Method, device and electronic device for obtaining account association information in application
By generating directed graphs and obtaining social relationship chains, determining the social correlation between accounts, the problem of content recommendations in the prior art is solved, and accurate screening of user relationships and diversity of content recommendations is achieved.
Patent Information
- Application Number
- CN201911185771.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2019-11-27
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2039-11-27
AI Technical Summary
When existing applications recommend content to users, they usually only rely on the user's behavioral habits and do not consider the user's social attributes, resulting in a relatively single type of recommended content.
By generating directed graphs, record the attention relationship and social closeness between accounts, obtain the social relationship chain, and determine the social correlation between accounts based on this, thereby filtering out related accounts and recommending richer content to users.
By accurately determining the social relationship chain and correlation degree of the account, the accurate screening of user relationships is achieved, and the diversity and pertinence of content recommendations is improved.
Smart Images

Figure CN112861015B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of multimedia technology, and in particular to a method, device and electronic device for obtaining account association information in an application. Background Art
[0002] With the development of computer technology and network technology, a large number of applications have emerged. Users can register accounts in applications and log in. When users log in to applications through registered accounts, the operators of applications can analyze the basic information filled in by users when registering accounts and the user behavior data corresponding to the accounts, obtain the user's behavior habits, and recommend content that meets the user's needs for the account, so that the user can view it easily.
[0003] In the process of implementing the present invention, the inventors found that there are at least the following problems in the related art:
[0004] When application operators recommend content to users, they usually only base it on the user's behavioral habits without considering the user's social attributes, that is, the relationship between various accounts. Therefore, in related technologies, the types of content recommended by application operators to accounts are relatively simple. Summary of the invention
[0005] In order to solve the technical problem that the operator of an application program recommends a relatively single type of content to an account in the related art, the present disclosure provides a method, device and electronic device for obtaining account association information in an application program. The technical solution of the present disclosure is as follows:
[0006] According to a first aspect of an embodiment of the present disclosure, a method for obtaining account association information in an application is provided, comprising:
[0007] Generate a directed graph according to the following relationship between accounts in the application, wherein the following relationship is used to represent the social association relationship actively or passively established between the accounts, and the directed graph has directed information for marking the following relationship, and the directed information is also used to record the degree of social closeness between two accounts with the following relationship;
[0008] According to the attention relationship and social closeness between the accounts in the directed information record, a social relationship chain of each account in the application is obtained, wherein the social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account;
[0009] Based on the social relationship chain of the acquired account, determine the social association between the accounts in the application;
[0010] Based on the determined social association degree between each account, the associated accounts of each account are screened out from other accounts of the application.
[0011] Optionally, obtaining the social relationship chain of each account in the application according to the attention relationship and social closeness between the accounts in the directed information record includes:
[0012] For any account in the application, a random walk sequence set corresponding to the account is determined by using a random walk method according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, wherein the random walk sequence set determined by the random walk method includes: a plurality of random walk sequences composed of account information, each random walk sequence is: a random walk sequence formed by determining a subsequent account information corresponding to the previous account information according to the attention relationship and social closeness between the accounts, and the account corresponding to the subsequent account information is: an account with close social connection corresponding to the previous account information;
[0013] For any account in the application, a random walk sequence in the random walk sequence set corresponding to the account is determined as the social relationship chain of the account.
[0014] Optionally, for any account in the application, according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, using random walk to construct a random walk sequence set corresponding to the account includes:
[0015] According to the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application;
[0016] According to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence, wherein the updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence;
[0017] Accordingly, according to the order of the account information put into the account information sequence, the following are determined in sequence: the account corresponding to the last account information put into the account information sequence, and the accounts with close social connections, until the number of account information put into the account information sequence reaches a first preset number;
[0018] A second preset number of account information sequences are determined as a set of random walk sequences corresponding to the first account.
[0019] Optionally, determining the social association between accounts in the application based on the acquired social relationship chain of the account includes:
[0020] Convert each account information in the first social relationship chain of the first account into a first sub-vector, and convert each account information in the second social relationship chain of other accounts in the application into a second sub-vector;
[0021] The first sub-vectors corresponding to the account information in the first social relationship chain are combined into a first vector; the second sub-vectors corresponding to the account information in the second social relationship chain are combined into a second vector;
[0022] The distance between the first vector and each second vector is calculated to obtain the social association degree between the first account and each other account.
[0023] Optionally, the method further includes:
[0024] recommending videos related to a fourth account to each account, wherein the fourth account is: other accounts among the corresponding associated accounts except the account having a direct social association relationship with the account;
[0025] or,
[0026] Recommending videos related to a fifth account to each account, wherein the fifth account is: an account currently broadcasting live among the corresponding associated accounts.
[0027] Optionally, each vertex in the directed graph corresponds to account information of an account, and each edge in the directed graph is used to characterize that: two accounts corresponding to the account information corresponding to two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to characterize: the degree of social closeness between two accounts corresponding to the account information corresponding to the two vertices of the edge.
