to each business item information through the posterior distribution of the item sampling information fully considers the actual efficiency of the item and adds uncertainty, giving potentially high-quality items more exposure opportunities. Furthermore, the posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information. By dividing the business item information into business item hierarchical categories, the growth status of the items can be fully reflected. Therefore, by using the business item hierarchy, a precise prior distribution of item sampling information can be generated. Then, based on the obtained item sampling information set, the business item information set is precisely allocated to corresponding business information, resulting in business item information groups for each business item. Furthermore, based on the obtained business item information group set, at least one recommended item can be accurately generated for the target item recommendation request. Finally, the at least one recommended item is sent to the user terminal corresponding to the target item recommendation request. In summary, by using the posterior distribution of item sampling information corresponding to business item information, the precise generation of at least one recommended item can be achieved, enabling its transmission to the user terminal corresponding to the target item recommendation request.

[0105] Further reference Figure 4 The diagram illustrates a flow 400 of another embodiment of the item recommendation method according to this disclosure. This item recommendation method includes the following steps:

[0106] Step 401: In response to receiving the target item recommendation request, for each piece of business item information in the business item information set, determine the item sampling information corresponding to the business item information based on the posterior distribution of the item sampling information corresponding to the business item information.

[0107] Step 402: Based on the obtained item sampling information set, allocate corresponding business information to the above business item information set to obtain business item information groups for each business information.

[0108] Step 403: Obtain the refined item information set.

[0109] In some embodiments, the executing entity (e.g. Figure 1 The electronic device 101 shown can acquire the refined item information set via wired or wireless means. The refined item information set can be the set of item information selected through refined ranking. Specifically, the refined item information set includes: a general candidate item information set corresponding to the refined ranking results and a Search EE candidate set. The general candidate item information set corresponding to the refined ranking results can be the set of item information selected by the refined ranking model. The Search EE candidate set can be the set of item information selected by the EE algorithm.

[0110] Step 404: Input the above-mentioned fine-ranked item information set and the above-mentioned business item information set into the user exploration score information generation model to generate a fine-ranked exploration score information set for the above-mentioned fine-ranked item information set and a business exploration score information set for the above-mentioned business item information set.

[0111] In some embodiments, the executing entity can input the fine-ranked item information set and the business item information set into a user exploration score generation model to generate a fine-ranked exploration score information set for the fine-ranked item information set and a business exploration score information set for the business item information set. The user exploration score generation model can be a model that generates user exploration score information for item information. User exploration score information can represent a user's liking for item information or the probability of a user viewing an item. A higher user exploration score indicates that the user is more likely to view the relevant content of the corresponding item information. The user exploration score generation model may include: a deep-learning neural network (DNN) + sparse variational Gaussian process (SVGP) + Thompson sampling. There is a one-to-one correspondence between the fine-ranked item information in the fine-ranked item information set and the fine-ranked exploration scores in the fine-ranked exploration score information set. Similarly, there is a one-to-one correspondence between the business item information in the business item information set and the business exploration score information in the business exploration score information set. The ranking exploration score represents a user's level of liking for ranked items, and can also represent the probability of a user viewing ranked items. Similarly, the business exploration score represents a user's level of liking for business items, and can also represent the probability of a user viewing business items.

[0112] Step 405: Determine the number of items to recommend for the target user.

[0113] In some embodiments, the executing entity may determine the number of items recommended to the target user. The number of items recommended may be the number of item information items recommended to the target user.

[0114] As an example, the aforementioned implementing entity can set the corresponding number of recommended items based on the target user's application login duration.

[0115] In some optional implementations of certain embodiments, determining the number of item recommendations for a target user may include the following steps:

[0116] The first step is to determine the user item exploration intention information corresponding to the target users mentioned above. This information represents the degree to which the target users are willing to explore item information. In practice, this information can be presented as a score. The higher the score, the greater the target user's willingness to explore item information.

[0117] As an example, the aforementioned executing entity can obtain the target user's recent historical behavior information. Then, this recent historical behavior information is input into a user item exploration intention information generation model to generate user item exploration intention information. The user item exploration intention information generation model can be a Transformer model.

[0118] The second step is to determine the number of items to recommend based on the user's willingness to explore items.

[0119] As an example, the aforementioned implementing entity can use a bucketing mechanism to determine the number of items recommended based on the user's item exploration intention information. The bucketing mechanism represents the number of item recommendations corresponding to each bucket. Each bucket also has a corresponding range of user item exploration intention values.

