Method, device and storage medium for recommending comment information of songs

Abstract

The present application discloses a method, device and storage medium for recommending comment information of songs, and belongs to the field of computer technology. The method comprises: receiving a comment acquisition request for a corresponding target song sent by a target account; for each comment information of the target song, determining the account relationship distance between the operating account of each operation corresponding to the comment information and the target account, and determining the association degree between the comment information and the target account according to the operation type and account relationship distance corresponding to each operation; determining the comment information to be recommended according to the association degree between each comment information and the target account; and sending the comment information to be recommended to the target account. The use of this application helps to increase the popularity of the comment area.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device and storage medium for recommending comment information of songs. Background Art

[0002] Playing songs through terminal devices has become a part of people's lives. When playing songs, the terminal can display the song's details page, and a comment area is generally set in the details page, where users can comment on the song and browse other users' comments.

[0003] In some applications, the comment area of ​​a song includes two areas, one is the wonderful comments area and the other is the all comments area. The wonderful comments area will recommend some comments that the user may be interested in, and generally the most popular comments are recommended.

[0004] In the process of implementing this application, the inventors found that the related technology has at least the following problems:

[0005] Highly popular comments are not necessarily the comment information that the current user is interested in. Users are less likely to conduct further browsing and operations in the comment area, which is not conducive to increasing the popularity of the comment area. Summary of the invention

[0006] The embodiment of the present application provides a method, device and storage medium for recommending song comment information, which can solve the problem that the related technology is not conducive to increasing the popularity of the comment area. The technical solution is as follows:

[0007] In a first aspect, a method for recommending comment information of a song is provided, the method comprising:

[0008] Receive a comment acquisition request for the corresponding target song sent by the target account;

[0009] For each comment information of the target song, determine the account relationship distance between the operation account of each operation corresponding to the comment information and the target account, and determine the association degree between the comment information and the target account according to the operation type corresponding to each operation and the account relationship distance;

[0010] Determining the review information to be recommended according to the relevance between each review information and the target account;

[0011] The review information to be recommended is sent to the target account.

[0012] In a possible implementation, determining the account relationship distance between the operating account of each operation corresponding to the comment information and the target account includes:

[0013] For the operating account of each operation corresponding to the comment information, if there is at least one connection path between the operating account and the target account in the account community relationship graph data, the number of connection edges contained in the shortest connection path in at least one connection path between the operating account and the target account is determined as the account relationship distance between the operating account and the target account; if there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account.

[0014] In a possible implementation, determining the relevance between the comment information and the target account according to the operation type and the account relationship distance corresponding to each operation includes:

[0015] Determine the correlation impact value corresponding to each operation according to the pre-stored correspondence between the operation type and the correlation impact value and the operation type corresponding to each operation;

[0016] Determine a weight value corresponding to each operation according to the account relationship distance corresponding to each operation, wherein the weight value is negatively correlated with the account relationship distance;

[0017] The relevance between the comment information and the target account is determined according to the relevance influence value and the weight value corresponding to each operation.

[0018] In a possible implementation, determining the relevance between the comment information and the target account according to the relevance influence value and the weight value corresponding to each operation includes:

[0019] The product of the relevance influence value and the weight value corresponding to each operation is determined, and all the obtained products are added together to obtain the relevance between the comment information and the target account.

[0020] In a possible implementation, the operation type includes one or more of a click-to-expand operation type, a like operation type, and a reply operation type.

[0021] In a possible implementation, when the operation type includes the click-to-expand operation type, the like operation type and the reply operation type, in the corresponding relationship, the association influence values ​​corresponding to the click-to-expand operation type, the like operation type and the reply operation type are all positive values, and the association influence value corresponding to the like operation type is greater than the association influence value corresponding to the click-to-expand operation type, and the association influence value corresponding to the reply operation type is greater than the association influence value corresponding to the like operation type.

[0022] In a possible implementation, determining the review information to be recommended according to the relevance between each review information and the target account includes:

[0023] Determine the selection probability corresponding to each comment information according to the relevance between each comment information and the target account;

[0024] According to the selection probability corresponding to each piece of comment information, a preset number of comment information are selected from the comment information of the target song as the comment information to be recommended.

