A like rate acquisition method, device and equipment

Abstract

The like rate obtaining method provided in the application comprises: obtaining attribute information of a target user and first historical interaction data of the target user in a first specified time period for the target user; determining a first target age bucket to which the target user belongs and at least one second target age bucket to which a to-side user involved in the first historical interaction data belongs according to the attribute information, the first historical interaction data and a pre-obtained group preference matrix; for each second target age bucket, finding a standard like rate of a from-side user of the first target age bucket to a to-side user of the second target age bucket and a standard display value of the from-side user of the first target age bucket to the to-side user of the second target age bucket from the group preference matrix and a pre-obtained smoothing parameter matrix; and calculating a like rate based on the standard like rate, the standard display value, a number of target display data and a number of target like data for the second target age bucket involved in the first historical interaction data.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular to a method, apparatus and device for obtaining likes rate. Background Technology

[0002] In recent years, with the widespread adoption of smartphones and mobile internet, social networking apps have become increasingly popular. In some social networking apps, users can swipe right to show their liking for recommended users. Furthermore, if the recommended user also swipes right back, the two can establish a connection and begin further chat and communication.

[0003] Currently, to improve user experience, promote social interaction, enhance the functionality of the social client, and support business decisions, it is necessary to calculate the proportion of right swipes to page views, i.e., the user's "like rate." Related technologies typically use historical user interaction data to calculate the like rate. However, when a user is a new user or has a low number of right swipes, the like rate calculated based on historical interaction data cannot accurately represent the user's preferences. Therefore, accurately obtaining the like rate has become a pressing issue that needs to be addressed. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus, and device for obtaining likes rate, so as to accurately obtain the user's likes rate.

[0005] Specifically, this application is implemented through the following technical solution:

[0006] The first aspect of this application provides a method for obtaining a like rate, the method comprising: for a target user, obtaining attribute information of the target user and first historical interaction data of the target user within a first specified time period; wherein, the first historical interaction data includes display data and like data based on the display data;

[0007] Based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket;

[0008] For each second target age bucket, the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket are found from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket.

[0009] For each second target age bucket, based on the standard like rate, the standard display value, and the number of target display data and target like data for that second target age bucket involved in the first historical interaction data, the like rate of the target user for the to-side user of that second target age bucket is calculated.

[0010] A second aspect of this application provides a like rate acquisition device, the device comprising an acquisition module, a determination module, a search module, and a calculation module; wherein...

[0011] The acquisition module is used to acquire attribute information of the target user and first historical interaction data of the target user within a first specified time period; wherein, the first historical interaction data includes display data and like data based on the display data.

[0012] The determining module is used to determine, based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket.

[0013] The search module is used to search, for each second target age bucket, the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket.

[0014] The calculation module is used to calculate the target user's like rate for each second target age bucket based on the standard like rate, the standard display value, and the number of target display data and target like data for that second target age bucket in the first historical interaction data.

[0015] A third aspect of this application provides a like rate acquisition device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods provided in the first aspect of this application.

[0016] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods provided in the first aspect of this application.

[0017] The liking rate acquisition method, apparatus, and device provided in this application, targeting a user, acquires the user's attribute information and first historical interaction data within a first specified time period. Then, based on the attribute information, the first historical interaction data, and a pre-acquired group preference matrix, it determines the first target age bucket to which the target user belongs, and at least one second target age bucket to which users on the "to" side of the first historical interaction data belong. For each second target age bucket, it obtains the standard liking rate and standard display value of users on the "from" side of the first target age bucket for users on the "to" side of that second target age bucket from the group preference matrix and a pre-acquired smoothing parameter matrix. Then, for each second target age bucket, based on the standard liking rate, the standard display value, and the number of target display data and target liking data points for that second target age bucket in the first historical interaction data, it calculates the target user's liking rate for users on the "to" side of that second age bucket. This allows for liking rate calculation based on the target user's historical interaction data and standard data, improving the accuracy of the calculated liking rate. Attached Figure Description