[0120] As another example, the aforementioned implementing entity can use a pre-determined association table that represents the range of values ​​for user item exploration intention information and the size of the number of item recommendations to determine the number of item recommendations based on the aforementioned user item exploration intention information.

[0121] Step 406: Based on the above-mentioned fine-ranking exploration score information set and the above-mentioned business exploration score information set, at least one item information in the above-mentioned fine-ranking item information set and business item information set is determined as at least one item information to be recommended.

[0122] In some embodiments, the executing entity may determine at least one item information from the fine-ranking item information set and the business item information set as at least one item information to be recommended, based on the fine-ranking exploration score information set and the business exploration score information set. The number of item information corresponding to the at least one item information to be recommended is the same as the number of recommended items.

[0123] In some optional implementations of certain embodiments, determining at least one item information from the fine-ranking item information set and the business item information set as at least one item information to be recommended, based on the fine-ranking exploration score information set and the business exploration score information set, may include the following steps:

[0124] The first step is for the aforementioned executing entity to filter out finely ranked exploration score information with values ​​greater than the aforementioned user item exploration intention information from the aforementioned finely ranked exploration score information set, thereby obtaining at least one finely ranked exploration score information.

[0125] The second step involves the aforementioned implementing entity determining the difference between the number of refined exploration score information items included in at least one refined exploration score information item and the number of recommended items.

[0126] Third, the aforementioned implementing entity can filter out the business exploration score information sets whose corresponding scores rank within the top target number from the aforementioned business exploration score information set. The number of business exploration score information pieces included in each business exploration score information set is the same as the aforementioned difference. That is, the target number is the same as the difference.

[0127] Fourth, the aforementioned executing entity can combine the business item information set corresponding to the aforementioned business exploration score information set with at least one finely ranked item information corresponding to at least one finely ranked exploration score information set to generate at least one item information to be recommended.

[0128] Step 407: Send at least one of the above-mentioned recommended item information to the user terminal corresponding to the above-mentioned target item recommendation request.

[0129] In some embodiments, the execution entity may send at least one of the recommended item information to the user terminal corresponding to the target item recommendation request.

[0130] In some embodiments, the specific implementation of steps 401-402 and 407 and their resulting technical effects can be found in [reference needed]. Figure 2 Steps 201-201 and 204 in the corresponding embodiments will not be repeated here.

[0131] from Figure 4 It can be seen from this that, with Figure 2 Compared to the description of some corresponding embodiments, Figure 4In some corresponding embodiments, the item recommendation method's process 400, based on the finely ranked item information set, utilizes a user exploration score information generation model to accurately obtain the finely ranked exploration score information set and the business exploration score information set. Based on the finely ranked exploration score information set and the business exploration score information set, at least one item information to be recommended can be accurately selected from these two sets.

[0132] Further reference Figure 5 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an item recommendation device, which are similar to... Figure 2 Corresponding to the method embodiments shown, the recommended device can be specifically applied to various electronic devices.

[0133] like Figure 5 As shown, an item recommendation device 500 includes: a determining unit 501, an allocation unit 502, a generating unit 503, and a sending unit 504. The determining unit 501 is configured to, in response to receiving a target item recommendation request, determine, for each piece of business item information in a business item information set, item sampling information corresponding to the business item information based on the posterior distribution of item sampling information corresponding to the business item information, wherein the posterior distribution of item sampling information is determined based on the prior distribution of item sampling information related to the hierarchical association of the business items corresponding to the business item information; the allocation unit 502 is configured to allocate corresponding business information to the business item information set according to the obtained item sampling information set, obtaining business item information groups for each business item; the generating unit 503 is configured to generate at least one recommended item information for the target item recommendation request based on the obtained business item information group set; and the sending unit 504 is configured to send the at least one recommended item information to the user terminal corresponding to the target item recommendation request.

[0134] In some optional implementations of certain embodiments, the allocation unit 502 may be further configured to: determine a set of business information to be recommended; determine the business item recommendation quota information corresponding to each business information in the business information set; for each business information in the business information set, perform the following business item information filtering steps: filter out business item information that has a business relationship with the business information from the business item information set, as target business item information, to obtain a subset of target business item information; according to the business item recommendation quota information corresponding to the business information, filter out target business item information whose corresponding item sampling information meets the preset sampling information conditions from the subset of target business item information, to obtain a group of business item information.

[0135] In some optional implementations of certain embodiments, the generation unit 503 may be further configured to: acquire a finely ranked item information set; input the finely ranked item information set and the business item information set into a user exploration score information generation model to generate a finely ranked exploration score information set for the finely ranked item information set and a business exploration score information set for the business item information set; determine the number of item recommendations for the target user; and, based on the finely ranked exploration score information set and the business exploration score information set, determine at least one item information from the finely ranked item information set and the business item information set as at least one item information to be recommended, wherein the number of item information corresponding to the at least one item information to be recommended is the same as the number of item recommendations.

