CN113221023B - Information pushing method and device - Google Patents
Information pushing method and device Download PDFInfo
- Publication number
- CN113221023B CN113221023B CN202110563315.7A CN202110563315A CN113221023B CN 113221023 B CN113221023 B CN 113221023B CN 202110563315 A CN202110563315 A CN 202110563315A CN 113221023 B CN113221023 B CN 113221023B
- Authority
- CN
- China
- Prior art keywords
- users
- user
- value
- moment
- state
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
Abstract
The application provides an information pushing method and device, wherein the method comprises the following steps: acquiring initial association relations of a plurality of users according to historical data of the plurality of users; for each user, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to initial association relations of a plurality of users; wherein the first time is earlier than the second time; classifying a plurality of users according to a plurality of groups of state values to obtain a plurality of groups of users; and generating recommendation information according to the historical data of each user in each group of users, and pushing the recommendation information to each user. Compared with the prior art, the method and the device for classifying the users have the advantages that the users are classified according to the state values of other users when each user is in an abnormal state, so that multiple groups of users are obtained, recommendation information is pushed to the users, the rationality of the classification method is improved, and the accuracy of information pushing is improved.
Description
Technical Field
The present application relates to the field of computer application technologies, and in particular, to an information pushing method and apparatus.
Background
By detecting the community structure in the bank customer relationship network, a plurality of important reference information can be mined. In general, clients included in the same community often play the same or similar roles in the network, and clients included in different communities are greatly different in business transaction. According to the community division result of the whole network, functional module division can be carried out on all clients, and further prediction of business requirements of the clients by banks is facilitated.
The existing customer classification method is that community division is carried out through a community detection technology, a corresponding customer relationship network is constructed according to customer basic data and business data, then customer classification is carried out through the community detection technology, classification results are obtained, and business demands of customers are predicted according to the classification results.
However, in the method for classifying clients by using the community detection technology, the edge betweenness between clients is calculated, and the connection line between the clients with high edge betweenness is removed, and as each time the connection edge is deleted, the betweenness of each of the remaining edges in the client relation network is required to be recalculated, the condition of repeatedly calculating the shortest path can occur, so that the time complexity of the method is too high; once the connection between two clients is split, it is always in two different communities and cannot be adjusted subsequently. Therefore, the existing scheme has the problem of inaccurate information pushing of the clients caused by unreasonable client classification methods.
Disclosure of Invention
The embodiment of the application provides an information pushing method and device, which are used for solving the problem of inaccurate information pushing of clients caused by unreasonable client classification methods in the prior art.
A first aspect of the present application provides an information pushing method, the method including:
acquiring initial association relations of a plurality of users according to historical data of the plurality of users;
for each user, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at a first moment according to initial association relations of the plurality of users; wherein the first time is earlier than the second time;
classifying the plurality of users according to the plurality of groups of state values to obtain a plurality of groups of users;
and generating recommendation information according to the historical data of each user in each group of users, and pushing the recommendation information to each user.
In an optional implementation manner, for each user, according to the initial association relationship of the plurality of users, a set of state values of other users at a second moment when the user is abnormal at a first moment is obtained through calculation, including:
setting the state value of the ith user from the first moment to the second moment as a first state value indicating an abnormal state for the ith user;
circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating moment information until calculating to obtain the state value of other users at the second moment;
The state values of the other users at the second moment are a group of state values of the other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users.
In an alternative embodiment, the state value of the other user at the next moment is obtained through calculation according to a first formula, wherein the first formula specifically includes:
wherein x is i (t) represents the state value of the ith user at time t, ε represents the coupling strength, f (x) i (t)) represents the value of the relationship function of xi (t), where the relationship function f (x) =4x (1-x), ai j The method comprises the steps of representing initial association relation between an ith user and a jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of connected edges connected to the ith user in the initial association relation.
In an optional embodiment, the classifying the plurality of users according to the plurality of sets of state values to obtain a plurality of sets of users includes:
calculating the dissimilarity value between any two users according to the multiple groups of state values;
classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users;
Calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result;
and obtaining the plurality of groups of users according to the classification result corresponding to the modularity value.
In an alternative embodiment, the calculating the dissimilarity value between any two sets of state values includes;
obtaining Euclidean distances of the arbitrary two sets of state values according to the arbitrary two sets of state values;
and obtaining the dissimilarity value between any two users according to the Euclidean distance.
In an optional embodiment, the obtaining the plurality of groups of users according to the modularity value includes:
and obtaining the plurality of groups of users according to the classification result corresponding to the maximum module degree value in the module degree values.
A second aspect of the present application provides an information pushing apparatus, the apparatus comprising:
the acquisition module is used for acquiring initial association relations of a plurality of users according to historical data of the plurality of users;
the processing module is used for calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to the initial association relation of the plurality of users for each user; wherein the first time is earlier than the second time; classifying the plurality of users according to the plurality of groups of state values to obtain a plurality of groups of users;
And the pushing module is used for generating recommendation information according to the historical data of each user in each group of users and pushing the recommendation information to each user.
In an optional implementation manner, the processing module is specifically configured to set, for an ith user, a state value of the ith user from the first moment to the second moment to a first state value indicating that the ith user is in an abnormal state; circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating moment information until calculating to obtain the state value of other users at the second moment; the state values of the other users at the second moment are a group of state values of the other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users.
