Node category determination method, related device, equipment and storage medium

Abstract

The embodiment of the present invention discloses a node category determination method, related devices, equipment and storage medium, the method comprising: a terminal device determines at least one neighbor node of a target node; calculates at least one first probability that any neighbor node of the target node propagates a target label to the target node based on various types of edges connected to the target node, so as to obtain a first probability set that each neighbor node propagates the target label to the target node based on various types of edges; determines a first probability vector of the target node receiving the target label based on the same type of edge based on the first probability set, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges; obtains a second probability of the target node receiving the target label based on the first probability vector set; and determines the node category of the target node based on the second probability. By adopting the embodiment of the present invention, the category of any node in any social network can be predicted, and the applicability is high.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a node category determination method, related devices, equipment and storage medium. Background Art

[0002] At present, social communication (social influence communication) has a wide range of application scenarios, and social influence is mainly achieved through label propagation in social networks. However, most traditional label propagation methods are based on homogeneous networks. Since the nodes in a homogeneous network belong to the same type of object (entity), the nodes are comparable, so the category of each node can be predicted based on the propagation of labels in the homogeneous network.

[0003] However, for heterogeneous networks, there are many types of different nodes and edges between nodes, and each type of node has different characteristics. The existing label propagation method cannot effectively determine whether each node in the heterogeneous network is affected by other nodes, and has great limitations. Summary of the invention

[0004] The embodiments of the present invention provide a node category determination method, related apparatus, device and storage medium, which can predict the category of any node in any social network and have high applicability.

[0005] In a first aspect, an embodiment of the present invention provides a method for determining a node category, the method comprising:

[0006] The terminal device determines at least one neighbor node of the target node, wherein the neighbor node is a node connected to the target node based on at least one type of edge;

[0007] The terminal device calculates at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each neighbor node propagates the target label to the target node based on various types of edges;

[0008] The terminal device determines, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges;

[0009] The terminal device obtains a second probability that the target node receives the target tag based on the first probability vector set;

[0010] The terminal device determines the node category of the target node based on the second probability, and outputs the node category of the target node to the user interaction interface of the terminal device for display to the user.

[0011] In combination with the first aspect, in a possible implementation, the terminal device calculates at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, including:

[0012] The terminal device obtains at least one edge attribute of any type of edge connecting any neighbor node and the target node;

[0013] The terminal device determines an attribute weight of each edge attribute in the at least one edge attribute;

[0014] The terminal device determines the edge weight of any of the above-mentioned types of edges based on the attribute weights of the above-mentioned edge attributes, and determines the edge weight as the first probability that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on any of the above-mentioned types of edges, so as to obtain the first probabilities that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on each type of edge between the above-mentioned neighboring nodes and the above-mentioned target node.

[0015] In combination with the first aspect, in a possible implementation, the terminal device determines, based on the first probability set, a first probability vector that the target node receives the target label based on the same type of edge, including:

[0016] The terminal device determines, from the first probability set, at least one first probability corresponding to edges whose edge types belong to the same type;

[0017] The terminal device constructs a first probability vector of the target node receiving the target label based on the same type of edge based on the at least one first probability.

[0018] In combination with the first aspect, in a possible implementation manner, the terminal device determines multiple neighbor nodes of the target node, including:

[0019] The terminal device obtains the association relationship between the nodes in at least one social network, determines at least one node that has an association relationship with the target node from the nodes based on the association relationship between the nodes, and determines the at least one node as a neighbor node of the target node.

[0020] In combination with the first aspect, in a possible implementation manner, the terminal device obtains the second probability that the target node receives the target tag based on the first probability vector set, including:

[0021] The terminal device determines the type weights corresponding to the edge types, wherein one type corresponds to one type weight;

[0022] The terminal device determines a second probability that the target node receives the target tag based on the weights of each type and the first probability vector set.

[0023] In combination with the first aspect, in a possible implementation manner, the terminal device determines the node category of the target node based on the second probability, including:

[0024] The terminal device compares the second probability with a preset threshold;

[0025] When the second probability is greater than or equal to the preset threshold, the terminal device determines that the node category of the target node is the node category marked by the target tag.

[0026] In combination with the first aspect, in a possible implementation, the neighbor nodes of the target node include a first-degree neighbor node of the target node and a second-degree neighbor node of the target node, the first-degree neighbor node is a node connected to the target node based on at least one type of edge, and the second-degree neighbor node is any node connected to the target node based on any first-degree neighbor node;

[0027] The above method further includes:

[0028] The terminal device determines a second probability vector of the target label propagated from each second-degree neighbor node based on the same type of edge reception of the target node, so as to obtain a set of second probability vectors of the target label propagated from each second-degree neighbor node based on multiple types of edge reception of the target node;

[0029] The terminal device obtains a second probability that the target node receives the target tag based on the first probability vector set and the second probability vector set.

[0030] In a second aspect, an embodiment of the present invention provides a node category determination device, the device comprising:

[0031] A node determination module, used to determine at least one neighbor node of a target node, wherein the neighbor node is a node connected to the target node based on at least one type of edge;

[0032] A first probability determination module is used to calculate at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each neighbor node propagates the target label to the target node based on various types of edges;

[0033] A probability vector determination module is used to determine, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges;

[0034] A second probability determination module, configured to obtain a second probability of the target node receiving the target label based on the first probability vector set;

[0035] A node category determination module is used to determine the node category of the target node based on the second probability, and output the node category of the target node to the user interaction interface of the terminal device for display to the user.

[0036] In conjunction with the second aspect, in a possible implementation manner, the first probability determination module includes:

[0037] An edge attribute acquisition unit, used to acquire at least one edge attribute of any type of edge connecting any neighbor node and the target node;

[0038] an attribute weight determination unit, used to determine the attribute weight of each edge attribute in the at least one edge attribute;

[0039] An edge weight determination unit is used to determine the edge weight of any of the above-mentioned types of edges based on the attribute weights of the above-mentioned edge attributes, and determine the above-mentioned edge weight as the first probability that any of the above-mentioned neighboring nodes propagates the target label to the above-mentioned target node based on any of the above-mentioned types of edges, so as to obtain the first probabilities that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on each type of edge between the above-mentioned neighboring node and the above-mentioned target node.

[0040] In conjunction with the second aspect, in a possible implementation manner, the probability vector determination module is used to:

[0041] Determine at least one first probability corresponding to edges of the same type from the first probability set;

[0042] A first probability vector of the target node receiving the target label based on the same type of edge is constructed based on the at least one first probability.

[0043] In conjunction with the second aspect, in a possible implementation manner, the node determination module is configured to:

[0044] Obtain association relationships between nodes in at least one social network, determine at least one node that has an association relationship with a target node from among the nodes based on the association relationships between the nodes, and determine the at least one node as a neighbor node of the target node.

[0045] In conjunction with the second aspect, in a possible implementation manner, the second probability determination module is configured to:

[0046] Determine the type weights corresponding to the edge types, where one type corresponds to one type weight;

[0047] A second probability of the target node receiving the target label is determined based on the above-mentioned various type weights and the above-mentioned first probability vector set.

[0048] In conjunction with the second aspect, in a possible implementation manner, the node category determination module includes:

[0049] A comparing unit, used for comparing the second probability with a preset threshold;

[0050] A category determination unit is used to determine that the node category of the target node is the node category marked by the target tag when the second probability is greater than or equal to the preset threshold.

[0051] In combination with the second aspect, in a possible implementation, the neighbor nodes of the target node include a first-degree neighbor node of the target node and a second-degree neighbor node of the target node, the first-degree neighbor node is a node connected to the target node based on at least one type of edge, and the second-degree neighbor node is any node connected to the target node based on any first-degree neighbor node;

[0052] The probability vector determination module is further used to determine a second probability vector of the target label propagated from each second-degree neighbor node by the target node based on the same type of edge reception, so as to obtain a set of second probability vectors of the target label propagated from each second-degree neighbor node by the target node based on multiple types of edge reception;

[0053] The second probability determination module is further used to obtain a second probability that the target node receives the target label based on the first probability vector set and the second probability vector set.

[0054] In a third aspect, an embodiment of the present invention provides a terminal device, the terminal device comprising a processor and a memory, the processor and the memory being connected to each other. The memory is used to store a computer program that supports the terminal device to execute the method provided by the first aspect and / or any possible implementation of the first aspect,

[0055] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method provided by the above-mentioned first aspect and / or any possible implementation method of the first aspect.