[0025] In a possible implementation, determining the review information to be recommended according to the relevance between each review information and the target account includes:

[0026] A preset number of review information with the highest correlation with the target account is obtained as the review information to be recommended.

[0027] In a possible implementation, the review information to be recommended includes multiple pieces, and the method further includes:

[0028] Determining display order information of the plurality of review information to be recommended according to the relevance between each review information to be recommended and the target account;

[0029] The display sequence information is sent to the target account.

[0030] In a second aspect, a device for recommending comment information of a song is provided, the device comprising:

[0031] A receiving module, used for receiving a comment acquisition request for a corresponding target song sent by a target account;

[0032] A correlation module, for determining, for each comment information of the target song, an account relationship distance between an operation account of each operation corresponding to the comment information and the target account, and determining a correlation degree between the comment information and the target account according to an operation type corresponding to each operation and the account relationship distance;

[0033] A recommendation module, used to determine the review information to be recommended according to the relevance between each review information and the target account;

[0034] The sending module is used to send the review information to be recommended to the target account.

[0035] In a possible implementation, the association module is used to:

[0036] For the operating account of each operation corresponding to the comment information, if there is at least one connection path between the operating account and the target account in the account community relationship graph data, the number of connection edges contained in the shortest connection path in at least one connection path between the operating account and the target account is determined as the account relationship distance between the operating account and the target account; if there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account.

[0037] In a possible implementation, the association module is used to:

[0038] Determine the correlation impact value corresponding to each operation according to the pre-stored correspondence between the operation type and the correlation impact value and the operation type corresponding to each operation;

[0039] Determine a weight value corresponding to each operation according to the account relationship distance corresponding to each operation, wherein the weight value is negatively correlated with the account relationship distance;

[0040] The relevance between the comment information and the target account is determined according to the relevance influence value and the weight value corresponding to each operation.

[0041] In a possible implementation, the association module is used to:

[0042] The product of the relevance influence value and the weight value corresponding to each operation is determined, and all the obtained products are added together to obtain the relevance between the comment information and the target account.

[0043] In a possible implementation, the operation type includes one or more of a click-to-expand operation type, a like operation type, and a reply operation type.

[0044] In a possible implementation, when the operation type includes the click-to-expand operation type, the like operation type and the reply operation type, in the corresponding relationship, the association influence values ​​corresponding to the click-to-expand operation type, the like operation type and the reply operation type are all positive values, and the association influence value corresponding to the like operation type is greater than the association influence value corresponding to the click-to-expand operation type, and the association influence value corresponding to the reply operation type is greater than the association influence value corresponding to the like operation type.

[0045] In a possible implementation, the recommendation module is used to:

[0046] Determine the selection probability corresponding to each comment information according to the relevance between each comment information and the target account;

[0047] According to the selection probability corresponding to each piece of comment information, a preset number of comment information are selected from the comment information of the target song as the comment information to be recommended.

[0048] In a possible implementation, the recommendation module is used to:

[0049] A preset number of review information with the highest correlation with the target account is obtained as the review information to be recommended.

[0050] In a possible implementation, the review information to be recommended includes multiple pieces;

[0051] The recommendation module is further used to determine the display order information of the plurality of review information to be recommended according to the relevance between each review information to be recommended and the target account;

[0052] The sending module is further used to send the display sequence information to the target account.

[0053] According to a third aspect, a computer device is provided, comprising a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the operations performed by the method for recommending comment information of songs as described in the first aspect.

[0054] In a fourth aspect, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the operations performed by the method for recommending comment information of songs as described in the first aspect.

