Group member recommendation method, device, server, storage medium and system
By calculating and evaluating the recommendation index of socially related accounts, the problem of screening group members in the existing technology is solved, and high-quality socially related accounts are recommended, which improves the social interaction experience between anchors and fans.
Patent Information
- Application Number
- CN202010358274.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-04-29
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2040-04-29
AI Technical Summary
It is difficult to effectively screen and recommend social group members in the existing technology, especially when the number of anchor fans is large, it is difficult to establish high-quality social relationships.
By obtaining the account attribute data of the target user and social-related accounts, calculating the recommendation index, selecting social-related accounts that meet the preset conditions as group members for recommendation, and assessing social interaction interest based on user behavior and identity portrait.
It has achieved the recommendation of fan accounts with high social interaction interest to the anchor, which has enhanced the quality and interaction effect of group members and improved the interactive experience of social groups.
Smart Images

Figure CN111581521B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of Internet technology, and in particular to a method, device, server, storage medium and system for recommending group members. Background Art
[0002] With the development of Internet technology, more and more application software have emerged for users to use to meet their needs for work, entertainment, etc. These application software usually include follow, subscribe or similar functions, through which users (called fans) can follow or subscribe to other users of interest, so as to obtain the content published by the followed or subscribed users (called anchors) in a timely manner.
[0003] In actual applications, when a host has a large number of fans, the host usually hopes to establish a social group to interact with fans with close social relationships and improve his or her influence. Summary of the invention
[0004] The present disclosure provides a method, device, server, storage medium and system for recommending group members, which can be used to recommend group members to anchors. The technical solution of the present disclosure is as follows:
[0005] According to a first aspect of an embodiment of the present disclosure, a method for recommending group members is provided, comprising:
[0006] In response to a group member recommendation triggering operation implemented by a target user account, obtaining a set of social associated accounts of the target user account;
[0007] For each social association account in the set of social association accounts, obtaining a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account;
[0008] A social associated account whose recommendation index meets a preset recommendation condition is selected as a target group member, and displayed to the target user account to trigger the target user account to establish a group relationship with the target group member.
[0009] According to a second aspect of an embodiment of the present disclosure, a method for recommending group members is provided, comprising:
[0010] Receive a recommendation index acquisition request for a target user account from a client;
[0011] In response to the recommendation index acquisition request, for each social association account in the set of social association accounts of the target user account, acquiring a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account;
[0012] The recommendation index of each social association account in the social association account set is sent to the client, so that the client can recommend group members to the target user account according to the recommendation index.
[0013] According to a third aspect of an embodiment of the present disclosure, there is provided a device for recommending group members, including:
[0014] a response unit configured to execute a group member recommendation trigger operation in response to a target user account, and obtain a set of social associated accounts of the target user account;
[0015] an acquisition unit configured to acquire, for each social association account in the set of social association accounts, a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account;
[0016] The display unit is configured to select social related accounts whose recommendation index meets the preset recommendation conditions as target group members, and display them to the target user account to trigger the target user account to establish a group relationship with the target group members.
[0017] According to a fourth aspect of an embodiment of the present disclosure, there is provided a device for recommending group members, including:
[0018] A request receiving unit, configured to execute receiving a recommendation index acquisition request of a target user account from a client;
[0019] a request response unit configured to execute, in response to the recommendation index acquisition request, acquiring, for each social association account in the set of social association accounts of the target user account, a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account;
[0020] The sending unit is configured to send the recommendation index of each social association account in the social association account set to the client, so that the client can recommend group members to the target user account according to the recommendation index.
[0021] According to a fifth aspect of an embodiment of the present disclosure, a client is provided, including:
[0022] processor;
[0023] a memory for storing instructions executable by the processor;
[0024] The processor is configured to execute the instructions to implement the method for recommending group members provided in the embodiment of the present disclosure.
[0025] According to a sixth aspect of an embodiment of the present disclosure, a server is provided, including:
[0026] processor;
[0027] a memory for storing instructions executable by the processor;
[0028] The processor is configured to execute the instructions to implement the group member recommendation method provided in the embodiment of the present disclosure.
[0029] According to a seventh aspect of an embodiment of the present disclosure, a group member recommendation system is provided, including the client provided by an embodiment of the present disclosure and the server provided by an embodiment of the present disclosure.
[0030] According to an eighth aspect of an embodiment of the present disclosure, a storage medium is provided. When instructions in the storage medium are executed by a processor of a server, the server is enabled to execute the method for recommending group members provided by an embodiment of the present disclosure; or, when instructions in the storage medium are executed by a processor of a client, the client is enabled to execute the method for recommending group members provided by an embodiment of the present disclosure.
[0031] According to a ninth aspect of an embodiment of the present disclosure, a computer program product is provided. When the computer program product is run on a computer, the computer is enabled to execute the method for recommending group members provided by an embodiment of the present disclosure.
[0032] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:
[0033] By adopting the group member recommendation method provided by the embodiment of the present disclosure, in response to the group member recommendation triggering operation implemented by the target user account, the social association account set of the target user account can be obtained, and then the account attribute data of the target user account and the account attribute data of each social association account in the social association account set can be obtained, so as to calculate the recommendation index of each social association account in the social association account set, and select the social association account whose recommendation index meets the preset recommendation condition as the target group member to display to the target user account, so as to trigger the target user account to establish a group relationship with the target group member. Since the account attribute data of the target user account can usually reflect the identity portrait or user behavior of the target user account, and the account attribute data of the social association account can also reflect the identity image or user behavior of the social association account, the recommendation index obtained based on the account attribute data of the target user account and the account attribute data of the social association account can reflect the interest of the social association account in social interaction with the target user account. Therefore, through this method, the social association account can be recommended to the target user account as a target group member, and the target user account can be guided to add the social association account with a high degree of social interaction interest to the group initiated by it, so as to achieve high-quality social group interaction.
[0034] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.
[0036] Figure 1 The figure is a schematic diagram of a specific flow chart of a method for recommending group members according to an exemplary embodiment.
[0037] Figure 2 It is a schematic diagram of a specific flow chart of a method for recommending group members according to another exemplary embodiment.
[0038] Figure 3 It is a schematic diagram of a specific structure of a device for recommending group members according to an exemplary embodiment.
[0039] Figure 4 It is a schematic diagram of a specific structure of a device for recommending another group member according to an exemplary embodiment.
[0040] Figure 5 It is a schematic diagram of a specific structure of a device for recommending another group member according to an exemplary embodiment.
[0041] Figure 6 It is a schematic diagram of a specific structure of a device for recommending another group member according to an exemplary embodiment.
[0042] Figure 7 It is a schematic diagram of a specific structure of a server according to an exemplary embodiment.
[0043] Figure 8 It is a schematic diagram of a specific structure of a group member recommendation system according to an exemplary embodiment.
[0044] Fig. 9 The figure is a schematic diagram of a specific flow chart of a group member recommendation method based on a group member recommendation system according to an exemplary embodiment. DETAILED DESCRIPTION
[0045] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.
[0046] It should be noted that the terms "first", "second", "third", "fourth", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein.
[0047] As mentioned above, when a host has a large number of fans, in order to facilitate interaction with fans, it is usually necessary to establish a social group, and then add some fans to the social group for interaction. Therefore, how to select some fans from many fans to recommend group members to the host.