[0136] In some optional implementations of certain embodiments, the generation unit 503 may be further configured to: determine the user item exploration intention information corresponding to the target user; and determine the number of item recommendations based on the user item exploration intention information.

[0137] In some optional implementations of certain embodiments, the generation unit 503 may be further configured to: filter out finely ranked exploration score information with values ​​greater than the user's item exploration intention information from the finely ranked exploration score information set to obtain at least one finely ranked exploration score information; determine the difference between the number of finely ranked exploration score information included in the at least one finely ranked exploration score information and the number of recommended items; filter out a set of business exploration score information corresponding to the top target number of score sizes from the business exploration score information set, wherein the number of business exploration score information included in the business exploration score information set is the same as the difference; combine the business item information set corresponding to the business exploration score information set and the at least one finely ranked item information corresponding to the at least one finely ranked exploration score information to generate at least one item information to be recommended.

[0138] It is understandable that the units described in the recommended device 500 are related to the reference. Figure 2 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to the item recommendation device 500 and the units contained therein, and will not be repeated here.

[0139] The following is for reference. Figure 6 It illustrates electronic devices suitable for implementing some embodiments of this disclosure (e.g., Figure 1 A schematic diagram of the structure of electronic device 101)600 in the middle. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0140] like Figure 6As shown, the electronic device 600 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory 602 or a program loaded from a storage device 608 into a random access memory 603. The random access memory 603 also stores various programs and data required for the operation of the electronic device 600. The processing unit 601, the read-only memory 602, and the random access memory 603 are interconnected via a bus 604. An input / output interface 605 is also connected to the bus 604.

[0141] Typically, the following devices can be connected to the input / output interface 605: input devices 606 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 608 including, for example, magnetic tape, hard disk, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 6 Each box shown can represent a device or multiple devices as needed.

[0142] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a read-only memory 602. When the computer program is executed by the processing device 601, it performs the functions defined above in the methods of some embodiments of this disclosure.

[0143] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0144] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0145] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: in response to receiving a target item recommendation request, for each piece of business item information in the business item information set, determine the corresponding item sampling information based on the posterior distribution of the item sampling information corresponding to the business item information, wherein the posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business items corresponding to the business item information; allocate corresponding business information to the business item information set according to the obtained item sampling information set, obtaining business item information groups for each piece of business information; generate at least one recommended item information for the target item recommendation request based on the obtained set of business item information groups; and send the at least one recommended item information to the user terminal corresponding to the target item recommendation request.

[0146] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0147] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0148] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a determining unit, an allocating unit, a generating unit, and a sending unit. The names of these units do not necessarily limit the specific unit; for example, an acquiring unit may also be described as "a unit that sends at least one of the recommended item information to the user terminal corresponding to the target item recommendation request."

[0149] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0150] Some embodiments of this disclosure also provide a computer program product, including a computer program that, when executed by a processor, implements any of the above-described item recommendation methods.

[0151] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for recommending items, including: In response to receiving a target item recommendation request, for each business item in the business item information set, the item sampling information corresponding to the business item information is determined based on the posterior distribution of the item sampling information corresponding to the business item information. The posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information. The business item hierarchy is the classification standard for business items, and the items in the business item information set are classified into hierarchical levels based on the sales volume of the business items. Based on the obtained item sampling information set, the business item information set is allocated with corresponding business information to obtain business item information groups for each business information; Based on the obtained set of business item information, at least one recommended item information is generated for the target item recommendation request, including: Obtain a set of refined item information; The fine-ranked item information set and the business item information set are input into the user exploration score information generation model to generate a fine-ranked exploration score information set for the fine-ranked item information set and a business exploration score information set for the business item information set. Determine the number of items to recommend to the target users; Based on the fine-ranking exploration score information set and the business exploration score information set, at least one item information in the fine-ranking item information set and the business item information set is determined as at least one item information to be recommended, wherein the number of item information corresponding to the at least one item information to be recommended is the same as the number of recommended items; The information of at least one recommended item is sent to the user terminal corresponding to the target item recommendation request.

