Information recommendation method and device, electronic equipment and computer program product

Abstract

The application relates to the field of mobile information, and provides an information recommendation method and device, electronic equipment and computer program product. The method comprises the following steps: determining a social relationship network of a current user, and determining initial grade values of relationship edges between initial nodes in the social relationship network; performing transition node setting on the social relationship network according to the initial grade values, so as to obtain a behavior relationship network; calculating social influence of each initial user in the behavior relationship network; determining a target user from the initial users based on the social influence, and performing information recommendation according to the target user. The application can improve the timeliness of recommended content, so that the recommended content can be accurately matched with the future behavior of the recommended user, and the accuracy of information recommendation is improved.

Description

Technical Field

[0001] This application relates to the field of mobile information technology, specifically to an information recommendation method, apparatus, electronic device, and computer program product. Background Technology

[0002] Currently, information recommendations are made based on users' social networks. For example, when viewing content online, recommendations are made based on the user's own browsing history and that of their social friends. For instance, when a user watches short videos, a playlist of videos is recommended based on the user's past consumption (shopping) and viewing behavior; similarly, when a user reads an ebook, a playlist of ebooks is recommended based on the reading history of their friends (book lovers); and when a user purchases a ringback tone, a list of ringback tones from users with whom the user has a social relationship (or who is currently on the phone) is recommended.

[0003] However, all of the above-mentioned information recommendation methods determine the recommended content based on historical browsing behavior, resulting in poor timeliness of the recommended content. Furthermore, due to the extremely rapid pace of content updates and iterations on the internet, the accuracy of information recommendations is currently low. Summary of the Invention

[0004] This application provides an information recommendation method, apparatus, electronic device, and computer program product to solve the technical problem of low accuracy in current information recommendation.

[0005] In a first aspect, embodiments of this application provide an information recommendation method, including:

[0006] Determine the current user's social relationship network, and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0007] Based on the initial level values, transition nodes are set in the social relationship network to obtain a behavioral relationship network;

[0008] Calculate the social influence of each initial user in the behavioral relationship network;

[0009] Based on the aforementioned social influence, target users are identified from the initial users, and information is recommended based on the target users.

[0010] In one embodiment, the step of setting transition nodes in the social relationship network based on each of the initial level values ​​to obtain a behavioral relationship network includes:

[0011] Determine whether there are adjacent relationship edges with unequal initial level values ​​in the social relationship network;

[0012] If such a network exists, a transition node is set on the adjacent relationship edges with unequal initial level values ​​in the social relationship network to obtain the behavioral relationship network.

[0013] In one embodiment, the step of setting transition nodes on adjacent relationship edges with unequal initial ranking values ​​in the social relationship network includes:

[0014] For each adjacent edge in the social relationship network with unequal initial ranking values, the following steps are performed respectively:

[0015] Determine the grade difference between the initial grade values ​​corresponding to the two current relation edges;

[0016] Set a transition node with the same value as the grade difference between the two current relation edges.

[0017] In one embodiment, the step of calculating the social influence of each initial user in the behavioral relationship network includes:

[0018] Traverse all nodes in the behavioral relationship network to obtain node attribute data;

[0019] The initial node is determined from all traversed nodes based on the node attribute data;

[0020] The social influence of the user corresponding to each initial node in the behavioral relationship network is calculated based on the node attribute data.

[0021] In one embodiment, the step of calculating the social influence of the user corresponding to each initial node in the behavioral relationship network based on the node attribute data includes:

[0022] The number of relation chains and edge data included in each initial node of the behavioral relation network are determined from the node attribute data;

[0023] The social influence of the user corresponding to each initial node in the behavioral relationship network is calculated based on the number of relationship chains and the edge data.

[0024] In one embodiment, the step of determining target users from the initial users based on the social influence of each of the initial users includes:

[0025] The social influences are sorted according to a preset sorting method, wherein each social influence corresponds to an initial user.

[0026] The initial users corresponding to the social influences that fall within the preset sorting range after sorting are identified as target users.

[0027] In one embodiment, the step of recommending information based on the target user includes:

[0028] For each of the aforementioned target users, perform the following steps respectively:

[0029] Determine the target level value of each relation edge corresponding to the relation chain of the current target user in the behavioral relation network;

[0030] Compare each of the target level values ​​with a preset level threshold;

[0031] Determine the target level value that is greater than the preset level threshold among all the target level values;

[0032] Information recommendations are made based on the users pointed to by the relation edges containing target level values ​​greater than the preset level threshold.

[0033] Secondly, embodiments of this application provide an information recommendation device, comprising:

[0034] The determination module is used to determine the current user's social relationship network and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0035] The prediction module is used to set transition nodes in the social relationship network based on the initial level values ​​to obtain a behavioral relationship network;

[0036] The calculation module is used to calculate the social influence of each initial user in the behavioral relationship network;

[0037] The recommendation module is used to determine target users from the initial users based on their social influence and to recommend information based on the target users.

