CN111881615A - Model and method for information cooperative propagation of individual sensitivity in multilayer network - Google Patents

Abstract

The invention discloses a model and a method for information cooperative propagation of individual sensitivity in a multilayer network, which overcome a series of realistic problems of the relation degree between individuals and the channel quantity of the individual for obtaining information sources in the real network in the prior art, and comprises a multilayer complex network modeling module, a numerical simulation module and a real network verification module which are connected in sequence; the multilayer complex network modeling module is used for constructing a network model for modeling; the numerical simulation module is used for carrying out simulation on the model constructed by the multilayer complex network modeling module; and the real network verification module simulates the process of realizing information propagation in the real network from the model. The invention adopts the process of researching information transmission from the angle of a complex network, controls adverse effects caused by malignant transmission to a certain extent through the improvement and optimization of the model, and can better provide a relatively safe social network environment for users.

Description

An Individual Sensitivity to Information Collaborative Propagation Model and Method in Multilayer Networks

technical field

The present invention relates to the technical field of complex network propagation dynamics, in particular to a multi-layer network information cooperative propagation model and method based on the individual sensitivity of the dynamic characteristics of information propagation in a multi-layer network.

Background technique

With the development of the information age, the communication methods of information are gradually diversifying, and the information exchange between infrastructure networks in the real world is more extensive. Different social factors will affect the scope of information dissemination. In the field of complex networks, using the ideas of statistical physics and other disciplines, some scholars have proposed some theoretical methods, such as mean field theory, etc., through which the propagation behavior of epidemics on complex networks can be studied. Propagation is strangled in the cradle, and benign propagation is spread as quickly as possible. Some scholars have proposed the SI model, the SIR model and the SIS model.

In the early days, the research results of communication dynamics on a single network at home and abroad were remarkable. However, in the study of communication dynamics on a single network, many factors affecting communication are inevitably ignored, such as the multi-channel nature of communication channels, social networks based on different interaction platforms (Facebook, Twitter, etc.), interaction methods (SMS, telephony, etc.) resulting in multipath propagation of information, etc.

SUMMARY OF THE INVENTION

The present invention is to overcome the problems caused by ignoring multiple factors affecting propagation in the prior art, and to provide a model and method for information cooperative propagation in a multi-layer network with individual sensitivity. The present invention utilizes the characteristics and propagation dynamics of real networks. , and proposed that "individual sensitivity", a social relationship that generally exists between different individuals in different social networks, more truly and effectively reflects the impact of the relationship between different individuals on the collaborative dissemination of information in multi-layer networks. It simulates the process and control of information cooperative dissemination, and solves a series of real problems such as the degree of relationship between individuals in the real network and the number of channels for individuals to obtain information sources.

In order to achieve the above object, the present invention adopts the following technical solutions:

An individual sensitivity to information cooperative propagation model in a multi-layer network, including a multi-layer complex network modeling module, a numerical simulation module and a real network verification module connected in sequence; the multi-layer complex network modeling module is used to build a network model for Modeling; the numerical simulation module is used to simulate the model constructed by the multi-layer complex network modeling module; the real network verification module simulates the process of information propagation in the real network from the model, and applies it to the actual network according to the principle of propagation dynamics In order to better understand the scope of information dissemination in the real network and the prevention and control and intervention of malicious information flooding in the network.

Preferably, the multi-layer complex network modeling module performs modeling based on the SI model and the SIR model, and the process involves the principle of propagation dynamics.

Preferably, in the initial state of the SI model network, the individual has two states, namely the susceptible state and the adoption state, and the information is propagated in the multi-layer network, and the propagation process includes the following steps:

A1: Assuming that a certain number of individuals have been infected in the initial state of the network, that is, the initial infection probability is unchanged, and the number of channels for individuals to obtain information is certain, so that information can be spread in the network, and the relationship between individuals and individuals in the network is defined. The degree of relationship between , that is, the individual sensitivity is θ;

A2: A user in the first-layer network is informed of a certain information, then the probability of the user being infected is P ₁ =(1/3)θ;

A3: A user in the second-layer network is informed of a certain information, then the probability of the user being infected is P ₂ =(2/3)θ;

A4: When the user is informed of the same information in the three networks, then the user must believe the information, with a probability of 1; the dynamic process when a susceptible individual is infected is:

S+I→2I

Among them, S is the susceptible state, and I is the adoption state.