[0055] The beneficial effects of the technical solution provided by the embodiment of the present application are:

[0056] In the embodiment of the present application, the relevance between the comment information and the target account is determined by the account relationship distance between the operating account of the comment information and the target account, as well as the operation type corresponding to the operation, and then the comment information to be recommended is determined based on the relevance, and recommended to the target account. Because accounts with close relationships are more likely to interact with each other, pushing comment information that users with close account relationship distance to the target account are concerned about to the target account makes it easier to stimulate interaction in the comment area, which helps to increase the popularity of the comment area. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0058] Figure 1 This is a schematic diagram of an interface display provided in an embodiment of the present application;

[0059] Figure 2 This is a schematic diagram of account community relationship graph data provided by an embodiment of the present application;

[0060] Figure 3 This is a flow chart of a method for recommending comment information of a song provided in an embodiment of the present application;

[0061] Figure 4 This is a flow chart of a method for recommending comment information of a song provided in an embodiment of the present application;

[0062] Figure 5 This is a schematic diagram of the structure of a device for recommending comment information of a song provided in an embodiment of the present application;

[0063] Figure 6 It is a structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0065] The embodiment of the application provides a method for recommending comment information of songs, and the execution subject of the method can be a server. The server can be a background server of a music application. The server can be a single server or a server group. If it is a single server, the server can be responsible for all the processing in the following scheme. If it is a server group, different servers in the server group can be responsible for different processing in the following scheme. The specific processing distribution can be arbitrarily set by the technician according to actual needs, and will not be repeated here.

[0066] The server may include components such as a processor, a memory, and a communication component. The processor is connected to the memory and the communication component respectively.

[0067] The processor may be a CPU (Central Processing Unit). The processor may be used to determine the account relationship distance between the operating account and the target account for each operation corresponding to the comment information, may be used to determine the relevance between the comment information and the target account according to the operation type and account relationship distance corresponding to each operation, may be used to determine the comment information to be recommended according to the relevance between each comment information and the target account, and so on.

[0068] The memory may include ROM (Read-Only Memory), RAM (Random Access Memory), CD-ROM (Compact Disc Read-Only Memory), magnetic disk, optical data storage device, etc. The memory may be used for pre-stored data, generated intermediate data, and generated result data during the processing of the comment information of the recommended songs, such as a large amount of comment data, operation data of different users on the comments, account relationship distance, and the correlation between the comment information and the target account.

[0069] The communication component may be a wired network connector, a WiFi (Wireless Fidelity) module, a Bluetooth module, a cellular network communication module, etc. The communication component may be used to transmit data with other devices, which may be other servers or terminals, etc. For example, the communication component may receive a comment acquisition request sent by a target account.

[0070] In the music application, there are various song lists, some of which are generated by the system and recommended to users, and some are edited by users themselves. In addition, song lists can be shared among friends. When users search for songs, the search results will also be presented in the form of song lists. Users select songs in the song list and click on the option of a song to play it. The terminal will enter the details page of the song. The details page shows the name of the song, the name of the singer, the name of the album, the type label, the cover image, the sound quality information, etc. There is a comment area in the details page, such as Figure 1 As shown, the comment area includes two areas, one is the wonderful comment area (or recommended comment area) and the other is the all comment area. The wonderful comment area recommends some comment information that the user may be interested in, and the all comment area includes all the comment information of the song. The embodiment of the present application provides a mechanism for selecting comment information from the wonderful comments.

[0071] Explanation of terms involved in the plan:

[0072] Account community relationship graph data, graph data of the friendship relationship between user-reflected accounts, such as Figure 2As shown in the figure, the account community relationship graph data includes a large number of nodes and connecting edges. Nodes represent accounts, and edges represent the friendship between accounts. If there is a connecting edge between two nodes, it means that the accounts corresponding to the two nodes have a friendship. If there is no connecting edge between two nodes, it means that the accounts corresponding to the two nodes do not have a friendship. In different application scenarios, the account community relationship graph data can be used to analyze the relationship between different accounts.

[0073] A connection path is a connection path between two nodes in the account community relationship graph data, which is a path connecting two nodes composed of one or more connection edges connected end to end. Figure 2 In the connection path between account A1 and account A4, account A1 and account A2 are friends, account A2 and account A3 are friends, and account A3 and account A4 are friends. There may be one or more connection paths between two accounts, or there may be no connection path.

[0074] Figure 3 is a flow chart of a method for recommending song comment information provided by an embodiment of the present application. Figure 3 , the embodiment comprises:

[0075] 301, receiving a comment acquisition request for a corresponding target song sent by a target account.