[0038] Thirdly, embodiments of this application provide an electronic device, including a processor and a memory storing a computer program, wherein the processor executes the program to implement the steps of the information recommendation method described in the first or second aspect.

[0039] Fourthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the information recommendation method described in the first or second aspect.

[0040] The information recommendation method, apparatus, electronic device, and computer program product provided in this application improve the timeliness of recommended content by setting transitional nodes in the current user's social relationship network based on the initial level values ​​of the relationship edges between the initial nodes in the current user's social relationship network to achieve user behavior prediction and obtain a behavioral relationship network. Then, based on the behavioral relationship network, the social influence of various user behaviors in the network is calculated. Target users are determined according to each social influence and information is recommended based on the target users. This improves the accuracy of information recommendation by ensuring that the recommended content accurately matches the future behavior of the recommended users. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating the information recommendation method provided in an embodiment of this application;

[0043] Figure 2 This is one of the scenario illustrations of the information recommendation method provided in the embodiments of this application;

[0044] Figure 3 This is a second scenario illustration of the information recommendation method provided in the embodiments of this application;

[0045] Figure 4 This is the third scenario illustration of the information recommendation method provided in the embodiments of this application;

[0046] Figure 5 This is the fourth scenario illustration of the information recommendation method provided in the embodiments of this application;

[0047] Figure 6 This is the fifth scenario illustration of the information recommendation method provided in the embodiments of this application;

[0048] Figure 7 This is a sixth scenario illustration of the information recommendation method provided in the embodiments of this application;

[0049] Figure 8 This is a functional module diagram of an embodiment of the information recommendation device of this application;

[0050] Figure 9 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] Figure 1 A flowchart illustrating the information recommendation method provided in this application embodiment. (Refer to...) Figure 1 This application provides an information recommendation method, which may include:

[0053] Step S100: Determine the current user's social relationship network and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0054] The information recommendation method in this embodiment can be applied to electronic devices such as smartphones, tablets, and PCs. In this embodiment, the electronic device can obtain the current user's friend relationships and the social relationships of each friend in those friend relationships, thereby determining the current user's social relationship network.

[0055] The user's current friend relationships can include online relationships and offline relationships. Online relationships come from various information platforms, such as social platforms for sharing short, real-time information and platforms for information interaction and transmission.

[0056] Offline relationships come from users' address books and communication records, such as: mobile phone address books, call and answer records, SMS sending and receiving records, express delivery records, etc.

[0057] It should be noted that the social relationships of friends are the same as the current user's friend relationships, and may include both online and offline relationships.

[0058] In a defined social relationship network of the current user, the current user, the current user's friends, and the friends of the user's friends can all be considered user nodes in the network. Each user node in the social relationship network is designated as an initial node. The edge formed between the current user and their friend's nodes uniquely represents the relationship between the user and their friend; this edge is called a relationship edge. For example... Figure 2 As shown, Figure 2 This is one of the scenario diagrams illustrating the information recommendation method provided in the embodiments of this application. Figure 2In the context of the relationship A, C, D, and E are friends, A->C means C is a friend of A (e.g., C is in A's contacts), and C->A means A is a friend of C (i.e., A is in C's contacts). Here, "contacts" includes all online and offline friend relationships. Similarly, in the relationship B->F->E, B->F means F is a friend of B, and F->E means E is a friend of F. It should be noted that... Figure 2 This is just a simple example; real users' social networks are enormous and complex.

[0059] After determining the current user's social network, the electronic device further acquires historical behavioral data of users (i.e., the current user's friends and their friends' friends) corresponding to each initial node in the current user's social network. Figure 2 Taking the social relationship network shown as an example, we obtain the historical behavioral data of users A, B, C, D, E, and F corresponding to nodes A, B, C, D, E, and F. This historical behavioral data may include behavioral attribute data such as behavior name, behavior identifier, and behavior time.

[0060] Taking short video scenarios as an example, historical behavior data can include browsing behavior (behavior identifier can be 01, behavior time can be the occurrence and duration of behavior recorded by the terminal), liking behavior (02), following behavior (03), tipping behavior (04), etc.

[0061] Taking video ringback tone scenarios as an example, historical behavior data can include browsing behavior on the homepage of the mall (11), ringback tone listening behavior (12), ringback tone adding behavior (13), called ringtone listening behavior (12-1), called ringback tone adding behavior (13-1), ringback tone payment behavior (14), etc.

[0062] Taking shopping and consumption scenarios as an example, historical behavioral data can include APP entry into product homepage behavior (21), product homepage browsing behavior (22), product search box search behavior (23), product browsing behavior (24), clicking promotional links to enter product browsing behavior (24-1), product addition behavior (25), clicking links to add to cart behavior (25-1), product payment behavior (26), etc.