[0048] Based on this, the embodiment of the present disclosure provides a method for recommending group members, which can be used to recommend group members to the anchor. Figure 1 The flowchart of the method is shown in Figure 1In the group member recommendation method shown, the execution subject can usually be a client on a mobile phone, computer or other device, and the target user (such as a host) can perform corresponding operations on the client to trigger the recommendation of group members. The specific steps of the method are as follows:
[0049] Step S11: In response to a group member recommendation triggering operation implemented by a target user account, a set of social associated accounts of the target user account is obtained.
[0050] The target user account here can usually be the user account of the anchor, so the target user account will be referred to as the anchor account in the following. The target user account's social association account set includes multiple social association accounts, which usually follow or subscribe to the target user account, that is, the social association account is a user account of the anchor's fan, so the social association account will be referred to as the fan account in the following.
[0051] In actual applications, the operation behavior of the anchor account can be detected to determine whether it has implemented the trigger operation of group member recommendation. For example, when a certain anchor has a large number of fans, it is necessary to screen some of the fans to establish a group. The recommendation operation of the group member can be implemented on terminals such as computers and mobile phones. For example, a group member recommendation button can be added to the live broadcast APP on the mobile phone for the anchor account, and the group member recommendation button can be used on the anchor account to trigger the recommendation of group members. Therefore, when it is determined that the anchor account has implemented the trigger operation of group member recommendation, in response to the operation, the social associated account set of the anchor account can be obtained.
[0052] Step S12: Obtain the account attribute data of the anchor account and the account attribute data of each fan account in the social associated account set.
[0053] The account attribute data of a streamer account can usually reflect the identity image or user behavior of the streamer account, such as the user behavior data of the streamer account or the streamer portrait of the streamer. The account attribute data of a fan account can usually reflect the identity image or user behavior of the fan account, such as the user behavior data of the fan account or the fan portrait of the fan.
[0054] In this step S12 to the subsequent step S14, the situation in which the account attribute number of the anchor account is the user behavior data of the anchor account and the account attribute number of the fan account is the user behavior data of the fan account can be explained.
[0055] The user behavior data of the anchor account can generally be the content publishing behavior data of the anchor account in a preset time period, which can reflect the frequency, number of times, duration, etc. of the anchor's content publishing in the preset time period. According to different scenarios, the content can be articles, audio and video, live broadcasts, etc., which are not limited here.
[0056] In actual applications, the content publishing behavior data of the anchor account may be the number of times the anchor account publishes content within a preset time period. For example, if the anchor account publishes 3 articles within the preset time period, the content publishing behavior data is 3. The content publishing behavior data may also be the number of days the anchor account publishes content within the preset time period. For example, if the anchor account uploads videos for 4 days within the preset time period, the content publishing behavior data is 4. In the live broadcast scenario, the content publishing behavior data may also be the number of times the anchor account broadcasts live within the preset time period, or the number of days the anchor account broadcasts live within the preset time period, or the length of time the anchor account broadcasts live within the preset time period, etc.
[0057] Usually, the number, frequency, and duration of fans' interactions with anchors or their comments on the content posted by anchor accounts can reflect the fans' interest in social interaction with anchors. For example, if a fan frequently interacts with an anchor or frequently comments on the content posted by an anchor account, it may indicate that the fan has a high interest in social interaction with the anchor. Depending on different scenarios, the interaction or evaluation behavior (hereinafter collectively referred to as evaluation behavior) here can be likes, comments, rewards, and / or red envelopes, etc., or other behaviors of interacting with the anchor.
[0058] Therefore, the user behavior data of the fan account may be the evaluation behavior data of the fan account, and the evaluation behavior data may be the number of times the fan account rewards, comments and / or likes the content posted by the anchor account within a preset time period. For example, within the preset time period, the fan account rewards 1 time, comments 2 times and likes 4 times on the content posted by the anchor account, then the evaluation behavior data may be 1 (only the number of rewards is counted), 2 (only the number of comments is counted), 4 (only the number of likes is counted), 3 (rewards plus comments), 5 (rewards plus likes), 6 (comments plus likes) or 7 (total number of rewards, comments and likes).
[0059] The evaluation behavior data of the fan account and the anchor account can also be the number of days that the fan account rewards, comments and / or likes the content posted by the anchor account within the preset time period. For example, within the preset time period, the fan account rewards, comments and / or likes the content posted by the anchor for 8 days, then the evaluation behavior data is 8. In the live broadcast scenario, the evaluation behavior data can also be the number of times the fan account rewards, comments and / or likes the live broadcast of the anchor account within the preset time period; or the number of days that the fan account rewards, comments and / or likes the live broadcast of the anchor account within the preset time period; or the length of time that the fan account rewards, comments and / or likes the live broadcast of the anchor account within the preset time period, etc.
[0060] It should be noted that the above-mentioned preset time period can be set according to actual needs. For example, it can be 1 year, 1 quarter, 30 days, 10 days, 20 days, 1 week, 1 day, etc., so as to obtain the content publishing behavior data of the anchor account within the preset time period and the evaluation behavior data of the fan account on the anchor.
[0061] For example, the number of days the anchor publishes content within 30 days, and the number of days the fans to be evaluated reward, comment and / or like the content published by the anchor within 30 days are obtained.
[0062] Step S13: For each fan account in the social association account set, the recommendation index of the fan account is obtained according to the account attribute data of the fan account and the account attribute data of the anchor account.
[0063] When the account attribute data of the anchor account is the user behavior data of the anchor account, and the account attribute data of the fan account is the user behavior data of the fan account, since the user behavior data of the anchor account is usually the content publishing behavior data of the anchor account, it can reflect the frequency, number of times, duration, etc. of content publishing by the anchor account, and the user behavior data of the fan account can be the evaluation behavior data of the fan account and the anchor account, which can reflect the number, frequency, duration, etc. of the interaction between the fan account and the anchor account. Therefore, the recommendation index of the fan account can be calculated based on the user behavior data of the anchor account and the user behavior data of the fan account, and the recommendation index can reflect the interest of the fan in social interaction with the anchor.
[0064] Since the recommendation index can reflect the fans' interest in social interaction with the anchor, the recommendation index is usually positively correlated with the user behavior data (evaluation behavior data) of the fan account, that is, the larger the evaluation behavior data, the larger the recommendation index; on the contrary, the recommendation index is usually inversely correlated with the user behavior data (content publishing behavior data) of the anchor account, that is, the larger the content publishing behavior data, the smaller the recommendation index. In general, the larger the evaluation behavior data and the smaller the content publishing behavior data, the smaller the frequency, number, and duration of the anchor account's content publishing, but the greater the number, frequency, and duration of the fan account's interaction with the anchor account, the higher the fans' interest in social interaction with the anchor, and the larger the recommendation index.
[0065] Therefore, when the account attribute data of the anchor account is the user behavior data of the anchor account, and the account attribute data of the fan account is the user behavior data of the fan account, the recommendation index of the fan account is obtained according to the user behavior data of the fan account and the user behavior data of the anchor account. One way to calculate the recommendation index is: recommendation index = P × (user behavior data of the fan account / user behavior data of the anchor account), where P is a preset constant, such as 100, 200 or other values. That is, the corresponding recommendation index is calculated according to the ratio of the user behavior data of the fan account to the user behavior data of the anchor account, and the ratio is multiplied by a preset constant.
[0066] For example, the user behavior data of the anchor account is the number of days that the anchor account publishes content within 30 days, and the user behavior data of the fan account is the number of days that the fan account comments on or likes the content published by the anchor account within 30 days, and when P is 100, the recommendation index = 100×(number of days that the fan account comments on or likes the content published by the anchor account within 30 days / number of days that the anchor account publishes content within 30 days).