In an optional implementation manner, the processing module is specifically configured to calculate and obtain a state value of the other user at a next moment according to a first formula, where the first formula specifically includes:
Wherein x is i (t) represents the state value of the ith user at time t, ε represents the coupling strength, f (x) i (t)) represents x i The value of the relation function of (t), wherein the relation function f (x) =4x (1-x), a ij The method comprises the steps of representing initial association relation between an ith user and a jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of connected edges connected to the ith user in the initial association relation.
In an optional implementation manner, the processing module is specifically configured to calculate a dissimilarity value between any two users according to the multiple sets of state values; classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users; calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result; and obtaining the plurality of groups of users according to the classification result corresponding to the modularity value.
In an optional implementation manner, the processing module is specifically configured to obtain the euclidean distance between the arbitrary two sets of state values according to the arbitrary two sets of state values; and obtaining the dissimilarity value between any two users according to the Euclidean distance.
In an optional implementation manner, the processing module is specifically configured to obtain the multiple groups of users according to the classification result corresponding to the largest module degree value among the module degree values.
A third aspect of the present application provides an electronic apparatus comprising: a processor and a memory;
the memory is used for storing a computer program;
the processor is configured to invoke and run a computer program stored in the memory for performing the method according to the first aspect.
A fourth aspect of the application provides a computer-readable storage medium storing a computer program for causing a computer to perform the method according to the first aspect.
A fifth aspect of the application provides a computer program product comprising a computer program which, when executed by a processor, implements the method according to the first aspect.
The information pushing method and device provided by the embodiment of the application comprise the following steps: acquiring initial association relations of a plurality of users according to historical data of the plurality of users; for each user, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to initial association relations of a plurality of users; wherein the first time is earlier than the second time; classifying a plurality of users according to a plurality of groups of state values to obtain a plurality of groups of users; and generating recommendation information according to the historical data of each user in each group of users, and pushing the recommendation information to each user. Compared with the prior art, the method and the device have the advantages that according to the state value of other users at the second moment when each user is in the abnormal state at the first moment, the influence on the other users when one user in the multiple users is in the abnormal state can be obtained, then the multiple users are classified according to the obtained multiple groups of state values, multiple groups of users are obtained, the rationality of the classification method is improved, and then recommendation information is pushed to the users according to the classification result of the multiple users obtained by the classification method, so that the accuracy of information pushing is improved.
Drawings
In order to more clearly illustrate the application or the technical solutions of the prior art, the following description of the embodiments or the drawings used in the description of the prior art will be given in brief, it being obvious that the drawings in the description below are some embodiments of the application and that other drawings can be obtained from them without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of an application scenario of an information push method according to an embodiment of the present application;
fig. 2 is a schematic flow chart of an information pushing method according to an embodiment of the present application;
fig. 3 is a schematic flow chart of another information pushing method according to an embodiment of the present application;
fig. 4 is a flowchart of another information pushing method according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an information pushing device according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
By detecting the community structure in the bank customer relationship network, a plurality of important reference information can be mined. In general, clients included in the same community often play the same or similar roles in the network, and clients included in different communities are greatly different in business transaction. According to the community division result of the whole network, functional module division can be carried out on all clients, and further prediction of business requirements of the clients by banks is facilitated. The existing customer classification method is that community division is carried out through a community detection technology, a corresponding customer relationship network is constructed according to customer basic data and business data, then customer classification is carried out through the community detection technology, classification results are obtained, and business demands of customers are predicted according to the classification results.
However, in the method for classifying clients by using the community detection technology, the edge betweenness between clients is calculated, and the connection line between the clients with high edge betweenness is removed, and as each time the connection edge is deleted, the betweenness of each of the remaining edges in the client relation network is required to be recalculated, the condition of repeatedly calculating the shortest path can occur, so that the time complexity of the method is too high; once the connection between two clients is split, it is always in two different communities and cannot be adjusted subsequently. Therefore, the existing scheme has the problem of inaccurate information pushing caused by unreasonable client classification method.
In order to solve the problems, the application provides an information pushing method and device, which are used for classifying a plurality of users according to the state values of other users when each user is in an abnormal state to obtain a plurality of groups of users, so as to push recommended information to the users, improve the rationality of the classification method and further improve the accuracy of information pushing.
First, the terms involved in the present application will be explained:
complex network theory: the method is a unified and simple expression mode provided by the pointer to the client relationship, namely the client relationship is abstracted into a network, and the network consists of nodes and connecting edges. Wherein clients are represented by nodes and business connections between clients are represented by links. The complexity of a network mainly refers to the complexity of the network structure, the diversity of network connections, and the interaction and connection laws between individuals hidden in the network.
Community structure: it is meant that all clients in a client relationship network are divided into a plurality of different communities, each community consisting of a group of clients. The clients in the same community are closely connected, and the clients in different communities are sparsely connected.
Degree value of node: refers to the number of edges connected to the node.
Node distance: refers to the number of edges included in the shortest path between two nodes.
Network diameter: refers to the maximum value of the distance between any two nodes in the network.
The application scenario of the present application is described below.