[0076] The target account can be any account, and the target song can be any type of song.

[0077] In practice, when a user wants to listen to a song, he can operate the terminal to start the music application. The music application is logged in with the user's account, which is the target account mentioned above. The terminal will display the main interface of the music application. The user can operate in the main interface and select the song he wants to play. For example, he can use the search function to search for songs using keywords and select a song in the search result list to play. Alternatively, he can directly click on the recommended songs in the main interface to play. Alternatively, the user can open his own playlist in the main interface and click on a song to play. After the user clicks on any song, the terminal starts playing the song, and the application will enter the details page of the song. This song is the target song mentioned above.

[0078] A comment area can be set in the details page, and when the application enters the details page, it can trigger the terminal to send a comment acquisition request for the target song to the server. Alternatively, the user operates in the details page, and when the details page scrolls to a position close to the comment area, the terminal is triggered to send a comment acquisition request for the target song to the server. Alternatively, the comment area may not be directly displayed in the details page, but a comment option is displayed. When the user clicks the comment option, it triggers the application to enter the comment interface of the target song. The comment interface is the above-mentioned comment area, which includes the wonderful comment area and the all comment area. When the comment interface is triggered, the terminal is also triggered to send a comment acquisition request for the target song to the server.

[0079] Comment acquisition request The user requests the server to load comment information. The comment acquisition request carries the ID of the target song and the ID of the target account.

[0080] When the server receives a comment acquisition request sent by a terminal logged in with a target account, it can parse the comment acquisition request to obtain the identifier of the target song and the identifier of the target account carried therein.

[0081] 302. For each comment information of the target song, determine the account relationship distance between the operation account of each operation corresponding to the comment information and the target account, and determine the association degree between the comment information and the target account according to the operation type corresponding to each operation and the account relationship distance.

[0082] Among them, the account relationship distance is a parameter reflecting the closeness of the friendship relationship between accounts. For example, in the first case, account A1 and account A2 are friends, so the friendship distance between account A1 and account A2 is the closest. In the second case, the friend of account A1 is the friend of account A2, and account A1 and account A2 are not friends, so the friendship distance between account A1 and account A2 is slightly farther than the first case. The account relationship distance can be the number of connecting edges of the shortest path between accounts in the account community relationship graph data.

[0083] Operation types may include click to expand, like, and dislike. Click to expand can expand the reply information corresponding to a comment information. Like indicates that other users have a positive attitude towards the comment information, such as agreeing or liking it. Dislike indicates that other users have a negative attitude towards the comment information, such as disagreeing or liking it.

[0084] In implementation, the server can traverse each comment information of the target song. For each comment information, the server can traverse each operation of the comment information and obtain the relevant information. The operation account corresponding to the operation can be obtained. The operation account and the target account are searched in the account community relationship graph data to determine all connection paths between the operation account and the target account.

[0085] If there is at least one connection path between the operating account and the target account, the shortest connection path among the at least one connection path between the operating account and the target account is determined. The length of the connection path can be represented by the number of connection edges, and the length of each connection path can be determined, and then the shortest connection path can be determined by comparison. The number of connection edges included in the shortest connection path is further determined as the account relationship distance between the operating account and the target account.

[0086] If there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account. The maximum account relationship distance here is a preset value, which is pre-set by the technical staff. When setting this parameter, the maximum path length between two accounts in the graph data can be considered. When setting the maximum account relationship distance, its value can be greater than the maximum path length, or equal to the maximum path length.

[0087] After determining the account relationship distance between the operating account and the target account for each operation, the correlation between the comment information and the target account is determined based on the operation type and account relationship distance corresponding to each operation. It can be considered that the operations of liking and clicking to expand have a positive impact on the correlation, while the operations of disliking have a negative impact on the correlation. In addition, the smaller the account relationship distance, the greater the positive or negative impact mentioned above. The specific processing will be introduced in the following content.

[0088] 303 , determining the comment information to be recommended according to the relevance between each comment information and the target account.