[0067] According to the user behavior data of the fan account and the user behavior data of the anchor account, the recommendation index of the fan account is obtained. Another way to calculate the recommendation index is: recommendation index = Y × (user behavior data of the fan account - user behavior data of the anchor account), where Y can be a preset constant that is the same as or different from X, such as 100 or other values. That is, according to the difference between the user behavior data of the fan account and the user behavior data of the anchor account, the difference is multiplied by a preset constant to calculate the corresponding recommendation index.
[0068] Of course, the recommendation index can also be calculated by other methods, as long as the calculation result is positively correlated with the user behavior data of the fan account and inversely correlated with the user behavior data of the anchor account. The specific calculation method is not limited here.
[0069] Step S14: Select fan accounts whose recommendation index meets the preset recommendation conditions as target group members and display them to the anchor account.
[0070] After calculating the recommendation index of each fan account, fan accounts that meet the preset recommendation conditions can be selected as target group members for display to the anchor account. For example, fan accounts with a recommendation index greater than a third preset threshold are selected as target group members for display to the anchor account to trigger the anchor account to establish a group relationship with the target group members.
[0071] By adopting the group member recommendation method provided by the embodiment of the present disclosure, in response to the group member recommendation trigger operation implemented by the anchor account, the social associated account set of the anchor account can be obtained, and then the account attribute data of the anchor account and the account attribute data of each fan account in the social associated account set can be obtained, so as to calculate the recommendation index of each fan account in the social associated account set, and select the fan account whose recommendation index meets the preset recommendation conditions as the target group member to display to the anchor account, so as to trigger the anchor account to establish a group relationship with the target group member. Since the account attribute data of the anchor account can usually reflect the identity portrait or user behavior of the anchor account, and the account attribute data of the fan account can also reflect the identity image or user behavior of the fan account, the recommendation index obtained based on the account attribute data of the anchor account and the account attribute data of the fan account can reflect the interest of the fan account in social interaction with the anchor account. Therefore, through this method, the social associated account can be recommended to the anchor account as a target group member, and the anchor account can be guided to add the fan account with a high degree of social interaction interest to the group initiated by it, so as to achieve high-quality social group interaction.
[0072] Based on Figure 1 The invention concept of the group member recommendation method shown in FIG. 1 is the same as that of the invention concept of the group member recommendation method shown in FIG. 1 . The embodiment of the present disclosure can also provide a group member recommendation method. The recommendation method is similar to Figure 1The difference of the group member recommendation method shown is that the execution subject of the recommendation method can usually be a server. For example, the anchor can perform corresponding operations on the client to trigger the recommendation of group members, so that for each social associated account in the social associated account set, the client sends a recommendation index request to the server to trigger the server to obtain and return the recommendation index of the social associated account. At this time, the server first receives the recommendation index acquisition request of the target user account from the client, and then responds to the recommendation index acquisition request, for each social associated account in the social associated account set of the target user account, according to the account attribute data of the social associated account and the account attribute data of the target user account, obtains the recommendation index of the social associated account; then sends the recommendation index of each social associated account in the social associated account set to the client, so that the client can recommend group members to the target user account according to the recommendation index.
[0073] Obviously, the method executed by the server is the same as Figure 1 The group member recommendation method shown in the figure (mainly performed by the client) is based on the same inventive concept, and thus can also solve the problems in the prior art. In addition, for the method performed by the server, if there is any unclear point, you can also refer to Figure 1 The relevant contents of the method for recommending group members shown will not be repeated here.
[0074] In actual applications, a fan account may usually follow multiple anchor accounts at the same time, and thus join the groups created by these anchor accounts. Often, when a fan joins groups created by multiple different anchor accounts, it may affect the fan's attention to a single anchor, which in turn leads to a decrease in the fan's interest in social interaction with the anchor. Therefore, the recommendation index decay value of the fan account can be calculated based on the number of groups the fan has joined and the activity in these groups, so as to correct the recommendation index.
[0075] Therefore, the network user quality assessment method provided by the embodiment of the present disclosure may also include: for each fan account in the social association account set, obtaining the number of groups joined by the fan account and the activity level in each group; then, according to the number of groups and the activity levels, calculating the recommendation index decay value of each fan account in the social association account set; after calculating the recommendation index decay value of each fan account, for each fan account in the social association account set, the recommendation index of the fan account can be obtained by combining the user behavior data of the fan account, the user behavior data of the anchor account and the recommendation index decay value. For example, the ratio of the number of days that the fan account rewards, comments and / or likes the content posted by the anchor account within a preset time period to the number of days that the anchor account posts content within the preset time period is calculated, and then the ratio is multiplied by 100, and then the corresponding recommendation index decay value is subtracted, so as to calculate the recommendation index of the fan account.
[0076] The recommendation index calculated by this method is combined with other information of the fan account, such as other groups joined and the activity level in these groups, and the recommendation index decay value is deducted, so that the final calculated recommendation index more accurately reflects the fans' interest in social interaction with the anchor.
[0077] Among them, the following methods can be used to determine the activity of fan accounts in each group:
[0078] Each group can be taken as the current group, and then the activity of the fan account in the current group can be calculated based on the total interaction time of the fan account in the current group (such as total days, total hours, etc.) and the total time the fan account has joined the current group (such as total days, total hours, etc.).
[0079] Generally, when the total interaction time of a fan account in the current group is longer and the total time of joining the current group is shorter, it means that the fan account is more active in the current group, that is, the more active it is; on the contrary, when the total interaction time of the fan account in the current group is shorter and the total time of joining the current group is longer, it means that the fan account is less active in the current group, that is, the less active it is.
[0080] Therefore, the total interaction time / total time of joining the current group can be determined as the activity of the fan account in the current group; or the total interaction time-total time of joining the current group can be determined as the activity of the fan account in the current group. The total interaction time of the fan account in the current group can be the total time of the fan account speaking, rewarding, sending red envelopes, and / or other interactive methods in the current group.
[0081] For example, if the number of groups created by other anchors that the fan account has joined is 3, you can set these 3 groups as the current groups respectively, and then calculate the activity of the fan account in the current group by the total interaction time of the fan account in the current group / the total time of joining the current group, and finally calculate the corresponding activity of the fan account in these 3 groups respectively.
[0082] After determining the number of groups of other anchors that each fan account has joined and the activity level in each group, the recommendation index decay value of each fan account in the social association account set can be calculated based on the number of groups and the activity levels. Generally, the more groups a fan account has and the greater the activity level, the more active the fan account is in other anchors, and the corresponding recommendation index decay value is larger. Therefore, the recommendation index decay value can be calculated by the following formula:
[0083] Recommended exponential decay value = a × number of groups + average activity of the group / first preset threshold + b;
[0084] That is, for each social association account in the set of social association accounts, the average value of the activity of each group joined by the social association account (referred to as the group average activity) can be obtained; then, when the maximum activity among each activity is greater than the second preset threshold, the recommended index decay value of the social association account is obtained according to the product of the number of groups of the social association account and the preset constant (a), (in this case, b is the second preset constant), the second preset constant, and the quotient of the group average activity and the first preset threshold; or, when the maximum activity among each activity is less than or equal to the second preset threshold, the recommended index decay value of the social association account is obtained according to the product of the number of groups of the social association account and the preset constant (a), (in this case, b is the third preset constant), the third preset constant, and the quotient of the group average activity and the first preset threshold, wherein the third preset constant is less than the second preset constant.