[0063] Furthermore, the electronic device determines the initial level values ​​of the relationship edges between initial nodes in the social relationship network based on the user's social relationship network and the acquired historical behavior data. Specifically, the initial level values ​​of the relationship edges between initial nodes correspond to user behavior. For example, user A's historical behavior includes browsing, adding items to cart, and making payments. Payment behavior indicates the user's highest interest in the relevant content, while browsing behavior indicates the user's lowest attention to the relevant content. Therefore, payment behavior can be selected as the benchmark behavior for calculating the level value of the relationship edge between two initial nodes. Then, the level value of the relationship edge is determined by the standard that the level value of the relationship edge including payment behavior is higher than the level value of the relationship edge including browsing behavior. This determines the initial level values ​​of the relationship edges between all initial nodes in the social relationship network.

[0064] Once the initial rank values ​​of the relationship edges are determined, the social relationship network is as follows: Figure 3 As shown, Figure 3 This is the second scenario illustration of the information recommendation method provided in the embodiments of this application. Figure 3 The initial level of the edge connecting user A and user C is 1; the initial level of the edge connecting user A and user D is 4; the initial level of the edge connecting user A and user E is 4; the initial level of the edge connecting user B and user D is 3; the initial level of the edge connecting user B and user C is 2; the initial level of the edge connecting user B and user F is 5; and the initial level of the edge connecting user E and user F is 2.

[0065] By determining the initial level values ​​of the relationship edges between each initial node in the social relationship network, subsequent transition node settings can be applied to the current user's social relationship network based on these initial level values. This allows for the prediction of user behavior and the creation of a behavioral relationship network. Then, based on this behavioral relationship network, the social influence of various user behaviors within the network is calculated. Target users are identified based on these social influences, and information recommendations are made based on these target users. This improves the timeliness of recommended content, ensuring a precise match between recommended content and the future behavior of the recommended users, thus enhancing the accuracy of information recommendations. Social influence, or social media influence, describes an individual's ability to influence the thoughts of others in online social communities. The greater an individual's influence, the more attractive they are to companies or individuals seeking to promote an idea or sell a product.

[0066] It should be noted that the initial level value and the subsequent target level value are only used in this embodiment to distinguish the level values ​​of the relation edges under different situations, conditions or time points.

[0067] Step S200: Set transition nodes for the social relationship network according to each of the initial level values ​​to obtain a behavioral relationship network;

[0068] After determining the initial level values ​​of the relationship edges between the initial nodes in the social relationship network, it's important to note that the user behavior relationships reflected in the social relationship network are determined based on each user's historical behavior data, representing only a historical state. If information recommendations are made based on this historical state, the recommended content may not meet the user's current interests.

[0069] Therefore, in this embodiment, the electronic device combines the behavioral relationships in the social relationship network after assigning level values ​​to predict the behavior of the behavioral relationship chains in the social relationship network. Specifically, when there are adjacent relationship edges with different initial level values ​​in the social relationship network, a transition node is set between these adjacent relationship edges with different initial level values ​​to expand the social relationship network, thereby predicting the behavioral relationship network. It should be noted that this embodiment only expands the adjacent relationship edges of the nodes forming the relationship chain according to the initial level value. For adjacent nodes in different relationship chains, no expansion is performed, and the original node connection relationship is maintained.

[0070] By setting transition nodes in the social relationship network to predict user behavior, a behavioral relationship network is obtained. This allows for the subsequent calculation of the social influence of various user behaviors within the network. Based on each social influence, target users are identified, and information is recommended based on these target users. This improves the timeliness of the recommended content, ensures that the recommended content accurately matches the future behavior of the recommended users, and enhances the accuracy of information recommendation.

[0071] Step S300: Calculate the social influence of each initial user in the behavioral relationship network;

[0072] After obtaining the behavioral relationship network, the electronic device traverses all nodes in the network to obtain node attribute data, which includes node identifiers, user identity identifiers, the number of relationship chains and node data for each node, and edge data for each node. Node identifiers are set in initial and transitional nodes to distinguish different node types; node data includes a set of node identifiers and the number of relationship chains; edge data can include the number of edges and their level values. Further, the initial node representing the user node is determined based on the node identifiers in the node attribute data, and the social influence of the initial user corresponding to the initial node is calculated based on the number of relationship chains and edge data of the initial node in the behavioral relationship network.

[0073] By traversing the behavioral relationship network and determining the initial nodes from the node attribute data obtained from the traversal, and then accurately calculating the social influence of the users corresponding to each initial node (also known as the initial users) based on the node attribute data, the target users can be accurately identified based on their social influence. This makes the recommended content based on the target users more accurate, improves the timeliness of the recommended content, and ensures that the recommended content can accurately match the future behavior of the recommended users, thereby improving the accuracy of information recommendation.