[0089] There are many ways to determine the review information to be recommended. Two feasible ways are given below:

[0090] Method 1: directly sort and recommend by relevance.

[0091] The corresponding processing is: obtaining a preset number of review information with the highest correlation with the target account as the review information to be recommended.

[0092] A preset number (eg, 4) of review information may be selected from all the review information in descending order of corresponding relevance as the review information to be recommended.

[0093] Method 2: Assign probability to comment information according to the relevance and perform random selection.

[0094] The corresponding processing is: according to the correlation between each comment information and the target account, the selection probability corresponding to each comment information is determined; according to the selection probability corresponding to each comment information, a preset number of comment information are selected from the comment information of the target song as the comment information to be recommended.

[0095] The selection probability of each comment information is positively correlated with its corresponding relevance. For example, the relevance is recorded as X, the probability is recorded as P, and the range of relevance is 0-1, then the conversion relationship between P and X can be: P = (X*100)%. If the relevance of a comment information with the target account is 0.8, then the selection probability corresponding to the comment information is 80%.

[0096] After determining the selection probability of each comment information, the comment information can be randomly selected according to the selection probability. The corresponding processing can be:

[0097] According to the selection probability, a corresponding number of values ​​to be extracted is assigned to each comment information. All the values ​​to be extracted assigned to all comment information are different. The number of values ​​to be extracted corresponding to each comment information is positively correlated with the selection probability corresponding to the comment information. For example, the number of values ​​to be extracted is recorded as M, and M=P*100 can be set. If the selection probability of a certain comment information is 80%, then the number of values ​​to be extracted corresponding to it is 80. If the selection probability of a certain comment information is 76%, then the number of values ​​to be extracted corresponding to it is 76. In this way, after all the comment information is assigned a number of different values ​​to be extracted, a random extraction is performed among all the assigned values ​​to be extracted. The comment information corresponding to the target value to be extracted is determined, which is the comment information to be recommended. Multiple extractions can be performed to obtain multiple comment information to be recommended. If no comment information is extracted, all the values ​​to be extracted corresponding to this comment information are deleted, and the next extraction is performed.

[0098] This method can make highly relevant comment information more likely to be recommended, while preventing comment information with less exposure from becoming less and less exposed, thus forming a vicious cycle.

[0099] 304, sending the review information to be recommended to the target account.

[0100] In implementation, there may be multiple comment information to be recommended. The display order information of the multiple comment information to be recommended may be determined according to the relevance between each comment information to be recommended and the target account. That is, the comment information is arranged in descending order according to the corresponding relevance to obtain the display order information. Then, the display order information and the comment information to be recommended may be sent together to the terminal logged in by the target account. After the terminal receives the comment information with recommendation and the display order information, it may arrange the multiple comment information according to the display order information and display them in the above-mentioned wonderful comment area.

[0101] In addition, the server can also send the contents of all comment areas to the terminal at the same time. In this case, the server can obtain a preset number of latest published comment information in the order of publication time and send them to the terminal. After receiving these comment information, the terminal can arrange these comment information in the order of publication time and display them in the above-mentioned all comment areas.

[0102] In an embodiment of the present application, for the case where the account community relationship graph data is an undirected graph, step 302 can adopt the above method. In another case, the account community relationship graph data is a directed graph, that is, graph data with a direction of the connecting edge. The node in the direction indicated by the connecting edge can be called the destination node corresponding to the connecting edge, and the node at the other end of the connecting edge can be called the source node corresponding to the connecting edge. For example, the connecting edge represents that the account of the source node follows the account of the destination node. In a directed graph, there can be two connecting edges between two nodes, and the directions of the two connecting edges are opposite, so that the two nodes are source nodes and destination nodes to each other. For example, the existence of two connecting edges between two nodes can indicate that the accounts corresponding to the two nodes follow each other. For the case of a directed graph, the specific processing of determining the account relationship distance between the operating account and the target account for each operation corresponding to the comment information in the above step 302 can be:

[0103] For the operating account of each operation corresponding to the comment information, if there is at least one connection path between the operating account and the target account in the account community relationship graph data, the path length of the shortest connection path in at least one connection path between the operating account and the target account is determined as the account relationship distance between the operating account and the target account; if there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account.