Preferably, the SIR model has three states of the individual in the initial state of the network, namely the susceptible state, the adoption state and the immune state. In the multi-layer network, the infected individual will recover with probability γ, and the infected individual will recover the pairwise information. The dynamic process of the state of immunity is:

I→R

Among them, I is the adoption state, R is the immune state, and the susceptible state is represented by S.

Preferably, the numerical simulation module includes social influence factors, and the social influence factors include individual sensitivity, the number of channels for individual perception information, the number of infected individuals in the initial state, and the infected individual has a certain recovery probability.

Preferably, the numerical simulation module includes the following steps for the analysis process of the cooperative propagation of information in the multi-layer network:

B1: Initialize the initial infection probability ω and individual sensitivity θ. In the case that the nodes in the network do not exist and become immune nodes, by adjusting the value of the number of channels δ through which individuals perceive information, when the information spread reaches a steady state, the network can be obtained. The function curve of the proportion of infected nodes in the middle and δ;

B2: Initialize the initial infection probability ω, the number of channels for individual perception information δ, and the recovery probability γ. By adjusting the individual sensitivity θ, the step size of the individual sensitivity θ is set to 0.02. When the information propagation reaches a steady state, the network can be obtained. The function curve of the number of nodes in the three states and θ, and the function curve of the proportion of immune nodes and θ;

B3: Initialize the initial infection probability ω, the number of channels of individual perception information δ, and the individual sensitivity θ. When some infected nodes in the network are restored to immune nodes, the step size of γ is set to 0.002 by adjusting the recovery probability γ. , when the information propagation reaches a steady state, the function curve of the number of nodes and γ in the three states in the network, and the function curve of the proportion of immune nodes and γ can be obtained;

B4: Calculate the variance of B1, B2 and B3 through 100 iterations and obtain the function curve;

B5: Set multiple different network layers, repeat B1-B4;

B6: Replace the ER network in the model with the BA network and repeat B1-B5.

An individual sensitivity for information cooperative propagation method in multi-layer network, adopting an individual sensitivity for information cooperative propagation model in multi-layer network, including the following steps:

S1: Generate a three-layer network with an initial infection probability ω;

S2: Initialize the number of individual perception information channels δ and recovery probability γ = 0, and iteratively update with the individual's sensitivity to information θ∈(0, 10);

S3: Judgment of iterative condition termination, update the number of infected nodes in the three-layer network, and count its probability as _PI and variance;

S4: Adjust the parameters of the model, re-initialize the number of individual perception information channels δ and the individual's sensitivity to information θ, and iteratively update with the recovery probability γ∈(0,1);

S5: iterative condition termination judgment, update the number of susceptible nodes, the number of infected nodes and the number of immune nodes in the three-layer network, and calculate the probability of immune nodes as _PR and variance;

S6: Verify the model of the present invention in an actual network system.

Therefore, the present invention has the following beneficial effects:

1. The present invention will simulate the process of realizing information dissemination in the real network from the model, and apply it to the actual network according to the principle of dissemination dynamics to better understand the scope of information dissemination in the real network and the prevention of malicious information flooding in the network. control and intervention;

2. The present invention utilizes the characteristics and propagation dynamics of the real network, more truly and effectively reflects the influence of the relationship between different individuals on the cooperative propagation of information in the multi-layer network, and technically simulates the process of information cooperative propagation and communication. Control, which solves a series of real problems such as the degree of relationship between individuals in the real network and the number of channels through which individuals obtain information sources;

3. The effect and advantage achieved by the present invention are that, from the perspective of complex network, the process of information dissemination is studied, and through the improvement and optimization of the model, combined with the characteristics of the real network and the propagation dynamics theory, the flood of information in the real network can be prevented. control and intervention; the present invention can effectively analyze how to promote the widespread dissemination of benign information and the suppression of unhealthy information in the network, so as to control the adverse consequences caused by malignant dissemination to a certain extent in real life, it can better It provides users with a relatively safe social networking environment.

Description of drawings

Fig. 1 is the working flow chart of the present invention.

Fig. 2 is a simulation result diagram of the SIR model of the present invention.

FIG. 3 is a simulation result diagram of the immune state of the present invention.

Fig. 4 is a structural block diagram of the present invention.

In the figure: 1-Multilayer complex network modeling module 2-Numerical simulation module 3-Real network verification module

Detailed ways

The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

This embodiment provides an individual sensitivity to information cooperative propagation model in a multi-layer network, as shown in Figures 1-3, including a multi-layer complex network modeling module 1, a numerical simulation module 2, and a real-world network verification that are connected in sequence Module 3; the multi-layer complex network modeling module is used to build network models for modeling; the numerical simulation module is used to simulate the model constructed by the multi-layer complex network modeling module; the real network verification module simulates the reality from the model The process of information dissemination in the network is applied to the actual network according to the principle of dissemination dynamics, so as to better understand the scope of information dissemination in the real network and the prevention and intervention of malicious information flooding in the network; multi-layer complex network modeling module Based on the SI model and the SIR model, the process involves the principle of propagation dynamics. The initial state of the SI model network has two states, namely the susceptible state and the adoption state. Information is propagated in the multi-layer network, and the propagation process It includes the following steps:

A1: Assuming that a certain number of individuals have been infected in the initial state of the network, that is, the initial infection probability is unchanged, and the number of channels for individuals to obtain information is certain, so that information can be spread in the network, and the relationship between individuals and individuals in the network is defined. The degree of relationship between , that is, the individual sensitivity is θ;

A2: A user in the first-layer network is informed of a certain information, then the probability of the user being infected is P ₁ =(1/3)θ;

A3: A user in the second-layer network is informed of a certain information, then the probability of the user being infected is P ₂ =(2/3)θ;

A4: When the user is informed of the same information in the three networks, then the user must believe the information, with a probability of 1; the dynamic process when a susceptible individual is infected is:

S+I→2I

Among them, S is the susceptible state, and I is the adoption state.

Among them, the SIR model has three states of the individual in the initial state of the network, namely the susceptible state, the adoption state and the immune state. In the multi-layer network, the infected individual will recover with probability γ, and the infected individual will recover the state of paired information immunity. The kinetic process is:

I→R

Among them, I is the adoption state, R is the immune state, and the susceptible state is represented by S.

Among them, the numerical simulation module includes social influencing factors, which include individual sensitivity, the number of channels through which individuals perceive information, the number of individuals who have been infected in the initial state, and the infected individuals have a certain recovery probability.

Among them, the analysis process of the numerical simulation module for the cooperative propagation of information in the multi-layer network includes the following steps: B1: Initialize the initial infection probability ω and individual sensitivity θ. Adjust the value of the number of channels δ through which individuals perceive information. When the information spread reaches a steady state, the function curve of the proportion of infected nodes in the network and δ can be obtained;

B2: Initialize the initial infection probability ω, the number of channels for individual perception information δ, and the recovery probability γ. By adjusting the individual sensitivity θ, the step size of the individual sensitivity θ is set to 0.02. When the information propagation reaches a steady state, the network can be obtained. The function curve of the number of nodes in the three states and θ, and the function curve of the proportion of immune nodes and θ;

B3: Initialize the initial infection probability ω, the number of channels of individual perception information δ, and the individual sensitivity θ. When some infected nodes in the network are restored to immune nodes, the step size of γ is set to 0.002 by adjusting the recovery probability γ. , when the information propagation reaches a steady state, the function curve of the number of nodes and γ in the three states in the network, and the function curve of the proportion of immune nodes and γ can be obtained;

B4: Calculate the variance of B1, B2 and B3 through 100 iterations and obtain the function curve;

B5: Set multiple different network layers, repeat B1-B4;

B6: Replace the ER network in the model with the BA network, repeat B1-B5;

A phenomenon can be obtained in this set of experiments. The ratio of immune nodes does not increase monotonically with the increase of recovery probability, which is not completely consistent with the monotonous increase predicted by theory, but presents a non-monotonic phenomenon.

This embodiment also provides a corresponding method for information cooperative propagation in a multi-layer network with individual sensitivity, using an individual sensitivity for information cooperative propagation model in a multi-layer network. The data in this embodiment comes from 2013 World Athletics Championship, 2013 World Athletics Championships data: a user is regarded as a node, various types of social relations between users are regarded as an edge, and the data of this championship is abstracted into an undirected network; the users in the above network In the three-layer network of RT, MT and RE respectively, the dynamic propagation mechanism of information in the multi-layer network is propagated according to the model algorithm proposed by the present invention, and compared with the model, mainly about the existence between users obtained on Twitter. There are 88,804 actual network nodes and 210,250 edges, including the following steps:

S1: Generate a three-layer network with an initial infection probability ω;

S2: Initialize the number of individual perception information channels δ and recovery probability γ = 0, and iteratively update with the individual's sensitivity to information θ∈(0, 10);

S3: Judgment of iterative condition termination, update the number of infected nodes in the three-layer network, and count its probability as _PI and variance;

S4: Adjust the parameters of the model, re-initialize the number of individual perception information channels δ and the individual's sensitivity to information θ, and iteratively update with the recovery probability γ∈(0,1);

S5: iterative condition termination judgment, update the number of susceptible nodes, the number of infected nodes and the number of immune nodes in the three-layer network, and calculate the probability of immune nodes as _PR and variance;

S6: Verify the model of the present invention in an actual network system.

The effects and advantages achieved by the present invention are that the process of information dissemination is studied from the perspective of complex networks, and through the improvement and optimization of the model, combined with the characteristics of the actual network and the theory of dissemination dynamics, the information flooding of the actual network can be prevented and controlled. The function of intervention; the present invention can effectively analyze how to promote the wide spread of benign information and the suppression of bad information in the network, and then apply it in real life to control the adverse consequences caused by malignant spread to a certain extent. Users are provided with a relatively safe social networking environment.

The specific description of the present invention in the above embodiments is only used to further illustrate the present invention, and should not be construed as a limitation on the protection scope of the present invention. Some non-essential improvements and adjustments made to the present invention by technical engineers in the field according to the content of the above invention are all within the scope of the present invention. into the protection scope of the present invention.

Claims (7)

1. A kind of individual sensitivity is for the information cooperation propagation model in the multi-layer network, it is characterized in that, comprises the multi-layer complex network modeling module, the numerical simulation simulation module and the realistic network verification module which are connected in turn; The multi-layer complex network modeling module It is used to build a network model for modeling; the numerical simulation module is used to simulate the model constructed by the multi-layer complex network modeling module; the real network verification module simulates the process of realizing information dissemination in the real network from the model, according to the dissemination power The principle of learning is applied to the actual network to better understand the scope of information dissemination in the real network and the prevention and control and intervention of malicious information flooding in the network. 2. a kind of individual sensitivity according to claim 1 is for information cooperative propagation model in multi-layer network, it is characterized in that, described multi-layer complex network modeling module is modeled based on SI model and SIR model, involves in the process. to the principles of propagation dynamics. 3. a kind of individual sensitivity according to claim 2 is for information cooperative propagation model in multi-layer network, it is characterized in that, individual has two kinds of states during described SI model network initial state, are respectively susceptible state and adopt state , information is propagated in a multi-layer network, and the propagation process includes the following steps: A1: Assuming that a certain number of individuals have been infected in the initial state of the network, that is, the initial infection probability is unchanged, and the number of channels for individuals to obtain information is certain, so that information can be spread in the network, and the relationship between individuals and individuals in the network is defined. The degree of relationship between , that is, the individual sensitivity is θ; A2: A user in the first-layer network is informed of a certain information, then the probability of the user being infected is P ₁ =(1/3) ^θ ; A3: A user in the second-layer network is informed of a certain information, then the probability of the user being infected is P ₂ =(2/3) ^θ ; A4: When the user is informed of the same information in the three networks, the user must believe the information, and the probability is 1; the dynamic process when a susceptible individual is infected is: S+I→2I Among them, S is the susceptible state, and I is the adoption state. 4. a kind of individual sensitivity according to claim 2 is for information cooperative propagation model in multi-layer network, it is characterized in that, described SIR model individual has three kinds of states when network initial state, are respectively susceptible state, adopt state and immune state, the infected individual in the multi-layer network will recover with probability γ, and the dynamic process of the infected individual recovering to the state of information immunity is: I→R Among them, I is the adoption state, R is the immune state, and the susceptible state is represented by S. 5. A kind of individual sensitivity according to claim 1 is for information cooperation propagation model in multi-layer network, it is characterized in that, described numerical simulation simulation module comprises social influence factor, and social influence factor comprises individual sensitivity, individual perception information The number of channels, the number of infected individuals in the initial state, and the infected individuals have a certain probability of recovery. 6. a kind of individual sensitivity according to claim 5 is for information cooperative propagation model in multi-layer network, it is characterized in that, described numerical simulation simulation module comprises the following steps for the analysis process of information cooperative propagation in multi-layer network: B1: Initialize the initial infection probability ω and individual sensitivity θ. In the case that the nodes in the network do not exist and become immune nodes, by adjusting the value of the number of channels δ through which individuals perceive information, when the information spread reaches a steady state, the network can be obtained. The function curve of the proportion of infected nodes in the middle and δ; B2: Initialize the initial infection probability ω, the number of channels for individual perception information δ, and the recovery probability γ. By adjusting the individual sensitivity θ, the step size of the individual sensitivity θ is set to 0.02. When the information propagation reaches a steady state, the network can be obtained. The function curve of the number of nodes in the three states and the function curve of the proportion of immune nodes and θ; B3: Initialize the initial infection probability ω, the number of channels of individual perception information δ, and the individual sensitivity θ. When some infected nodes in the network are restored to immune nodes, the step size of γ is set to 0.002 by adjusting the recovery probability γ. , when the information propagation reaches a steady state, the function curve of the number of nodes and γ in the three states in the network, and the function curve of the proportion of immune nodes and γ can be obtained; B4: Calculate the variance of B1, B2 and B3 through 100 iterations and obtain the function curve; B5: Set multiple different network layers, repeat B1-B4; B6: Replace the ER network in the model with the BA network and repeat B1-B5. 7. A kind of individual sensitivity is for the information cooperative propagation method in the multi-layer network, adopts a kind of individual sensitivity described in any one of claim 1-6 for the information cooperative propagation model in the multi-layer network, it is characterized in that, comprises the following: step: S1: Generate a three-layer network with an initial infection probability ω; S2: Initialize the number of individual perception information channels δ and recovery probability γ=0, and iteratively update the individual’s sensitivity to information θ∈(0, 10); S3: Judgment of iterative condition termination, update the number of infected nodes in the three-layer network, and count its probability as _PI and variance; S4: Adjust the parameters of the model, re-initialize the number of individual perception information channels δ and the individual's sensitivity to information θ, and iteratively update with the recovery probability γ∈(0,1); S5: Iterative condition termination judgment, update the number of susceptible nodes, the number of infected nodes and the number of immune nodes in the three-layer network, and calculate the probability of immune nodes as _PR and variance; S6: Verify the model of the present invention in an actual network system.

Images