[0074] Step S400: Based on the social influence of each user, target users are determined from the initial users, and information is recommended based on the target users.

[0075] After calculating the social influence of each initial user in the behavioral relationship network, the electronic device identifies users whose social influence falls within a preset ranking range as target users. This preset ranking range is a range set based on actual needs or environmental requirements. Furthermore, the electronic device uses the level value of each relationship edge corresponding to the target user's position in the behavioral relationship network as the target level value. This target level value is compared with a preset level threshold. Information is recommended to users pointed to by relationship edges containing target level values ​​greater than the preset level threshold. It should be noted that users pointed to by relationship edges containing target level values ​​greater than the preset level threshold include both the target user and other users. The preset level threshold is a value set based on actual needs or environmental requirements.

[0076] When recommending information based on the user pointed to by the relation edge of the target level value that is greater than the preset level threshold, the information recommendation can be made based on the user's interest content. This includes users in the relation chain of the user pointed to by the relation edge of the target level value that is greater than the preset level threshold. The interest content can be determined based on the user's historical behavior data (such as browsing behavior, purchasing behavior, etc.).

[0077] Furthermore, the step of determining target users from the initial users based on their respective social influence includes:

[0078] Step S401: Sort each of the social influences according to a preset sorting method, wherein each social influence corresponds to an initial user.

[0079] Step S402: Determine the initial users corresponding to the social influences that fall within the preset sorting range from the sorted social influences as target users.

[0080] After calculating the social influence corresponding to each initial user in the behavioral relationship network, the electronic device sorts these social influences according to a preset sorting method, which can be either descending or ascending. Further, the electronic device identifies the social influences within the preset sorting range and determines the initial users corresponding to these influences as target users. This accurate identification of target users based on social influences makes the recommended content more accurate, improves the timeliness of the recommended content, and ensures a precise match between the recommended content and the future behavior of the recommended users, thus enhancing the accuracy of information recommendations.

[0081] For example, sort 10 social influences from largest to smallest, and identify the initial users corresponding to the top 5 social influences as target users.

[0082] For example, sort 20 social influences in ascending order, and identify the initial users corresponding to the bottom 5 social influences as target users.

[0083] The information recommendation method provided in this embodiment improves the timeliness of recommended content by setting transitional nodes in the current user's social relationship network based on the initial level values ​​of the relationship edges between initial nodes in the current user's social relationship network to achieve user behavior prediction and obtain a behavior relationship network. Then, based on the behavior relationship network, the social influence of various user behaviors in the network is calculated. Target users are determined according to each social influence and information is recommended based on the target users. This improves the accuracy of information recommendation by ensuring that the recommended content can accurately match the future behavior of the recommended users.

[0084] In one embodiment, the step of setting transition nodes in the social relationship network based on each of the initial level values ​​to obtain a behavioral relationship network includes:

[0085] Step S201: Determine whether there are adjacent relationship edges with unequal initial level values ​​in the social relationship network;

[0086] Step S202: If it exists, set a transition node on the adjacent relationship edges with unequal initial level values ​​in the social relationship network to obtain the behavioral relationship network.

[0087] After determining the initial level values ​​of the relationship edges between the initial nodes in the social relationship network, it's important to note that the user behavior relationships reflected in the social relationship network are determined based on each user's historical behavior data, representing only a historical state. If information recommendations are made based on this historical state, the recommended content may not meet the user's current interests.

[0088] Therefore, in this embodiment, the electronic device can combine the behavioral relationships in the social relationship network after assigning level values ​​to predict the behavior of the behavioral relationship chain in the social relationship network. Specifically, it can determine whether there are adjacent relationship edges in the social relationship network with unequal initial level values ​​by calculating whether the initial level values ​​between two adjacent relationship edges in the social relationship network are equal.

[0089] If it is determined that there are adjacent edges with unequal initial rank values ​​in the social relationship network, transition nodes are set between these edges to expand the network and predict the behavioral relationship network. Specifically, the rank difference between the initial rank values ​​of two adjacent edges with unequal initial rank values ​​is determined, and transition nodes with the same rank difference value are added to these edges. When transition nodes are set on all adjacent edges with unequal initial rank values ​​in the social relationship network, the user behavior prediction is complete, and the behavioral relationship network is obtained.

[0090] It should be noted that this embodiment only expands the adjacent relationship edges of the nodes that form a relationship chain according to the initial level value. For adjacent nodes that are in different relationship chains, no expansion is performed, and the original node connection relationship is maintained.

[0091] By setting transition nodes in the social relationship network to predict user behavior, a behavioral relationship network is obtained. This network allows for the calculation of the social influence of various user behaviors within the network. Target users are then identified based on their social influence, and information recommendations are made accordingly. This improves the timeliness of recommended content, ensures a precise match between recommended content and the future behavior of the recommended users, and enhances the accuracy of information recommendations.

