CN109558540B - Method, device, equipment and storage medium for determining user influence - Google Patents

Method, device, equipment and storage medium for determining user influence Download PDF

Info

Publication number
CN109558540B
CN109558540B CN201811456146.1A CN201811456146A CN109558540B CN 109558540 B CN109558540 B CN 109558540B CN 201811456146 A CN201811456146 A CN 201811456146A CN 109558540 B CN109558540 B CN 109558540B
Authority
CN
China
Prior art keywords
node
influence
tdag
social network
nodes
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811456146.1A
Other languages
Chinese (zh)
Other versions
CN109558540A (en
Inventor
王浩然
周效军
宋国栋
任化强
刘长龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Migu Cultural Technology Co Ltd
Original Assignee
Migu Cultural Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Migu Cultural Technology Co Ltd filed Critical Migu Cultural Technology Co Ltd
Priority to CN201811456146.1A priority Critical patent/CN109558540B/en
Publication of CN109558540A publication Critical patent/CN109558540A/en
Application granted granted Critical
Publication of CN109558540B publication Critical patent/CN109558540B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The embodiment of the application discloses a method, a device, equipment and a storage medium for determining user influence, wherein the method comprises the following steps: acquiring historical information propagated by Q1 nodes in a social network within a preset time period, wherein Q1 is an integer greater than or equal to 2; determining topic information of the social network and influence topic information of each node based on historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics; and determining the influence of each node in the social network based on the topic information of the social network and the influence topic information of each node.