[0104] In this case, the path length of each connection path between the operating account and the target account is determined, and then the shortest path length is obtained as the account relationship distance between the operating account and the target account.

[0105] When determining a connection path, connection edges in any direction can be considered as part of the connection path. If there are two connection edges between two nodes, the two connection edges will be counted together in the connection path (i.e., the two connection edges will not be counted separately in different connection paths).

[0106] For the calculation of path length, the path lengths corresponding to different connection situations between adjacent nodes can be preset. If there is only one connection edge between adjacent nodes, and the direction of the connection edge matches the direction from the operating account to the target account, the path length between the adjacent nodes is the first value. If there is only one connection edge between adjacent nodes, and the direction of the connection edge matches the direction from the target account to the operating account, the path length between the adjacent nodes is the second value. If there are two connection edges between adjacent nodes, the path length between the adjacent nodes is the third value. The third value can be set to be smaller than the second value, and the second value can be set to be smaller than the first value.

[0107] When calculating the path length, for each connection path, the connection status between every two adjacent nodes on the connection path can be obtained. The connection status of any two adjacent nodes should belong to one of the above three situations. According to the above settings, the path length corresponding to the connection status between every two adjacent nodes can be determined, and then the path lengths between all adjacent nodes on the connection path are added to obtain the path length of the connection path. In this way, the path length of each connection path can be calculated.

[0108] In the embodiment of the present application, the process of determining the relevance between the comment information and the target account according to the operation type and account relationship distance corresponding to each operation can be as follows: Figure 4 As shown, the following steps are included:

[0109] 401 , determining the correlation influence value corresponding to each operation according to the pre-stored correspondence between the operation type and the correlation influence value and the operation type corresponding to each operation.

[0110] The operation type includes one or more of the click-to-expand operation type, the like operation type, the reply operation type, and the dislike operation type. Different operation types can reflect different relevance between the comment information and the target account, which is the relevance impact value. For example, the comment information liked by the friends of the target account has a higher relevance to the target account, so the relevance impact value corresponding to the like operation type is set to a positive value.

[0111] When the operation types include a click-to-expand operation type, a like operation type, and a reply operation type, in the corresponding relationship, the association influence values ​​corresponding to the click-to-expand operation type, the like operation type, and the reply operation type are all positive values, and the association influence value corresponding to the reply operation type is greater than the association influence value corresponding to the like operation type, and the association influence value corresponding to the like operation type is greater than the association influence value corresponding to the click-to-expand operation type.

[0112] When the operation type includes a tap operation type, in the corresponding relationship, the correlation influence value corresponding to the tap operation type is a negative value.

[0113] In implementation, a score (i.e., correlation influence value) is set in advance for different operation types, and the score corresponding to the click-to-expand operation type is Y1, the score corresponding to the like operation type is Y2, the score corresponding to the dislike operation type is Y3, and so on. For example, the correlation influence value corresponding to the like operation type is 2, the correlation influence value corresponding to the dislike operation type is -2, and the correlation influence value corresponding to the click-to-expand operation type is 1.

[0114] 402. Determine a weight value corresponding to each operation according to the account relationship distance corresponding to each operation, wherein the weight value is negatively correlated with the account relationship distance.

[0115] In implementation, there are many ways to determine the weight value. The reciprocal of the account relationship distance can be determined as the weight value, or the sum of the reciprocal of the account relationship distance and a fixed value can be determined as the weight value, or the reciprocal of the square of the account relationship distance can be determined as the weight value.

[0116] Here is an example, taking the comment R1 of song B as an example, assuming that two accounts have operated on the comment R1, account A1 has clicked to expand R1, and account A2 has liked R1. At this time, the account relationship distance S1 between account A1 and target account A0 can be determined, and the account relationship distance S2 between account A2 and account A0 can be determined, and then the weights a1=1 / S1 and a2=1 / S2 can be determined.

[0117] 403 , determining the relevance between the comment information and the target account according to the relevance influence value and the weight value corresponding to each operation.