[0018] Figure 1 A flowchart of Embodiment 1 of the like rate acquisition method provided in this application;

[0019] Figure 2 A flowchart of Embodiment 2 of the like rate acquisition method provided in this application;

[0020] Figure 3 A flowchart of Embodiment 3 of the liking rate acquisition method provided in this application;

[0021] Figure 4 Hardware structure diagram of the like rate acquisition device provided in this application;

[0022] Figure 5 A schematic diagram of the structure of the like rate acquisition device according to Embodiment 1 provided in this application;

[0023] Figure 6 This is a schematic diagram of the structure of Embodiment 2 of the like rate acquisition device provided in this application. Detailed Implementation

[0024] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0025] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0026] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0027] In view of this, this application provides a method, apparatus, and device for obtaining likes rate, so as to accurately obtain the user's likes rate.

[0028] The liking rate acquisition method, apparatus, and device provided in this application, targeting a user, acquires the user's attribute information and first historical interaction data within a first specified time period. Then, based on the attribute information, the first historical interaction data, and a pre-acquired group preference matrix, it determines the first target age bucket to which the target user belongs, and at least one second target age bucket to which users on the "to" side of the first historical interaction data belong. For each second target age bucket, it obtains the standard liking rate and standard display value of users on the "from" side of the first target age bucket for users on the "to" side of that second target age bucket from the group preference matrix and a pre-acquired smoothing parameter matrix. Then, for each second target age bucket, based on the standard liking rate, the standard display value, and the number of target display data and target liking data points for that second target age bucket in the first historical interaction data, it calculates the target user's liking rate for users on the "to" side of that second target age bucket. This allows for liking rate calculation based on the target user's historical interaction data and standard data, improving the accuracy of the calculated liking rate.

[0029] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0030] Figure 1 The flowchart is for Embodiment 1 of the like rate acquisition method provided in this application. Please refer to... Figure 1 The method provided in this embodiment includes:

[0031] S101. For a target user, obtain the target user's attribute information and the target user's first historical interaction data within a first specified time period; wherein, the first historical interaction data includes display data and like data for the display data.

[0032] Specifically, the target user refers to the user whose like rate is to be calculated, which can be any registered user on the social client. Furthermore, the target user's attribute information includes age, occupation, hobbies, etc. Additionally, the first specified time period is set according to actual needs, and this embodiment does not limit it. For example, in one possible implementation, the first specified time period could be 90 days.

[0033] It should be noted that the first historical interaction data includes display data and like data related to the display data. The display data corresponds to the operation of showing the data from the "to" side to the "from" side. Furthermore, the like data related to the display data corresponds to the "from" side user performing a right swipe operation on the displayed "to" side user after the "to" side user is shown to the "from" side user. For example, in one embodiment, the target user is user 1. One piece of display data could be the data corresponding to showing the data from the "to" side user 3 to user 1, and one piece of like data could be the data corresponding to user 1 performing a right swipe operation on the "to" side user 3.

[0034] S102. Based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, determine the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket.

[0035] It should be noted that the group preference matrix represents the standard liking rate of users from the first age group ("from") towards users from the second age group ("to") in each age group. It should also be noted that the group preference matrix is ​​obtained based on massive amounts of historical user interaction data. For example, in one possible implementation, the first and second age groups can be pre-defined, and then the standard liking rate of users from the first age group ("from") towards users from the second age group ("to") can be calculated based on the massive amounts of historical user interaction data. Furthermore, in this implementation, when dividing the first and second age groups, users with similar ages can be grouped into the same bucket.