[0028] According to a second aspect of an embodiment of the present disclosure, a device for acquiring account association information in an application is provided, comprising:
[0029] A directed graph generation module is configured to generate a directed graph according to a follow relationship between accounts in an application, wherein the follow relationship is used to represent a social association relationship actively or passively established between accounts, and the directed graph has directed information for marking the follow relationship, and the directed information is also used to record the degree of social closeness between two accounts having a follow relationship;
[0030] A social relationship chain acquisition module is configured to acquire the social relationship chain of each account in the application according to the attention relationship and social closeness between the accounts recorded in the directed information, wherein the social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account;
[0031] A social association degree determination module is configured to determine the social association degree between the accounts in the application based on the acquired social relationship chain of the accounts;
[0032] The associated account determination module is configured to perform filtering of associated accounts of each account from other accounts of the application based on the determined social association between each account.
[0033] Optionally, the social relationship chain acquisition module includes:
[0034] The random walk sequence set determination unit is configured to perform, for any account in the application, determining the random walk sequence set corresponding to the account by using a random walk method according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, wherein the random walk sequence set determined by the random walk method includes: a plurality of random walk sequences composed of account information, each random walk sequence being: a random walk sequence formed by determining a subsequent account information corresponding to the previous account information according to the attention relationship and social closeness between the accounts, and the account corresponding to the subsequent account information is: an account with close social connection corresponding to the previous account information;
[0035] The social relationship chain determination unit is configured to determine, for any account in the application, a random walk sequence in a random walk sequence set corresponding to the account as the social relationship chain of the account.
[0036] Optionally, the random walk sequence set determination unit is configured to execute:
[0037] According to the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application;
[0038] According to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence, wherein the updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence;
[0039] Accordingly, according to the order of the account information put into the account information sequence, the following are determined in sequence: the account corresponding to the last account information put into the account information sequence, and the accounts with close social connections, until the number of account information put into the account information sequence reaches a first preset number;
[0040] A second preset number of account information sequences are determined as a set of random walk sequences corresponding to the first account.
[0041] Optionally, the social association degree determination module is configured to execute:
[0042] Convert each account information in the first social relationship chain of the first account into a first sub-vector, and convert each account information in the second social relationship chain of other accounts in the application into a second sub-vector;
[0043] The first sub-vectors corresponding to the account information in the first social relationship chain are combined into a first vector; the second sub-vectors corresponding to the account information in the second social relationship chain are combined into a second vector;
[0044] The distance between the first vector and each second vector is calculated to obtain the social association degree between the first account and each other account.
[0045] Optionally, the device further comprises:
[0046] The video recommendation module is configured to recommend videos related to a fourth account for each account, where the fourth account is: other accounts among the corresponding associated accounts except for accounts with a direct social association relationship with the account; or, to recommend videos related to a fifth account for each account, where the fifth account is: an account among the corresponding associated accounts that is currently broadcasting live.
[0047] Optionally, each vertex in the directed graph corresponds to account information of an account, and each edge in the directed graph is used to characterize that: two accounts corresponding to the account information corresponding to two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to characterize: the degree of social closeness between two accounts corresponding to the account information corresponding to the two vertices of the edge.
[0048] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:
[0049] processor;
[0050] a memory for storing instructions executable by the processor;
[0051] The processor is configured to execute the instructions to implement the method for obtaining account-associated information in an application as described in the first aspect.
[0052] According to a fourth 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 an electronic device, the electronic device can execute the method for obtaining account association information in an application described in the first aspect.
[0053] According to a fifth aspect of an embodiment of the present disclosure, there is provided a computer program product comprising instructions, which, when executed on a computer, enables the computer to implement the method for obtaining account association information in an application as described in the first aspect.
[0054] The technical solution provided by the embodiment of the present disclosure generates a directed graph according to the attention relationship between accounts in the application when acquiring account association information in the application; acquires the social relationship chain of each account in the application according to the attention relationship and social closeness between accounts recorded in the directed information in the directed graph; and determines the social association degree between each account in the application based on the acquired social relationship chain of the account; and finally, based on the determined social association degree between each account, screens out the associated accounts of each account from other accounts in the application.
[0055] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend rich types of content to each account based on the association relationship of each account in the application. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] Figure 1 is a flowchart of a method for obtaining account association information in an application according to an exemplary embodiment;
[0057] Figure 2 yes Figure 1 A flowchart of a specific implementation of step S12 in the illustrated embodiment;
[0058] Figure 3 yes Figure 2 A flowchart of a specific implementation of step S121 in the illustrated embodiment;
[0059] Figure 4 yes Figure 1 A flowchart of a specific implementation of step S13 in the illustrated embodiment;
[0060] Figure 5 is a flowchart of another method for obtaining account association information in an application according to an exemplary embodiment;
[0061] Figure 6 is a block diagram of a device for acquiring account association information in an application according to an exemplary embodiment;
[0062] Figure 7 is a block diagram of an electronic device according to an exemplary embodiment;
[0063] Figure 8 The present invention is a block diagram of a device for acquiring account association information in an application according to an exemplary embodiment. DETAILED DESCRIPTION
[0064] 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.