[0118] In implementation, the product of the correlation influence value and the weight value corresponding to each operation is determined, and all the products are added together to obtain the correlation between the comment information and the target account.

[0119] L can be calculated as L=Y1*a1+Y2*a2 to get the total score L, which is the correlation between comment R1 and user 0. Based on this method, the correlation between each comment of song B and user 0 can be calculated, and then the comment information to be displayed can be determined based on the correlation.

[0120] In the embodiment of the present application, the relevance between the comment information and the target account is determined by the account relationship distance between the operating account of the comment information and the target account, as well as the operation type corresponding to the operation, and then the comment information to be recommended is determined based on the relevance, and recommended to the target account. Because accounts with close relationships are more likely to interact with each other, pushing comment information that users with close account relationship distance to the target account are concerned about to the target account makes it easier to stimulate interaction in the comment area, which helps to increase the popularity of the comment area.

[0121] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0122] The present application also provides a device for recommending comment information of songs. The device can be applied to the server in the above embodiment, such as Figure 5 As shown, the device comprises:

[0123] Receiving module 510, used to receive a comment acquisition request for a corresponding target song sent by a target account;

[0124] The association module 520 is used to determine, for each comment information of the target song, the account relationship distance between the operation account of each operation corresponding to the comment information and the target account, and determine the association degree between the comment information and the target account according to the operation type corresponding to each operation and the account relationship distance;

[0125] A recommendation module 530, configured to determine the review information to be recommended according to the relevance between each review information and the target account;

[0126] The sending module 540 is used to send the review information to be recommended to the target account.

[0127] In a possible implementation, the association module 520 is used to:

[0128] For the operating account of each operation corresponding to the comment information, if there is at least one connection path between the operating account and the target account in the account community relationship graph data, the number of connection edges contained in the shortest connection path in at least one connection path between the operating account and the target account is determined as the account relationship distance between the operating account and the target account; if there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account.

[0129] In a possible implementation, the association module 520 is used to:

[0130] Determine the correlation impact value corresponding to each operation according to the pre-stored correspondence between the operation type and the correlation impact value and the operation type corresponding to each operation;

[0131] Determine a weight value corresponding to each operation according to the account relationship distance corresponding to each operation, wherein the weight value is negatively correlated with the account relationship distance;

[0132] The relevance between the comment information and the target account is determined according to the relevance influence value and the weight value corresponding to each operation.

[0133] In a possible implementation, the association module 520 is used to:

[0134] The product of the relevance influence value and the weight value corresponding to each operation is determined, and all the obtained products are added together to obtain the relevance between the comment information and the target account.

[0135] In a possible implementation, the operation type includes one or more of a click-to-expand operation type, a like operation type, and a dislike operation type;

[0136] When the operation type includes the click-to-expand operation type and the like operation type, in the corresponding relationship, the association influence values ​​corresponding to the click-to-expand operation type and the like operation type are both positive values, and the association influence value corresponding to the like operation type is greater than the association influence value corresponding to the click-to-expand operation type;

[0137] When the operation type includes the tap operation type, in the corresponding relationship, the correlation influence value corresponding to the tap operation type is a negative value.

[0138] In a possible implementation, the recommendation module 530 is used to:

[0139] Determine the selection probability corresponding to each comment information according to the relevance between each comment information and the target account;

[0140] According to the selection probability corresponding to each piece of comment information, a preset number of comment information are selected from the comment information of the target song as the comment information to be recommended.

[0141] In a possible implementation, the recommendation module 530 is used to:

[0142] A preset number of review information with the highest correlation with the target account is obtained as the review information to be recommended.

[0143] In a possible implementation, the review information to be recommended includes multiple pieces;

[0144] The recommendation module 530 is further configured to determine display order information of the plurality of review information to be recommended according to the relevance between each review information to be recommended and the target account;

[0145] The sending module 540 is further configured to send the display sequence information to the target account.

[0146] In the embodiment of the present application, the relevance between the comment information and the target account is determined by the account relationship distance between the operating account of the comment information and the target account, as well as the operation type corresponding to the operation, and then the comment information to be recommended is determined based on the relevance, and recommended to the target account. Because accounts with close relationships are more likely to interact with each other, pushing comment information that users with close account relationship distance to the target account are concerned about to the target account makes it easier to stimulate interaction in the comment area, which helps to increase the popularity of the comment area.

[0147] It should be noted that: the device for recommending song comment information provided in the above embodiment only uses the division of the above functional modules as an example when recommending song comment information. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device for recommending song comment information provided in the above embodiment and the method embodiment for recommending song comment information belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0148] Figure 6 600 may have relatively large differences due to different configurations or performances, and may include one or more processors 601 and one or more memories 602, wherein the memory 602 stores at least one instruction, and the at least one instruction is loaded and executed by the processor 601 to implement the methods provided by the above-mentioned various method embodiments. Of course, the server may also have components such as a wired or wireless network interface, a keyboard, and an input and output interface for input and output, and the server may also include other components for implementing device functions, which will not be described in detail here.

[0149] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including instructions, and the instructions can be executed by a processor in a terminal to complete the method of recommending comment information of songs in the above embodiment. The computer-readable storage medium can be non-transitory. For example, the computer-readable storage medium can be a ROM (Read-Only Memory), a RAM (Random Access Memory), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0150] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0151] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for recommending comment information of a song, characterized in that: The method comprises: Receive a comment acquisition request for the corresponding target song sent by the target account; For each comment information of the target song, determine the account relationship distance between the operating account of each operation corresponding to the comment information and the target account, determine the association influence value corresponding to each operation according to the pre-stored correspondence between the operation type and the association influence value, and the operation type corresponding to each operation, determine the weight value corresponding to each operation according to the account relationship distance corresponding to each operation, determine the product of the association influence value corresponding to each operation and the weight value, add all the obtained products, and obtain the association degree between the comment information and the target account, wherein the account relationship distance reflects the closeness of the friend relationship between the operating account and the target account. , the weight value is negatively correlated with the account relationship distance, the operation type includes one or more of a click-to-expand operation type, a like operation type, and a reply operation type. When the operation type includes the click-to-expand operation type, the like operation type, and the reply operation type, in the corresponding relationship, the association influence values ​​corresponding to the click-to-expand operation type, the like operation type, and the reply operation type are all positive values, and the association influence value corresponding to the like operation type is greater than the association influence value corresponding to the click-to-expand operation type, and the association influence value corresponding to the reply operation type is greater than the association influence value corresponding to the like operation type; Determine the selection probability corresponding to each comment information according to the relevance between each comment information and the target account; Allocating a corresponding number of values ​​to be extracted to each piece of comment information according to the selection probability, wherein the number of values ​​to be extracted corresponding to each piece of comment information is positively correlated with the selection probability corresponding to the comment information; Randomly draw from all the assigned values ​​to be extracted, and use the comment information corresponding to the drawn target value to be extracted as the comment information to be recommended. After each comment information is drawn, all the values ​​to be extracted corresponding to this comment information are deleted, and then the next draw is performed; The review information to be recommended is sent to the target account.

2. The method according to claim 1, characterized in that The determining the account relationship distance between the operating account of each operation corresponding to the comment information and the target account includes: For the operating account of each operation corresponding to the comment information, if there is at least one connection path between the operating account and the target account in the account community relationship graph data, the number of connection edges contained in the shortest connection path in at least one connection path between the operating account and the target account is determined as the account relationship distance between the operating account and the target account; if there is no connection path between the operating account and the target account in the account community relationship graph data, the pre-stored maximum account relationship distance is obtained as the account relationship distance between the operating account and the target account.

3. The method according to any one of claims 1 or 2, characterized in that: The review information to be recommended includes a plurality of pieces, and the method further includes: Determining display order information of the plurality of review information to be recommended according to the relevance between each review information to be recommended and the target account; The display sequence information is sent to the target account.

4. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the operations performed by the method for recommending comment information of a song as described in any one of claims 1 to 3.

5. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the operation performed by the method for recommending comment information of a song as described in any one of claims 1 to 3.

Images