[0036] Furthermore, specific embodiments will be given below to describe in detail the process of obtaining the group preference matrix, which will not be elaborated here. For example, in one possible implementation, the group preference matrix can be represented as Table 1:

[0037] Table 1 Group Preference Matrix

[0038]

[0039] Specifically, the first target age bucket to which the target user belongs can be determined based on the age and group preference matrix in the target user's attribute information. In practice, the target age bucket containing the target user's age is searched from each first age bucket in the group preference matrix, and then the found target age bucket is determined as the first target age bucket.

[0040] Accordingly, for each user on the "to" side involved in the first historical interaction data, the second target age bucket to which the user belongs is determined based on the user's age and group preference matrix. In specific implementation, the target age bucket containing the user's age is searched from each second age bucket in the group preference matrix, and then the found target age bucket is determined as the second target age bucket.

[0041] Referring to Table 1, for example, target user 1 is 18 years old. The first historical interaction data for target user 1 shows that target user 1 has two swipe records (i.e., right swipe data). These two right swipe records show that target user 1 swiped right to 19-year-old user 2 (to-side) and 22-year-old user 3 (to-side). At this point, by looking up the table, we know that the first target age bucket to which target user 1 belongs is age bucket 1 within the first age bucket (the age bucket corresponding to the user from the side). Further, for user 2 (to-side), the second target age bucket to which it belongs is determined to be age bucket 2 within the second age bucket (the age bucket corresponding to the user from the side); for user 3 (to-side), the second target age bucket to which it belongs is determined to be age bucket 4 within the second age bucket (the age bucket corresponding to the user from the side).

[0042] S103. For each second target age bucket, find the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket.

[0043] It should be noted that the smoothing parameter matrix is ​​obtained based on massive amounts of historical user interaction data. The acquisition process will be described in detail in the following embodiments and will not be repeated here. For example, in one possible implementation, the smoothing parameter matrix can be represented as shown in Table 2:

[0044] Table 2 Smoothing parameter matrix

[0045]

[0046] In practice, for each second target age bucket, the standard like rate of users from the first target age bucket to users from the second target age bucket is found in the group preference matrix, and the standard display value of users from the first target age bucket to users from the second target age bucket is found in the smoothing parameter matrix.

[0047] Combining the examples above, and referring to Tables 1 and 2, for instance, regarding the first second target age group (19-20 years old), a search reveals that the standard like rate for users from the first target age group (18-19 years old) on the "from" side of this second target age group (19-20 years old) is 0.5375, and the standard display value is 45. Furthermore, regarding the second target age group (22 years old), a search reveals that the standard like rate for users from the first target age group (18-19 years old) on the "from" side of this second target age group (22 years old) is 0.5325, and the standard display value is 30.

[0048] S104. For each second target age bucket, based on the standard like rate, the standard display value, and the number of target display data and the number of target like data for that second target age bucket involved in the first historical interaction data, calculate the like rate of the target user for the to-side user of that second target age bucket.

[0049] In practice, the specific implementation process of this step may include:

[0050] (1) The sum of the number of target display data and the standard display value is determined as the updated display value.

[0051] (2) The product of the standard liking rate and the standard display value is determined as the standard liking value.

[0052] (3) The sum of the standard like value and the number of target like data is determined as the updated like value.

[0053] (4) The ratio of the updated like value to the updated display value is determined as the like rate of the target user to the to-side users of the second target age bucket.

[0054] In other words, for each second target age group, when calculating the liking rate of target users for that second target age group's users on the "to" side, it can be calculated using the following formula:

[0055] Like rate = (Number of target likes + Standard like rate * Standard display value) / (Number of target display data + Standard display value)

[0056] For example, referring to the example above, for the first second target age group (19-20 years old), after searching, it was determined that the standard like rate of users from the first target age group (18-19 years old) to users from the second target age group (19-20 years old) was 0.5375, and the standard display value was 45. Further, based on the first historical interaction data, the number of target display data was 50, and the number of target like data was 10. At this point, the updated display value was 95, the standard like value was 24.1875, the updated like value was 34.1875, and the target user's like rate to users from the second target age group (19-20 years old) was 0.36.