Fig. 1 is an application scenario schematic diagram of an information push method provided by an embodiment of the present application. As shown in fig. 1, includes: terminal device 001 and server 002. The terminal device 001 obtains the historical data of a plurality of users input by the users, the terminal device 001 processes the user data, classifies the plurality of users to obtain a plurality of groups of users, wherein each group of users forms a community, the terminal device 001 can send the historical data or the classification result to the server 002 for storage, and in the subsequent use, the terminal device 001 can obtain the data from the server 002 for use and further analysis.
Optionally, the terminal device 001 obtains historical data input by a user and for a plurality of users, the terminal device 001 sends the historical data to the server 002, the server 002 processes the user data, classifies the plurality of users to obtain a plurality of groups of users, wherein each group of users forms a community, the server 002 sends the obtained results of the plurality of groups of users to the terminal device 001, the terminal device 001 displays the classification results, that is, the obtained plurality of groups of users, the server 002 can store the historical data or the classification results, and in subsequent use, the terminal device 001 can obtain the data from the server 002 for use and further analysis.
The terminal device may be a computer with a wireless transceiving function, a Virtual Reality (VR) terminal device, an augmented reality (augmented reality, AR) terminal device, a mobile phone (mobile phone), a tablet (pad), a wireless terminal in industrial control (industrial control), a wireless terminal in teleoperation (remote medical surgery), a wireless terminal in smart grid (smart grid), a wireless terminal in smart home (smart home), and the like.
In the embodiment of the present application, the device for implementing the information push function may be a terminal device, or may be a device capable of supporting the implementation of the function, for example, a chip system, and the device may be installed in the terminal device. In the embodiment of the application, the chip system can be composed of chips, and can also comprise chips and other discrete devices.
It should be noted that, the application scenario of the technical solution of the present application may be the scenario in fig. 1, but is not limited thereto, and may be applied to other scenarios where information pushing is required.
It may be understood that the information pushing method may be implemented by using the information pushing apparatus provided by the embodiment of the present application, where the information pushing apparatus may be part or all of a certain device, for example, the terminal device or a chip of the terminal device.
The following takes an information pushing device integrated with or installed with related execution codes as an example, and specific embodiments are used for describing the technical scheme of the embodiment of the present application in detail. The following embodiments may be combined with each other, and some embodiments may not be repeated for the same or similar concepts or processes.
Fig. 2 is a flow chart of an information pushing method provided by the embodiment of the application, and the embodiment relates to a specific information pushing process. As shown in fig. 2, the method includes:
s101, obtaining initial association relations of a plurality of users according to historical data of the plurality of users.
The historical data may include basic data of the user and service data of the user, where the basic data is used to identify the user, and the basic data may be identification data of the user, for example.
In the embodiment of the application, the initial association relationship between the users is established through the service data of the users, and for example, if the user 1 and the user 2 purchase the same product or transact the same service at the same time, the association relationship between the user 1 and the user 2 can be established.
The method for representing the initial association relation in the embodiment of the application is not limited, and the initial association relation can be represented by an adjacent matrix by way of example. Specifically, a user relationship network is established according to a complex network theory, one user node in the user relationship network represents one user, and the association relationship between the users is represented by the connection edges between the user nodes. Complex network theory is a tool for network research and analysis, and represents a network in the form of nodes and edges, and is exemplified by a user relationship network G (V, E) with N user nodes, wherein V= { V 1 ，v 2 ，...，v N E.e.V.times.V.represents a set of connected edges in the user relationship network, E ij ＝(v i ，v j )，e ij Representing user node v i And v j Between which are locatedThere is one continuous edge. Using adjacency matrix a to represent the connection relationship between user nodes:
wherein, when A ij When=1, it indicates the user node v i And v j Is connected by a connecting edge, when A ij When=0, it indicates the user node v i And v j There is no edge connection between them.
S102, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to initial association relations of a plurality of users for each user.
Wherein the first time is earlier than the second time.
In this step, after the terminal device obtains the initial association relationships of the plurality of users according to the historical data of the plurality of users, for each user, a set of state values of other users at the second moment when the user is abnormal at the first moment are obtained by calculation according to the initial association relationships of the plurality of users.
It can be understood that after the terminal device obtains the initial association relationships of the plurality of users, for example, the above-mentioned adjacency matrix, the state value of other users when each user is abnormal is calculated according to the adjacency matrix. For example, the state value of other users may be calculated according to the successive fault model when each user is abnormal, and when one user is abnormal in the successive fault model, the state value of other users may also be abnormal when the state value of other users is calculated.
Specifically, for each user, in order to make the user at the first moment in an abnormal state, adding a disturbance value to the state value of the user at the first moment to make the state value of the user greater than a state threshold value, and making the user in an abnormal state; at this time, the state value of the user is recorded, and the influence of the abnormality of the user on other users can be obtained by setting the state value of the user as a state threshold in the subsequent calculation; and calculating the state values of the other users when the user is abnormal, and if the state values of one or more of the other users are larger than the state threshold in the calculation process, namely the abnormality of the user causes the abnormality of the other users, obtaining the state values of the one or more users, setting the state values of the one or more users as the state threshold in the calculation process, so that the influence of the abnormality of the user on the other users can be obtained. By the method, when all the users are abnormal, the state values of other users are obtained.
The setting of the state threshold is not limited in the present application, and may be set to 1 by way of example.