2. The method according to claim 1, wherein, The step of allocating corresponding business information to the business item information set based on the obtained item sampling information set to obtain business item information groups for each business information includes: Determine the set of business information to be recommended; Determine the recommended quota information for each business item corresponding to each business item in the business information set; For each piece of business information in the aforementioned business information set, perform the following business item information filtering steps: From the set of business item information, select business item information that has a business relationship with the business information, and use it as target business item information to obtain a subset of target business item information; Based on the business item recommendation quota information corresponding to the business information, target business item information that meets the preset sampling information conditions is selected from the target business item information subset to obtain a business item information group.

3. The method according to claim 1, wherein, The posterior distribution of the item sampling information corresponding to each business item information in the business item information set is determined through the following steps: Determine the business item level corresponding to the business item information at the current time, and use it as the target business item level; Determine the prior distribution of item sampling information corresponding to the target business item level; Based on the prior distribution of the item sampling information, the posterior distribution of the item sampling information is determined.

4. The method according to claim 3, wherein, Determining the posterior distribution of the item sampling information based on the prior distribution of the item sampling information includes: Obtain the click-through rate and exposure information of the business items within a predetermined time period; The item click pass rate information and the item exposure information are substituted into the item sampling information posterior distribution formula corresponding to the item sampling information prior distribution to generate the item sampling information posterior distribution.

5. The method according to claim 3, wherein, Determining the business item level corresponding to the business item information at the current time includes: For each piece of business item information in the business item information set, perform the following generation steps: Determine the previous time period corresponding to the current time as the target historical time period; Determine the click-through rate and conversion rate of the business item information within the target historical time period; Based on the click pass rate and the conversion rate, generate item efficiency information for the business item information; The obtained set of item efficiency information is sorted to obtain a sequence of item efficiency information; The business item level corresponding to the business item information is determined based on the position of the item efficiency information corresponding to the business item information in the item efficiency information sequence.

6. The method according to claim 1, wherein, Determining the number of item recommendations for the target user includes: Determine the user's willingness to explore items corresponding to the target user; Determine the number of items to recommend based on the user's item exploration intention information.

7. The method according to claim 6, wherein, The step of determining at least one item information from the fine-ranking item information set and the business item information set as at least one item information to be recommended, based on the fine-ranking exploration score information set and the business exploration score information set, includes: Filter out at least one refined exploration score from the refined exploration score set whose value is greater than the user's item exploration intention information. Determine the difference between the number of fine-ranking exploration score information included in the at least one fine-ranking exploration score information and the number of item recommendations; From the business exploration score information set, select the business exploration score information set whose corresponding score is among the top target number, wherein the number of business exploration score information included in the business exploration score information set is the same as the difference; The business item information set corresponding to the business exploration score information set and the at least one finely ranked item information corresponding to at least one finely ranked exploration score information are combined to generate at least one item information to be recommended.

8. The method according to claim 1, wherein, The business item information set is generated through the following steps: Obtain the offline dataset corresponding to the initial business item information set; Based on the offline dataset, the initial business item information set is filtered to generate a filtered item information set; Based on the real-time dataset corresponding to the filtered item information set, remove item information whose corresponding exposure information meets the preset exposure conditions from the filtered item information set to obtain the removed item information set, which serves as the business item information set.

9. The method according to claim 8, wherein, The initial business item information set includes: a self-driven mining business item information set and a business input item information set.

10. An item recommendation device, comprising: The determining unit is configured to, in response to receiving a target item recommendation request, for each business item information in the business item information set, determine the item sampling information corresponding to the business item information based on the posterior distribution of the item sampling information corresponding to the business item information, wherein the posterior distribution of the item sampling information is determined based on the prior distribution of the item sampling information associated with the hierarchical relationship of the business item information, wherein the business item hierarchy is the classification standard of business items, and the hierarchical classification is performed on each item in the business item information set according to the sales volume of the business items; The allocation unit is configured to allocate corresponding business information to the business item information set according to the obtained item sampling information set, so as to obtain business item information groups for each business information. The generation unit is configured to generate at least one recommended item information for the target item recommendation request based on the obtained set of business item information, including: Obtain a set of refined item information; The fine-ranked item information set and the business item information set are input into the user exploration score information generation model to generate a fine-ranked exploration score information set for the fine-ranked item information set and a business exploration score information set for the business item information set. Determine the number of items to recommend to the target users; Based on the fine-ranking exploration score information set and the business exploration score information set, at least one item information in the fine-ranking item information set and the business item information set is determined as at least one item information to be recommended, wherein the number of item information corresponding to the at least one item information to be recommended is the same as the number of recommended items; The sending unit is configured to send the at least one recommended item information to the user terminal corresponding to the target item recommendation request.

11. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-9.

12. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-9.

13. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-9.

Images