Description

Method, device, equipment and storage medium for determining user influence
Technical Field
The embodiment of the application relates to computer technology, in particular to a method, a device, equipment and a storage medium for determining user influence.
Background
With the continuous development of internet information technology, networks bring practical changes to users in the aspect of people's daily life. And the birth of the social network greatly facilitates the friend making of people in daily life. Generally, a social network is often composed of a large number of users and complex relationships among the users (including relatives, friends, classmates, and working relationships), and unlike a conventional network, the propagation and diffusion of information in the social network depend on the relationships among the users. Therefore, how to accurately measure the influence of the user in the social network (referred to as user influence for short) is to select the user with the largest influence from the user influence, so that when the information is transmitted by the user with the largest influence, the information can be received by as many users as possible in the social network, and the method becomes a hot point of research on the current social network and application thereof.
Currently, in research on how to determine user influence, the user influence is generally determined through an independent cascading information propagation model, that is, the node influence (i.e., user influence) is calculated mainly based on the number of neighbor nodes of the node, so that for users with high influence, because of low activity, the influence is not really exerted to transmit information, and these nodes with high influence should not be elected to transmit information. For example, for advertisement or public opinion control, when selecting users with influence, only the influence of the users themselves is considered, and the influence of other advertisements or opposite public opinions on the influence of the users is not considered, so that the final influence effect is limited by considering one side. That is, when measuring the influence of the node, only the network topology such as the number of connections of the node is considered, and the characteristics of the node are not sufficiently considered, so that the finally determined user influence cannot reflect the actual propagation capacity of the user.
Disclosure of Invention
In view of this, embodiments of the present application provide a method and apparatus, a device, and a storage medium for determining user influence to solve at least one problem in the related art.
The technical scheme of the embodiment of the application is realized as follows:
in a first aspect, an embodiment of the present application provides a method for determining user influence, where the method includes:
acquiring historical information propagated by Q1 nodes in a social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
determining topic information of the social network and influence topic information of each node based on historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
and determining the influence of each node in the social network based on the topic information of the social network and the influence topic information of each node.
In a second aspect, an embodiment of the present application provides an apparatus for determining user influence, including:
the information acquisition module is configured to acquire historical information propagated by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
a vector determination module configured to determine topic information of the social network and influence topic information of each of the nodes based on the history information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
and the influence determination module is configured to determine the influence of each node in the social network based on the topic information of the social network and the influence topic information of each node.
In a third aspect, an embodiment of the present application provides a computer device, including a memory and a processor, where the memory stores a computer program executable on the processor, and the processor implements the steps in the method for determining the influence of a user when executing the program.
In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps in the method for determining the influence of a user.
In an embodiment of the present application, there is provided a method based on determining user influence, in which method, because the influence topic information represents the influence probability of the corresponding node on the corresponding neighbor node under the N topics, namely the influence topic information can truly represent the preference of the corresponding user in the process of spreading information, therefore, based on the topic information of the social network and the influence topic information of each node, the determined influence of each node in the social network is more practical, so that when the node with the largest influence is selected as the propagation node based on the influence of each node in the social network, the selected propagation node can be received by more users in the social network when the information is propagated, so that the selected propagation node can exert the maximum propagation effect.
Drawings
Fig. 1 is a schematic flow chart illustrating an implementation of a method for determining user influence according to an embodiment of the present application;
fig. 2A is a schematic flow chart illustrating an implementation of another method for determining user influence according to the embodiment of the present application;
FIG. 2B is a schematic diagram of a first TDAG diagram according to an embodiment of the present disclosure;
fig. 3A is a schematic flow chart illustrating an implementation process of a method for determining user influence according to another embodiment of the present application
Fig. 3B is a schematic structural diagram of another first TDAG diagram according to an embodiment of the present disclosure;
FIG. 3C is a schematic diagram of a structure of a first TDAG diagram according to an embodiment of the present application;
FIG. 4 is a schematic diagram of an implementation process of a method for determining user influence according to another embodiment of the present application
Fig. 5A is a schematic flow chart illustrating an implementation of another method for determining user influence according to the embodiment of the present application;
fig. 5B is a schematic structural diagram of another first TDAG diagram according to an embodiment of the present disclosure;
fig. 6 is a schematic flow chart illustrating an implementation of a method for determining user influence according to another embodiment of the present application;
fig. 7A is a schematic flowchart of an implementation of a method for determining user influence according to an embodiment of the present application;
FIG. 7B is a schematic structural diagram of a first TDAG diagram according to an embodiment of the present application;
fig. 7C is a schematic structural diagram of another first TDAG diagram according to an embodiment of the present disclosure;
fig. 8 is a schematic flowchart illustrating an implementation process of another method for determining user influence according to an embodiment of the present application;
fig. 9 is a schematic flowchart of an implementation of a method for determining user influence according to an embodiment of the present application;
fig. 10 is a schematic flowchart of an implementation of a method for determining user influence according to another embodiment of the present application;
FIG. 11 is a diagram illustrating a specific model for solving topic vectors between nodes according to an embodiment of the present application;
fig. 12 is a schematic structural diagram of a network topology according to an embodiment of the present application;
FIG. 13 is a schematic diagram illustrating the structure of a TDAG chart according to an embodiment of the present invention;
FIG. 14A is a schematic diagram illustrating a structure of an apparatus for determining user influence according to an embodiment of the present disclosure;
FIG. 14B is a schematic diagram illustrating a structure of another apparatus for determining user influence according to an embodiment of the present disclosure;
FIG. 14C is a schematic diagram illustrating a structure of a device for determining user influence according to an embodiment of the present disclosure;
fig. 15 is a schematic diagram of a hardware entity of a computer device according to an embodiment of the present application.
Detailed Description
The technical solution of the present application is further elaborated below with reference to the drawings and the embodiments.
The embodiment of the application provides a method for determining user influence, which is applied to a computer device, and generally, the computer device may be various types of devices capable of determining user influence in an implementation process, for example, the computer device may include a personal computer, a server, or a network device. The functions implemented by the method may be implemented by calling program code by a processor in a computer device, which may, of course, be stored in a computer storage medium, which may comprise at least a processor and a storage medium.
Fig. 1 is a schematic flow chart illustrating an implementation process of a method for determining user influence according to an embodiment of the present application, as shown in fig. 1, the method includes steps S101 to S103:
s101, acquiring historical information spread by Q1 nodes in a social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
in general, one node represents one user, and the Q1 nodes may be all users in the social network.
S102, determining topic information of the social network and influence topic information of each node based on historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics, N is an integer greater than or equal to 1 and is used for reading books generally, and N is greater than or equal to 2;
in practical applications, when determining the topic information of the social network and the influence topic information of each node, the historical information of the Q1 nodes may be input into a preset topic model, for example, the historical information of the Q1 nodes is input into an implicit dirichlet distribution model, so as to output the topic information of the social network and the influence topic information of each node. It should be noted that, generally, the history information of the Q1 nodes is input into the topic model as a whole, so that the topics corresponding to the topic information and the influence topic information of each node can be consistent.
It is understood that the neighboring nodes of the node are actually nodes having a direct social relationship with the node, for example, two neighboring nodes are friends with each other. The influence topic information represents influence probability of each node on corresponding neighbor nodes under the N topics, so that in practical application, the influence topic information can be expressed in a vector form, and how many influence topic vectors are corresponding to how many neighbor nodes are in one node. For exampleThe neighboring nodes of the node u0 include u1, u2 and u3, and correspondingly, the influence topic vector of the node u0 includes: alpha is alphau0,u10.6 for childbearing (basketball: 0.3; automobile: 0.1), alphau0,u20.3 for childbearing (basketball: 0.5; automobile: 0.2), alphau0,u3= (child bearing: 0.2; basketball: 0.6; car: 0.2); alpha is alphau0,u1When the subject is "child bearing", "basketball", or "car", the term means that the node u0 has an influence probability of 0.6, 0.3, or 0.1 on the node u1, and α is 0.6, 0.3, or 0.1, respectivelyu0,u2And alphau0,u3The same is true.
S103, determining the influence of each node in the social network based on the topic information of the social network and the influence topic information of each node.
In an embodiment of the application, a method based on determining a user influence is provided, in which method, because the influence topic information represents the influence probability of the corresponding node on the corresponding neighbor node under the N topics, namely the influence topic information can truly represent the preference of the corresponding user in the process of spreading information, therefore, based on the topic information of the social network and the influence topic information of each node, the determined influence of each node in the social network is more practical, so that when the node with the largest influence is selected as the propagation node based on the influence of each node in the social network, the selected propagation node can be received by more users in the social network when the information is propagated, so that the selected propagation node can exert the maximum propagation effect.
An embodiment of the present application provides another method for determining user influence, and fig. 2A is a schematic flowchart illustrating an implementation process of the another method for determining user influence according to the embodiment of the present application, as shown in fig. 2A, the method includes steps S201 to S205:
s201, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s202, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s203, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
it should be noted that the influence topic information of each node has a one-to-one correspondence relationship with its neighboring nodes, so that here, actually, the influence of the corresponding node on the neighboring node corresponding to the influence topic information is determined based on the topic information of the social network and the influence topic information of each node. For example, the influence Neig of the node u on the neighbor node v of the node u may be calculated based on the following formula (1)u,v：
Figure BDA0001887781560000061
In the formula, λzRepresenting the probability of topic Z being propagated in the social network,
Figure BDA0001887781560000062
representing the influence probability of the node u on the neighbor node v under the topic Z, and k representing the number of topics.
S204, acquiring a first TDAG (time domain access graph) composed of each node as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
it should be understood that each node corresponds to a first TDAG graph, for example, fig. 2B is a schematic diagram of a composition structure of the first TDAG graph 20 corresponding to the node v1, where the node v1 is a central node of the first TDAG graph, and the nodes e, d, a, f, and c are associated nodes of the node v, that is, nodes having an influence on the node v greater than a preset threshold.