Optionally, for the ith user, setting a state value of the ith user from the first moment to the second moment as a first state value indicating that the ith user is in an abnormal state; and circularly executing from the first moment, calculating and obtaining the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating moment information until calculating and obtaining the state value of other users at the second moment.
The state values of other users at the second moment are a group of state values of other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users.
Specifically, when calculating the state value of the ith user at the first moment, the state value of the ith user at the second moment is made to be the first state value by adding disturbance to the state value, the first state value is larger than the state threshold, the ith user is in an abnormal state at the second moment, and then the state values of other users except the ith user in the abnormal state of the ith user are calculated. It is known that, when the i-th user is in an abnormal state, that is, when the state value of the other user is calculated, the state value of the i-th user is greater than the state threshold value, so that the other user is in an abnormal state, at this time, in addition to calculating the state value of the i-th user in an abnormal state, the state value of the user in an abnormal state in the other users needs to be set as the state threshold value in the subsequent calculation, thereby more reasonably obtaining the influence on the state value of the other user when the certain user is in an abnormal state, and improving the rationality of user classification.
Further, when the ith user is in an abnormal state, the state values of other users at the next moment are obtained through calculation, and then loop iteration is carried out to obtain the state values of other users at the next moment when all the users are in the abnormal state. Thus, a plurality of sets of state values can be obtained.
The method for representing the obtained multiple sets of state values in the present application is not limited, and for example, a state matrix may be obtained by taking each set of state values as a column, for example, if there are N users, the size of the state matrix is n×n.
The method for obtaining the state value is not limited in the application, and the state value can be obtained by calculating a state value calculation formula in a coupling image grid model (Coupled Map Lattice, CML) in the successive fault model by way of example; optionally, calculating according to a first formula to obtain a state value of other users at a next moment, where the first formula specifically includes:
wherein x is i (t) represents the state value of the ith user at time t, ε represents the coupling strength, f (x) i (t)) represents x i The value of the relation function of (t), wherein the relation function f (x) =4x (1-x), a ij The method comprises the steps of representing initial association relation between an ith user and a jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of connected edges connected to the ith user in the initial association relation.
Specifically, for the relation function f (x), if 0 is less than or equal to x is less than or equal to 1, 0 is less than or equal to f (x) is less than or equal to 1, so that if the state value of the user at the time t+1 is between 0 and 1, the user is in a normal state, and if the state value of the user at the time t+1 is greater than 1, the user is in an abnormal state.
The value of the coupling strength epsilon is not limited in the application, and can be exemplified by 0.5.
In the embodiment of the application, in order to enable the ith user to be in an abnormal state at the t+1th moment, a disturbance R is applied to the first user at the t moment, wherein R is more than or equal to 1. Specifically, at the t-th moment, calculating a state value of the i-th user according to a second formula, where the second formula specifically includes:
specifically, disturbance is applied to the ith user at the t moment through a second formula, so that the ith user is in an abnormal state at the t+1 moment, and then the state values of other users are calculated through the first formula at the t+2 moment. The users directly connected to the ith user are affected by the state value of the ith user at the time t+1, and when the state values of the users are calculated through the first formula, the calculated result may exceed 1, so that an abnormal state of a new round of users is caused.
When calculating the state value of the ith user in an abnormal state, if any one of the other users is in the abnormal state at the next moment, setting the state value of the user in the abnormal state as 1 in the subsequent calculation, wherein the state value comprises the state value of the ith user and the state value of the user in the abnormal state influenced by the abnormal state of the ith user, so that the influence of the abnormal state of any one of the users on the states of the other users can be obtained more accurately, and the users can be classified more reasonably.
Further, the method is iterated for D/2 times, D is the diameter of the user relation network, and the state values of other users when the ith user is in an abnormal state can be obtained. Repeating the above steps, when each user in the user relation network is set to be in abnormal state in turn,the state values of other users result in a state matrix X of size N, wherein each column vector in the state matrix, e.g. the ith column vector X i And when the ith user is in an abnormal state, the state values of other users are represented after D/2 iterations.
Optionally, at the first moment, all users in the user relationship network are in a normal state 0; at a second moment, selecting a user and applying an external disturbance thereto; at a third moment, the state value of the user is calculated by using the second formula, the user is in an abnormal state at the moment and is set to be 1, and the state value of the user is always kept to be 1 in the subsequent moment. In addition, the neighbor users connected to the user are affected by the abnormal state, the state values of the neighbor users can be calculated through the first formula, some neighbor users of the user may be in the abnormal state, and the state values of the users are always maintained to be 1 in the following moments. In the above iterative manner, the number of iterations is set to D/2, where D is the diameter of the network. Each user in the user relationship network may be sequentially subjected to an external disturbance R to trigger an anomaly of the user relationship network once. By repeating the above steps, we can obtain an N×N state matrix X, each column vector X in the matrix i Representing the state value of each user after D/2 iterations when the ith user is in an abnormal state.
S103, classifying the plurality of users according to the plurality of groups of state values to obtain a plurality of groups of users.
In the embodiment of the application, the multiple groups of users are classified according to the service data to obtain the multiple groups of users, and the community division is completed.
The embodiment of the application does not limit the classification modes of a plurality of users according to the state values in the obtained state matrix, and the embodiment of the application calculates the dissimilarity value between any two users according to a plurality of groups of state values by way of example; classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users; calculating a module degree value of the classification result, wherein the module degree value table is used for classifying the structural strength corresponding to the classification result; and obtaining a plurality of groups of users according to the classification result corresponding to the modularity value.