[0092] Furthermore, the step of setting transition nodes on adjacent relationship edges with unequal initial rank values ​​in the social relationship network includes:

[0093] Step S2021: For each adjacent relationship edge in the social relationship network with unequal initial level values, execute steps S2022-S2023 respectively;

[0094] Step S2022: Determine the level difference between the initial level values ​​corresponding to the two current relation edges;

[0095] Step S2023: Set a transition node with the same value as the grade difference between the two current relation edges.

[0096] Specifically, for adjacent edges with unequal initial rank values ​​in a social relationship network, multiple groups of adjacent edges with unequal initial rank values ​​can be sorted, and transition nodes can be set according to the sorting order. Alternatively, transition nodes can be set for multiple groups of adjacent edges with unequal initial rank values ​​simultaneously. When setting transition nodes for adjacent edges with unequal initial rank values, the rank difference between the initial rank values ​​of the two edges can be calculated first. A transition node with the same value as the rank difference is then set between the two edges; that is, the rank difference is the number of transition nodes set. After setting transition nodes for all groups of adjacent edges with unequal initial rank values, the user behavior prediction is completed, resulting in the behavioral relationship network.

[0097] In one instance of this application, reference is made to Figure 4 , Figure 4 This is the third scenario illustration of the information recommendation method provided in the embodiments of this application. Figure 4 In the example of the relationship chain User 1 (U1) -> User 2 (U2) -> User 3 (U3) -> User 4 (U4) -> User 5 (U5) -> User 1 (U1), for User 2, the initial level value corresponding to the relationship edge (User 1 -> User 2) is 5 (assuming the corresponding behavior is payment), and the initial level value corresponding to the relationship edge (User 2 -> User 3) is 3 (assuming the corresponding behavior is liking). Since the two relationship edges are adjacent and their initial level values ​​are not equal, it can be predicted that there is a browsing behavior with an initial level value of 4 among User 1, User 2, and User 3. To expand the social relationship network, a number of transition nodes corresponding to the level difference of 2 are added to the relationship chain User 1 -> User 2 -> User 3. The expanded relationship chain (i.e., the social relationship network) (e.g., User 1 -> User 2 -> User 3) is further expanded. Figure 5 As shown, Figure 5 The fourth scenario diagram of the information recommendation method provided in the embodiments of this application is as follows: User 1 (U1) -> Transition Node 1 (T1) -> User 2 (U2) -> Transition Node 2 (T1) -> User 3 (U3), wherein the initial level value of the relationship edge between User 1 (U1) -> Transition Node 1 (T1) is 5, the initial level value of the relationship edge between Transition Node 1 (T1) -> User 2 (U2) is 4, the initial level value of the relationship edge between User 2 (U2) -> Transition Node 2 (T1) is 4, and the initial level value of the relationship edge between Transition Node 2 (T1) -> User 3 (U3) is 3.

[0098] Similarly, for the relationship chain: User 3 -> User 4 -> User 5, the expanded relationship chain is (e.g.) Figure 6 As shown, Figure 6The fifth scenario diagram of the information recommendation method provided in this application embodiment is as follows: User 3 (U3) -> Transition Node 3 (T3) -> User 4 (U4) -> Transition Node 4 (T4) -> User 5 (U5), wherein the initial level value of the relationship edge between User 3 (U3) and Transition Node 3 (T3) is 2, the initial level value of the relationship edge between Transition Node 3 (T3) and User 4 (U4) is 3, the initial level value of the relationship edge between User 4 (U4) and Transition Node 4 (T4) is 3, the initial level value of the relationship edge between Transition Node 4 (T4) and User 5 (U5) is 4, and other information is... Figure 5 same.

[0099] For the relationship chain: User 4 -> User 5 -> User 3, since the initial level difference between the relationship edges (User 4 -> User 5) and (User 5 -> User 3) is 3, a total of 3 transition nodes need to be added to the relationship edges (User 4 -> User 5) and (User 5 -> User 3). Also, due to the aforementioned... Figure 6 In the relationship chain expansion, a transition node 4 was added between user 4 and user 5. Therefore, the relationship chain can be adjusted based on this as follows (e.g., Figure 7 As shown, Figure 7 The following is a schematic diagram of a scenario for the information recommendation method provided in this application embodiment: User 4 (U4) -> Transition Node 4 (T4) -> Transition Node 5 (T5) -> User 5 (U5) -> Transition Node 6 (T6) -> User 3 (U3), wherein the initial level value of the relationship edge (User 4 -> Transition Node 4) is 4, the initial level value of the relationship edge (Transition Node 4 -> Transition Node 5) is 3, the initial level value of the relationship edge (Transition Node 5 -> User 5) is 2, the initial level value of the relationship edge (User 5 -> Transition Node 6) is 2, and the initial level value of the relationship edge (Transition Node 6 -> User 3) is 1. Other information is... Figure 6 same.