[0056] In the embodiment of the present invention, by converting different types of node relationships in different social networks into the same type of node relationships, the restrictions on node relationships in different social networks can be eliminated, so that the node category of any node in different social networks can be predicted. By calculating the attribute weights of each edge attribute of each type of edge, the corresponding proportion of each edge attribute in the edge of its corresponding type can be determined. Therefore, the edge weight of each type of edge can be accurately and reasonably determined based on the attribute weights of each edge attribute, and the applicability is high. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0058] Figure 1 is a schematic diagram of the architecture of a social network provided by an embodiment of the present invention;

[0059] Figure 2 is a flowchart of a method for determining a node category provided by an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram of node association relationship transformation provided by an embodiment of the present invention;

[0061] Figure 4 is a flow chart of a method for calculating a first probability provided by an embodiment of the present invention;

[0062] Figure 5 is a schematic diagram of a scenario of neighbor nodes provided by an embodiment of the present invention;

[0063] Figure 6 is a schematic diagram of the structure of a device for determining a node category provided by an embodiment of the present invention;

[0064] Figure 7 It is a schematic diagram of the structure of a terminal device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] See also Figure 1 , Figure 1 is a schematic diagram of the architecture of a social network provided by an embodiment of the present invention. The network architecture includes multiple users and devices and networks for connecting multiple users in series. Figure 1 As shown in , multiple users can form a social network 1 based on terminals 100 and 200, some users can form a social network 3 based on Wifi300 and Wifi400, and another part of users can form a social network 2 based on other terminals, such as mobile phones, tablet computers, laptops, PDAs, mobile internet devices (MID), wearable devices (such as smart watches, smart bracelets, etc.). Among them, users in social networks 1, 2 and 3 can spread social information based on a certain social method, for example, they can spread social information based on friend circles and friend circle advertisements. After seeing user B's social behavior (posting friend circles, sharing public account texts, liking advertisements), user A may interact with user B in social information, such as commenting on B's friend circle, forwarding articles shared by B, liking advertisements liked by B, etc. In short, each user in a social network can be regarded as each node in the social network. Each node will spread labels to other nodes associated with it to assign the label to other nodes, thereby changing the node category of the node receiving the label spread in the social network.

[0067] See also Figure 2 , Figure 2 1 is a flow chart of a node category determination method provided by an embodiment of the present invention. The node category determination method provided by an embodiment of the present invention may include the following steps 101-104:

[0068] 101. A terminal device determines at least one neighbor node of a target node.

[0069] In some feasible implementations, since the actual social network is composed of connections and / or behavioral relationships between multiple devices and users, and different users can be associated with other users through multiple devices and networks. Therefore, for any node in the social network, only a part of the other nodes in the social network have multiple associations with the node. It can be seen that when determining the node category of a certain node (for the convenience of description, the certain node can be represented by a target node, which will not be repeated below), the terminal device can obtain the association relationship between each node in at least one social network where the above target node is located, and determine at least one node with which the above target node has an association based on the association relationship between each node, and determine the at least one node as a neighbor node of the above target node, so that the category of the above target node can be determined based on the probability of the neighbor node of the above target node propagating the label to the target node.

[0070] In some feasible implementations, since the target node may exist in different social networks and there are differences between the types of node associations in different social networks, the node relationships between two different social networks cannot be directly compared under normal circumstances. Moreover, the types of associations between nodes in different social networks are not necessarily the same type of associations. Therefore, before determining the neighbor nodes of the target node, the associations between nodes in different social networks may be converted to the same type of associations, so as to avoid the limitation that the neighbor nodes cannot be determined due to different associations between nodes in different social networks. For details, please refer to Figure 3 , Figure 3 Schematic diagram of node association transformation provided by an embodiment of the present invention. Figure 2 As shown, nodes B, C, and D are all associated with node A through a certain association relationship, that is, nodes B, C, and D are associated with each other through node A. If nodes B, C, and D belong to nodes in different social networks, although nodes B, C, and D have a certain association relationship with each other through node A, the association relationship between nodes B, C, and D cannot be directly determined. At this time, nodes B, C, and D can be formed into a fully connected graph of association relationships to represent the association relationship established between nodes B, C, and D based on node A. For example, when node A is a certain device, and nodes B, C, and D are respectively associated with three users of the device, the user association relationship established between the three users based on the above device can be represented based on the fully connected graph of the user association relationship between the above three users.

[0071] 102. The terminal device calculates at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node.

[0072] In some feasible implementations, after the terminal device determines at least one neighbor node of the target node, it can calculate at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node. Thus, a first probability set of the table node propagating the target label to the target node based on various types of edges can be obtained. Among them, one type of edge between any neighbor node and the target node corresponds to one first probability. Specifically, the calculation method of the above first probability can be referred to Figure 4 , Figure 4 201 to 204:

[0073] 201. A terminal device obtains at least one edge attribute of any type of edge connecting any neighbor node and a target node.

[0074] In some feasible implementations, the target node and any neighbor node may be connected to each other based on at least one type of edge, and in different social scenarios of different types of social networks, the type of edge between the target node and any neighbor node is also different, which can be determined based on the actual application scenario, and is not limited here. For example, when a certain type of edge connected to the target node and any node can represent the transfer relationship between the two nodes, can represent the wifi connection relationship between the nodes, can also represent the chat relationship between the two nodes, etc., which is not limited here. For any type of edge between the target node and any neighbor node, the terminal device can obtain different edge attributes in any type of edge, so that it can be based on different edge attributes in any type of edge. For example, for an edge representing a transfer relationship, the edge attributes of the edge may include edge attributes such as transfer amount and transfer frequency. For an edge representing a wifi connection relationship between nodes, the edge attributes of the edge may include edge attributes such as connection duration and connection frequency. The type and quantity of edge attributes of each type of edge can be determined based on the actual type of edge, and are not limited here.

[0075] 202. The terminal device determines an attribute weight of each edge attribute in at least one edge attribute.

[0076] In some feasible implementations, after the terminal device determines at least one edge attribute of any type of edge connecting any neighbor node to the target node, it can determine the edge values ​​corresponding to different edge attributes and then calculate the attribute weight of each edge attribute. Among them, since the edge values ​​of different edge attributes have different dimensions, after determining the edge values ​​corresponding to different edge attributes, the terminal device can perform data normalization processing on the edge values ​​of different dimensions to eliminate the impact caused by the different dimensions of different edge values. Among them, the specific data normalization processing method can be based on Implementation, where k is the normalization coefficient, and x is the edge value corresponding to each edge attribute. Based on the normalized edge values, the attribute weight of each edge attribute can be determined based on the normalized value of each edge value, where the attribute weight of each edge attribute is

[0077] 203. The terminal device determines an edge weight of any type of edge based on the attribute weights of each edge attribute.

[0078] In some feasible implementations, after the terminal device calculates the weights of each edge attribute of any type of edge between the target node and any neighboring node, the terminal device may determine the edge weight of any type of edge based on the attribute weights of each edge attribute. Specifically, when any type of edge has only one edge attribute, the attribute weight of the edge attribute may be determined as the edge weight of any type of edge. When any type of edge has two edge attributes, the edge weight of any type of edge may be determined based on W=(w1+w2)-(w1*w2), wherein w1 and w2 are the edge weights of any type of edge, respectively, and W represents the edge weight of any type of edge. It is not difficult to derive that when any type of edge has n edge attributes, it may be determined based on To derive the edge weight of any of the above types of edges.

[0079] 204. The terminal device determines the edge weight as the first probability that any neighbor node propagates the target label to the target node based on any type of edge, so as to obtain each first probability that any neighbor node propagates the target label to the target node based on various types of edges between the neighbor node and the target node.

[0080] In some feasible implementations, after obtaining the edge weight W of any of the above-mentioned types of edges, the terminal device may determine the edge weight W of any of the above-mentioned types of edges as the first probability that any of the above-mentioned neighboring nodes propagates the target label to the above-mentioned target node based on any type of edge. Based on the above method, the terminal device can obtain multiple first probabilities that any of the above-mentioned neighboring nodes propagates the target label to the target node based on various types of edges between the target node, and then obtain the first probability set that each neighboring node propagates the target label to the target node based on various types of edges. In short, the first probability set contains the first probability that each neighboring node propagates the target label to the target node based on each type of edge.

[0081] In the embodiment of the present invention, by calculating the attribute weights of each edge attribute of each type of edge, the corresponding weight of each edge attribute in the edge of its corresponding type can be determined. Therefore, the edge weight of each type of edge can be accurately and reasonably determined based on the attribute weights of each edge attribute, and the applicability is high.

[0082] 103. The terminal device determines, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges.

[0083] In some feasible implementations, the terminal device may determine at least one first probability corresponding to edges of the same type from the above-mentioned first probability set, and the above-mentioned at least one first probability may include multiple first probabilities that different neighboring nodes forward the target label to the target node based on different edges of the same edge type. That is, the terminal device may classify each first probability in the above-mentioned first probability set based on the type of edge to obtain multiple sub-first probability sets, wherein at least one first probability included in each sub-first probability set is the first probability corresponding to the same type of edge. At the same time, the terminal device may construct the first probability vector p of the above-mentioned target node receiving the target label based on the same type of edge based on each sub-first probability set. s , where s represents the type of edge, such as the first probability vector of receiving the target label based on the first type of edge is p s1 , the first probability vector of receiving the target label based on the second type of edge is p s2 , thus obtaining the first probability vector set P of the target node receiving the target label based on multiple types of edges.

[0084] 104. The terminal device obtains a second probability that the target node receives the target label based on the first probability vector set.

[0085] In some feasible implementations, after the terminal device determines that the target node receives the target label based on multiple types of edges, the first probability vector combination P, due to the different importance of different nodes in the social network, nodes of different importance may affect the probability of label propagation in the social network. Therefore, a new first probability vector set S can be obtained based on the importance of the target label, where S = αP + (1-α) V, P is the first probability vector set obtained above, V represents the importance of the target node, and V = (1-α), 0≤α≤1. Based on the terminal device, the type weights corresponding to the various types of the above-mentioned edge types can be determined, and each type corresponds to a type weight. Thus, the terminal device can be based on A second probability of the target node receiving the target label is determined, where Y is the type weight corresponding to each edge type, X is the probability of the target node receiving the target label based on the same type of edge, and m is the number of edge types.

[0086] In some feasible implementations, the neighbor nodes of the target node may include multi-degree neighbor nodes, see Figure 5 , Figure 5 FIG. 1 is a schematic diagram of a scenario of neighbor nodes provided by an embodiment of the present invention. Figure 5, when node A is the target node, and the neighbor nodes include the first-degree neighbor nodes (node ​​B1, node B2, node B3, and node B4) and second-degree neighbor nodes (node ​​C1 and node C2) of the target node, the terminal device may first calculate the probability vector of the target node receiving the target label propagated from each first-degree neighbor node (node ​​B1, node B2, node B3, and node B4) based on the same type of edge, thereby obtaining the probability vector set of the target node receiving the target label propagated from each first-degree neighbor node (node ​​B1, node B2, node B3, and node B4) based on multiple types of edges. The terminal device may calculate the probability vector of the target node receiving the target label propagated from each second-degree neighbor node (node ​​C1 and node C2) based on the same type of edge, thereby obtaining the probability vector set of the target node receiving the target label propagated from each second-degree neighbor node (node ​​C1 and node C2) based on multiple types of edges. Among them, the probability vector of the target node A receiving the propagation of the second-degree neighbor node C2 based on multiple types of edges can be determined by the propagation probability of the first-degree neighbor node B3 receiving the propagation based on multiple types of edges between node B2 and the propagation probability of the first-degree neighbor node propagating the target label to the target node A based on multiple types of edges, and the specific method is not limited here. Based on the above two probability vector sets, the second probability vector set of the target label propagated by each neighbor node based on multiple types of edges is obtained, and then the second probability of the target node receiving the target label is obtained based on the second probability vector set. Among them, the specific calculation method can be based on the implementation method shown above, which is not limited here.

[0087] 105. The terminal device determines a node category of the target node based on the second probability, and outputs the node category of the target node to a user interaction interface of the terminal device for display to the user.

[0088] In some feasible implementations, after the terminal device determines the second probability that the target node receives the target label, the second probability can be compared with a preset threshold. When the second probability is greater than or equal to the preset threshold, the terminal device can determine that the node category of the target node is the node category marked by the target label. Optionally, when the second probability is less than the preset threshold, it can be determined that the target label propagated to the target node is a negative label (such as gambling, fraud, etc.). At the same time, after the terminal device determines the node category of the target node, the node category of the target node can be output to the user interaction interface of the terminal device for display to the user.

[0089] In the embodiment of the present invention, by converting different types of node relationships in different social networks into the same type of node relationships, the restrictions on node relationships in different social networks can be eliminated, so that the node category of any node in different social networks can be predicted. By calculating the attribute weights of each edge attribute of each type of edge, the corresponding proportion of each edge attribute in the edge of its corresponding type can be determined. Therefore, the edge weight of each type of edge can be accurately and reasonably determined based on the attribute weights of each edge attribute, and the applicability is high.

[0090] See also Figure 6 , Figure 6 : is a schematic diagram of the structure of a node category determination device provided in an embodiment of the present invention. The node category determination device provided in an embodiment of the present invention includes:

[0091] A node determination module 31 is used to determine at least one neighbor node of a target node, wherein the neighbor node is a node connected to the target node based on at least one type of edge;

[0092] A first probability determination module 32 is used to calculate at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each neighbor node propagates the target label to the target node based on various types of edges;

[0093] A probability vector determination module 33 is used to determine, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges;

[0094] A second probability determination module 34, configured to obtain a second probability of the target node receiving the target label based on the first probability vector set;

[0095] The node category determination module 35 is used to determine the node category of the target node based on the second probability, and output the node category of the target node to the user interaction interface of the terminal device for display to the user.