In practical applications, there may be no node having an influence on the central node greater than a predetermined threshold, that is, M is equal to 0, and the first TDAG map only includes the central node.
S205, determining the influence of each node in the social network based on the Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes.
In fact, in a social network, as the distance between nodes increases, the influence between nodes also becomes smaller. Therefore, when the influence of each node in the social network is calculated, the nodes with smaller influence can be removed, and only the nodes with the influence on the central node larger than the preset threshold value are calculated, that is, only the influence of the associated nodes in the social network is calculated, so that the influence of all the nodes in the social network is avoided being required to be calculated, and the calculation efficiency is improved.
An embodiment of the present application provides a further method for determining user influence, and fig. 3A is a schematic flowchart illustrating an implementation process of the further method for determining user influence according to the embodiment of the present application, as shown in fig. 3A, the method includes steps S301 to S307:
s301, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s302, determining topic information of the social network and influence topic information of each node based on historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics, N is an integer greater than or equal to 1 and is used for reading books generally, and N is greater than or equal to 2;
s303, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s304, acquiring a first TDAG graph consisting of each node serving as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
s305, based on the Q1 first TDAG graphs, selecting Q2 first TDAG graphs to which a node i belongs, wherein the node i is any one of Q1 nodes, and Q2 is less than or equal to Q1;
s306, in the Q2 first TDAG graphs, determining the influence of each node i on each central node based on the influence of each node on the corresponding neighbor node;
s307, in the Q2 first TDAG graphs, determining the influence of the node i in the social network based on the influence of the node i on each central node.
For example, the influence inf (i) of the node i in the social network may be determined based on the following formula (2):
inf(i)＝∑v∈infset(i)αv(i) (2)
wherein, infiset (i) represents all the first TDAG graphs where the node i is located, and α v (i) represents the influence of the node i on the central node v in the first TDAG graph corresponding to the central node v, wherein, the calculation algorithm of α v (i), for example, as shown in fig. 3B, takes calculating the influence α v1(d) of the node d on the central node v1 in the first TDAG graph 30 as an example, as can be seen from the figure, the path from the node d to the central node v1 includes two paths: d → f → v1, d → c → v1, therefore, α v1(d) ═ Neigd,f*Neigf,v1)+(Neigd,c*Neigf,v1). That is, α v (i) is equal to the sum of the influences of all paths from node i to the central node v in the first TDAG map in which node i is located.
Still taking node d of the first TDAG graph 30 as an example, for determining the influence inf (d) of node d in the social network, for example, as shown in fig. 3C, the first TDAG graph where node d is located includes: the first TDAG map 30 with the node v1 as the center node, the first TDAG map 31 with the node v2 as the center node, and the first TDAG map 32 with the node v3 as the center node, then inf (d) ═ α v1(d) + α v2(d) + α v3(d) can be determined according to formula (2).
In the embodiment of the application, a method for determining user influence is provided, wherein when influence of a node i in the social network is determined, Q2 first TDAG maps to which the node i belongs are selected from Q1 first TDAG maps, and Q2 is less than or equal to Q1; then, in the Q2 first TDAG graphs, the influence of the node i in the social network is determined based on the influence of the node i on each central node. In this way, the influence of the node i in the social network is determined based on only the Q2 first TDAG maps where the node i is located, and the complexity of the algorithm is greatly reduced.
An embodiment of the present application provides a further method for determining user influence, and fig. 4 is a schematic flowchart illustrating an implementation process of the further method for determining user influence according to the embodiment of the present application, as shown in fig. 4, the method includes steps S401 to S406:
s401, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s402, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s403, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s404, with each node as a central node, based on the influence of each central node on the corresponding neighbor node, searching M associated nodes with the influence on the central node larger than a preset threshold value on a path where the central node is located, wherein M is an integer larger than or equal to 0;
s405, forming a first TDAG graph corresponding to the central node by using the central node and the M associated nodes;
it is understood that if M is equal to 0, the first TDAG map corresponding to the central node is composed only of the central node. For example, if the central node is the node d, and the first TDAG map corresponding to the node d as the central node includes only the node d and does not include other nodes, when determining the influence of the node d on the central node based on the first TDAG map corresponding to the node d as the central node, the influence of the node d on itself is actually determined, and may generally be regarded as 0 by default.
S406, determining the influence of each node in the social network based on the Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes.
An embodiment of the present application provides another method for determining user influence, and fig. 5A is a schematic flowchart illustrating an implementation process of the another method for determining user influence according to the embodiment of the present application, as shown in fig. 5A, the method includes steps S501 to S506:
s501, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s502, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s503, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s504, with each node as a central node, determining the influence of the associated node of the central node on the central node step by step on the path where the central node is located based on the influence of each central node on the corresponding neighbor node, and obtaining M associated nodes with the influence on the central node larger than a preset threshold;
for example, if the influence of the current associated node on the central node is greater than the preset threshold, determining the current associated node as a node in the corresponding first TDAG graph, and continuing to determine the influence of the associated node of the next level on the path where the current associated node is located on the central node; if the influence of the current associated node on the central node is smaller than or equal to the preset threshold, ending the search on the path where the current associated node is located, and gradually determining the influence of the associated node of the central node on other paths where the central node is located to obtain M associated nodes of which the influence on the central node is larger than the preset threshold;
s505, forming a first TDAG graph corresponding to the central node by using the central node and the M associated nodes;
it can be understood that, in the social network, as the distance between nodes increases, the influence between the nodes may gradually decrease, so to reduce the complexity of the algorithm, when searching for M associated nodes whose influence on the central node is greater than a preset threshold on a path where the central node is located with each of the nodes as the central node, a step-by-step search method may be adopted to perform the search. For example, taking the social network 50 shown in fig. 5B as an example, when determining the first TDAG map 51 corresponding to the node v1 as a central node, first determining whether the influence of the node c, the node f, the node a, and the node g closest to the node v1 on the node v1 is less than or equal to the preset threshold one by one, for example, if the influence of the node g on the node v1 is less than the preset threshold, the search on the path where the node g is located is ended, and the influence of the node h on the node v1 at the next level is no longer determined. For another example, if the influence of the node f on the node v1 is greater than the preset threshold, the node f is determined as the node in the first TDAG map 51, whether the influence of the node y at the next level on the path where the node f is located on the node v1 is smaller than or equal to the preset threshold is continued, and if the influence of the node y on the node v1 is smaller than or equal to the preset threshold, the search on the path where the node y is located is ended. The finally obtained first TDAG graph 51 includes a node v1, a node f, a node c, a node d, a node e, and a node a, where the node v1 is a central node of the first TDAG graph 51, and the node f, the node c, the node d, the node e, and the node a are associated nodes of the node v 1.
Here, it should be noted that, the step-by-step determination of the influence of the node associated with the central node on the central node is actually performed by taking the central node as a reference point, and determining the influence of the node on the path where the central node is located on the central node from near to far step by step.
Alternatively, in other embodiments, the first TDAG map corresponding to each node may be searched through a breadth-first search algorithm.
S506, determining the influence of each node in the social network based on the Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes.
An embodiment of the present application provides a further method for determining user influence, fig. 6 is a schematic view of an implementation flow of the further method for determining user influence according to the embodiment of the present application, and as shown in fig. 6, the method includes steps S601 to S608:
s601, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s602, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s603, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s604, acquiring a first TDAG graph consisting of each node serving as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
s605, determining the influence of each node in the social network based on Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes;
s606, based on the influence of each node in the social network, selecting the node with the largest influence as a propagation node;
in practical applications, the propagation node is used for propagating information specified by the social network, for example, propagating advertisement or public opinion information through the propagation node.
S607, updating the influence of the root node on the path where the propagation node is located in the social network based on the Q1 first TDAG graphs;
in fact, a plurality of nodes with the largest influence need to be selected as the propagation nodes, so after one propagation node is selected, the influence of other nodes in the social network except the propagation node needs to be updated. And in the first TDAG graph, if the propagation node is deleted, only the influence of the root node on the path of the propagation node on the center node of the corresponding first TDAG graph is affected, so in order to reduce the amount of computation, only the influence of the root node on the path of the propagation node on the first TDAG graph in the social network is updated.
S608, based on the updated influence of the root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network, selecting the node with the largest influence as the next propagation node.
Fig. 7A is a schematic flow chart illustrating an implementation of the method for determining user influence according to the embodiment of the present application, and as shown in fig. 7A, the method includes steps S701 to S712:
s701, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s702, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s703, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s704, acquiring a first TDAG (time domain access graph) composed of each node as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
s705, determining the influence of each node in the social network based on Q1 first TDAG graphs and the influence of each node on corresponding neighbor nodes;
s706, based on the influence of each node in the social network, selecting the node with the largest influence as a propagation node, and then entering the step S707;
s707, based on the Q1 first TDAG graphs, selecting Q3 first TDAG graphs to which the propagation node belongs, wherein Q3 is less than or equal to Q1;
s708, determining a root node on a path where the propagation node is located from the Q3 first TDAG graphs;
it can be understood that after one propagation node is selected from the Q1 first TDAG graphs, only the influence of the root node on the path of the propagation node in the first TDAG graph on the corresponding central node is affected, and further the influence of the root node in the social network is affected, so that when the influences of all nodes except the propagation node in the social network are updated, only the influence of the root node in the social network needs to be updated, which can greatly reduce the amount of computation and reduce the time complexity of the algorithm.
For example, as shown in fig. 7B, assuming that a node f is selected as the propagation node, and the first TDAG maps of the node f are the first TDAG map 70, the first TDAG map 71, and the first TDAG map 72, respectively, then, in the first TDAG map 70, after the node f is selected as the propagation node, the influence of the node e and the node d on the path thereof on the central node v1 of the first TDAG map 70 is affected; in the first TDAG map 71, after removing the node f, the influence of the node h on the path of the node f on the central node v2 of the first TDAG map 71 is affected; in the first TDAG map 72, since the node f itself is the root node of the first TDAG map 72, the influence of the other nodes in the first TDAG map 72 on the central node v3 is not affected after the node f is removed. Based on the first TDAG graph 70, the first TDAG graph 71, and the first TDAG graph 72, when the node f is selected as the propagation node, the root node on the path where the node f is located includes the node e, the node d, and the node h, so that the influence of the node e, the node d, and the node h in the social network only needs to be updated.