[0057] The liking rate acquisition method provided in this embodiment, targeting a target user, obtains the target user's attribute information and the target user's first historical interaction data within a first specified time period. Then, based on the attribute information, the first historical interaction data, and a pre-acquired group preference matrix, it determines the target user's first target age bucket and at least one second target age bucket to which the "to" side users involved in the first historical interaction data belong. For each second target age bucket, it obtains the standard liking rate and standard display value of the "from" side users of the first target age bucket for the "to" side users of that second target age bucket from the group preference matrix and a pre-acquired smoothing parameter matrix. Then, for each second target age bucket, based on the standard liking rate, the standard display value, and the number of target display data and target liking data points for that second target age bucket involved in the first historical interaction data, it calculates the target user's liking rate for the "to" side users of that second target age bucket. This allows for liking rate calculation based on the target user's historical interaction data and standard data, improving the accuracy of the calculated liking rate.

[0058] Figure 2 The flowchart for Embodiment 2 of the like rate acquisition method provided in this application is shown below. Please refer to... Figure 2 Based on the above embodiments, the group preference matrix is ​​obtained through the following method:

[0059] S201. Obtain massive amounts of second historical interaction data of users within a second specified time period, and based on the second historical interaction data, obtain the standard liking rate of users from the first age group on the to side of each second age group.

[0060] It should be noted that a large number of users can refer to all registered users of the social client. Furthermore, the second specified time period is set according to actual needs, and this embodiment does not limit it. For example, in one possible implementation, the second specified time period could be 90 days.

[0061] Specifically, the second historical interaction data of each user in the massive user base includes display data and like data based on the display data.

[0062] It should be noted that the first and second age groups are set according to actual needs, and they can be the same or different. In this embodiment, no limitations are imposed on the first and second age groups. For example, in one embodiment, each age point within the allowed age range of the social client is defined as an age group. Referring to the example above, if the allowed age range of the social client is 18 to 100 years old, then 18, 19, ..., 100 years old are defined as age groups.

[0063] The following example illustrates this: "The allowed age range for social clients is 18 to 22 years old. The first age group and the second age group are the same. The first age group includes 18, 19, 20, 21, and 22 years old, and the second age group includes 18, 19, 20, 21, and 22 years old." In this step, the standard liking rate of users from the "from" side of each first age group to users from the "to" side of each second age group is obtained. For example, in one embodiment, the results are shown in Table 3:

[0064] Table 3

[0065]

[0066] It should be noted that, in specific implementation, for each user in the massive user base, the like rate of that user to multiple second age groups involved in the user's second historical interaction data can be calculated first. This yields the like rate of the "from" side users of the corresponding first age group to the "to" side users of the aforementioned multiple second age groups. Furthermore, when calculating the standard like rate of a certain second age group's "from" side users to a certain first age group, the like rates of all the "from" side users of the first age group to the "to" side users of the second age group are found from all the calculated like rates. The average of all found like rates is then determined as the standard like rate of the "from" side users of the first age group to the "to" side users of the second age group.

[0067] S202. According to a preset aggregation rule, the standard liking rate of users from the first age group to users from the second age group is aggregated to obtain the group preference matrix used to characterize the standard liking rate of users from the first age group to users from the second age group; the number of first age groups is less than the number of first age groups and / or the number of second age groups is less than the number of first age groups.

[0068] Specifically, the preset aggregation rules are set according to actual needs, and this embodiment does not limit them.

[0069] For example, in one possible implementation, the preset aggregation rules include a first aggregation rule and / or a second aggregation rule. The first aggregation rule is as follows: for each second age group's "to" side users, when the age difference between the "from" side users of the first target age group and the "from" side users of the second target age group is less than a first preset threshold, and the difference between the standard liking rate of the "from" side users of the first target age group for each second age group's "to" side users and the standard liking rate of the "from" side users of the second target age group for each second age group's "to" side users is less than a second preset threshold, the first target age group and the second target age group are aggregated into a first age bucket.