[0065] It should be noted that the terms "first", "second", 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 data 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. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0066] In order to solve the technical problem in the related art that the content type recommended by the operator of an application for an account is relatively single, the present disclosure provides a method, device and electronic device for obtaining account association information in an application.
[0067] First, a method for obtaining account association information in an application provided by an embodiment of the present disclosure is introduced in detail below.
[0068] It should be noted that the execution subject of the method for obtaining account association information in an application provided by the embodiment of the present disclosure is an account association information obtaining device in the application, and the account association information obtaining device in the application runs in an electronic device, which may be a server.
[0069] Figure 1is a flowchart of a method for obtaining account association information in a visual application according to an exemplary embodiment. Figure 1 As shown, the method may include the following steps.
[0070] In step S11, a directed graph is generated according to the attention relationship between accounts in the application.
[0071] Among them, the follow-up relationship is used to represent the social association relationship established actively or passively between accounts. The directed graph has directed information for marking the follow-up relationship. The directed information is also used to record the social closeness between two accounts with a follow-up relationship.
[0072] Specifically, any two accounts in the application can follow each other. When one of the two accounts follows the other account, the two accounts have a follow relationship, thereby establishing a social association relationship between the two accounts. For an account that actively follows another account, the account actively establishes a social association relationship; similarly, for another account that is followed, the other account passively establishes a social association relationship.
[0073] The electronic device can generate a directed graph based on the following relationship between accounts in the application. In the directed graph, the following relationship between any two accounts and the degree of social closeness between the two accounts with the following relationship can be recorded with directed information. The degree of social closeness can be used to indicate the degree of attention of the two accounts with the following relationship.
[0074] For example, two accounts with a following relationship are account A and account B. The degree of social closeness between account A and account B can be measured by: the percentage of clicks on videos posted by account A to account B; or the percentage of viewing time of videos posted by account A to account B. It can be understood that the higher the percentage of clicks or the percentage of viewing time, the higher the degree of social closeness between account A and account B.
[0075] In order to describe the solution clearly, the directed graph will be described in detail in the following embodiments.
[0076] In step S12, the social relationship chain of each account in the application is obtained according to the attention relationship and social closeness between the accounts in the directed information record.
[0077] The social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account;
[0078] Specifically, since the directed information records the following relationships and social closeness between accounts in the application, for any account in the application, the social relationship chain of the account can be obtained based on the following relationships and social closeness between the account and other accounts.
[0079] The social relationship chain of an account may be composed of account information of accounts that have direct social association or indirect social association with the account. It is understandable that an account that has a direct social association with the account may be an account that has a direct follow-up relationship with the account; an account that has an indirect social association with the account may be an account that has an indirect follow-up relationship with the account.
[0080] In order to describe the solution clearly, the specific implementation method of "obtaining the social relationship chain of each account in the application according to the attention relationship and social closeness between accounts in the directed information record" will be explained in detail in the following embodiment.
[0081] In step S13, based on the acquired social relationship chains of the accounts, the social associations between the accounts in the application are determined.
[0082] It is understandable that, for any two accounts in an application, if the social relationship chains of the two accounts are more similar, the social association degree of the two accounts is higher. If the social relationship chains of the two accounts are less similar, that is, the greater the difference, the social association degree of the two accounts is lower. Therefore, after obtaining the social relationship chains of each account in the application, the social association degree between the accounts in the application can be determined based on the social relationship chains of the obtained accounts.
[0083] In order to describe the solution clearly, in the following embodiments, a specific implementation method for determining the social association between accounts in an application based on the acquired social relationship chain of the accounts will be described in detail.
[0084] In step S14, based on the determined social association degree between each account, the associated accounts of each account are screened out from other accounts of the application.
[0085] If the social association between two accounts is higher, it means that the probability that the two accounts are associated accounts is higher; similarly, the lower the social association between two accounts is, the lower the probability that the two accounts are associated accounts. Therefore, after determining the social association between each account, the associated account of each account can be determined based on the determined social association between each account.
[0086] Specifically, for any account in the application, an account with a social association degree greater than a preset social association degree with the account can be determined as an associated account of the account. The preset association degree can be determined according to actual conditions, and the embodiment of the present invention does not specifically limit the size of the preset association degree.
[0087] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend rich types of content to each account based on the association relationship of each account in the application.
[0088] exist Figure 1 On the basis of the illustrated embodiment, in order to describe the solution clearly, the directed graph will be described in detail in the following embodiments.
[0089] In one embodiment, each vertex in a directed graph corresponds to account information of an account, and each edge in the directed graph is used to represent that two accounts corresponding to the account information corresponding to two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to represent the degree of social closeness between the two accounts corresponding to the account information corresponding to the two vertices of the edge.
[0090] In this implementation, each vertex in the directed graph may have a vertex identifier, and the vertex identifier is used to identify an account, that is, the vertex identifier may be account information, and the account information may be an ID number or name of the account, etc. The present disclosure embodiment does not specifically limit the account information.
[0091] Assuming that there is a follow relationship between two accounts, then an edge can be used to connect the two vertices corresponding to the two accounts; otherwise, the two vertices are not connected. It can be understood that an edge can be directed or undirected.
[0092] In practical applications, assuming that an account with account information A follows an account with account information B, then in the directed graph, a directed edge can be formed from the vertex corresponding to account information A to the vertex corresponding to account information B.
[0093] Assume that an account with account information B follows an account with account information A. Then, in the directed graph, a directed edge can be formed from the vertex corresponding to account information B to the vertex corresponding to account information A.
[0094] If an account with account information A follows an account with account information B, and an account with account information B also follows an account with account information A, that is, the account with account information A and the account with account information B follow each other, then the edge connecting the vertex corresponding to account information A and the vertex corresponding to account information B can be an edge without direction.
[0095] In addition, each edge in the directed graph has an edge weight, which is used to represent: the account information corresponding to the two vertices of the edge, and the degree of social closeness between the two corresponding accounts.
[0096] For example, assume that an account with account information A follows an account with account information B.
[0097] As an optional implementation method, the edge weight of the edge connecting the vertex corresponding to account information A and the vertex corresponding to account information B can be: the proportion of clicks on videos posted by the account with account information A to the account with account information B.
[0098] As another optional implementation method, the edge weight of the edge connecting the vertex corresponding to account information A and the vertex corresponding to account information B can be: the viewing time ratio of the account with account information A to the video posted by the account with account information B.
[0099] As another optional implementation, the edge weight of the edge connecting the vertex corresponding to the account information A and the vertex corresponding to the account information B may be: a weighted value obtained by weighted calculation of the percentage of clicks and the percentage of viewing time.
[0100] It is understandable that the edge weight is usually a number between 0 and 1. In order to improve the accuracy of the edge weight, the edge weight can be calculated in a normalized manner. For example, the account with account information A follows three accounts, namely, the account with account information B, the account with account information C, and the account with account information D. Then, the edge weight of the edge connecting the vertex corresponding to account information A and the vertex corresponding to account information B can be: the number of clicks on the video published by the account with account information A to the account with account information B, divided by the total number of clicks on the video published by the account with account information A to the account with account information B, the account with account information C, and the account with account information D.
[0101] Of course, the edge weights may also be estimated through a model, and the embodiment of the present invention does not specifically limit the calculation method of the edge weights.
[0102] It can be seen that, through the directed graph provided by this embodiment, the follow relationship between accounts and the degree of social closeness between two accounts with a follow relationship can be accurately recorded.
[0103] In an optional embodiment, a specific implementation method of obtaining the social relationship chain of each account in the application according to the attention relationship and social closeness between accounts in the directed information record is as follows.
[0104] Specifically, Figure 1 Step S12 of the illustrated embodiment may include the following two steps: Figure 2 As shown, they are: S121 and S122.
[0105] S121, for any account in the application, according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, a random walk method is used to determine the random walk sequence set corresponding to the account.
[0106] Among them, the random walk sequence set determined by the random walk method includes: multiple random walk sequences composed of account information, each random walk sequence is: a random walk sequence formed by determining the next account information corresponding to the previous account information according to the attention relationship and social closeness between accounts, and the account corresponding to the next account information is: an account with close social connection corresponding to the previous account information.
[0107] Random walk is also called random walk. The core concept is that the conservation quantity carried by any irregular walker corresponds to a diffusion transport law, which is close to Brownian motion and is the ideal mathematical state of Brownian motion. For any account in the application, according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, multiple random walk sequences can be generated by random walk, and these multiple random walk sequences constitute a random walk sequence set. Among them, the number of random walk sequences included in the random walk sequence set can be determined according to the actual situation, and the length of each random walk sequence can also be determined according to the actual situation.
[0108] As an implementation method of the embodiment of the present disclosure, S121 may include the following four steps: Figure 3 As shown, they are: S1211 to S1214.
[0109] S1211, based on the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application.
[0110] Specifically, the first account may have a follow relationship with one account or multiple accounts. When the first account has a follow relationship with only one account, the second account is: the account that has a follow relationship with the first account. When the first account has a follow relationship with multiple accounts, an account is randomly selected from the multiple accounts according to the degree of social closeness between the first account and the multiple accounts and determined as the second account.
[0111] After the first account and the second account are determined, the account information of the first account and the account information of the second account may be sequentially placed into an empty account information sequence.
[0112] S1212, according to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence. The updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence.
[0113] Specifically, the second account may have a follow relationship with one or more accounts other than the first account. When the second account has a follow relationship with only one account, the third account is: an account that has a follow relationship with the second account. When the second account has a follow relationship with multiple accounts, an account is randomly selected from the multiple accounts according to the degree of social closeness between the second account and the multiple accounts and determined as the third account. After being determined as the third account, the account information of the third account may be placed in the account information sequence.
[0114] S1213, accordingly, according to the order of the account information put into the account information sequence, determine in sequence: the account corresponding to the last account information put into the account information sequence, the closely social accounts, until the number of account information put into the account information sequence reaches the first preset number.
[0115] Specifically, after the account information of the third account is placed in the account information sequence, the account closely associated with the third account is determined based on the third account and the updated other accounts, and the account information of the determined account is placed in the account information sequence until the number of account information placed in the account information sequence reaches the first preset number, which can be determined according to actual conditions and is not specifically limited in the present disclosure.
[0116] S1214: Determine a second preset number of account information sequences as a set of random walk sequences corresponding to the first account.
[0117] Specifically, a second preset number of account information sequences may be determined in a random walk manner, and the second preset number of account information sequences may be determined as a random walk set corresponding to the first account.
[0118] For example, in a directed graph, the first vertex corresponding to the first account information of the first account is x. When generating a random walk sequence set for the first account, starting from the first vertex x, the edge weight of the edge containing the first vertex x is used as a probability random to select the second vertex y connected to the first vertex x through an edge, and then starting from the second vertex y, the edge weight of the edge containing the second vertex y is used as a probability random to select the third vertex z connected to the second vertex y through an edge, and this is repeated to obtain a random walk sequence x→y→z→… with a preset length L. Repeat the above process for a preset number of times r, and you can get the random walk sequence set of the first account, which includes: r random walk sequences of length L.
[0119] It can be understood that, for each account corresponding to a vertex in a directed graph, the above method can be used to generate a set of random walk sequences corresponding to the account, thereby obtaining a set of random walk sequences corresponding to each account in the application.
[0120] S122: For any account in the application, determine the random walk sequence in the random walk sequence set corresponding to the account as the social relationship chain of the account.
[0121] The random walk sequence set corresponding to each account is determined through step S121. For the random walk sequence set corresponding to each account, since the random walk sequences in the random walk sequence set are determined based on the attention relationship and social association degree between the account and other accounts, the random walk sequences in the random walk sequence set can be determined as the social relationship chain of the account.
[0122] It can be seen that through the technical solution provided by this embodiment, the social relationship chain of each account in the application can be accurately determined based on the attention relationship and social closeness between accounts in the directed information record.
[0123] In an optional embodiment, based on the acquired social relationship chain of the account, a specific implementation method of determining the social association between the accounts in the application is as follows.
[0124] Specifically, Figure 1 S13 in the illustrated embodiment may include the following three steps: Figure 4 As shown, they are S131 to S133 respectively.
[0125] S131, converting each account information in a first social relationship chain of the first account into a first sub-vector, and converting each account information in a second social relationship chain of other accounts in the application into a second sub-vector.
[0126] Among them, the word2vec algorithm can be used to map each word to a vector, which can be used to represent the relationship between words. In order to more accurately determine the social association between different accounts in the subsequent steps, the word2vec algorithm can be used to convert each account information in the first social relationship chain into a first sub-vector; and each account information in the second social relationship chain of other accounts in the application is converted into a second sub-vector. Among them, in the word2vec algorithm, except for the dimension, window size (how large a window is used to construct a training sample), and the selection of skip_gram (a method of constructing samples by predicting other words in the window with the center word), the parallel number needs to be specified, and other word2vec algorithm parameters such as learning rate learning_rate can use the default values of the word2vec algorithm. It can be seen that the dimension of the vector converted from each account information in the first social relationship chain can be pre-set. The present disclosure does not make specific restrictions on this.
[0127] S132, the first sub-vectors corresponding to the account information in the first social relationship chain are combined into a first vector; the second sub-vectors corresponding to the account information in the second social relationship chain are combined into a second vector.
[0128] Specifically, since the first social relationship chain includes multiple account information, after each account information included in the first social relationship chain is converted into a first sub-vector, multiple first sub-vectors corresponding to the multiple account information can form a first vector. Similarly, since the second social relationship chain also includes multiple account information, after each account information included in the second social relationship chain is converted into a second sub-vector, multiple second sub-vectors corresponding to the multiple account information can form a second vector.
[0129] S133, calculating the distance between the first vector and each second vector, and obtaining the social association degree between the first account and each other account.
[0130] By calculating the distance between the first vector and each second vector, the distance between the first social relationship chain of the first account and the second social relationship chain of each other account can be determined, and the distance between the first social chain and the second social chain of other accounts can reflect the social association between the first account and each other account. It can be understood that the smaller the distance between the first vector and a second vector, the higher the social association between the first account and the account corresponding to the second vector, otherwise, the lower the social association between the first account and the account corresponding to the vector.
[0131] Specifically, there are multiple ways to calculate the distance between the first vector and the second vector.
[0132] The first method is: calculate the cosine value of the first vector and the second vector. The larger the cosine value of the two vectors, the closer the distance between the two vectors.
[0133] The second method is: calculating the Euclidean distance between the first vector and the second vector, and determining the calculated Euclidean distance as the distance between the first vector and the second vector.
[0134] The third method is: using the distributed computing framework Hadoop to calculate the distance between the first vector and the second vector, or using an algorithm with a complexity of O(n) to obtain the distance between the first vector and the second vector.
[0135] Of course, the method of calculating the distance between the first vector and the second vector is not limited to the above three methods, and the present disclosure does not make any specific limitation on this.
[0136] It can be seen that through the technical solution provided by this embodiment, the social association between each account in the application can be accurately determined through the social relationship chain of each account.
[0137] In order to recommend videos with rich types to an account, the embodiment of the present disclosure also provides a method for obtaining account association information in an application program, such as Figure 5 As shown, the method comprises the following steps:
[0138] In step S51, a directed graph is generated according to the attention relationship between accounts in the application.
[0139] In step S52, the social relationship chain of each account in the application is obtained according to the attention relationship and social closeness between the accounts in the directed information record.
[0140] In step S53, based on the acquired social relationship chains of the accounts, the social associations between the accounts in the application are determined.
[0141] In step S54, based on the determined social association degree between each account, the associated accounts of each account are screened out from other accounts of the application.
[0142] Since steps S51 to S54 are the same as steps S11 to S14, Figure 1 In the illustrated embodiment, steps S11 to S14 have been described in detail, and therefore, steps S51 to S54 will not be described in detail.
[0143] In step S55, videos related to the fourth account are recommended to each account, or videos related to the fifth account are recommended to each account.
[0144] Among them, the fourth account is: other accounts in the corresponding associated accounts except the accounts that have a direct social association relationship with the account; the fifth account is: the account that is currently broadcasting live in the corresponding associated accounts.
[0145] It is understandable that if two accounts are associated accounts, then the probability that one of the two accounts likes to watch videos related to the other account is higher, so when recommending videos to an account, the account associated with the account can be recommended. For example, the first account and the second account are associated accounts. Assuming that the first account usually likes to watch idol dramas, and the second account has recently watched basketball games, basketball games can be recommended to the first account, so that the first account can watch more comprehensive types of videos.
[0146] Furthermore, in one embodiment, in the process of recommending videos, it is taken into account that among the accounts associated with an account, there are accounts with which the account has a direct social association relationship, that is, since the videos posted by the accounts directly followed by the account are directly viewable by the account, therefore, in order to recommend more diverse videos to the account, videos related to other accounts among the associated accounts, excluding accounts with which the account has a direct social association relationship, may be recommended to the account.
[0147] In another embodiment, in order to enable an account to watch the latest live video, the real-time index of the recommendation system can determine which accounts are live streaming and recommend to the account the videos that are currently being live streamed by the account in the associated accounts.
[0148] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the follow-up relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of videos to each account based on the association relationship of each account in the application.
[0149] According to a second aspect of an embodiment of the present disclosure, a device for obtaining account association information in an application is provided. Figure 6 As shown, including:
[0150] The directed graph generation module 610 is configured to generate a directed graph based on the follow relationship between accounts in the application, wherein the follow relationship is used to represent the social association relationship actively or passively established between accounts, and the directed graph has directed information for marking the follow relationship, and the directed information is also used to record the social closeness between two accounts with a follow relationship.
[0151] In an optional embodiment, each vertex in the directed graph corresponds to the account information of an account, and each edge in the directed graph is used to characterize that: two accounts corresponding to the account information corresponding to two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to characterize: the degree of social closeness between the two accounts corresponding to the account information corresponding to the two vertices of the edge.
[0152] It can be seen that, through the directed graph provided by this embodiment, the follow-up relationship between accounts and the degree of social closeness between two accounts having a follow-up relationship can be accurately recorded.
[0153] The social relationship chain acquisition module 620 is configured to acquire the social relationship chain of each account in the application according to the attention relationship and social closeness between the accounts recorded in the directed information, wherein the social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account;
[0154] The social association degree determination module 630 is configured to determine the social association degree between the accounts in the application based on the acquired social relationship chain of the accounts;
[0155] The associated account determination module 640 is configured to filter out associated accounts of each account from other accounts of the application based on the determined social association between each account.
[0156] In an optional implementation, the social relationship chain acquisition module 620 may include:
[0157] The random walk sequence set determination unit is configured to perform, for any account in the application, determining the random walk sequence set corresponding to the account by using a random walk method according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, wherein the random walk sequence set determined by the random walk method includes: a plurality of random walk sequences composed of account information, each random walk sequence being: a random walk sequence formed by determining a subsequent account information corresponding to the previous account information according to the attention relationship and social closeness between the accounts, and the account corresponding to the subsequent account information is: an account with close social connection corresponding to the previous account information;
[0158] The social relationship chain determination unit is configured to determine, for any account in the application, a random walk sequence in a random walk sequence set corresponding to the account as the social relationship chain of the account.
[0159] In an optional implementation, the random walk sequence set determination unit is configured to execute:
[0160] According to the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application;
[0161] According to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence, wherein the updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence;
[0162] Accordingly, according to the order of the account information put into the account information sequence, the following are determined in sequence: the account corresponding to the last account information put into the account information sequence, and the accounts with close social connections, until the number of account information put into the account information sequence reaches a first preset number;
[0163] A second preset number of account information sequences are determined as a set of random walk sequences corresponding to the first account.
[0164] It can be seen that through the technical solution provided by this embodiment, the social relationship chain of each account in the application can be accurately determined based on the attention relationship and social closeness between accounts in the directed information record.
[0165] In an optional implementation, the social association degree determination module is configured to execute:
[0166] Convert each account information in the first social relationship chain of the first account into a first sub-vector, and convert each account information in the second social relationship chain of other accounts in the application into a second sub-vector;
[0167] The first sub-vectors corresponding to the account information in the first social relationship chain are combined into a first vector; the second sub-vectors corresponding to the account information in the second social relationship chain are combined into a second vector;
[0168] The distance between the first vector and each second vector is calculated to obtain the social association degree between the first account and each other account.
[0169] It can be seen that through the technical solution provided by this embodiment, the social association between each account in the application can be accurately determined through the social relationship chain of each account.
[0170] In an optional embodiment, the device may further include:
[0171] The video recommendation module is configured to recommend videos related to a fourth account for each account, where the fourth account is: other accounts among the corresponding associated accounts except for accounts with a direct social association relationship with the account; or, to recommend videos related to a fifth account for each account, where the fifth account is: an account among the corresponding associated accounts that is currently broadcasting live.
[0172] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of content to each account based on the association relationship of each account in the application, and the content can be video.
[0173] According to a third aspect of the embodiments of the present disclosure, there is provided an electronic device, such as Figure 7 As shown, including:
[0174] Processor 710;
[0175] A memory 720 for storing instructions executable by the processor;
[0176] The processor is configured to execute the instructions to implement the method for obtaining account-associated information in an application as described in the first aspect.
[0177] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of content to each account based on the association relationship of each account in the application, and the content can be video.
[0178] Figure 8 8 is a block diagram of an apparatus 800 for acquiring account-related information in an application according to an exemplary embodiment. For example, the apparatus 800 may be provided as a server. Figure 8The apparatus 800 includes a processing component 822, which further includes one or more processors, and a memory resource represented by a memory 832, for storing instructions executable by the processing component 822, such as an application. The application stored in the memory 832 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 822 is configured to execute instructions to execute the account association information acquisition method in any application described in the first aspect.
[0179] The device 800 may also include a power supply component 826 configured to perform power management of the device 800, a wired or wireless network interface 850 configured to connect the device 800 to a network, and an input / output (I / O) interface 858. The device 800 may operate based on an operating system stored in the memory 832, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or the like.
[0180] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of content to each account based on the association relationship of each account in the application, and the content can be video.
[0181] According to a fourth 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 an electronic device, the electronic device can execute the method for obtaining account association information in an application described in the first aspect.
[0182] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of content to each account based on the association relationship of each account in the application, and the content can be video.
[0183] According to a fifth aspect of an embodiment of the present disclosure, there is provided a computer program product comprising instructions, which, when executed on a computer, enables the computer to implement the method for obtaining account association information in an application as described in the first aspect.
[0184] It can be seen that through the technical solution provided by the embodiment of the present disclosure, when determining the associated accounts of an account in an application, the social relationship chain of the account can be accurately determined based on the attention relationship and social closeness between the account and other accounts in the application, and then the social association between the account and other accounts can be determined based on the social relationship chain of the account, so as to accurately screen out the associated accounts associated with the account. After determining the associated accounts of each account in the application, the association relationship between each account in the application is determined, so that the operator of the application can recommend a variety of content to each account based on the association relationship of each account in the application, and the content can be video.
[0185] It should be noted that the account or user information involved in this application is collected and subsequently processed and analyzed with the authorization of the user or account.
[0186] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application 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 specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.
[0187] 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 obtaining account association information in an application. It is characterized in that include: Generate a directed graph according to the following relationship between accounts in the application, wherein the following relationship is used to represent the social association relationship actively or passively established between the accounts, and the directed graph has directed information for marking the following relationship, and the directed information is also used to record the social closeness of the two accounts with the following relationship; the social closeness is used to represent the degree of attention of the two accounts with the following relationship; According to the attention relationship and social closeness between the accounts in the directed information record, a social relationship chain of each account in the application is obtained, wherein the social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account; Based on the acquired social relationship chains of the accounts, the social association degree between the accounts in the application is determined; the similarity degree of the social relationship chains of any two accounts in the application represents the social association degree; Based on the determined social association between each account, filter out the associated accounts of each account from other accounts of the application; The step of obtaining the social relationship chain of each account in the application according to the attention relationship and social closeness between the accounts in the directed information record includes: For any account in the application, a random walk sequence set corresponding to the account is determined by using a random walk method according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, wherein the random walk sequence set determined by the random walk method includes: a plurality of random walk sequences composed of account information, each random walk sequence is: a random walk sequence formed by determining a subsequent account information corresponding to the previous account information according to the attention relationship and social closeness between the accounts, and the account corresponding to the subsequent account information is: an account with close social connection corresponding to the previous account information; For any account in the application, a random walk sequence in a random walk sequence set corresponding to the account is determined as a social relationship chain of the account; Determining the social association between accounts in the application based on the acquired social relationship chain of the account includes: Convert each account information in a first social relationship chain of the first account into a first sub-vector, and convert each account information in a second social relationship chain of other accounts in the application into a second sub-vector; The first sub-vectors corresponding to the account information in the first social relationship chain are combined into a first vector; the second sub-vectors corresponding to the account information in the second social relationship chain are combined into a second vector; The distance between the first vector and each second vector is calculated to obtain the social association degree between the first account and each other account.
2. The method according to claim 1, It is characterized in that For any account in the application, according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, a random walk sequence set corresponding to the account is constructed by using random walk, including: According to the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application; According to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence, wherein the updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence; Accordingly, according to the order of the account information put into the account information sequence, the following are determined in sequence: the account corresponding to the last account information put into the account information sequence, and the accounts with close social connections, until the number of account information put into the account information sequence reaches a first preset number; A second preset number of account information sequences are determined as a set of random walk sequences corresponding to the first account.
3. The method according to any one of claims 1 to 2, It is characterized in that The method further comprises: recommending videos related to a fourth account to each account, wherein the fourth account is: other accounts among the corresponding associated accounts except for accounts having a direct social association relationship with each account; or, Recommend videos related to a fifth account to each account, where the fifth account is: an account currently broadcasting live among the corresponding associated accounts.
4. The method according to any one of claims 1 to 2, It is characterized in that Each vertex in the directed graph corresponds to the account information of an account, and each edge in the directed graph is used to represent that: the two accounts corresponding to the account information corresponding to the two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to represent: the degree of social closeness between the two accounts corresponding to the account information corresponding to the two vertices of the edge.
5. A device for obtaining account-related information in an application, It is characterized in that include: A directed graph generation module is configured to generate a directed graph according to a following relationship between accounts in an application, wherein the following relationship is used to represent a social association relationship actively or passively established between accounts, and the directed graph has directed information for marking the following relationship, and the directed information is also used to record the social closeness of two accounts with the following relationship; the social closeness is used to represent the degree of attention of two accounts with the following relationship; A social relationship chain acquisition module is configured to acquire the social relationship chain of each account in the application according to the attention relationship and social closeness between the accounts recorded in the directed information, wherein the social relationship chain is used to represent: account information of accounts that have a social association relationship with the corresponding account; A social association degree determination module is configured to determine the social association degree between the accounts in the application based on the acquired social relationship chain of the account; the similarity degree of the social relationship chain between any two accounts in the application represents the social association degree; A related account determination module, configured to perform filtering of related accounts of each account from other accounts of the application based on the determined social association between each account; The social relationship chain acquisition module includes: a random walk sequence set determination unit, configured to determine, for any account in the application, a random walk sequence set corresponding to the account according to the attention relationship and social closeness between the account and other accounts recorded in the directed information, using a random walk method, wherein the random walk sequence set determined by the random walk method includes: a plurality of random walk sequences composed of account information, each random walk sequence being: a random walk sequence formed by determining a subsequent account information corresponding to a previous account information according to the attention relationship and social closeness between the accounts, the account corresponding to the subsequent account information being: a socially close account corresponding to the previous account information; the social relationship chain determination unit is configured to determine, for any account in the application, a random walk sequence in the random walk sequence set corresponding to the account as the social relationship chain of the account; The social association degree determination module is configured to execute: converting each account information in the first social relationship chain of the first account into a first sub-vector, and converting each account information in the second social relationship chain of other accounts in the application into a second sub-vector; forming a first vector with the first sub-vectors corresponding to each account information in the first social relationship chain; forming a second vector with the second sub-vectors corresponding to each account information in the second social relationship chain; and calculating the distance between the first vector and each second vector to obtain the social association degree between the first account and each other account.
6. The device according to claim 5, It is characterized in that The random walk sequence set determination unit is configured to execute: According to the attention relationship and social closeness between the first account and other accounts, determine a second account that is socially close to the first account, and put the account information of the first account and the account information of the second account into an empty account information sequence in sequence, where the first account is any account in the application; According to the degree of social closeness between the second account and the updated other accounts, determine a third account that is socially close to the second account, and put the account information of the third account into the account information sequence, wherein the updated other accounts are: other accounts except the accounts corresponding to the account information in the account information sequence; Accordingly, according to the order of the account information put into the account information sequence, the following are determined in sequence: the account corresponding to the last account information put into the account information sequence, and the accounts with close social connections, until the number of account information put into the account information sequence reaches a first preset number; A second preset number of account information sequences are determined as a set of random walk sequences corresponding to the first account.
7. The device according to any one of claims 5 to 6, It is characterized in that The device also includes: The video recommendation module is configured to recommend videos related to a fourth account for each account, where the fourth account is: other accounts among the corresponding associated accounts except accounts with a direct social association relationship with the account; or, to recommend videos related to a fifth account for each account, where the fifth account is: an account among the corresponding associated accounts that is currently broadcasting live.
8. The device according to any one of claims 5 to 6, It is characterized in that Each vertex in the directed graph corresponds to the account information of an account, and each edge in the directed graph is used to represent that: the two accounts corresponding to the account information corresponding to the two vertices of the edge have a follow-up relationship, and each edge in the directed graph has an edge weight, and the edge weight is used to represent: the degree of social closeness between the two accounts corresponding to the account information corresponding to the two vertices of the edge.
9. An electronic device, It is 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 obtaining account-associated information in an application according to any one of claims 1 to 4.
10. A storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the method for obtaining account-related information in an application according to any one of claims 1 to 4.
11. A computer program product comprising instructions, which, when executed on a computer, enables the computer to implement the method for obtaining account-related information in an application according to any one of claims 1 to 4.