[0096] In some feasible implementations, the first probability determination module 32 includes:

[0097] An edge attribute acquisition unit 321 is used to acquire at least one edge attribute of any type of edge connecting any neighbor node and the target node;

[0098] An attribute weight determination unit 322, configured to determine an attribute weight of each edge attribute in the at least one edge attribute;

[0099] The edge weight determination unit 323 is used to determine the edge weight of any of the above-mentioned types of edges based on the attribute weights of the above-mentioned edge attributes, and determine the above-mentioned edge weight as the first probability that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on any of the above-mentioned types of edges, so as to obtain the first probabilities that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on each type of edge between the above-mentioned neighboring nodes and the above-mentioned target node.

[0100] In some feasible implementations, the probability vector determination module 33 is used to:

[0101] Determine at least one first probability corresponding to edges of the same type from the first probability set;

[0102] A first probability vector of the target node receiving the target label based on the same type of edge is constructed based on the at least one first probability.

[0103] In some feasible implementations, the node determination module 31 is used to:

[0104] Obtain association relationships between nodes in at least one social network, determine at least one node that has an association relationship with a target node from among the nodes based on the association relationships between the nodes, and determine the at least one node as a neighbor node of the target node.

[0105] In some feasible implementations, the second probability determination module 34 is configured to:

[0106] Determine the type weights corresponding to the edge types, where one type corresponds to one type weight;

[0107] A second probability of the target node receiving the target label is determined based on the above-mentioned various type weights and the above-mentioned first probability vector set.

[0108] In some feasible implementations, the node category determination module 35 includes:

[0109] A comparing unit 351 is used to compare the second probability with a preset threshold;

[0110] The category determination unit 352 is configured to determine that the node category of the target node is the node category marked by the target tag when the second probability is greater than or equal to the preset threshold.

[0111] In some feasible implementations, the neighbor nodes of the target node include first-degree neighbor nodes of the target node and second-degree neighbor nodes of the target node, the first-degree neighbor nodes are nodes connected to the target node based on at least one type of edge, and the second-degree neighbor nodes are any nodes connected to the target node based on any first-degree neighbor nodes;

[0112] The probability vector determination module 33 is further used to determine a second probability vector of the target label propagated from each second-degree neighbor node by the target node based on the same type of edge reception, so as to obtain a set of second probability vectors of the target label propagated from each second-degree neighbor node by the target node based on multiple types of edge reception;

[0113] The second probability determination module 34 is further configured to obtain a second probability that the target node receives the target label based on the first probability vector set and the second probability vector set.

[0114] In a specific implementation, the node category determination device can execute the above-mentioned operations through its built-in modules and / or units. Figures 1 to 5 The implementation methods provided in each step will not be repeated here.

[0115] In the embodiment of the present invention, by converting different types of node relationships in different social networks into the same type of node relationships, the restrictions on node relationships in different social networks can be eliminated, so that the node category of any node in different social networks can be predicted. By calculating the attribute weights of each edge attribute of each type of edge, the corresponding proportion of each edge attribute in the edge of its corresponding type can be determined. Therefore, the edge weight of each type of edge can be accurately and reasonably determined based on the attribute weights of each edge attribute, and the applicability is high.

[0116] See also Figure 7 , Figure 7It is a structural diagram of the terminal device provided by an embodiment of the present invention. The electronic device 1000 may include: a processor 1001, a network interface 1004 and a memory 1005. In addition, the electronic device 1000 may also include: a user interface 1003, and at least one communication bus 1002. The communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 1005 may optionally be at least one storage device located away from the aforementioned processor 1001. As Figure 7 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a device control application program.

[0117] exist Figure 7 In the electronic device 1000 shown, the network interface 1004 can provide a network communication function; the user interface 1003 is mainly used to provide an input interface for the user; and the processor 1001 can be used to call the device control application stored in the memory 1005 to achieve:

[0118] Determine at least one neighbor node of the target node, wherein the neighbor node is a node connected to the target node based on at least one type of edge;

[0119] Calculate at least one first probability that any neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each neighbor node propagates the target label to the target node based on various types of edges;

[0120] Determine, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges;

[0121] Obtaining a second probability of the target node receiving the target label based on the first probability vector set;

[0122] The node category of the target node is determined based on the second probability, and the node category of the target node is output to the user interaction interface of the terminal device to be displayed to the user.

[0123] In some feasible implementations, the processor 1001 is configured to:

[0124] Obtain at least one edge attribute of any type of edge connecting any neighbor node and the target node;

[0125] Determining an attribute weight of each edge attribute in the at least one edge attribute;

[0126] Based on the attribute weights of the above-mentioned edge attributes, the edge weight of any of the above-mentioned types of edges is determined, and the above-mentioned edge weight is determined as the first probability that any of the above-mentioned neighboring nodes propagates the target label to the above-mentioned target node based on any of the above-mentioned types of edges, so as to obtain the first probabilities that any of the above-mentioned neighboring nodes propagate the target label to the above-mentioned target node based on each type of edge between the above-mentioned neighboring node and the above-mentioned target node.

[0127] In some feasible implementations, the processor 1001 is configured to:

[0128] Determine at least one first probability corresponding to edges of the same type from the first probability set;

[0129] A first probability vector of the target node receiving the target label based on the same type of edge is constructed based on the at least one first probability.

[0130] In some feasible implementations, the processor 1001 is configured to:

[0131] Obtain association relationships between nodes in at least one social network, determine at least one node that has an association relationship with a target node from among the nodes based on the association relationships between the nodes, and determine the at least one node as a neighbor node of the target node.

[0132] In some feasible implementations, the processor 1001 is configured to:

[0133] Determine the type weights corresponding to the edge types, where one type corresponds to one type weight;

[0134] A second probability of the target node receiving the target label is determined based on the above-mentioned various type weights and the above-mentioned first probability vector set.

[0135] In some feasible implementations, the processor 1001 is configured to:

[0136] Comparing the second probability with a preset threshold;

[0137] When the second probability is greater than or equal to the preset threshold, it is determined that the node category of the target node is the node category marked by the target tag.

[0138] In some feasible implementations, the neighbor nodes of the target node include first-degree neighbor nodes of the target node and second-degree neighbor nodes of the target node, the first-degree neighbor nodes are nodes connected to the target node based on at least one type of edge, and the second-degree neighbor nodes are any nodes connected to the target node based on any first-degree neighbor nodes;

[0139] The processor 1001 is further configured to:

[0140] Determine a second probability vector of the target label propagated from each second-degree neighbor node by the target node based on the same type of edge reception, so as to obtain a set of second probability vectors of the target label propagated from each second-degree neighbor node by the target node based on multiple types of edge reception;

[0141] A second probability of the target node receiving the target label is obtained based on the first probability vector set and the second probability vector set.

[0142] It should be understood that in some feasible implementations, the processor 1001 may be a central processing unit (CPU), and the processor 1001 may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0143] The memory 1005 may include a read-only memory and a random access memory, and provide instructions and data to the processor 1001. A portion of the memory 1005 may also include a non-volatile random access memory. For example, the memory 1005 may also store information about the device type.

[0144] In a specific implementation, the above terminal device can execute the above-mentioned functions through its built-in functional modules. Figures 1 to 5 For the implementation methods provided in each step, please refer to the implementation methods provided in the above steps for details, which will not be repeated here.

[0145] In the embodiment of the present invention, by converting different types of node relationships in different social networks into the same type of node relationships, the restrictions on node relationships in different social networks can be eliminated, so that the node category of any node in different social networks can be predicted. By calculating the attribute weights of each edge attribute of each type of edge, the corresponding proportion of each edge attribute in the edge of its corresponding type can be determined. Therefore, the edge weight of each type of edge can be accurately and reasonably determined based on the attribute weights of each edge attribute, and the applicability is high.

[0146] The embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program and is executed by a processor to implement Figures 1 to 5 For the methods provided in each step, please refer to the implementation methods provided in the above steps for details, which will not be repeated here.

[0147] The computer-readable storage medium may be the task processing device provided in any of the aforementioned embodiments or the internal storage unit of the terminal device, such as a hard disk or memory of an electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device. The computer-readable storage medium may also include a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. Further, the computer-readable storage medium may also include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0148] The terms "first", "second", etc. in the claims, specification and drawings of the present invention are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. Mentioning "embodiment" in this article means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present invention. Displaying the phrase at various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments. The term "and / or" used in the specification and the appended claims of the present invention refers to any combination of one or more of the items listed in association and all possible combinations, and includes these combinations.

[0149] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Professional and technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0150] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for determining node categories, It is characterized in that The method comprises: The terminal device obtains association relationships between nodes in multiple social networks, determines at least one node that has an association relationship with a target node from the nodes based on the association relationships between the nodes, and determines the at least one node as a neighbor node of the target node; wherein the neighbor node is a node connected to the target node based on at least one type of edge; wherein each type of edge corresponds to an association relationship between nodes; the neighbor nodes of the target node include a first-degree neighbor node of the target node and a second-degree neighbor node of the target node, wherein the first-degree neighbor node is a node connected to the target node based on at least one type of edge, and the second-degree neighbor node is any node connected to the target node based on any first-degree neighbor node; The terminal device calculates at least one first probability that any first-degree neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each first-degree neighbor node propagates the target label to the target node based on various types of edges; The terminal device determines, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges; The terminal device determines a second probability vector of the target label propagated by the target node from each second-degree neighbor node based on edges of the same type, so as to obtain a set of second probability vectors of the target label propagated by the target node from each second-degree neighbor node based on edges of multiple types; The terminal device obtains a second probability that the target node receives the target label based on the first probability vector set and the second probability vector set; The terminal device determines the node category of the target node based on the second probability, and outputs the node category of the target node to a user interaction interface of the terminal device for display to the user.

2. The method according to claim 1, It is characterized in that The terminal device calculates at least one first probability that any first-degree neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, including: The terminal device obtains at least one edge attribute of any type of edge connecting any first-degree neighbor node and the target node; The terminal device determines an attribute weight of each edge attribute in the at least one edge attribute; The terminal device determines the edge weight of any type of edge based on the attribute weights of the various edge attributes, and determines the edge weight as the first probability that any first-degree neighbor node propagates the target label to the target node based on any type of edge, so as to obtain the first probabilities that any first-degree neighbor node propagates the target label to the target node based on each type of edge between the node and the target node.

3. The method according to claim 1, It is characterized in that The terminal device determines, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, including: The terminal device determines, from the first probability set, at least one first probability corresponding to edges whose edge types belong to the same type; The terminal device constructs, based on the at least one first probability, a first probability vector of the target node receiving the target label based on the same type of edge.

4. The method according to any one of claims 1 to 3, It is characterized in that The terminal device obtains a second probability that the target node receives the target label based on the first probability vector set and the second probability vector set, including: The terminal device determines each type weight corresponding to each type of edge type, wherein one type corresponds to one type weight; The terminal device determines a second probability that the target node receives the target label based on the weights of each type and the first probability vector set and the second probability vector set.

5. The method according to any one of claims 1 to 3, It is characterized in that The terminal device determines the node category of the target node based on the second probability, including: The terminal device compares the second probability with a preset threshold; When the second probability is greater than or equal to the preset threshold, the terminal device determines that the node category of the target node is the node category marked by the target tag.

6. A node category determination device, It is characterized in that The device comprises: A node determination module, used to obtain association relationships between nodes in multiple social networks, determine at least one node that has an association relationship with a target node from the nodes based on the association relationships between the nodes, and determine the at least one node as a neighbor node of the target node, wherein the neighbor node is a node connected to the target node based on at least one type of edge; wherein each type of edge corresponds to an association relationship between nodes; the neighbor nodes of the target node include a first-degree neighbor node of the target node and a second-degree neighbor node of the target node, wherein the first-degree neighbor node is a node connected to the target node based on at least one type of edge, and the second-degree neighbor node is any node connected to the target node based on any first-degree neighbor node; A first probability determination module is used to calculate at least one first probability that any first-degree neighbor node of the target node propagates the target label to the target node based on various types of edges connected to the target node, wherein one type of edge corresponds to one first probability, so as to obtain a first probability set that each first-degree neighbor node propagates the target label to the target node based on various types of edges; A probability vector determination module, configured to determine, based on the first probability set, a first probability vector of the target node receiving the target label based on the same type of edge, so as to obtain a first probability vector set of the target node receiving the target label based on multiple types of edges; The probability vector determination module is further used to determine a second probability vector of the target label propagated from each second-degree neighbor node by the target node based on the same type of edge reception, so as to obtain a second probability vector set of the target label propagated from each second-degree neighbor node by the target node based on multiple types of edge reception; A second probability determination module, configured to obtain a second probability of the target node receiving the target label based on the first probability vector set and the second probability vector set; A node category determination module is used to determine the node category of the target node based on the second probability, and output the node category of the target node to a user interaction interface of a terminal device for display to a user.

7. A terminal device, It is characterized in that include: Processor and memory; The processor is connected to a memory, wherein the memory is used to store program codes, and the processor is used to call the program codes to execute the method according to any one of claims 1 to 5.

8. A computer storage medium, It is characterized in that The computer storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the method according to any one of claims 1 to 5 is executed.

Images