Where the dissimilarity value indicates the dissimilarity between users, for example, if the i-th user and the j-th user are similar, the dissimilarity value will be small, and conversely, if the i-th user and the j-th user are different, the dissimilarity value will be large.
The embodiment of the application does not limit the calculation mode of the dissimilarity value, and the dissimilarity value is calculated according to the distance between users, and optionally, the Euclidean distance between any two groups of state values is obtained according to any two groups of state values; and obtaining the dissimilarity value between any two users according to the Euclidean distance.
Specifically, any two columns of state vectors X in the state matrix X are calculated according to a third formula i And X j The Euclidean distance between the two formulas specifically comprises:
d ij ＝‖X i -X j ‖
wherein II is the Euclidean norm d ij Representing a state vector X i And X j The euclidean distance between the two state vectors can be obtained through a third formula, and the distance between the two users is reflected.
In the embodiment of the application, the mode of obtaining the dissimilarity value between any two users according to the Euclidean distance is not limited, and the dissimilarity value between any two users is calculated according to the dissimilarity formula by way of example.
The dissimilarity formula specifically includes:
where Λ (i, j) represents a dissimilarity value between the i-th user and the j-th user.
Further, after calculating the dissimilarity value between any two users in the N users, the terminal equipment classifies the users according to the dissimilarity value to obtain classification results of the users.
Specifically, the terminal device sequentially removes the connection between the two users with the highest dissimilarity value until the connection between all the users is removed, and in the process, the terminal device records a corresponding classification result of the plurality of users once every time the connection is removed, wherein each classification result corresponds to one community division, the plurality of users are divided into a plurality of communities, and each community comprises a group of users.
It can be understood that, in the embodiment of the present application, for each classification result, a module degree value of the classification result is calculated once, and then multiple groups of users are obtained according to the classification result corresponding to the module degree value.
In the application, the mode of obtaining multiple groups of users according to the classification result corresponding to the module degree value is not limited, and the multiple groups of users are obtained according to the classification result corresponding to the maximum module degree value in the module degree values by way of example.
According to a modularity formula, calculating a modularity value of the classification result, wherein the modularity value formula specifically comprises:
wherein Q represents the modularity value of the classification result C, s is a group of users in the classification result, one group of users corresponds to one community, the classification result C comprises a plurality of groups of users, the plurality of groups of users correspond to a plurality of communities, l s Represents the number of edges in the community s, d s And representing the sum of the degree values among the user nodes in the communities s, wherein M is the total edge number in the user relationship network. The structural strength of the community structure corresponding to the classification result is represented by the modular value, which can be understood as the sum of the closeness of the relationship between the users in each group of users, so that the larger the modular value is, the better the corresponding community structure is.
Optionally, if a group of individual user nodes appears in the final classification result, the individual user nodes are added into the community with the highest connection strength.
Illustratively, the edge bets of the user nodes in the single user node communities and other communities are calculated according to the adjacency matrix, and the single user node communities are added into communities with the largest edge bets in the communities connected with the single user node communities according to the edge bets
Wherein the edge medium marks the proportion of the number of paths passing through the edge in all the shortest paths in the user relation network to the total number of the shortest paths. In the embodiment of the application, the edge betweenness is calculated by an edge betweenness formula, and the edge betweenness formula specifically comprises:
wherein B is ij Characterizing passing edge e in all shortest paths in a user relationship network ij Ratio of the number of N lm (e ij ) Representing the shortest path between the first user node and the mth user node passing edge e ij Number of bars, N lm The number of shortest paths between the ith user node and the mth user node is characterized.
S104, generating recommendation information according to the historical data of each user in each group of users, and pushing the recommendation information to each user.
After the terminal equipment obtains a plurality of groups of users corresponding to the classification result, the relevance among the users in each group is high, and recommendation information can be generated according to the historical data of each user in each group of users and pushed to each user.
Specifically, the users in the same group have a high possibility of purchasing the same product or transacting the same service, and the terminal device can generate corresponding recommendation information according to the service data of the users to recommend the users in the group.
The information pushing method provided by the embodiment of the application comprises the following steps: acquiring initial association relations of a plurality of users according to historical data of the plurality of users; for each user, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to initial association relations of a plurality of users; wherein the first time is earlier than the second time; classifying a plurality of users according to a plurality of groups of state values to obtain a plurality of groups of users; and generating recommendation information according to the historical data of each user in each group of users, and pushing the recommendation information to each user. Compared with the prior art, the method and the device for classifying the users have the advantages that the users are classified according to the state values of other users when each user is in an abnormal state, so that multiple groups of users are obtained, recommendation information is pushed to the users, and the rationality of the classification method is improved.
On the basis of the above embodiments, the information pushing method provided in the present application is further described below. Fig. 3 is a flow chart of another information pushing method provided in an embodiment of the present application, as shown in fig. 3, the method includes:
s201, obtaining initial association relations of a plurality of users according to historical data of the plurality of users.
S202, setting a state value of the ith user from the first moment to the second moment as a first state value indicating an abnormal state for the ith user.
S203, performing loop execution from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating the moment information until calculating to obtain the state value of other users at the second moment.