S709, based on the Q1 first TDAG graphs, selecting Q4 first TDAG graphs belonging to the root node, wherein Q4 is less than or equal to Q1;
s710, deleting the propagation nodes and paths connecting the propagation nodes from the Q4 first TDAG graphs to obtain Q5 second TDAG graphs, wherein Q5 is less than or equal to Q4;
s711, in the Q5 second TDAG graphs, updating the influence of each node on the corresponding neighbor nodes in the social network based on the influence of the root node on the corresponding neighbor nodes;
still taking the example shown in fig. 7B as an example, based on steps S707 and S708, it is determined that the root node on the path where the propagation node f is located includes node e, node d, and node h. For example, when the influence of node e in the social network is updated, according to step S709: based on the Q1 first TDAG maps, selecting the first TDAG map to which the node e belongs, as shown in fig. 7C, including: first TDAG map 70 and first TDAG map 73; then, according to step S710: from the first TDAG map 70 and the first TDAG map 73, the path connecting the node f and the node f is deleted, and as can be seen from the first TDAG map 70, the path connecting the node f includes the path connecting the node e and the node f, the path connecting the node d and the node f, and the path connecting the node f and the node v1, so that after the path connecting the node f and the node f is deleted, the obtained second TDAG map 74 is as shown in fig. 7C, and the second TDAG map 74 is composed of a central node v1, and associated nodes e, a, d, and C; as can be seen from the first TDAG map 73, there is no node f in the map, so the resulting second TDAG map 75 is the same as the first TDAG map 73. After obtaining the second TDAG map 74 and the second TDAG map 75 of the node e, the influence of the node e in the social network may be determined again according to step S711, that is, in the second TDAG map 74 and the second TDAG map 75, the influence of the node e in the social network is updated based on the influence of each node on the corresponding neighbor node.
S712, based on the updated influence of the root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network, selecting the node with the largest influence as the next propagation node.
Here, still taking the 3 first TDAG graphs shown in fig. 7B as an example, the node f is a selected propagation node, and the root node on the path where the node f is located includes the node e, the node d, and the node h, so that a node with the largest influence may be selected as a next propagation node from the nodes based on the influence of the updated node e, node d, and node h in the social network, and the influence of other nodes except the node e, node d, node h, and node f before updating in the social network.
An embodiment of the present application provides another method for determining user influence, fig. 8 is a schematic view illustrating an implementation flow of the another method for determining user influence according to the embodiment of the present application, and as shown in fig. 8, the method includes steps S801 to S815:
s801, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s802, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s803, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s804, acquiring a first TDAG (time domain access graph) composed of each node serving as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
s805, determining the influence of each node in the social network based on Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes;
s806, selecting a node with the largest influence from the influence of each node in the social network as a propagation node, and then entering the step S807;
s807, based on the Q1 first TDAG graphs, selecting Q3 first TDAG graphs to which the propagation node belongs, wherein Q3 is less than or equal to Q1;
s808, determining a root node on a path where the propagation node is located from the Q3 first TDAG graphs;
s809, based on the Q1 first TDAG graphs, selecting Q4 first TDAG graphs to which the root node belongs, wherein Q4 is less than or equal to Q1;
s810, deleting the propagation nodes and paths connecting the propagation nodes from the Q4 first TDAG maps to obtain Q5 second TDAG maps, wherein Q5 is equal to or less than Q4, and then proceeding to step S811;
s811, determining the activation probability of the root node;
it should be noted that the activation probability of the root node is used to characterize the probability that the root node activates other nodes to execute the information propagated by the root node. For example, node u publishes an advertisement message about a certain commodity in the social network, and if node v purchases the commodity after seeing the advertisement message, it is considered that the node is activated by the advertisement message.
S812, in each second TDAG graph, determining the influence of each root node on each central node based on the influence of each node on the corresponding neighbor node;
s813, determining an influence increment of the root node in the social network based on the activation probability of the root node and the influence of the root node on a central node of each second TDAG graph;
for example, the influence increment Δ infv (u) of the root node u in the social network may be determined according to the following formula (3):
ΔInfv(u)＝Av(u)*(1-ap(u)) (3)
in the formula, Av(u) a linear coefficient representing the node u, which is actually determined based on the influence of the root node u on the central node of each of the located second TDAG maps, that is, in fact, the influence of the root node u on the central node of each of the located second TDAG maps has a linear relationship; ap (u) represents the activation probability of the root node u.
S814, updating the influence of the root node in the social network based on the influence increment and the influence of the root node in the social network before updating;
for example, the influence of the root node u in the social network is re-determined according to the following formula (4):
Inf(u)′＝Inf(u)+ΔInfv(u) (4)
wherein inf (u)' represents the influence of the updated root node u in the social network, inf (u) represents the influence of the root node u in the social network before updating, and Δ infv (u) represents the influence increment of the root node u in the social network.
S815, based on the influence of the updated root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network, selecting the node with the largest influence as the next propagation node.
An embodiment of the present application provides a further method for determining a user influence, fig. 9 is a schematic flowchart illustrating an implementation process of the further method for determining a user influence according to the embodiment of the present application, and as shown in fig. 9, the method includes steps S901 to S917:
s901, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s902, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s903, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s904, acquiring a first TDAG (time domain access graph) composed of each node as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
s905, determining the influence of each node in the social network based on Q1 first TDAG graphs and the influence of each node on corresponding neighbor nodes;
s906, based on the influence of each node in the social network, selecting the node with the largest influence as a propagation node, and then entering the step S907;
s907, based on the Q1 first TDAG graphs, selecting Q3 first TDAG graphs to which the propagation nodes belong, wherein Q3 is less than or equal to Q1;
s908, determining a root node on a path where the propagation node is located from the Q3 first TDAG graphs;
s909, based on the Q1 first TDAG graphs, selecting Q4 first TDAG graphs belonging to the root node, wherein Q4 is less than or equal to Q1;
s910, deleting the propagation nodes and the paths connecting the propagation nodes from the Q4 first TDAG graphs to obtain Q5 second TDAG graphs, wherein Q5 is less than or equal to Q4, and then entering the step S911;
s911, determining the activation probability of the central node of each second TDAG graph;
s912, determining the influence of the central node of each second TDAG graph on the root node;
s913, determining the activation probability of the root node based on the activation probability of the central node of each second TDAG graph and the influence of the central node corresponding to the activation probability on the root node, and then entering step S914;
for example, the activation probability ap (u) of the root node u is determined according to the following formula (5):
ap(u)＝∑ap(v)*w(v,u) (5)
wherein ap (v) represents the activation probability of the central node v of each of the second TDAG graphs to which the root node u belongs; w (v, u) represents the influence of the central node v of each second TDAG graph to which the root node u belongs on the root node u; as can be seen from equation (5), the activation probability of the root node is equal to the sum of the activation probability of the central node v of each second TDAG map and the product of the influence of the central node corresponding to the activation probability on the root node. For example, assume that the second TDAG map corresponding to the root node u includes: and if the second TDAG map with the node v1 as the center node, the second TDAG map with the node v2 as the center node, and the second TDAG map with the node v3 as the center node, the activation probability ap (u) ═ ap (v1) × w (v1, u) + ap (v2) × w (v2, u) + ap (v3) × w (v3, u) of the root node u is determined according to the formula (5).
S914, in each second TDAG graph, determining the influence of each central node by the root node based on the influence of each node on the corresponding neighbor node;
step S914 may be executed first, and then step S911 to step S913 may be executed.
S915, determining an influence increment of the root node in the social network based on the activation probability of the root node and the influence of the root node on a central node of each second TDAG graph;
s916, updating the influence of the root node in the social network based on the influence increment and the influence of the root node in the social network before updating;
for example, after determining the activation probability of the root node and the influence of the root node on the central node of each of the second TDAG maps, the influence of the root node in the social network may be updated by the above formula (3) and formula (4).
S917, based on the influence of the updated root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network, selecting the node with the largest influence as the next propagation node.
An embodiment of the present application provides a further method for determining user influence, and fig. 10 is a schematic flowchart illustrating an implementation process of the further method for determining user influence according to the embodiment of the present application, as shown in fig. 10, the method includes steps S1001 to S1020:
s1001, acquiring historical information spread by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
s1002, determining topic information of the social network and influence topic information of each node based on the historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
s1003, determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
s1004, with each node as a central node, based on the influence of each central node on a corresponding neighbor node, determining the influence of the associated node of the central node on the central node step by step on the path where the central node is located, and obtaining M associated nodes with the influence on the central node larger than a preset threshold;
for example, if the influence of the current associated node on the central node is greater than the preset threshold, determining the current associated node as a node in the corresponding first TDAG graph, and continuing to determine the influence of the associated node of the next level on the path where the current associated node is located on the central node; if the influence of the current associated node on the central node is smaller than or equal to the preset threshold, ending the search on the path where the current associated node is located, and gradually determining the influence of the associated node of the central node on other paths where the central node is located to obtain the M associated nodes;
s1005, forming a first TDAG graph corresponding to the central node by using the central node and the M associated nodes;
s1006, based on the Q1 first TDAG graphs, selecting Q2 first TDAG graphs to which a node i belongs, wherein the node i is any one of the Q1 nodes;
s1007, in the Q2 first TDAG graphs, determining the influence of each node i on each central node based on the influence of each node on the corresponding neighbor node;
s1008, in the Q2 first TDAG graphs, determining the influence of the node i in the social network based on the influence of the node i on each central node;
s1009, based on the influence of each node in the social network, selecting the node with the largest influence as a propagation node, and then entering S1010;
s1010, based on the Q1 first TDAG graphs, selecting Q3 first TDAG graphs to which the propagation node belongs, wherein Q3 is less than or equal to Q1;
s1011, determining a root node on a path where the propagation node is located from the Q3 first TDAG graphs;
s1012, based on the Q1 first TDAG graphs, selecting Q4 first TDAG graphs to which the root node belongs, wherein Q4 is less than or equal to Q1;
s1013, deleting the propagation nodes and the paths connecting the propagation nodes from the Q4 first TDAG graphs to obtain Q5 second TDAG graphs, wherein Q5 is less than or equal to Q4;
s1014, determining the activation probability of the central node of each second TDAG graph, and then entering the step S1015;
s1015, determining the influence of the central node of each second TDAG graph on the root node;
s1016, determining the activation probability of the root node based on the activation probability of the central node of each second TDAG graph and the influence of the central node corresponding to the activation probability on the root node, and then entering S1017;