[0070] The second aggregation rule is as follows: For users on the "from" side of each first age group, when the age difference between users on the "to" side of the third target age group and users on the "to" side of each second age group is less than the first preset threshold, and the difference between the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the third target age group and the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the fourth target age group is less than the second preset threshold, the third target age group and the fourth target age group are aggregated into a second age bucket.

[0071] Specifically, the first and second preset thresholds are set according to actual needs, and this embodiment does not limit them. For example, the first preset threshold is 3 and the second preset threshold is 0.1.

[0072] The following explanation uses a preset aggregation rule that includes a first aggregation rule and a second aggregation rule as an example. Referring to Table 3, for example, after first aggregating users on the "from" side according to the first aggregation rule, the aggregation results are shown in Table 4:

[0073] Table 4

[0074]

[0075] Furthermore, after aggregating the results shown in Table 4 according to the second aggregation rule, the aggregated group preference matrix is ​​shown in Table 1.

[0076] The liking rate acquisition method provided in this embodiment acquires massive amounts of second historical interaction data of users within a second specified time period. Based on this second historical interaction data, it obtains the standard liking rate of users from each first age group on the "from" side for users from each second age group on the "to" side. Then, according to a preset aggregation rule, it aggregates the standard liking rates of users from each first age group on the "from" side for users from each second age group on the "to" side, obtaining the group preference matrix representing the standard liking rate of users from each first age bucket on the "from" side for users from each second age bucket on the "to" side for users from each second age bucket. The number of first age buckets is less than the number of first age groups and / or the number of second age buckets is less than the number of first age groups. In this way, users with similar ages and similar liking rates can be aggregated into the same age bucket, avoiding the problem of data sparsity and improving the accuracy of the calculated liking rate.

[0077] Optionally, in one possible embodiment, the smoothing parameter matrix is ​​obtained by the following method:

[0078] Based on the second historical interaction data, the standard display values ​​of the "from" side users of each first age bucket to the "to" side users of each second age bucket are obtained, and the smoothing parameter matrix is ​​obtained.

[0079] It should be noted that, in specific implementation, for each user in the massive user base, the display values ​​for multiple second-age buckets involved in that user's second historical interaction data can be calculated. This yields the display values ​​from the "from" side of the corresponding first-age bucket to the "to" side of the aforementioned multiple second-age buckets. Furthermore, when calculating the standard display value for a certain second-age bucket from the "from" side of a certain first-age bucket, all the display values ​​calculated above are used to find the display values ​​from the "from" side of that first-age bucket to the "to" side of that second-age bucket. These values ​​are then sorted in ascending order, and the median is taken as the standard display value from the "from" side of that first-age bucket to the "to" side of that second-age bucket. Finally, the matrix formed by the display values ​​from the "from" side of each first-age bucket to the "to" side of each second-age bucket constitutes the smoothing parameter matrix.

[0080] The like rate acquisition method provided in this embodiment, based on the second historical interaction data, obtains the smoothing parameter matrix by acquiring the standard display values ​​of users from the "from" side of each first age bucket to users from the "to" side of each second age bucket. This allows the like rate to be calculated based on the smoothing parameter matrix, avoiding data sparsity issues and improving the accuracy of the calculated like rate.

[0081] Figure 3 The flowchart for Embodiment 3 of the like rate acquisition method provided in this application is shown below. Please refer to... Figure 3 The method provided in this embodiment, before obtaining the standard like rate and the standard display value, may further include:

[0082] 301. Based on preset filtering rules, filter out abnormal users that meet the preset filtering rules from the second historical interaction data.

[0083] 302. Delete the abnormal historical interaction data corresponding to the abnormal user from the second historical interaction data.

[0084] Specifically, the preset filtering rules are set according to actual needs, and this embodiment does not limit them. For example, in one possible implementation, the preset filtering rule may be for a specific user whose second historical interaction data reflects that the user's preferences for users in each age group do not match the user's age-related preferences.