S204, calculating the dissimilarity value between any two users according to the multiple groups of state values.
S205, classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users.
S206, calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result.
S207, obtaining a plurality of groups of users according to the classification result corresponding to the modularity value.
The technical terms, effects, features, and alternative embodiments of S201-S207 may be understood with reference to S101-104 shown in fig. 2, and will not be further described herein for repeated matters.
On the basis of the above embodiment, the following is a case of obtaining multiple groups of users according to the embodiment of the present application. Fig. 4 is a flow chart of another information pushing method according to an embodiment of the present application, as shown in fig. 4, where the method includes:
s301, obtaining initial association relations of a plurality of users according to historical data of the plurality of users.
S302, setting a state value of the ith user from the first moment to the second moment as a first state value indicating an abnormal state for the ith user.
S303, circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating the moment information until calculating to obtain the state value of other users at the second moment.
S304, calculating Euclidean distance between any two groups of state values.
S305, calculating the dissimilarity value between any two users according to the Euclidean distance.
S306, removing the initial association relation between the two users with the highest dissimilarity value, and recording a classification result of removing the initial association relation.
In the step, after the terminal equipment calculates the dissimilarity value between any two users according to the Euclidean distance, removing the initial association relationship between the two users with the highest dissimilarity value, and recording the classification result of removing the initial association relationship; and in step S308, the terminal device determines whether all the initial association relationships are removed, and if not, continues to remove the initial association relationship between the two users with the highest dissimilarity value, and records the classification result of removing the initial association relationship.
S307, calculating the modularity value of the classification result.
In this step, the terminal device removes the initial association relationship between the two users with the highest dissimilarity value, records the classification result of the removed initial association relationship, and calculates the modularity value of the classification result.
S308, judging whether the initial association relation is completely removed.
In this step, when the terminal device removes the initial association relationship between the two users with the highest dissimilarity value, and records the classification result of removing the initial association relationship and the module degree value of calculating the classification result, it determines whether all the initial association relationship is removed, if not, step S306 is continuously executed, and if yes, step S309 is executed.
S309, obtaining a plurality of groups of users according to the classification result corresponding to the maximum module degree value in the module degree values.
In this step, after the terminal device determines whether all the initial association relationships are removed, if yes, multiple groups of users are obtained according to the classification result corresponding to the largest module value among the module values.
The technical terms, effects, features, and alternative embodiments of S301-S309 may be understood with reference to S101-S104 shown in fig. 2, and will not be described in detail herein for repeated content.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware associated with program instructions, where the foregoing program may be stored in a computer readable storage medium, and when executed, the program performs steps including the above method embodiments; and the aforementioned storage medium includes: various media that can store program code, such as ROM, RAM, magnetic or optical disks.
Fig. 5 is a schematic structural diagram of an information pushing device according to an embodiment of the present application, where the information pushing device may be implemented by software, hardware, or a combination of both, so as to execute the information pushing method in the foregoing embodiment. As shown in fig. 5, the information pushing apparatus 400 includes: an acquisition module 401, a processing module 402 and a push module 403.
An obtaining module 401, configured to obtain initial association relationships of a plurality of users according to historical data of the plurality of users;
a processing module 402, configured to calculate, for each user, a set of state values of other users at a second moment when the user is abnormal at the first moment according to initial association relationships of a plurality of users; wherein the first time is earlier than the second time; classifying a plurality of users according to a plurality of groups of state values to obtain a plurality of groups of users;
and the pushing module 403 is configured to generate recommendation information according to the history data of each user in each group of users, and push the recommendation information to each user.
In an alternative embodiment, the processing module 402 is specifically configured to set, for the ith user, a state value of the ith user from the first time to the second time to be a first state value indicating that the ith user is in an abnormal state; circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of a plurality of users and the state value of each user at the current moment, and updating moment information until calculating to obtain the state value of other users at the second moment; the state values of other users at the second moment are a group of state values of other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users.
In an alternative embodiment, the processing module 402 is specifically configured to calculate and obtain a state value of the other user at a next moment according to a first formula, where the first formula specifically includes:
/>
wherein x_i (t) represents a state value of the ith user at time t, represents a coupling strength, f (x_i (t)) represents a value of a relationship function of x_i (t), wherein the relationship function f (x) =4x (1-x), a_ij represents an initial association relationship between the ith user and the jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of continuous edges connected to the ith user in the initial association relationship.
In an alternative embodiment, the processing module 402 is specifically configured to calculate a dissimilarity value between any two users according to multiple sets of state values; classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users; calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result; and obtaining a plurality of groups of users according to the classification result corresponding to the modularity value.
In an alternative embodiment, the processing module 402 is specifically configured to obtain the euclidean distance between any two sets of state values according to any two sets of state values; and obtaining the dissimilarity value between any two users according to the Euclidean distance.
In an alternative embodiment, the processing module 402 is specifically configured to obtain multiple groups of users according to the classification result corresponding to the largest module degree value among the module degree values.
It should be noted that, the information pushing device provided in the embodiment of the present application may be used to execute the method provided in any of the foregoing embodiments, and the specific implementation manner and technical effects are similar, and are not repeated here.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 6, the electronic device may include: at least one processor 501 and a memory 502. Fig. 6 shows an electronic device, for example a processor.