s1017, in each second TDAG graph, determining the influence of each root node on each central node based on the influence of each node on the corresponding neighbor node;
s1018, determining an influence increment of the root node in the social network based on the activation probability of the root node and the influence of the root node on a central node of each second TDAG graph;
s1019, updating the influence of the root node in the social network based on the influence increment and the influence of the root node in the social network before updating;
s1020, based on the updated influence of the root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network, selecting the node with the largest influence as the next propagation node.
With the continuous development of internet information technology, networks bring practical changes to users in the aspect of people's daily life. And the birth of the social network greatly facilitates the friend making of people in daily life.
Generally, a social network is often composed of a large number of users and complex relationships among the users (including relatives, friends, classmates, work relationships, and the like), and unlike a conventional network, the propagation and diffusion of information in the social network depend on the relationships among the users. Therefore, how to make information received by as many users as possible in the network, namely the problem of maximizing social network influence, is a hot research focus of current social networks and applications thereof.
At present, corresponding algorithms proposed by researchers include two major categories, one is a greedy hill climbing algorithm with an influence range close to the optimum, and the other is a heuristic algorithm with superior timeliness; in contrast, heuristic algorithms are the main direction of current research because they are better suited to large social networks. The existing heuristic algorithm is based on two influence propagation models, namely an Independent Cascade (IC) model and a Linear Threshold (LT) model, wherein the IC model is a model mainly used by the algorithm, and an activation rule is the core of the propagation model. The activation rule adopted in the IC model considers that, in a state where the probability of the neighbor node being activated is p (u, v), a node in an activated state autonomously performs an activation behavior to the neighbor node by 100%.
However, as can be seen from practical analysis, the selection effect of the initial node (i.e. the propagation node in the above embodiment) depends on two aspects, firstly, how to define the influence of the node (i.e. the influence of the node in the social network); secondly, the effect of algorithm execution is also very significant. Currently, a commonly used Independent Cascade (IC) information propagation model is researched, node influence is generally calculated based on two attributes, namely the number of neighbor nodes of a node and the activation probability of the node to the neighbor nodes, and the calculation method has two problems: the first problem is that the users with high influence are not really influenced to transmit information due to low activity, and the nodes with high influence should not be selected as a seed node set, namely a set of nodes for transmitting the designated information of the social network; the second problem is that the transmission force of the active user is limited, and the traditional IC model considers that the node in the active state can perform the reactivation action on the neighbor nodes by 100%, but the fact is that the reactivation force of the active user is far lower than 100%, which is related to when the whole propagation process is finished.
For advertisement or public opinion control, when selecting users with influence, the traditional IC model only considers the influence of the users, does not consider the influence of other advertisements or opponent public opinions on the influence of the users, considers one side, and has limited effect of final influence. At present, only the network topology structures such as the connection number of nodes and the like are considered when the influence of the nodes is measured, and the characteristics of the nodes are not fully considered.
In order to solve the above problems of the conventional IC model, an influence maximization algorithm based on user preferences, that is, a DAG algorithm, is provided in the embodiments of the present application, and node influences are calculated and evaluated mainly by using topology information and node attributes of a network, so as to establish a reasonable quantization mechanism to measure the magnitude of the node influences.
In the DAG algorithm, all users in a social network and relationships between users are mapped into a network topology, in the network topology, one node represents one user, and a connection line between nodes represents that a social relationship exists between two nodes (i.e., users).
When the influence of each node in the social network is calculated, firstly, influence topic vectors among the nodes are calculated, the influence topic vectors among the nodes measure the influence of the nodes under different topics, and each vector represents the influence of the nodes under different topics. For example, for an edge from node u to node v, the impact topic vector on the edge is < IT: 0.6; basketball: 0.3; an automobile: 0.1>, which indicates that under the topics "IT", "basketball" and "car", the probability that the node v is affected by the node u is 0.6, 0.3 and 0.1 respectively.
For solving the influence distribution among the nodes under different topics, in a DAG algorithm, the hidden Dirichlet distribution model is used for calculating the influence topic vector among the nodes.
All documents (namely, historical information propagated by all nodes in a preset time period) form a document set, and the document set is input into the hidden Dirichlet distribution model as a whole, so that the topics of all the documents can be guaranteed to be the same. Because in the model, the theme is automatically generated according to the input document, when the input documents are different, the last generated theme may be different, so that all documents need to be input as a whole, and thus it can be ensured that the theme of each node generated finally is the same, and a concrete model of the theme vector between the nodes is solved, as shown in fig. 11, X in the model 110 represents a node; z represents a topic; w represents a word in the document; yi represents a neighboring node of the node X; α represents the topic distribution of the document (i.e., the influence topic vector); beta represents the topic distribution of words (i.e. the information topic vector of the social network). Each document in the model 110 corresponds to a document between the node X and the adjacent node Y, and the distribution of influence between the nodes under different subjects can be obtained by learning joint probability distribution parameters by a gibbs sampling learning method. And finally, outputting results of the model as alpha and beta, wherein the theme distribution alpha of the document is the distribution of the influence of the node Yi on the node X under different themes.
When solving the influence topic vector in the model 110, the number of topics needs to be set, and how good the effect is when different topic numbers are measured intentionally through the confusion, the smaller the confusion is, the better the effect is, so the number of topic vectors between nodes can be obtained through the confusion comparison.
The relationship between the nodes is modeled into a network topology diagram with topic awareness, that is, each edge corresponds to two different topic vectors, for example, fig. 12 is a schematic diagram of a composition structure of a network topology diagram according to an embodiment of the present application, as shown in fig. 12, each edge in the diagram corresponds to two vectors, for example, an edge between a node U1 and a node U2 in the diagram corresponds to two vectors (0, 0.5, 0.5) and (0.3, 0.5, 0.2), which respectively represent that the influence probabilities of a node U2 on a node U1 under three topics are 0, 0.5, and the influence probability of a node U1 on a node U2 under three topics is 0.3, 0.5, and 0.2.
Influence Neig between adjacent nodes based on subjectu,vThe calculation method of (2) is shown in equation (6):
Figure BDA0001887781560000241
as can be seen from equation (6), the influence force Neig between adjacent nodes is calculatedu,vWhen the influence is large or small, the influence under different subjects needs to be considered. In the case of a number of subjects of k, Neigu,vDenotes the influence of node u on its neighbor node v, λzRepresenting the probability of the propagation information being topic Z (i.e. the probability of topic Z being propagated in the social network),
Figure BDA0001887781560000242
representing the probability of influence of node u on node v under topic Z. Then, in the case where the influence topic vector and the information topic vector between the nodes are known, it is easy to calculate the influence between the adjacent nodes according to the formula (6).
When the mutual influence between two nodes is calculated, the influence is reduced as the distance between the nodes is increased. Therefore, when the influence of each node is calculated, the nodes with smaller influence can be removed, and only the influence of the nodes with influence within a certain threshold value in the social network is calculated. By removing points with small influence, the network topological graph can be reduced, and the calculation efficiency is improved.
For example, for the node v, a width-first search algorithm is used to calculate all nodes whose influence on the node v is within a preset threshold from among all nodes n (v) on the path where the node v is located, and those nodes whose influence on the node v is smaller are removed, so as to form a subject-based TDAG (v) graph (i.e., the first TDAG graph) with the node v as a center node. After obtaining the tdag (v) graph of the node v, the node v is considered to be influenced only by the nodes in the tdag (v) graph. Therefore, when the influence of the node v in the social network is calculated, the influence of the node in the tdag (v) graph in the social network only needs to be calculated.
Preferably, the algorithm that generates the TDAG map: for each node in the network topology map, for example, taking the node v as an example, the influence of all nodes connected to the node v on the node v is calculated, and when the influence is greater than a threshold value, the corresponding node is retained to form a TDAG map taking the node v as a center node. And calculating all nodes by a breadth-first search algorithm to obtain the TDAG graph of each node. For example, as shown in fig. 13, a TDAG map 130 is obtained with the node v as a center node.
After the TDAG graphs of all the nodes are obtained, the influence inf (u) of the node u is calculated according to the following formula (7), wherein the node u is any node in the social network.
inf(u)＝∑v∈infset(u)αv(u) (7)
Where, infset (u) represents the set of all TDAG maps where the node u is located, and α v (u) is the influence of the node u on the node v. Here, α v (u) actually indicates the influence of the node u on the center node v in the TDAG map corresponding to the center node v.
The algorithm for calculating α v (u) is the sum of the influence of all paths from node u to node v. For all adjacent nodes x and wx of the node u, the influence of the node u on the node x is wx × fx, and fx is the influence of the node x on v, so that the influence of the node u on the node v is wx × fx.
When selecting a node, when one node s is selected as an initial node (i.e., the propagation node), the influence of each of the other nodes needs to be recalculated. However, when updating the influence of the nodes, it is not necessary to perform calculation on all the nodes, which makes the algorithm inefficient. When the influence of the node is updated, only the influence of the node in the TDAG graph with the node in Infset(s) as the root node in the social network needs to be updated, wherein Infset(s) is a set of all the TDAG graphs where the node s is located.
For all nodes in the tdag (v) graph, when a certain node s is selected as an initial node, if the greedy algorithm is used for calculation, the influence of all nodes except the node s on the corresponding central node v is low in calculation efficiency. In fact, in the TDAG graph, the influence between two nodes has a linear relationship. E.g. in a social network of recalculating node uWhen the influence is exerted, only the linear relation coefficient of the node u and the activated probability ap (u) of the node u are obtained through calculation; wherein, the linear coefficient A of the node uv(u) the coefficient is actually determined based on the influence of the root node u corresponding to the initial node s on the central node of each located TDAG graph, that is, in fact, the influence of the root node u on the central node of each located TDAG graph has a linear relationship;
when calculating ap (u), it can be determined according to the following equation (8):
ap(u)＝∑ap(v)*w(v,u) (8)
in the formula, ap (v) represents the activation probability of the central node v of the TDAG graph after deleting the initial node s to which the node u belongs; w (v, u) represents the influence of the central node v of the TDAG graph after the initial node s is deleted to which the node u belongs on the node u; in general, if node v is in the initial node set, that is, if node v has been selected as the initial node, node v is activated with a probability of 1, i.e., ap (v) ═ 1.
In calculating Av(u) and ap (u), the incremental influence of node u, Δ Infv (u), may then be calculated according to equation (9):
ΔInfv(u)＝Av(u)*(1-ap(u)) (9)
after calculating the influence increment Δ infv (u) of the node u, the influence inf (u) of the root node u in the social network is re-determined according to the following formula (10):
Inf(u)′＝Inf(u)+ΔInfv(u) (10)
wherein inf (u)' represents the influence of the updated node u in the social network, inf (u) represents the influence of the node u in the social network before updating, and Δ infv (u) represents the influence increment of the node u in the social network.
And when the next initial node is selected, selecting the node with the maximum influence after updating. Preferably, in order to reduce the amount of computation, for example, after the node s is selected as the initial node, the influence of the root node on the path of the node s in Infset(s) in the social network is updated. And then selecting the node with the largest influence from the updated influence of the root node in the social network and the influence of all other nodes except the root node and the initial node in the social network before updating as the next propagation node.