[0100] As an alternative embodiment, the extended relationship chain can also be: User 4 -> Transition Node 4 -> User 5 -> Transition Node 5 -> Transition Node 6 -> User 3, with the initial level value of relationship edge (User 4 -> Transition Node 4) being 4, the initial level value of relationship edge (Transition Node 4 -> User 5) being 3, the initial level value of relationship edge (User 5 -> Transition Node 5) being 3, the initial level value of relationship edge (Transition Node 5 -> Transition Node 6) being 2, and the initial level value of relationship edge (Transition Node 6 -> User 3) being 1.

[0101] This embodiment achieves user behavior prediction and obtains a behavior relationship network by setting transition nodes in the social relationship network. This allows for the subsequent calculation of the social influence of various user behaviors in the network based on the behavior relationship network. Target users are then identified based on their social influence, and information is recommended based on these target users. This improves the timeliness of the recommended content and ensures that the recommended content accurately matches the future behavior of the recommended users, thereby increasing the accuracy of information recommendation.

[0102] In one embodiment, the step of calculating the social influence of each initial user in the behavioral relationship network includes:

[0103] Step S301: Traverse all nodes in the behavioral relationship network to obtain node attribute data;

[0104] Step S302: Determine the initial node from all traversed nodes based on the node attribute data;

[0105] Step S303: Calculate the social influence of the user corresponding to each initial node in the behavioral relationship network based on the node attribute data.

[0106] After obtaining the behavioral relationship network, the electronic device traverses all nodes in the network. Specifically, this can involve traversing the node types of each node and setting corresponding node identifiers for different node types, identifying the user identity identifier of the user node as the initial node, the node data of the user node's relationship chain in the behavioral relationship network, and the edge data of each node. After the traversal is completed, node attribute data is obtained, including the node identifier of each node, the user identity identifier of the corresponding user of each node, the number of relationship chains and node data of each node in the behavioral relationship network, and the edge data of each node. The node data includes the set of node identifiers and the number of relationship chains; the edge data can include the number of edges and the level value of the edges.

[0107] It should be noted that the above relationship chain can be one or more. When there are multiple relationship chains, the node identifier set in the node data includes multiple sets, and each set of node identifiers corresponds to one relationship chain. Similarly, the edge data also includes multiple sets, and each set of edge data corresponds to one relationship chain and is associated with a set of node identifiers.

[0108] Furthermore, the electronic device determines the initial node representing the user node based on the node identifier in the node attribute data. Specifically, it identifies the node identifier corresponding to each node in the node attribute information. If the node type corresponding to the node identifier is an initial node, it calculates the social influence of the initial user corresponding to the initial node based on the number of relationship chains and edge data of the initial node in the behavioral relationship network in the node attribute data.

[0109] By traversing the behavioral relationship network and determining the initial nodes from the node attribute data obtained from the traversal, and then accurately calculating the social influence of the users corresponding to each initial node based on the node attribute data, the target users can be accurately identified based on their social influence. This makes the recommended content based on the target users more accurate, improves the timeliness of the recommended content, and ensures that the recommended content can accurately match the future behavior of the recommended users, thereby improving the accuracy of information recommendation.

[0110] Furthermore, the step of calculating the social influence of the user corresponding to each initial node in the behavioral relationship network based on the node attribute data includes:

[0111] Step S3031: Determine the number of relation chains and edge data included in each initial node of the behavioral relation network from the node attribute data;

[0112] Step S3032: Calculate the social influence of the user corresponding to each initial node in the behavioral relationship network based on the number of relationship chains and the edge data.

[0113] After identifying all initial nodes (i.e., user nodes) in the behavioral relationship network, the electronic device adds each user node and its information in the node attribute data (including user identification, node data and edge data of the corresponding user's relationship chain in the behavioral relationship network) to a pre-set set in traversal order. This set is initially empty. The resulting set can be: S u ={S u1 ,S u2 ,S u3 ,……,S un}, where S u1 ,S u2 ,S u3 ,……,S un They are user 1, user 2, user 3... user n, specifically, S ui =(u i L ui L V ), where u i L serves as a user identifier. ui For user u i The node data of the relationship chain in the behavioral relationship network includes a set of node identifiers S. m With the number of relational chains j; L V The edge data can include the number of edges, L. V_t The level value L of the edge V_q ;

[0114] Furthermore, based on the number of relationship chains and edge data of each initial node in the behavioral relationship network, the social influence of the initial user corresponding to each initial node is calculated. Specifically, the number of relationship chains and edge data of each initial node in the behavioral relationship network are input into a preset social influence calculation formula to calculate the social influence of the user corresponding to each initial node. After completing the calculation of the social influence of all users corresponding to the initial nodes, a social influence set, such as K, can be generated. u ={K u1 ,K u2 ,K u3 ,……,K un}, K u1 ,K u2 ,K u3 ,……,K un This refers to the social influence of user 1, user 2, user 3... user n. The preset formula for calculating social influence is as follows:

[0115]

[0116] Among them, K ui For user u i Social influence, j represents user u i The number of relationship chains included, R ui Indicates user u i The number of user nodes included in the relationship chain, where w represents the discount factor, to avoid a single chain corresponding to a user having a high social influence. L V_t The number of edges and the edge rank value L. V_q .

[0117] This embodiment traverses the behavioral relationship network and determines initial nodes from the node attribute data obtained during the traversal. Then, it accurately calculates the social influence of the user corresponding to each initial node based on the node attribute data. This enables the determination of the social influence of various user behaviors based on the distance between the reachable endpoint of the behavior in the behavioral relationship network and the user's behavior. This allows for the accurate identification of target users based on each social influence, making the recommended content more accurate and improving the timeliness of the recommended content. It also ensures that the recommended content can accurately match the future behavior of the recommended user, thereby improving the accuracy of information recommendation.

[0118] In one embodiment, the step of recommending information based on the target user includes:

[0119] Step S403: For each of the target users, execute steps S404-S407 respectively;

[0120] Step S404: Determine the target level value of each relation edge corresponding to the relation chain of the current target user in the behavioral relation network;

[0121] Step S405: Compare each of the target level values ​​with a preset level threshold.

[0122] Step S406: Determine the target level value that is greater than the preset level threshold among all the target level values;

[0123] Step S407: Recommend information to users pointed to by the relation edges of target level values ​​that are greater than the preset level threshold.

[0124] After identifying the target users, for each target user, you can first sort the users and then recommend information based on that sorted user order, or you can simultaneously recommend information to multiple target users separately. Specifically, when recommending information based on target users...

[0125] First, determine the target level value for each relation edge corresponding to the current target user's relation chain in the behavioral relation network. This means the current target user may exist in multiple relation chains, and may be on one or more relation edges in each chain. Then, determine the level value of the target user on each relation edge in each relation chain as the target level value. Next, obtain a preset level threshold and compare each target level value corresponding to the target user with the preset level threshold to determine the relationship between each target level value and the preset level threshold. This further determines whether any target level value is greater than the preset level threshold.

[0126] Furthermore, if among the target level values ​​corresponding to the target user, there is a target level value greater than a preset level threshold, then information recommendation is made based on the user pointed to by the relation edge containing the target level value greater than the preset level threshold, wherein the user pointed to by the relation edge containing the target level value greater than the preset level threshold can include the target user and other users.

[0127] When recommending information based on the user pointed to by the relation edge of the target level value that is greater than the preset level threshold, the information recommendation can be made based on the user's interest content. This includes users in the relation chain of the user pointed to by the relation edge of the target level value that is greater than the preset level threshold. The interest content can be determined based on the user's historical behavior data (such as browsing behavior, purchasing behavior, etc.).

[0128] This embodiment can determine the final user for information recommendation based on the target level value of each relation edge corresponding to the target user's relation chain in the behavioral relation network. This can improve the timeliness of the recommended content, make the recommended content accurately match the future behavior of the recommended user, and improve the accuracy of information recommendation.

[0129] Furthermore, this application also provides an information recommendation device.

[0130] Reference Figure 8 , Figure 8 This is a schematic diagram of the functional modules of an embodiment of the information recommendation device of this application.

[0131] The information recommendation device includes:

[0132] The determination module 100 is used to determine the current user's social relationship network and determine the initial level value of the relationship edge between each initial node in the social relationship network;

[0133] Prediction module 200 is used to set transition nodes in the social relationship network according to the initial level values ​​to obtain a behavioral relationship network;

[0134] Calculation module 300 is used to calculate the social influence of each initial user in the behavioral relationship network;

[0135] The recommendation module 400 is used to determine target users from the initial users based on their social influence and to recommend information based on the target users.

[0136] Figure 9 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 9 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call a computer program in the memory 830 to execute the steps of the information recommendation method, such as including:

[0137] Determine the current user's social relationship network, and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0138] Based on the initial level values, transition nodes are set in the social relationship network to obtain a behavioral relationship network;

[0139] Calculate the social influence of each initial user in the behavioral relationship network;

[0140] Based on the aforementioned social influence, target users are identified from the initial users, and information is recommended based on the target users.

[0141] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] On the other hand, embodiments of this application also provide a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the steps of the information recommendation method provided in the above embodiments, such as including:

[0143] Determine the current user's social relationship network, and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0144] Based on the initial level values, transition nodes are set in the social relationship network to obtain a behavioral relationship network;

[0145] Calculate the social influence of each initial user in the behavioral relationship network;

[0146] Based on the aforementioned social influence, target users are identified from the initial users, and information is recommended based on the target users.

[0147] On the other hand, embodiments of this application also provide a processor-readable storage medium storing a computer program for causing a processor to perform the steps of the methods provided in the above embodiments, such as including:

[0148] Determine the current user's social relationship network, and determine the initial level value of the relationship edges between each initial node in the social relationship network;

[0149] Based on the initial level values, transition nodes are set in the social relationship network to obtain a behavioral relationship network;

[0150] Calculate the social influence of each initial user in the behavioral relationship network;

[0151] Based on the aforementioned social influence, target users are identified from the initial users, and information is recommended based on the target users.

[0152] The processor-readable storage medium can be any available medium or data storage device that the processor can access, including but not limited to magnetic memory (e.g., floppy disk, hard disk, magnetic tape, magneto-optical disk (MO)), optical memory (e.g., CD, DVD, BD, HVD), and semiconductor memory (e.g., ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)).

[0153] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0154] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An information recommendation method, characterized in that, include: Determine the current user's social relationship network, and determine the initial level value of the relationship edges between each initial node in the social relationship network; Based on the initial level values, transition nodes are set in the social relationship network to obtain a behavioral relationship network; Calculate the social influence of each initial user in the behavioral relationship network; Based on the aforementioned social influence, target users are identified from the initial users, and information is recommended based on the target users; The step of setting transition nodes in the social relationship network according to each of the initial level values ​​to obtain the behavioral relationship network includes: Determine whether there are adjacent relationship edges with unequal initial level values ​​in the social relationship network; If they exist, transition nodes are set on adjacent relationship edges with unequal initial level values ​​in the social relationship network to obtain a behavioral relationship network; The step of setting transition nodes on adjacent relationship edges with unequal initial ranking values ​​in the social relationship network includes: For each adjacent edge in the social relationship network with unequal initial ranking values, the following steps are performed respectively: Determine the grade difference between the initial grade values ​​corresponding to the two current relation edges; Set a transition node with the same value as the grade difference between the two current relation edges.

2. The information recommendation method according to claim 1, characterized in that, The steps for calculating the social influence of each initial user in the behavioral relationship network include: Traverse all nodes in the behavioral relationship network to obtain node attribute data; The initial node is determined from all traversed nodes based on the node attribute data; The social influence of the user corresponding to each initial node in the behavioral relationship network is calculated based on the node attribute data.

3. The information recommendation method according to claim 2, characterized in that, The step of calculating the social influence of the user corresponding to each initial node in the behavioral relationship network based on the node attribute data includes: The number of relation chains and edge data included in each initial node of the behavioral relation network are determined from the node attribute data; The social influence of the user corresponding to each initial node in the behavioral relationship network is calculated based on the number of relationship chains and the edge data.

4. The information recommendation method according to claim 1, characterized in that, The step of determining target users from each of the initial users based on each of the aforementioned social influences includes: The social influences are sorted according to a preset sorting method, wherein each social influence corresponds to an initial user. The initial users corresponding to the social influences that fall within the preset sorting range after sorting are identified as target users.

5. The information recommendation method according to claim 1, characterized in that, The step of recommending information based on the target user includes: For each of the aforementioned target users, perform the following steps respectively: Determine the target level value of each relation edge corresponding to the relation chain of the current target user in the behavioral relation network; Compare each of the target level values ​​with a preset level threshold; Determine the target level value that is greater than the preset level threshold among all the target level values; Information recommendations are made based on the users pointed to by the relation edges containing target level values ​​greater than the preset level threshold.

6. An information recommendation device, characterized in that, include: The determination module is used to determine the current user's social relationship network and determine the initial level value of the relationship edges between each initial node in the social relationship network; The prediction module is used to set transition nodes in the social relationship network based on the initial level values ​​to obtain a behavioral relationship network; The calculation module is used to calculate the social influence of each initial user in the behavioral relationship network; The recommendation module is used to determine target users from the initial users based on their social influence, and to recommend information based on the target users. The prediction module is also used to determine whether there are adjacent relationship edges with unequal initial level values ​​in the social relationship network; if so, a transition node is set on the adjacent relationship edges with unequal initial level values ​​in the social relationship network to obtain a behavioral relationship network. The prediction module is further configured to perform the following steps for each adjacent relation edge in the social relation network that has unequal initial ranking values: Determine the grade difference between the initial grade values ​​corresponding to the two current relation edges; Set a transition node with the same value as the grade difference between the two current relation edges.

7. An electronic device comprising a processor and a memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the information recommendation method according to any one of claims 1 to 5.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the information recommendation method according to any one of claims 1 to 5.

Images