A memory 502 for storing a program. In particular, the program may include program code including computer-operating instructions.
The memory 502 may comprise high-speed RAM memory or may further comprise non-volatile memory (non-volatile memory), such as at least one disk memory.
The processor 501 is configured to execute computer-executable instructions stored in the memory 502 to implement the information pushing method described above;
the processor 501 may be a central processing unit (Central Processing Unit, abbreviated as CPU), or an application specific integrated circuit (Application Specific Integrated Circuit, abbreviated as ASIC), or one or more integrated circuits configured to implement embodiments of the present application.
Alternatively, in a specific implementation, if the communication interface, the memory 502, and the processor 501 are implemented independently, the communication interface, the memory 502, and the processor 501 may be connected to each other through a bus and perform communication with each other. The bus may be an industry standard architecture (Industry Standard Architecture, abbreviated ISA) bus, an external device interconnect (Peripheral Component, abbreviated PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, abbreviated EISA) bus, among others. Buses may be divided into address buses, data buses, control buses, etc., but do not represent only one bus or one type of bus.
Alternatively, in a specific implementation, if the communication interface, the memory 502, and the processor 501 are integrated on a chip, the communication interface, the memory 502, and the processor 501 may complete communication through an internal interface.
The present application also provides a computer-readable storage medium, which may include: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, etc., in which program codes can be stored, and specifically, the computer-readable storage medium stores program information for the above information push method.
The embodiment of the application also provides a program which is used for executing the information pushing method provided by the embodiment of the method when being executed by a processor.
The present application also provides a program product, such as a computer readable storage medium, having instructions stored therein, which when run on a computer, cause the computer to perform the information push method provided by the above method embodiments.
In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions in accordance with embodiments of the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another, for example, by wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)), or wireless (e.g., infrared, wireless, microwave, etc.) means from one website, computer, server, or data center. Computer readable storage media can be any available media that can be accessed by a computer or data storage devices, such as servers, data centers, etc., that contain an integration of one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), etc.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.
Claims (10)
1. An information pushing method, characterized in that the method comprises:
acquiring initial association relations of a plurality of users according to historical data of the plurality of users;
for each user, calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at a first moment according to initial association relations of the plurality of users; wherein the first time is earlier than the second time;
classifying the plurality of users according to the plurality of groups of state values to obtain a plurality of groups of users;
generating recommendation information according to historical data of each user in each group of users, and pushing the recommendation information to each user;
For each user, calculating to obtain a set of state values of other users at a second moment when the user is abnormal at a first moment according to initial association relations of the plurality of users, wherein the set of state values comprises the following steps:
setting the state value of the ith user from the first moment to the second moment as a first state value indicating an abnormal state for the ith user;
circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating moment information until calculating to obtain the state value of other users at the second moment;
the state values of other users at the second moment are a group of state values of other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users;
calculating according to a first formula to obtain state values of other users at the next moment, wherein the first formula specifically comprises:
wherein x is i (t) represents the state value of the ith user at time t, ε represents the coupling strength, f (x) i (t)) represents x i The value of the relation function of (t), wherein the relation function f (x) =4x (1-x), a ij The method comprises the steps of representing initial association relation between an ith user and a jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of connected edges connected to the ith user in the initial association relation.
2. The method of claim 1, wherein classifying the plurality of users according to the plurality of sets of status values to obtain a plurality of sets of users comprises:
calculating the dissimilarity value between any two users according to the multiple groups of state values;
classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users;
calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result;
and obtaining the plurality of groups of users according to the classification result corresponding to the modularity value.
3. The method of claim 2, wherein calculating a dissimilarity value between any two users based on the plurality of sets of state values comprises;
obtaining Euclidean distances of the arbitrary two sets of state values according to the arbitrary two sets of state values;
and obtaining the dissimilarity value between any two users according to the Euclidean distance.
4. The method of claim 2, wherein said obtaining said plurality of groups of users from said modularity value comprises:
and obtaining the plurality of groups of users according to the classification result corresponding to the maximum module degree value in the module degree values.
5. An information pushing apparatus, characterized in that the apparatus comprises:
the acquisition module is used for acquiring initial association relations of a plurality of users according to historical data of the plurality of users;
the processing module is used for calculating and obtaining a group of state values of other users at a second moment when the user is abnormal at the first moment according to the initial association relation of the plurality of users for each user; wherein the first time is earlier than the second time; classifying the plurality of users according to the plurality of groups of state values to obtain a plurality of groups of users;
the pushing module is used for generating recommendation information according to the historical data of each user in each group of users and pushing the recommendation information to each user;
the processing module is specifically configured to set, for an ith user, a state value of the ith user from the first moment to the second moment as a first state value indicating that the ith user is in an abnormal state; circularly executing from the first moment, calculating to obtain the state value of other users at the next moment according to the initial association relation of the plurality of users and the state value of each user at the current moment, and updating moment information until calculating to obtain the state value of other users at the second moment; the state values of other users at the second moment are a group of state values of other users at the second moment when the user is abnormal at the first moment, the other users are users except the ith user, i is a positive integer, i is less than or equal to N, and N is the total number of users;
The processing module is specifically configured to calculate and obtain a state value of the other user at a next moment according to a first formula, where the first formula specifically includes:
wherein x is i (t) represents the state value of the ith user at time t, ε represents the coupling strength, f (x) i (t)) represents x i The value of the relation function of (t), wherein the relation function f (x) =4x (1-x), a ij The method comprises the steps of representing initial association relation between an ith user and a jth user, and k (i) represents a degree value of the ith user, wherein the degree value represents the number of connected edges connected to the ith user in the initial association relation.
6. The apparatus of claim 5, wherein the processing module is specifically configured to calculate a dissimilarity value between any two users based on the plurality of sets of state values; classifying the plurality of users according to the dissimilarity value to obtain classification results of the plurality of users; calculating a modularity value of the classification result, wherein the modularity value represents the structural strength corresponding to the classification result; and obtaining the plurality of groups of users according to the classification result corresponding to the modularity value.
7. The apparatus according to claim 6, wherein the processing module is specifically configured to obtain the euclidean distance of the arbitrary two sets of state values according to the arbitrary two sets of state values; and obtaining the dissimilarity value between any two users according to the Euclidean distance.
8. The apparatus of claim 6, wherein the processing module is specifically configured to obtain the plurality of groups of users according to the classification result corresponding to the largest module degree value among the module degree values.
9. An electronic device, comprising: a processor and a memory;
the memory is used for storing a computer program;
the processor is configured to invoke and run a computer program stored in the memory to perform the method of any of claims 1-4.
10. A computer readable storage medium storing a computer program for causing a computer to perform the method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110563315.7A CN113221023B (en) | 2021-05-24 | 2021-05-24 | Information pushing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110563315.7A CN113221023B (en) | 2021-05-24 | 2021-05-24 | Information pushing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113221023A CN113221023A (en) | 2021-08-06 |
| CN113221023B true CN113221023B (en) | 2023-09-15 |
Family
ID=77099292
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110563315.7A Active CN113221023B (en) | 2021-05-24 | 2021-05-24 | Information pushing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113221023B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107621997A (en) * | 2017-09-28 | 2018-01-23 | 联想(北京)有限公司 | A kind of processing method and processing device |
| CN111083005A (en) * | 2018-10-22 | 2020-04-28 | 瞻博网络公司 | Extensible visualization of health data for network devices |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012160795A (en) * | 2011-01-31 | 2012-08-23 | Hitachi Consumer Electronics Co Ltd | Content recommendation apparatus |
| US9465881B2 (en) * | 2012-10-31 | 2016-10-11 | Yahoo! Inc. | User displays using N-way paginated merge of information from diverse sources |
| US20180089193A1 (en) * | 2016-09-26 | 2018-03-29 | SWACK Holdings Inc. | Category-based data analysis system for processing stored data-units and calculating their relevance to a subject domain with exemplary precision, and a computer-implemented method for identifying from a broad range of data sources, social entities that perform the function of Social Influencers |
-
2021
- 2021-05-24 CN CN202110563315.7A patent/CN113221023B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107621997A (en) * | 2017-09-28 | 2018-01-23 | 联想(北京)有限公司 | A kind of processing method and processing device |
| CN111083005A (en) * | 2018-10-22 | 2020-04-28 | 瞻博网络公司 | Extensible visualization of health data for network devices |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113221023A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020192289A1 (en) | Method and device for determining graph node vectors in relational network graph | |
| CN110473083B (en) | Tree risk account identification method, device, server and storage medium | |
| CN112766500B (en) | Method and device for training graph neural network | |
| CN107770783B (en) | Base station capacity expansion transformation scheme design method and related equipment | |
| CN112085172A (en) | Method and device for training graph neural network | |
| CN111695965B (en) | Product screening method, system and equipment based on graphic neural network | |
| CN112052404B (en) | Group discovery method, system, equipment and medium of multi-source heterogeneous relation network | |
| CN110555172B (en) | User relationship mining method and device, electronic equipment and storage medium | |
| CN112085615A (en) | Method and device for training graph neural network | |
| CN110020124A (en) | The method and device excavated for Related product | |
| Li et al. | On multi-source networks: Enumeration, rate region computation, and hierarchy | |
| CN112529477A (en) | Credit evaluation variable screening method, device, computer equipment and storage medium | |
| CN111898247A (en) | Landslide displacement prediction method, equipment and storage medium | |
| CN112182370A (en) | Method and device for pushing item category information, electronic equipment and medium | |
| US8407255B1 (en) | Method and apparatus for exploiting master-detail data relationships to enhance searching operations | |
| WO2018135515A1 (en) | Information processing device, neural network design method, and recording medium | |
| CN113221023B (en) | Information pushing method and device | |
| CN111427915A (en) | Information processing method and device, storage medium and electronic equipment | |
| CN114417161B (en) | Virtual article time sequence recommendation method, device, medium and equipment based on special-purpose map | |
| CN115834257A (en) | Cloud electric power data safety protection method and protection system | |
| CN113435900A (en) | Transaction risk determination method and device and server | |
| CN111797282A (en) | Product label weight determination method and device, electronic equipment and readable storage medium | |
| CN114529108B (en) | Tree model based prediction method, apparatus, device, medium, and program product | |
| CN113837843B (en) | Product recommendation method and device, medium and electronic equipment | |
| CN110457572B (en) | Commodity information recommendation method based on graph network and electronic equipment |
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 |