Repeating the above process until p initial nodes are selected, wherein the algorithm is as follows: firstly, generating a TDAG graph of each node according to a TDAG graph generation algorithm; and then calculating the influence of each node in the graph in the social network in the TDAG graph of each node, and selecting the initial node with the maximum influence from the TDAG graph. After an initial node is selected, calculating the influence of a root node on a path where the initial node is located in a TDAG graph after one node is selected as the initial node according to the influence updating algorithm, and selecting a node with the largest influence as a next propagation node based on the influence of the updated root node in the social network and the influence of all other nodes except the root node and the initial node in the social network before updating. And repeating the steps p-1 times to select p initial nodes.
After p initial nodes are selected, the initial nodes are used as an initial node set. These nodes (i.e., users) can act as initial propagators of advertisements or public opinions, thereby achieving the effect of affecting more users.
The influence maximization algorithm based on the user preference, namely the DAG algorithm, is used for solving the problem that influence results are limited due to the fact that only the influence of the user is considered when the IC model is adopted to select the user with the influence, influence of other advertisements or opposition public opinion on the influence of the user is not considered, and one side is considered. Adopting a DAG algorithm, mapping all users and relations among the users into a network topological graph, wherein one edge identifies the relation among the users, one node represents one user in the network topological graph, and when the influence of each node in a social network is calculated, firstly calculating influence topic vectors among the nodes, wherein the influence topic vectors among the nodes are used for measuring the influence of the nodes under different topics, and each vector represents the influence of the nodes under different topics; then, based on the information topic vector of the social network and the influence topic vector of each node, the influence of each node in the social network is determined, so that the influence of each node in the social network is more practical, and then when the node with the largest influence is selected as a propagation node based on the influence of each node in the social network, the selected propagation node can be received by more users in the social network when the information is propagated, so that the selected propagation node can exert the largest propagation effect.
Based on the foregoing embodiments, the present application provides an apparatus for determining user influence, where the apparatus includes modules and units included in the modules, and the modules may be implemented by a processor in a computer device; of course, the implementation can also be realized through a specific logic circuit; in implementation, the processor may be a Central Processing Unit (CPU), a Microprocessor (MPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), or the like.
Fig. 14A is a schematic structural diagram of a device for determining influence of a user according to an embodiment of the present application, and as shown in fig. 14A, the device 1400 includes an information obtaining module 1401, a vector determining module 1402, and an influence determining module 1403, where:
an information obtaining module 1401, configured to obtain history information propagated by Q1 nodes in a social network within a preset time period, where Q1 is an integer greater than or equal to 2;
a vector determination module 1402, configured to determine topic information of the social network and influence topic information of each node based on the history information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
an influence determination module 1403 configured to determine influence of each node in the social network based on the topic information of the social network and the influence topic information of each node.
In other embodiments, the influence determination module 1403 includes:
the inter-node influence determining unit is configured to determine the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
a first TDAG map obtaining unit configured to obtain a first TDAG map composed of each node as a central node and M associated nodes, where the associated nodes are nodes whose influence on the central node is greater than a preset threshold, and M is an integer greater than or equal to 0;
a node influence determining unit configured to determine influence of each node in the social network based on the Q1 first TDAG graphs and influence of each node on a corresponding neighbor node.
In other embodiments, the node influence determination unit includes:
a selecting subunit, configured to select, based on the Q1 first TDAG graphs, Q2 first TDAG graphs to which a node i belongs, where the node i is any one of the Q1 nodes, and Q2 is equal to or less than Q1;
a first determining subunit configured to determine, in the Q2 first TDAG graphs, an influence of each node i on each central node based on an influence of each node on a corresponding neighbor node;
a second determining subunit configured to determine, in the Q2 first TDAG graphs, an influence of the node i in the social network based on the influence of the node i on each central node.
In other embodiments, as shown in fig. 14B, the apparatus 1400 further comprises:
the associated node searching module 1404 is configured to search, by taking each node as a central node, M associated nodes, on a path where the central node is located, of which the influence on the central node is greater than a preset threshold value, based on the influence of each central node on a corresponding neighbor node;
a first TDAG map determining module 1405, configured to compose, with the central node and the M associated nodes, a first TDAG map corresponding to the central node.
In other embodiments, the first TDAG map determination module 1405 is configured to:
each node is taken as a central node, and the influence of the associated node of the central node on the central node is determined step by step on the path where the central node is located based on the influence of each central node on the corresponding neighbor node;
if the influence of the current associated node on the central node is greater than the preset threshold, determining the current associated node as a node in a corresponding first TDAG graph, and continuously determining the influence of the associated node of the next level on the path where the current associated node is located on the central node;
if the influence of the current associated node on the central node is smaller than or equal to the preset threshold, ending the search on the path where the current associated node is located, and gradually determining the influence of the associated node of the central node on other paths where the central node is located to obtain the M associated nodes.
In other embodiments, as shown in fig. 14C, the apparatus 1400 further comprises:
a propagation node selection module 1406 configured to select a node with the largest influence from the nodes as a propagation node based on the influence of each node in the social network;
an updating module 1407 configured to update influence of a root node on a path where the propagation node is located in the social network based on the Q1 first TDAG graphs;
the propagation node selecting module 1406 is further configured to select a node with the largest influence from the updated influence of the root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network as a next propagation node.
In other embodiments, the update module 1407 includes:
a first selecting unit configured to select, based on the Q1 first TDAG maps, Q3 first TDAG maps to which the propagation node belongs, wherein Q3 is equal to or less than Q1;
a root node determination unit configured to determine a root node on a path where the propagation node is located from the Q3 first TDAG maps;
a second selecting unit configured to select, based on the Q1 first TDAG maps, Q4 first TDAG maps to which the root node belongs, wherein Q4 is less than or equal to Q1;
a TDAG map processing unit configured to delete the propagation node and a path connecting the propagation node from the Q4 first TDAG maps to obtain Q5 second TDAG maps, wherein Q5 is equal to or less than Q4;
an updating unit configured to update, in the Q5 second TDAG graphs, influence of each node on a corresponding neighbor node in the social network.
In other embodiments, the update unit includes:
a third determining subunit configured to determine an activation probability of the root node;
a fourth determining subunit, configured to determine, in each of the second TDAG maps, an influence of each node on each central node based on the influence of each node on a corresponding neighbor node;
a fifth determining subunit configured to determine an influence increment of the root node in the social network based on the activation probability of the root node and the influence of the root node on a central node of each of the second TDAG maps;
and the updating subunit is configured to update the influence of the root node in the social network based on the influence increment and the influence of the root node in the social network before updating.
In other embodiments, the third determining subunit is configured to:
determining an activation probability of a central node of each of the second TDAG maps;
determining an influence of a center node of each of the second TDAG graphs on the root node;
and determining the activation probability of the root node based on the activation probability of the central node of each second TDAG graph and the influence of the central node corresponding to the activation probability on the root node.
The above description of the apparatus embodiments, similar to the above description of the method embodiments, has similar beneficial effects as the method embodiments. For technical details not disclosed in the embodiments of the apparatus of the present application, reference is made to the description of the embodiments of the method of the present application for understanding.
It should be noted that, in the embodiment of the present application, if the method for determining the influence of the user is implemented in the form of a software functional module and is sold or used as a standalone product, it may also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially implemented or portions thereof contributing to the related art may be embodied in the form of a software product stored in a storage medium, and including several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read Only Memory (ROM), a magnetic disk, or an optical disk. Thus, embodiments of the present application are not limited to any specific combination of hardware and software.
Correspondingly, an embodiment of the present application provides a computer device, fig. 15 is a schematic diagram of a hardware entity of the computer device in the embodiment of the present application, and as shown in fig. 15, the hardware entity of the computer device 1500 includes: comprising a memory 1501 and a processor 1502, said memory 1501 storing a computer program operable on the processor 1502, said processor 1502 realizing the steps in the method of determining user influence provided in the above embodiments when executing said program.
The Memory 1501, which is configured to store instructions and applications executable by the processor 1502 and may also buffer data (e.g., image data, audio data, voice communication data, and video communication data) to be processed or already processed by the processor 1502 and modules in the computer device 1500, may be implemented by a FLASH Memory (FLASH) or a Random Access Memory (RAM).
Correspondingly, embodiments of the present application provide a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps in the method for determining user influence provided in the above embodiments.
Here, it should be noted that: the above description of the storage medium and device embodiments is similar to the description of the method embodiments above, with similar advantageous effects as the method embodiments. For technical details not disclosed in the embodiments of the storage medium and apparatus of the present application, reference is made to the description of the embodiments of the method of the present application for understanding.
It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. It should be understood that, in the various embodiments of the present application, the sequence numbers of the above-mentioned processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present application. The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described device embodiments are merely illustrative, for example, the division of the unit is only a logical functional division, and there may be other division ways in actual implementation, such as: multiple units or components may be combined, or may be integrated into another system, or some features may be omitted, or not implemented. In addition, the coupling, direct coupling or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection between the devices or units may be electrical, mechanical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units; can be located in one place or distributed on a plurality of network units; some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be separately regarded as one unit, or two or more units may be integrated into one unit; the integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: various media that can store program codes, such as a removable Memory device, a Read Only Memory (ROM), a magnetic disk, or an optical disk.
Alternatively, the integrated units described above in the present application may be stored in a computer-readable storage medium if they are implemented in the form of software functional modules and sold or used as independent products. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially implemented or portions thereof contributing to the related art may be embodied in the form of a software product stored in a storage medium, and including several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a removable storage device, a ROM, a magnetic or optical disk, or other various media that can store program code.
The above description is only for the embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (9)