[0085] In the above formula: a is a preset constant, such as 0.05, 0.1, 0.2 or other values;
[0086] When the maximum activity among all the activities is greater than the second preset threshold, the value of b is the second preset constant; and when the maximum activity among all the activities is less than or equal to the second preset threshold, the value of b is the third preset constant which is less than the second preset constant. When the maximum activity among all the activities is greater than the second preset threshold, it means that the fan account is more active in a group of other anchors, and the corresponding value of b is the relatively large second preset constant, so that the recommended exponential decay value is relatively large. Otherwise, the value of b is the relatively small third preset constant. In practical applications, the second preset constant can be 10, 15, 20, 25, etc., and the third preset constant can be 0, 5, 8.
[0087] In addition, the first preset threshold can usually be determined based on the activity of each network user in the application software platform. For example, the activity of each network user in the application software platform can be determined first, and then the average value can be calculated, and the average value can be used as the first preset threshold; or, a portion of network users can be randomly selected from the network users of the application software platform, and then the average value of the activity of the selected portion of network users can be calculated as the first preset threshold; or, the activity of each network user in the application software platform can be determined first, and then the median or mode of each activity can be determined as the first preset threshold.
[0088] The second preset threshold value can also be generally based on the activity of network users in the application software platform. For example, after determining the activity of each network user in the application software platform, these activities are arranged in order from large to small, and the maximum activity is selected and multiplied by 70% (or 60%, 65%, 50%, etc.) as the second preset threshold value; or, these activities are arranged in order from large to small, and the activity of the 70th digit in the sequence (that is, 30% of the activity is greater than this activity, and 70% of the activity is less than this activity) is selected as the second preset threshold value; or, the second preset threshold value can also be determined by other data statistics and analysis methods based on the activity of network users in the application software platform.
[0089] In actual applications, when the account attribute data of the anchor account is the anchor portrait of the anchor account, and the account attribute data of the fan account is the fan portrait of the fan account, the recommendation index of each fan account can also be obtained through the following group member recommendation method. Figure 2 The flowchart of the method is shown in Figure 2 In the group member recommendation method shown, the execution subject can also generally be a client on a mobile phone, computer or other device. For example, the anchor can perform corresponding operations on the client to trigger the recommendation of group members. The specific steps of the method are as follows:
[0090] Step S21: In response to a group member recommendation trigger operation implemented by the anchor account, a first similarity between the anchor classification category to which the anchor account belongs and the target anchor classification category is obtained by screening the anchor classification categories of other anchors followed by each fan account in the social association account set.
[0091] Among them, the anchor classification category to which the anchor account belongs can usually reflect the classification characteristics of the anchor account. For example, the common anchor account classification categories may include: appearance anchor, game anchor, food anchor, etc. When the anchor classification category to which a certain anchor account belongs is game anchor, it means that the anchor mainly provides games or related content. Therefore, the anchor classification category of the anchor account can usually be determined based on the characteristics of the content published by the anchor account, or the classification category to which it belongs can be obtained from the anchor portrait of the anchor account, where the anchor portrait usually includes the anchor classification category to which it belongs and the target fan classification category.
[0092] The target anchor classification category can be obtained by screening from the anchor classification categories of other anchors followed by each fan account in the social association account set. For example, among the anchor classification categories of other anchors followed by the fan account, the anchor classification category with the largest number is food anchor, and the food anchor can be determined as the target anchor classification category; of course, the anchor classification categories can also be sorted from large to small according to the number of anchor classification categories, and the top 3 or top 5 anchor classification categories can be selected, and the selected anchor classification categories can all be determined as target anchor classification categories.
[0093] When determining the first similarity between the anchor classification category to which the anchor account belongs and the target anchor classification category, the anchor classification category to which the anchor account belongs and the target anchor classification category can be converted into vectors, and then the cosine similarity between the vectors can be calculated as the first similarity, or the distance between the vectors (such as Euclidean distance) can be calculated, and the reciprocal of the distance can be used as the first similarity. Of course, the first similarity can also be determined based on the semantic similarity between the anchor classification category to which the anchor account belongs and the target anchor classification category.
[0094] The first similarity calculated by the above method can reflect the similarity between the anchor account and other anchors followed by the fan account.
[0095] Step S22: for each fan account in the social association account set, obtain a second similarity between the fan classification category to which the fan account belongs and the target fan classification category, wherein the target fan classification category is obtained by screening the fan classification categories to which each fan account in the social association account set belongs.
[0096] The fan classification category to which the fan account belongs can be obtained by analyzing the fan portrait of the fan account. The fan portrait of the fan account may include: the gender of the fan, the region of the fan, the type of content preferred by the fan, the anchor classification category to which other anchor accounts followed by the fan belong, etc. The fan portrait can be used to outline the fan account and reflect the bias of certain demands or interests of the fan account.
[0097] When analyzing the fan portrait of the fan account to obtain the fan classification category to which it belongs, the fan portrait of the fan account can be input into a pre-trained fan classification model, so as to analyze the fan portrait of the fan account through the fan classification model to obtain the fan classification category to which it belongs. The fan classification model can process the input fan portrait and output the corresponding fan classification category.
[0098] In practical applications, the fan classification model can be trained in the following way: first, collect fan portraits of fans that meet the preset collection conditions, where the preset collection conditions include following anchors, being active as normal users, and having registered for a preset time threshold; then, take each collected fan portrait as a sample, train the sample based on machine learning and neural networks, and verify it based on the verification sample, and finally obtain the fan classification model. Among them, some fan portraits are manually marked as verification samples according to the preset classification rules.
[0099] The target fan classification category can be obtained by screening from the fan classification categories to which each fan account in the social association account set belongs. For example, the fan classification categories to which each fan account of the anchor account belongs can be determined first, and then the top N fan classification categories with the largest number of fans are selected as the target fan classification categories, where N can be 1, 2 or other positive integers.
[0100] When determining the second similarity between the fan classification category to which a certain fan account belongs and the target fan classification category, the fan classification category to which the fan account belongs and the target fan classification category can also be converted into vectors, and then the cosine similarity between the vectors can be calculated as the second similarity, or the distance between the vectors (such as Euclidean distance) can be calculated, and the inverse of the distance can be used as the second similarity; of course, the second similarity can also be determined based on the semantic similarity between the fan classification category to which the fan account belongs and the target fan classification category.
[0101] Therefore, the second similarity calculated by this method can reflect the similarity between the fan account and the main fans of the anchor account.
[0102] Step S23: Obtaining a recommendation index of each social association account in the social association account set according to the corresponding first similarity and second similarity.
[0103] Since the first similarity can reflect the similarity between the anchor account and other anchors followed by the fan account, and the second similarity can reflect the similarity between the fan account and the main fans of the anchor account, the corresponding recommendation index calculated based on the weighted average of the first similarity and the second similarity can also reflect the interest of the fan account in social interaction with the anchor account, and can therefore be used to recommend group members, such as selecting fan accounts whose recommendation indexes meet preset recommendation conditions as target group members to display to the anchor account. When obtaining the recommendation index of each social associated account in the set of social associated accounts based on the corresponding first similarity and second similarity, the recommendation index can be calculated based on the weighted average of the first similarity and the second similarity.