[0085] For example, for a user who is 18 years old, their preference should be to like users of similar age, such as those aged 18 to 20. If the user's second historical interaction data reflects that the user's preference for users in each age group is that they prefer users aged 30, then this user is considered an abnormal user, and their abnormal historical interaction data is deleted.

[0086] The like rate acquisition method provided in this embodiment, before calculating the standard like rate and the standard display value, filters out abnormal users that meet the preset filtering rules from the second historical interaction data based on preset filtering rules, and then deletes the abnormal historical interaction data corresponding to the abnormal users from the second historical interaction data. This solves the confidence problem affected by outliers, improves the accuracy of the calculated group preference matrix and smoothing parameter matrix, and thus improves the accuracy of the calculated like rate.

[0087] Corresponding to the aforementioned embodiments of the like rate acquisition method, this application also provides embodiments of the like rate acquisition device.

[0088] The embodiment of the like rate acquisition device provided in this application can be applied to like rate acquisition devices. The device embodiment can be implemented through software, hardware, or a combination of both. Taking software implementation as an example, as a logical device, it is formed by the processor of the like rate acquisition device loading the corresponding computer program instructions from non-volatile memory into memory for execution. From a hardware perspective, such as... Figure 4 The diagram shown is a hardware structure diagram of the like rate acquisition device provided in this application, except for... Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown, the like rate acquisition device in the embodiment may also include other hardware depending on the actual function of the like rate acquisition device, which will not be described in detail here.

[0089] Figure 5 This is a schematic diagram of the structure of Embodiment 1 of the like rate acquisition device provided in this application. Please refer to... Figure 5 The apparatus provided in this embodiment may include an acquisition module 510, a determination module 520, a search module 530, and a calculation module 540; wherein,

[0090] The acquisition module 510 is used to acquire, for a target user, attribute information of the target user and first historical interaction data of the target user within a first specified time period; wherein, the first historical interaction data includes display data and like data based on the display data;

[0091] The determining module 520 is used to determine, based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket.

[0092] The search module 530 is used to search, for each second target age bucket, the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket.

[0093] The calculation module 540 is used to calculate the target user's liking rate for each second target age bucket based on the standard liking rate, the standard display value, and the number of target display data and target liking data for that second target age bucket involved in the first historical interaction data.

[0094] The apparatus provided in this embodiment can be used to perform... Figure 1 The technical methods of the illustrated embodiments are similar in principle and effect, and will not be described again here.

[0095] The like rate acquisition device provided in this embodiment, targeting a target user, acquires the target user's attribute information and the target user's first historical interaction data within a first specified time period. Then, based on the attribute information, the first historical interaction data, and a pre-acquired group preference matrix, it determines the target user's first target age bucket and at least one second target age bucket to which the "to" side users involved in the first historical interaction data belong. For each second target age bucket, it obtains the standard like rate and standard display value of the "from" side users of the first target age bucket for the "to" side users of that second target age bucket from the group preference matrix and a pre-acquired smoothing parameter matrix. Then, for each second target age bucket, based on the standard like rate, the standard display value, and the number of target display data and target like data related to that second target age bucket in the first historical interaction data, it calculates the target user's like rate for the "to" side users of that second target age bucket. This allows for the calculation of the like rate based on the target user's historical interaction data and standard data, improving the accuracy of the calculated like rate.

[0096] Optionally, the group preference matrix is ​​obtained by the following method:

[0097] Obtain massive amounts of second historical interaction data from users within a second specified time period, and based on the second historical interaction data, obtain the standard liking rate of users from the first age group on the to side of each second age group.

[0098] According to a preset aggregation rule, the standard liking rates of users from the first age group to users from the second age group are aggregated to obtain the group preference matrix, which represents the standard liking rates of users from the first age group to users from the second age group; the number of first age groups is less than the number of first age groups and / or the number of second age groups is less than the number of first age groups.

[0099] Optionally, the smoothing parameter matrix is ​​obtained by the following method:

[0100] Based on the second historical interaction data, the standard display values ​​of the "from" side users of each first age bucket to the "to" side users of each second age bucket are obtained, and the smoothing parameter matrix is ​​obtained.

[0101] Optionally, the computing module 540 is specifically used for:

[0102] The sum of the number of target display data and the standard display value is determined as the updated display value;

[0103] Calculate the standard like value based on the standard like rate and the standard display value;

[0104] The sum of the standard like value and the number of target like data is determined as the updated like value;

[0105] The ratio of the updated like value to the updated display value is determined as the like rate of the target user for the users on the to side of the second target age bucket.

[0106] Optionally, the preset aggregation rule includes a first aggregation rule and / or a second aggregation rule;

[0107] The first aggregation rule is as follows: for users on the "to" side of each second age group, when the age difference between users on the "from" side of the first target age group and users on the "from" side of the second target age group is less than a first preset threshold, and the difference between the standard liking rate of users on the "from" side of the first target age group for users on the "to" side of each second age group and the standard liking rate of users on the "from" side of the second target age group for users on the "to" side of each second age group is less than a second preset threshold, the first target age group and the second target age group are aggregated into a first age bucket.

[0108] The second aggregation rule is as follows: For users on the "from" side of each first age group, when the age difference between users on the "to" side of the third target age group and users on the "to" side of each second age group is less than the first preset threshold, and the difference between the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the third target age group and the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the fourth target age group is less than the second preset threshold, the third target age group and the fourth target age group are aggregated into a second age bucket.

[0109] Figure 6 This is a schematic diagram of the structure of Embodiment 2 of the like rate acquisition device provided in this application. Please refer to... Figure 6 The device provided in this embodiment, based on the above-described device, further includes a filtering module 550, wherein...

[0110] The filtering module 550 is used to filter out abnormal users that meet the preset filtering rules from the second historical interaction data based on preset filtering rules, and delete the abnormal historical interaction data corresponding to the abnormal users from the second historical interaction data.

[0111] The apparatus provided in this embodiment can be used to perform... Figure 3 The technical methods of the illustrated embodiments are similar in principle and effect, and will not be described again here.

[0112] The like rate acquisition device provided in this embodiment filters out abnormal users that meet the preset filtering rules from the second historical interaction data, thereby deleting the abnormal historical interaction data corresponding to the abnormal users from the second historical interaction data. This solves the confidence problem affected by outliers and improves the accuracy of the calculated group preference matrix and smoothing parameter matrix.

[0113] Please continue to refer to Figure 4 This application also provides a like rate acquisition device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the program to implement the steps of any of the methods provided in the first aspect of this application.

[0114] Furthermore, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods provided in the first aspect of this application.

[0115] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0116] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0117] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for obtaining like rate, characterized in that, The method includes: For a target user, obtain the target user's attribute information and the target user's first historical interaction data within a first specified time period; wherein, the first historical interaction data includes display data and like data related to the display data; Based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket; For each second target age bucket, the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket are found from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket. For each second target age bucket, based on the standard like rate, the standard display value, and the number of target display data and target like data for that second target age bucket involved in the first historical interaction data, the like rate of the target user for the to-side user of that second target age bucket is calculated.

2. The method according to claim 1, characterized in that, The group preference matrix is ​​obtained through the following method: Obtain massive amounts of second historical interaction data from users within a second specified time period, and based on the second historical interaction data, obtain the standard liking rate of users from the first age group on the to side of each second age group. According to a preset aggregation rule, the standard liking rates of users from the first age group to users from the second age group are aggregated to obtain the group preference matrix, which represents the standard liking rates of users from the first age group to users from the second age group; the number of first age groups is less than the number of first age groups and / or the number of second age groups is less than the number of second age groups.

3. The method according to claim 2, characterized in that, The smoothing parameter matrix is ​​obtained through the following method: Based on the second historical interaction data, the standard display values ​​of the "from" side users of each first age bucket to the "to" side users of each second age bucket are obtained, and the smoothing parameter matrix is ​​obtained.

4. The method according to claim 2 or 3, characterized in that, Before obtaining the standard like rate and the standard display value, the method further includes: Based on preset filtering rules, abnormal users that match the preset filtering rules are filtered out from the second historical interaction data; Delete the abnormal historical interaction data corresponding to the abnormal user from the second historical interaction data.

5. The method according to claim 1, characterized in that, The step of calculating the target user's liking rate for the users on the "to" side of the second target age bucket based on the standard liking rate, the standard display value, and the number of target display data and target liking data related to the second target age bucket in the first historical interaction data includes: The sum of the number of target display data and the standard display value is determined as the updated display value; Calculate the standard like value based on the standard like rate and the standard display value; The sum of the standard like value and the number of target like data is determined as the updated like value; The ratio of the updated like value to the updated display value is determined as the like rate of the target user for the users on the to side of the second target age bucket.

6. The method according to claim 2, characterized in that, The preset aggregation rules include a first aggregation rule and / or a second aggregation rule; The first aggregation rule is as follows: for users on the "to" side of each second age group, when the age difference between users on the "from" side of the first target age group and users on the "from" side of the second target age group is less than a first preset threshold, and the difference between the standard liking rate of users on the "from" side of the first target age group for users on the "to" side of each second age group and the standard liking rate of users on the "from" side of the second target age group for users on the "to" side of each second age group is less than a second preset threshold, the first target age group and the second target age group are aggregated into a first age bucket. The second aggregation rule is as follows: For users on the "from" side of each first age group, when the age difference between users on the "to" side of the third target age group and users on the "to" side of each second age group is less than the first preset threshold, and the difference between the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the third target age group and the standard liking rate of users on the "from" side of the first age group for users on the "to" side of the fourth target age group is less than the second preset threshold, the third target age group and the fourth target age group are aggregated into a second age bucket.

7. A device for obtaining a liking rate, characterized in that, The device includes an acquisition module, a determination module, a search module, and a calculation module; wherein... The acquisition module is used to acquire, for a target user, attribute information of the target user and first historical interaction data of the target user within a first specified time period; wherein, the first historical interaction data includes display data and like data based on the display data; The determining module is used to determine, based on the attribute information, the first historical interaction data, and the pre-acquired group preference matrix, the first target age bucket to which the target user belongs, and at least one second target age bucket to which the to-side users involved in the first historical interaction data belong; wherein, the group preference matrix represents the standard liking rate of the from-side users of each first age bucket to the to-side users of each second age bucket; The search module is used to search, for each second target age bucket, the standard like rate and standard display value of the "from" side users of the first target age bucket to the "to" side users of the second target age bucket from the group preference matrix and the pre-acquired smoothing parameter matrix; wherein, the smoothing parameter matrix represents the standard display value of the "from" side users of each first age bucket to the "to" side users of each second age bucket. The calculation module is used to calculate the target user's like rate for each second target age bucket based on the standard like rate, the standard display value, and the number of target display data and target like data for that second target age bucket in the first historical interaction data.

8. The apparatus according to claim 7, characterized in that, The group preference matrix is ​​obtained through the following method: Obtain massive amounts of second historical interaction data from users within a second specified time period, and based on the second historical interaction data, obtain the standard liking rate of users from the first age group on the to side of each second age group. According to a preset aggregation rule, the standard liking rates of users from the first age group to users from the second age group are aggregated to obtain the group preference matrix, which represents the standard liking rates of users from the first age group to users from the second age group; the number of first age groups is less than the number of first age groups and / or the number of second age groups is less than the number of second age groups.

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

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

Images