1. A method of determining user influence, the method comprising:
acquiring historical information propagated by Q1 nodes in a social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
determining topic information of the social network and influence topic information of each node based on historical information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
determining the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
acquiring a first TDAG (time domain access graph) composed of each node serving as a central node and M associated nodes, wherein the associated nodes are nodes with influence on the central node larger than a preset threshold value, and M is an integer larger than or equal to 0;
determining the influence of each node in the social network based on the Q1 first TDAG graphs and the influence of each node on the corresponding neighbor nodes;
based on the influence of each node in the social network, selecting the node with the largest influence as a propagation node;
updating the influence of a root node on a path where the propagation node is located in the social network based on the Q1 first TDAG graphs;
wherein the process of composing the first TDAG map comprises:
taking each node as a central node, and searching M associated nodes with influence on the central node larger than a preset threshold value on a path where the central node is located based on the influence of each central node on the corresponding neighbor node, wherein the searching comprises the following steps:
each node is taken as a central node, and the influence of the associated node of the central node on the central node is determined step by step on the path where the central node is located based on the influence of each central node on the corresponding neighbor node;
if the influence of the current associated node on the central node is greater than the preset threshold, determining the current associated node as a node in a corresponding first TDAG graph, and continuously determining the influence of the associated node of the next level on the path where the current associated node is located on the central node;
if the influence of the current associated node on the central node is smaller than or equal to the preset threshold, ending the search on the path where the current associated node is located, and gradually determining the influence of the associated node of the central node on other paths where the central node is located to obtain the M associated nodes;
and forming a first TDAG graph corresponding to the central node by using the central node and the M associated nodes.
2. The method of claim 1, wherein determining the influence of each node in the social network based on the Q1 first TDAG maps and the influence of each node on the corresponding neighbor nodes comprises:
based on Q1 first TDAG graphs, Q2 first TDAG graphs to which a node i belongs are selected, wherein the node i is any one of Q1 nodes, and Q2 is less than or equal to Q1;
in the Q2 first TDAG graphs, determining the influence of the node i on each central node based on the influence of each node on the corresponding neighbor node;
in the Q2 first TDAG graphs, determining the influence of the node i in the social network based on the influence of the node i on each central node.
3. The method according to claim 1, wherein after updating the influence of the root node on the path of the propagation node in the social network based on the Q1 first TDAG maps, the method further comprises:
and selecting a node with the largest influence from the updated influence of the root node in the social network and the influence of the nodes except the root node and the propagation node in the Q1 nodes before updating in the social network as a next propagation node.
4. The method according to claim 1, wherein the updating the influence of the root node on the path of the propagation node in the social network based on the Q1 first TDAG maps comprises:
based on Q1 first TDAG graphs, Q3 first TDAG graphs belonging to the propagation node are selected, wherein Q3 is less than or equal to Q1;
determining a root node on a path where the propagation node is located from the Q3 first TDAG graphs;
based on Q1 first TDAG graphs, Q4 first TDAG graphs belonging to the root node are selected, wherein Q4 is less than or equal to Q1;
deleting the propagation nodes and paths connecting the propagation nodes from the Q4 first TDAG graphs to obtain Q5 second TDAG graphs, wherein Q5 is less than or equal to Q4;
in the Q5 second TDAG graphs, the influence of the root node in the social network is updated based on the influence of each node on the corresponding neighbor node.
5. The method of claim 4, wherein the updating, in the Q5 second TDAG graphs, the influence of the root node in the social network based on the influence of each node on the corresponding neighbor node comprises:
determining an activation probability of the root node;
determining the influence of each node on each central node based on the influence of each node on the corresponding neighbor node in each second TDAG graph;
determining an influence increment of the root node in the social network based on the activation probability of the root node and the influence of the root node on a central node of each of the second TDAG graphs;
updating the influence of the root node in the social network based on the influence delta and the influence of the root node in the social network before updating.
6. The method of claim 5, wherein determining the activation probability of the root node comprises:
determining an activation probability of a central node of each of the second TDAG maps;
determining an influence of a center node of each of the second TDAG graphs on the root node;
and determining the activation probability of the root node based on the activation probability of the central node of each second TDAG graph and the influence of the central node corresponding to the activation probability on the root node.
7. An apparatus for determining user influence, comprising:
the information acquisition module is configured to acquire historical information propagated by Q1 nodes in the social network within a preset time period, wherein Q1 is an integer greater than or equal to 2;
a vector determination module configured to determine topic information of the social network and influence topic information of each of the nodes based on the history information of the Q1 nodes; the topic information is used for representing the probability that N topics are respectively transmitted in the social network, and the influence topic information is used for representing the influence probability of each node on the corresponding neighbor node under the N topics;
the inter-node influence determining unit is configured to determine the influence of the corresponding node on the corresponding neighbor node based on the topic information of the social network and the influence topic information of each node;
the propagation node selection module is configured to select a node with the largest influence from the nodes as a propagation node based on the influence of each node in the social network;
an updating module configured to update influence of a root node on a path where the propagation node is located in the social network based on the Q1 first TDAG graphs;
a first TDAG map obtaining unit configured to obtain a first TDAG map composed of each node as a central node and M associated nodes, where the associated nodes are nodes whose influence on the central node is greater than a preset threshold, and M is an integer greater than or equal to 0;
a node influence determination unit configured to determine influence of each node in the social network based on the Q1 first TDAG maps and influence of each node on a corresponding neighbor node;
wherein the first TDAG map determination module is configured to:
each node is taken as a central node, and the influence of the associated node of the central node on the central node is determined step by step on the path where the central node is located based on the influence of each central node on the corresponding neighbor node;
if the influence of the current associated node on the central node is greater than the preset threshold, determining the current associated node as a node in a corresponding first TDAG graph, and continuously determining the influence of the associated node of the next level on the path where the current associated node is located on the central node;
if the influence of the current associated node on the central node is smaller than or equal to the preset threshold, ending the search on the path where the current associated node is located, and gradually determining the influence of the associated node of the central node on other paths where the central node is located to obtain the M associated nodes;
and forming a first TDAG graph corresponding to the central node by using the central node and the M associated nodes.
8. A computer device comprising a memory and a processor, the memory storing a computer program operable on the processor, wherein the processor when executing the program performs the steps in the method of determining user influence of any of claims 1 to 6.
9. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method for determining a user influence of any one of claims 1 to 6.
CN201811456146.1A 2018-11-30 2018-11-30 Method, device, equipment and storage medium for determining user influence Active CN109558540B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811456146.1A CN109558540B (en) 2018-11-30 2018-11-30 Method, device, equipment and storage medium for determining user influence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811456146.1A CN109558540B (en) 2018-11-30 2018-11-30 Method, device, equipment and storage medium for determining user influence