[0104] Specifically, the recommendation index can be calculated according to the following formula:
[0105] Recommendation index=(a1×first similarity+a2×second similarity)×Z, wherein a1 and a2 are both positive numbers greater than 0 and less than 1, and a1+a2=1; Z is a preset constant, such as 100, 200 or other values.
[0106] Step S24: Select fan accounts whose recommendation index meets the preset recommendation conditions as target group members and display them to the anchor account.
[0107] For step S21 and step S22 in the network user quality assessment method, their specific execution order is not limited here. For example, they can be executed in the order of S21, S22, step S23 and step S24, or in the order of S22, S21, step S23 and step S24. Step S21 and step S22 can also be executed at the same time, or in other execution orders.
[0108] In addition, in practical applications, after the recommendation index (referred to as the second recommendation index) is calculated by step S23, the third recommendation index can be calculated based on the recommendation index and the recommendation index obtained by the above step S13 (referred to as the first recommendation index), so that it can also be used to recommend group members. For example, for each social association account in the social association account set, after obtaining its second recommendation index and first recommendation index respectively by step S23 and step S13, the third recommendation index is obtained according to the weighted summation result of the second recommendation index and the first recommendation index. Thus, the fan account whose third recommendation index meets the preset recommendation conditions is selected as the target group member to be displayed to the anchor account. Since the third recommendation index is calculated by the first recommendation index and the second recommendation index, and the first recommendation index and the second recommendation index are calculated based on two different concepts, the third recommendation index is usually more able to reflect the fans' interest in social interaction with a certain anchor. Of course, in the process of weighted summation, the respective weights can be determined according to the actual situation.
[0109] In practical applications, the first recommendation index, the second recommendation index, and the third recommendation index can be calculated by any of the above methods, so as to recommend group members to the anchor account. For example, the server can calculate the first recommendation index (second recommendation index or third recommendation index) of each fan account of anchor account A for a certain anchor account (referred to as anchor account A), and recommend fan accounts to anchor account A based on the first recommendation index corresponding to each fan account.
[0110] Of course, based on Figure 2 The invention concept of the group member recommendation method shown in FIG. 1 is the same as that of the invention concept of the group member recommendation method shown in FIG. 1 . The embodiment of the present disclosure can also provide a group member recommendation method. The recommendation method is similar to Figure 2The difference between the method for recommending group members shown is that the execution subject of the recommendation method can also usually be a server. For example, the anchor can perform corresponding operations on the client to trigger the recommendation of group members, so that for each social associated account in the set of social associated accounts, the client sends a recommendation index request to the server to trigger the server to obtain and return the recommendation index of the social associated account. At this time, the server first receives the recommendation index acquisition request of the target user account from the client; then, in response to the recommendation index acquisition request, obtains the first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category, and for each social associated account in the set of social associated accounts of the target user account, obtains the second similarity between the fan classification category to which the social associated account belongs and the target fan classification category; then, based on the corresponding first similarity and second similarity, obtains the recommendation index of each social associated account in the set of social associated accounts; then sends the recommendation index of each social associated account in the set of social associated accounts to the client, so that the client can recommend group members to the target user account based on these recommendation indices. Obviously, the method executed by the server is different from Figure 2 The group member recommendation method shown in the figure (executed by the client) is based on the same inventive concept, and thus can also solve the problems in the prior art. In addition, for the method executed by the server, if there is any unclear point, you can also refer to Figure 1 The relevant contents of the method for recommending group members shown will not be repeated here.
[0111] Based on the embodiments of the present disclosure Figure 1 The present disclosure also provides a group member recommendation device based on the same inventive concept as the group member recommendation method provided in the disclosure. The device includes a response unit, an acquisition unit and a display unit, wherein:
[0112] a response unit configured to execute a group member recommendation trigger operation in response to a target user account, and obtain a set of social associated accounts of the target user account;
[0113] an acquisition unit configured to acquire, for each social association account in the set of social association accounts, a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account;
[0114] The display unit is configured to select social related accounts whose recommendation index meets the preset recommendation conditions as target group members, and display them to the target user account to trigger the target user account to establish a group relationship with the target group members.
[0115] Since the device adopts the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, it can also solve the problems in the prior art. In addition, for the device, if there is any unclear point, you can refer to the corresponding content in the method embodiment, and no further description will be given here.
[0116] In addition, in actual applications, the account attribute data of the anchor account can be the user behavior data of the anchor account or the identity portrait of the anchor, and the account attribute data of the fan account can be the user behavior data of the fan account or the identity portrait of the fan. When the account attribute data of the anchor account is the user behavior data of the anchor account, and the account attribute data of the fan account is the user behavior data of the fan account, another group member recommendation device can be provided. Figure 3 FIG. 4 is a block diagram of a device for recommending group members according to an exemplary embodiment. Figure 3 The device 30 includes: a second response unit 301, a second acquisition unit 302 and a second display unit 303, wherein:
[0117] The second response unit 301 is configured to execute a group member recommendation trigger operation in response to a target user account, and obtain a set of social associated accounts of the target user account;
[0118] The second acquisition unit 302 is configured to execute, for each social association account in the social association account set, acquiring a recommendation index of the social association account according to user behavior data of the social association account and user behavior data of the target user account;
[0119] The second display unit 303 is configured to select social related accounts whose recommendation index meets the preset recommendation condition as target group members, and display them to the target user account to trigger the target user account to establish a group relationship with the target group members.
[0120] Since the device 30 adopts the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, it can also solve the problems in the prior art. In addition, for the device 30, if there is any unclear point, please refer to the corresponding content in the method embodiment, and no further description will be given here.
[0121] Similarly, when the account attribute number of the anchor account is the user behavior data of the anchor account, and the account attribute number of the fan account is the user behavior data of the fan account, based on the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, the present disclosure also provides a group member recommendation device. Figure 4 FIG. 4 is a block diagram of a device for recommending group members according to an exemplary embodiment. Figure 4 The device 40 includes: a request receiving unit 401, a request responding unit 402 and a sending unit 403, wherein:
[0122] The request receiving unit 401 is configured to receive a recommendation index acquisition request of a target user account from a client;
[0123] The request response unit 402 is configured to execute, in response to the recommendation index acquisition request, acquiring, for each social association account in the set of social association accounts of the target user account, a recommendation index of the social association account according to user behavior data of the social association account and user behavior data of the target user account;
[0124] The sending unit 403 is configured to send the recommendation index of each social association account in the social association account set to the client, so that the client can recommend group members to the target user account according to the recommendation index.
[0125] Since the device 40 also adopts the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, it can also solve the problems in the prior art. In addition, for the device 40, if there is any unclear point, please refer to the corresponding content in the method embodiment, and no further description will be given here.
[0126] Based on the embodiments of the present disclosure Figure 2 Based on the same inventive concept as the method for recommending group members provided in the disclosure, the present disclosure also provides a device for recommending group members. Figure 5 FIG. 4 is a block diagram of a device for recommending group members according to an exemplary embodiment. Figure 5 The device 50 includes: a similarity acquisition unit 501, a recommendation index acquisition unit 502 and a third display unit 503, wherein:
[0127] The similarity acquisition unit 501 is configured to execute a group member recommendation trigger operation in response to a target user account, acquire a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category, and acquire a second similarity between the fan classification category to which the social association account belongs and the target fan classification category for each social association account in the social association account set of the target user account, wherein the target anchor classification category is obtained by screening from the anchor classification categories to which other user accounts followed by each social association account in the social association account set of the target user account respectively belong, and the target fan classification category is obtained by screening from the fan classification categories to which each social association account in the social association account set respectively belongs;
[0128] A recommendation index acquisition unit 502 is configured to acquire a recommendation index of each social association account in the social association account set according to the corresponding first similarity and second similarity;
[0129] The third display unit 503 is configured to select social related accounts whose recommendation index meets the preset recommendation condition as target group members, and display them to the target user account to trigger the target user account to establish a group relationship with the target group members.
[0130] Since the device 50 also adopts the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, it can also solve the problems in the prior art. In addition, for the device 50, if there is any unclear point, please refer to the corresponding content in the method embodiment, and no further description will be given here.
[0131] Based on the same inventive concept as the group member recommendation method provided in the embodiment of the present disclosure, the present disclosure also provides a group member recommendation device. Figure 6 FIG. 4 is a block diagram of a device for recommending group members according to an exemplary embodiment. Figure 6 The device 60 includes: a second request receiving unit 601, a second request responding unit 602, a second recommendation index obtaining unit 603 and a second sending unit 604, wherein:
[0132] The second request receiving unit 601 is configured to receive a recommendation index acquisition request of a target user account from a client;
[0133] The second request response unit 602 is configured to execute, in response to the recommendation index acquisition request, obtaining a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category, and obtaining, for each social associated account in the social associated account set of the target user account, a second similarity between the fan classification category to which the social associated account belongs and the target fan classification category, wherein the target anchor classification category is obtained by screening from the anchor classification categories to which other user accounts followed by each social associated account in the social associated account set of the target user account respectively belong, and the target fan classification category is obtained by screening from the fan classification categories to which each social associated account in the social associated account set respectively belongs;
[0134] A second recommendation index acquisition unit 603 is configured to acquire a recommendation index of each social association account in the social association account set according to the corresponding first similarity and second similarity;
[0135] The second sending unit 604 is configured to send the recommendation index of each social association account in the set of social association accounts to the client, so that the client can recommend group members to the target user account according to the recommendation index.
[0136] In addition, the embodiments of the present disclosure may also provide a server. Figure 7 is a block diagram of a server 70 according to an exemplary embodiment. The server 70 includes a processor 701 and a memory 702 for storing instructions executable by the processor 701. The processor 701 is configured to execute the instructions to implement the method for recommending group members in the embodiment of the present disclosure.
[0137] In actual applications, the server 70 may also include a network interface 703, an I / O controller 704, a large-capacity storage device 705, and a bus 706 for connecting them.
[0138] In an exemplary embodiment, the present disclosure further provides a storage medium including instructions, such as a memory 702 including instructions, and the instructions can be executed by a processor 701 of a server 70 to complete the above method. The storage medium can be a non-transitory computer-readable storage medium, for example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.
[0139] The embodiment of the present disclosure may also provide a computer program product, which, when executed on a computer, enables the computer to execute the method for recommending group members provided by the embodiment of the present disclosure.
[0140] In addition, the embodiment of the present disclosure may also provide a group member recommendation system, the recommendation system 80 may include a client 90 and Figure 7 The server 70 shown. The client 90 may also include a processor and a memory, the memory is used to store the processor executable instructions, and the processor is configured to execute the instructions to implement any group member recommendation method provided in the embodiments of the present disclosure.
[0141] For ease of understanding, the present embodiment of the application can also be combined with the recommendation system 80 to illustrate the interaction between the client 90 and the server 70 in the group member recommendation process. In actual applications, when the anchor needs to filter out some fans from the fans as group members, the client 90 can be used to perform operations, such as by touching the corresponding controls on the client 90, so as to promote the recommendation of group members. Fig. 9 As shown, the group member recommendation method includes:
[0142] Step S101: the host triggers the recommendation of group members by performing a group member recommendation triggering operation on the client.
[0143] Step S102: the client sends a recommendation index acquisition request to the server in response to the group member recommendation triggering operation implemented by the anchor account.
[0144] Step S103: In response to the recommendation index acquisition request, the server acquires the recommendation index of each fan account in the social associated account set of the anchor account according to the account attribute data of the fan account and the account attribute data of the anchor account.
[0145] Among them, the account attribute data of the fan account can reflect the identity image or user behavior of the fan account, and the account attribute data of the anchor account can reflect the identity portrait or user behavior of the anchor account.
[0146] For example, the account attribute data of a fan account is the user behavior data of the fan account, and the account attribute data of a host account is the user behavior data of the host account. At this time, for each fan account in the social associated account set, the recommendation index of the fan account can be obtained based on the user behavior data of the fan account and the user behavior data of the host account.
[0147] In addition, the user behavior data of the fan account at least includes the evaluation behavior data of the fan account evaluating the content published by the anchor account within a preset time period; and the user behavior data of the anchor account at least includes the content publishing behavior data of the anchor account publishing content within a preset time period; then, for each fan account in the set of social associated accounts, based on the user behavior data of the fan account and the user behavior data of the anchor account, obtaining the recommendation index of the fan account can be specifically obtaining the recommendation index of the social associated account based on the evaluation behavior data of the fan account and the content publishing behavior data of the anchor account.
[0148] Furthermore, in actual applications, when a fan joins groups created by multiple different anchor accounts, it may affect the fan's attention to a single anchor, thereby reducing the fan's interest in social interaction with the anchor. Therefore, the recommendation index decay value of the fan account can be calculated based on the number of groups the fan has joined and the activity in these groups, so as to correct the recommendation index. The specific method is as follows:
[0149] For each fan account in the set of social associated accounts, the number of groups joined by the fan account and the activity level in each group are obtained; and, based on the number of groups joined by the fan account and the activity level in each group, the recommendation index decay value of the fan account is obtained; then, the difference or ratio between the evaluation behavior data of the fan account and the content publishing behavior data of the anchor account can be obtained, and based on the difference or ratio and the recommendation index decay value, the recommendation index of the social associated account can be obtained.
[0150] The activity of a fan account in each group it joins can usually be obtained by the following method: each group it joins is taken as the current group, and the activity of the fan account in the current group is obtained based on the difference or ratio between the total interaction time of the social-related account in the current group and the total time the fan account has joined the current group.
[0151] Among them, obtaining the recommendation index decay value of the fan account based on the number of groups joined by the fan account and the activity in each group can specifically include: for each fan account in the set of socially associated accounts, obtaining the average activity of each group joined by the fan account; when the maximum activity among each of the activities is greater than the second preset threshold, obtaining the recommendation index decay value of the fan account based on the product of the number of groups of the fan account and a preset constant, the second preset constant and the sum of the quotient of the average value and the first preset threshold; or, when the maximum activity among the activities is less than or equal to the second preset threshold, obtaining the recommendation index decay value of the fan account based on the product of the number of groups of the fan account and the preset constant, the third preset constant and the sum of the quotient of the average value and the first preset threshold, wherein the third preset constant is less than the second preset constant.
[0152] In actual applications, when the account attribute data of the anchor account is the anchor portrait of the anchor account, and the account attribute data of the fan account is the fan portrait of the fan account, in step S103, the server responds to the recommendation index acquisition request, and for each fan account in the social associated account set of the anchor account, according to the account attribute data of the fan account and the account attribute data of the anchor account, obtains the recommendation index of the fan account by: obtaining the first similarity between the anchor classification category to which the anchor account belongs and the target anchor classification category through the anchor portrait of the anchor account; obtaining the fan classification category to which the fan account belongs and obtaining the second similarity between the fan classification category and the target fan classification category for each fan account in the social associated account set through the fan portrait of the fan account; and obtaining the recommendation index of the fan account for each fan account in the social associated account set according to the corresponding first similarity and second similarity.
[0153] Step S104: the server sends the recommendation index of each fan account in the social association account set to the client.
[0154] Step S105: The client recommends group members to the anchor account according to the recommendation index of each fan account in the social association account set.
[0155] In this way, in actual applications, the anchor can implement the group member recommendation trigger operation on the client 90, and finally recommend group members to the anchor account through the client 90 and the server 70. Since the account attribute data of the anchor account can usually reflect the identity portrait or user behavior of the target user account, and the account attribute data of the fan account can also reflect the identity image or user behavior of the social association account, the recommendation index obtained based on the account attribute data of the anchor account and the account attribute data of the fan account can reflect the interest of the fan account in social interaction with the anchor account, and can guide the anchor account to add these recommended fan accounts to the group it initiated, so as to achieve high-quality social group interaction.
[0156] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present disclosure is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a server, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means to another website site, computer, server, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk Solid State Disk (SSD)), etc.
[0157] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.
[0158] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A method for recommending group members, characterized in that: include: In response to a group member recommendation triggering operation implemented by a target user account, obtaining a set of social associated accounts of the target user account; For each social association account in the set of social association accounts, obtaining a recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account, and the recommendation index is used to reflect the interest of the social association account in social interaction with the target user account; Selecting social related accounts whose recommendation index meets the preset recommendation conditions as target group members, and displaying them to the target user account to trigger the target user account to establish a group relationship with the target group members; Wherein, the method further comprises: Obtain the number of groups that the social association account has joined and the activity level in each group; According to the number of groups joined by the social association account and the activity level in each group, obtaining the recommendation index decay value of the social association account, wherein the more groups joined by the social association account and the greater the activity level in each group, the greater the recommendation index decay value of the social association account; The acquiring, according to the account attribute data of the social association account and the account attribute data of the target user account, a recommendation index of the social association account specifically includes: Obtain a difference or ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtain a recommendation index of the social association account based on the difference or ratio and the recommendation index decay value; the user behavior data of the social association account includes evaluation behavior data of the social association account evaluating the content published by the target user account within a preset time period, the user behavior data of the target user account includes content publishing behavior data of the target user account publishing content within a preset time period, and the recommendation index of the social association account is inversely proportional to the recommendation index decay value of the social association account.
2. The method according to claim 1, characterized in that The activity of social related accounts in each group can be obtained through the following methods: Each group is taken as the current group respectively, and the activity of the social associated account in the current group is obtained according to the difference or ratio between the total interaction time of the social associated account in the current group and the total time of the social associated account joining the current group.
3. The method according to claim 1, characterized in that According to the number of groups and the activity levels, obtaining the recommendation index decay value of the social association account specifically includes: For each social association account in the social association account set, obtaining an average value of activity of each group joined by the social association account; When the maximum activity among the activeness is greater than the second preset threshold, the recommended index decay value of the social associated account is obtained according to the product of the number of groups of the social associated account and the preset constant, the second preset constant and the sum of the quotient of the average value and the first preset threshold; or, when the maximum activity among the activeness is less than or equal to the second preset threshold, the recommended index decay value of the social associated account is obtained according to the product of the number of groups of the social associated account and the preset constant, the third preset constant and the sum of the quotient of the average value and the first preset threshold, wherein the third preset constant is less than the second preset constant.
4. The method according to claim 1, characterized in that: The method further comprises: Obtaining a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category through the anchor portrait of the target user account, wherein the anchor portrait includes the anchor classification category to which the target user account belongs, and the target anchor classification category is obtained by screening the anchor classification categories to which other user accounts followed by each social associated account in the set of social associated accounts belong respectively; For each social association account in the set of social association accounts, the fan classification category to which the social association account belongs is obtained through the fan portrait of the social association account, and the second similarity between the fan classification category and the target fan classification category is obtained, wherein the fan portrait reflects the bias of the interest of the social association account, and the target fan classification category is obtained by screening from the fan classification categories to which each social association account in the set of social association accounts belongs respectively; then, For each social association account in the social association account set, obtaining a recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account, specifically including: For each social association account in the set of social association accounts, a recommendation index of the social association account is obtained according to the user behavior data of the social association account, the user behavior data of the target user account, the first similarity, and the second similarity.
5. The method according to claim 4, characterized in that For each social association account in the set of social association accounts, obtaining a recommendation index of the social association account according to the user behavior data of the social association account, the user behavior data of the target user account, the first similarity, and the second similarity, specifically includes: For each social association account in the social association account set, obtaining a difference or a ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtaining a weighted sum result of the first similarity and the second similarity; The recommendation index of the social association account is obtained according to the difference or ratio and the weighted summation result.
6. The method according to claim 4, characterized in that Using the fan portrait of the social association account specifically includes: obtaining the fan classification category to which the social association account belongs by inputting the fan portrait of the social association account into a fan classification model.
7. The method according to claim 1, characterized in that For each social association account in the social association account set, obtaining a recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account, specifically including: For each social association account in the set of social association accounts, a recommendation index request is sent to the server, triggering the server to obtain the recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account.
8. The method according to claim 1, characterized in that The identity image of the social association account includes a fan portrait of the social association account; the identity portrait of the target user account includes a host portrait of the target user account; for each social association account in the social association account set, according to the account attribute data of the social association account and the account attribute data of the target user account, obtaining the recommendation index of the social association account, specifically including: Obtaining a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category through the anchor portrait of the target user account, wherein the anchor portrait includes the anchor classification category to which the target user account belongs, and the target anchor classification category is obtained by screening the anchor classification categories to which other user accounts followed by each social associated account in the set of social associated accounts belong respectively; For each social association account in the set of social association accounts, obtaining a fan classification category to which the social association account belongs through a fan portrait of the social association account, and obtaining a second similarity between the fan classification category and a target fan classification category, wherein the fan portrait reflects the bias of the interest of the social association account, and the target fan classification category is obtained by screening from the fan classification categories to which each social association account in the set of social association accounts belongs respectively; For each social association account in the social association account set, a recommendation index of the social association account is obtained according to the corresponding first similarity and second similarity.
9. A method for recommending group members, characterized in that: The method comprises: Receive a recommendation index acquisition request for a target user account from a client; In response to the recommendation index acquisition request, for each social association account in the set of social association accounts of the target user account, according to the account attribute data of the social association account and the account attribute data of the target user account, the recommendation index of the social association account is acquired, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account, and the recommendation index is used to reflect the interest of the social association account in social interaction with the target user account; Sending the recommendation index of each social association account in the social association account set to the client, so that the client can recommend group members to the target user account according to the recommendation index; Wherein, the method further comprises: Obtain the number of groups that the social association account has joined and the activity level in each group; According to the number of groups joined by the social association account and the activity level in each group, obtaining the recommendation index decay value of the social association account, wherein the more groups joined by the social association account and the greater the activity level in each group, the greater the recommendation index decay value of the social association account; The acquiring, according to the account attribute data of the social association account and the account attribute data of the target user account, a recommendation index of the social association account specifically includes: Obtain a difference or ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtain a recommendation index of the social association account based on the difference or ratio and the recommendation index decay value; the user behavior data of the social association account includes evaluation behavior data of the social association account evaluating the content published by the target user account within a preset time period, the user behavior data of the target user account includes content publishing behavior data of the target user account publishing content within a preset time period, and the recommendation index of the social association account is inversely proportional to the recommendation index decay value of the social association account.
10. The method according to claim 9, characterized in that The activity of social related accounts in each group can be obtained through the following methods: Each group is taken as the current group respectively, and the activity of the social associated account in the current group is obtained according to the difference or ratio between the total interaction time of the social associated account in the current group and the total time of the social associated account joining the current group.
11. The method according to claim 9, characterized in that According to the number of groups and the activity levels, obtaining the recommendation index decay value of the social association account specifically includes: For each social association account in the social association account set, obtaining an average value of activity of each group joined by the social association account; When the maximum activity among the activeness is greater than the second preset threshold, the recommended index decay value of the social associated account is obtained according to the product of the number of groups of the social associated account and the preset constant, the second preset constant and the sum of the quotient of the average value and the first preset threshold; or, when the maximum activity among the activeness is less than or equal to the second preset threshold, the recommended index decay value of the social associated account is obtained according to the product of the number of groups of the social associated account and the preset constant, the third preset constant and the sum of the quotient of the average value and the first preset threshold, wherein the third preset constant is less than the second preset constant.
12. The method according to claim 9, characterized in that The method further comprises: Obtaining a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category through the anchor portrait of the target user account, wherein the anchor portrait includes the anchor classification category to which the target user account belongs, and the target anchor classification category is obtained by screening the anchor classification categories to which other user accounts followed by each social associated account in the set of social associated accounts belong respectively; For each social association account in the set of social association accounts, the fan classification category to which the social association account belongs is obtained through the fan portrait of the social association account, and the second similarity between the fan classification category and the target fan classification category is obtained, wherein the fan portrait reflects the bias of the interest of the social association account, and the target fan classification category is obtained by screening from the fan classification categories to which each social association account in the set of social association accounts belongs respectively; then, For each social association account in the social association account set, obtaining a recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account, specifically including: For each social association account in the set of social association accounts, a recommendation index of the social association account is obtained according to the user behavior data of the social association account, the user behavior data of the target user account, the first similarity, and the second similarity.
13. The method according to claim 12, characterized in that For each social association account in the set of social association accounts, obtaining a recommendation index of the social association account according to the user behavior data of the social association account, the user behavior data of the target user account, the first similarity, and the second similarity, specifically includes: For each social association account in the social association account set, obtaining a difference or a ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtaining a weighted sum result of the first similarity and the second similarity; The recommendation index of the social association account is obtained according to the difference or ratio and the weighted summation result.
14. The method according to claim 9, characterized in that The identity image of the social association account includes a fan portrait of the social association account; the identity portrait of the target user account includes a host portrait of the target user account; for each social association account in the social association account set, according to the account attribute data of the social association account and the account attribute data of the target user account, obtaining the recommendation index of the social association account, specifically including: Obtaining a first similarity between the anchor classification category to which the target user account belongs and the target anchor classification category through the anchor portrait of the target user account, wherein the anchor portrait includes the anchor classification category to which the target user account belongs, and the target anchor classification category is obtained by screening the anchor classification categories to which other user accounts followed by each social associated account in the set of social associated accounts belong respectively; For each social association account in the set of social association accounts, obtaining a fan classification category to which the social association account belongs through a fan portrait of the social association account, and obtaining a second similarity between the fan classification category and a target fan classification category, wherein the fan portrait reflects the bias of the interest of the social association account, and the target fan classification category is obtained by screening from the fan classification categories to which each social association account in the set of social association accounts belongs respectively; For each social association account in the social association account set, a recommendation index of the social association account is obtained according to the corresponding first similarity and second similarity.
15. A device for recommending group members, characterized in that: include: a response unit configured to execute a group member recommendation trigger operation in response to a target user account, and obtain a set of social associated accounts of the target user account; an acquisition unit configured to acquire, for each social association account in the set of social association accounts, a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account, and the recommendation index is used to reflect the interest of the social association account in social interaction with the target user account; A display unit is configured to select social related accounts whose recommendation index meets preset recommendation conditions as target group members, and display them to the target user account to trigger the target user account to establish a group relationship with the target group members; The acquisition unit is further configured to acquire the number of groups joined by the social association account and the activity level in each group; The acquisition unit is further configured to acquire the recommendation index decay value of the social association account according to the number of groups joined by the social association account and the activity level in each group, wherein the more groups the social association account joins and the greater the activity level in each group, the greater the recommendation index decay value of the social association account; The acquisition unit is configured to acquire the recommendation index of the social association account according to the account attribute data of the social association account and the account attribute data of the target user account, specifically including: Obtain a difference or ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtain a recommendation index of the social association account based on the difference or ratio and the recommendation index decay value; the user behavior data of the social association account includes evaluation behavior data of the social association account evaluating the content published by the target user account within a preset time period, the user behavior data of the target user account includes content publishing behavior data of the target user account publishing content within a preset time period, and the recommendation index of the social association account is inversely proportional to the recommendation index decay value of the social association account.
16. A device for recommending group members, characterized in that: include: A request receiving unit, configured to execute receiving a recommendation index acquisition request of a target user account from a client; a request response unit configured to execute, in response to the recommendation index acquisition request, acquiring, for each social association account in the set of social association accounts of the target user account, a recommendation index of the social association account according to account attribute data of the social association account and account attribute data of the target user account, wherein the account attribute data of the social association account reflects the identity image or user behavior of the social association account, and the account attribute data of the target user account reflects the identity portrait or user behavior of the target user account, and the recommendation index is used to reflect the interest of the social association account in social interaction with the target user account; a sending unit configured to send the recommendation index of each social association account in the social association account set to the client, so that the client can recommend group members to the target user account according to the recommendation index; The request response unit is further configured to execute acquisition of the number of groups joined by the social association account and the activity level in each group; The request response unit is further configured to execute, according to the number of groups joined by the social association account and the activity level in each group, obtaining the recommendation index decay value of the social association account, wherein the more groups joined by the social association account and the greater the activity level in each group, the greater the recommendation index decay value of the social association account; The request response unit is configured to execute, according to the account attribute data of the social association account and the account attribute data of the target user account, obtaining the recommendation index of the social association account, specifically including: Obtain a difference or ratio between the user behavior data of the social association account and the user behavior data of the target user account, and obtain a recommendation index of the social association account based on the difference or ratio and the recommendation index decay value; the user behavior data of the social association account includes evaluation behavior data of the social association account evaluating the content published by the target user account within a preset time period, the user behavior data of the target user account includes content publishing behavior data of the target user account publishing content within a preset time period, and the recommendation index of the social association account is inversely proportional to the recommendation index decay value of the social association account.
17. A client, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for recommending group members as described in any one of claims 1 to 8.
18. A server, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for recommending group members as described in any one of claims 9 to 14.
19. A group member recommendation system, comprising the client according to claim 17 and the server according to claim 18.
20. A storage medium, when instructions in the storage medium are executed by a processor of a client, enables the client to execute the method for recommending group members according to any one of claims 1 to 8; or, When the instructions in the storage medium are executed by a processor of a server, the server is enabled to execute the method for recommending group members according to any one of claims 9 to 14.
21. A computer program product, characterized in that When the computer program product is executed on a computer, the computer is caused to execute the method for recommending group members according to any one of claims 1 to 14.
Citation Information
Patent Citations
Method and device for adding group members
CN104639426A
Mobile terminal live-broadcasting information pushing method and pushing device thereof
CN105872058A
Method and system for creating friendly neighbor group and server
CN110311795A