Publications (2)

Publication Number Publication Date
CN109558540A CN109558540A (en) 2019-04-02
CN109558540B true CN109558540B (en) 2021-10-29

Family

ID=65868232

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811456146.1A Active CN109558540B (en) 2018-11-30 2018-11-30 Method, device, equipment and storage medium for determining user influence

Country Status (1)

Country Link
CN (1) CN109558540B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111767473B (en) * 2020-07-30 2023-11-14 腾讯科技（深圳）有限公司 Object selection method and computer readable storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104008182A (en) * 2014-06-10 2014-08-27 盐城师范学院 Measuring method of social network communication influence and measure system thereof
CN107341270A (en) * 2017-07-28 2017-11-10 东北大学 Towards the user feeling influence power analysis method of social platform
CN107566249A (en) * 2016-10-28 2018-01-09 中国科学院计算技术研究所 A kind of method for training the model for predicting social network user forwarding message

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6580911B2 (en) * 2015-09-04 2019-09-25 Kddi株式会社 Speech synthesis system and prediction model learning method and apparatus thereof
US20170300823A1 (en) * 2016-04-13 2017-10-19 International Business Machines Corporation Determining user influence by contextual relationship of isolated and non-isolated content
CN107818514B (en) * 2016-09-12 2022-01-14 腾讯科技（深圳）有限公司 Method, device and terminal for controlling information propagation of online social network
CN108363752B (en) * 2018-01-30 2021-01-05 中国人民解放军战略支援部队信息工程大学 User social influence analysis method based on microblog propagation scale prediction

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104008182A (en) * 2014-06-10 2014-08-27 盐城师范学院 Measuring method of social network communication influence and measure system thereof
CN107566249A (en) * 2016-10-28 2018-01-09 中国科学院计算技术研究所 A kind of method for training the model for predicting social network user forwarding message
CN107341270A (en) * 2017-07-28 2017-11-10 东北大学 Towards the user feeling influence power analysis method of social platform

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
社交网络中基于竞争的影响力最大化研究;王浩然;《中国优秀硕士学位论文全文数据库 信息科技辑》;20170315(第3期);I139-374 *

Also Published As

Publication number Publication date
CN109558540A (en) 2019-04-02

Similar Documents

Publication Publication Date Title
Qi et al. Finding all you need: web APIs recommendation in web of things through keywords search
Ma et al. Distributed optimization with arbitrary local solvers
Regis et al. Combining radial basis function surrogates and dynamic coordinate search in high-dimensional expensive black-box optimization
CN110781407B (en) User tag generation method, device and computer readable storage medium
Panti Invariant measures in free MV-algebras
TW201939366A (en) Recommendation system construction method and device
JP7457125B2 (en) Translation methods, devices, electronic equipment and computer programs
CN110489558A (en) Polymerizable clc method and apparatus, medium and calculating equipment
CN109886699A (en) Activity recognition method and device, electronic equipment, storage medium
CN112581185B (en) Advertisement user gender and age estimation method, system, computer and storage medium
CN113722510A (en) Knowledge graph complex problem generation method and system based on graph neural network
CN109558540B (en) Method, device, equipment and storage medium for determining user influence
JP2013190838A (en) Information anonymization system, information loss determination method, and information loss determination program
KR20210043894A (en) Electronic apparatus and method of providing sentence thereof
CN116340648A (en) Knowledge graph attention network recommendation method based on graph collaborative filtering
JP2019200737A (en) Information processing device, information processing method, and information processing program
Jahandideh et al. Allocating duplicate copies for IoT data in cloud computing based on harmony search algorithm
Verstraete et al. Consistent iterative algorithm for stochastic dynamic traffic assignment with a stable route set
CN117061365B (en) Node selection method, device, equipment and readable storage medium
JP2024502400A (en) Automatic depiction and extraction of tabular data in portable document formats using graph neural networks
CN113535912A (en) Text association method based on graph convolution network and attention mechanism and related equipment
CN113269603A (en) Recommendation system-oriented space-time graph convolution method and system
CN110827078B (en) Information recommendation method, device, equipment and storage medium
CN117273018A (en) Information processing method, information processing apparatus, computer device, and storage medium
CN106447504A (en) Friend making method and apparatus

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant