Complex network information spreading method based on individual hesitant common characteristics

Abstract

The invention discloses a complex network information dissemination method based on individual hesitant common characteristics, relates to the technical field of social network information dissemination, and discusses hesitant common HECO characteristics of a group in a weighted social network by constructing a dissemination network model based on unassociated configuration. According to the method, in the aspect of group adoption heterogeneity, a group is divided into a hesitated state group and a common state group, and the propagation mechanisms of individuals in the hesitated state group and the common state group are quantified respectively, so that the information propagation mechanism of the individuals in the hesitated state group and the common state group is quantified, and the information propagation mechanism of the individuals in the hesitated state group and the common state group is quantified. Therefore, an accurate propagation mechanism is formed to perform information propagation on the complex network.

Description

Technical Field

[0001] The present invention relates to the technical field of social network information dissemination, and in particular to a complex network information dissemination method based on individual hesitant common characteristics. Background Art

[0002] With the rapid development of social media such as TikTok, WeChat, Twitter and other social tools, social networks play an increasingly important role in human life. Social media can help people quickly receive and share a variety of information, bringing great convenience to people's lives and work; at the same time, the communication links within these social media also constitute a huge information dissemination network; one of the main means of information dissemination in social networks is through connections between users; when the initial information is generated, the initial publisher disseminates the information. During the first dissemination, there are multiple receivers at the same time. If the receiver successfully adopts the information and becomes the next publisher, the information will eventually become shared information for the majority of people.

[0003] Considering the unfavorable factors that affect information dissemination, users' behaviors on social networks are diverse and heterogeneous, thus producing different information dissemination patterns; in real social networks, individuals with similar interests or preferences are more likely to interact with each other; studies have shown that individuals adopt heterogeneity and have different acceptance attitudes towards the same information, and individuals' attitudes will change as the amount of information obtained changes.

[0004] At present, the information diffusion in complex networks usually studies the degree of adoption between individuals in order to achieve information diffusion in complex networks. However, due to different personalities of individuals, their degree of adoption of information varies, and the impact of group adoption heterogeneity is not considered, making it difficult to apply the formed diffusion mechanism to the information diffusion of complex networks. Summary of the invention

[0005] The embodiment of the present invention provides a complex network information dissemination method based on the general characteristics of individual hesitation, which can solve the problem in the prior art that due to different individual personalities, the degree of information adoption varies, and the influence of group adoption heterogeneity is not considered, making it difficult to apply the formed dissemination mechanism to the information dissemination of complex networks.

[0006] The embodiment of the present invention provides a complex network information propagation method based on individual hesitant common characteristics, comprising the following steps:

[0007] According to the complex network theory, a double-layer SAR propagation network model is constructed in a complex network propagation environment;

[0008] Based on the edge partitioning theory of edge weight and individual hesitant-ordinary HECO characteristics, the propagation of complex network information between individuals in the double-layer SAR propagation network model is separated into network information propagation between hesitant groups, network information propagation between ordinary groups, and network information propagation between hesitant ordinary groups, and the network information propagation mechanism between hesitant groups, ordinary groups, and hesitant ordinary groups is quantified.

[0009] According to the differences in group hesitation and ordinary HECO characteristics and the degree of information adoption between the hesitant group and the ordinary group corresponding to the network information propagation mechanism, the information propagation mechanism corresponding to the complex network is quantified, and the information propagation of the complex network is carried out in the social network according to the information propagation mechanism corresponding to the complex network.

[0010] Preferably, the information propagation mechanism corresponding to the quantized complex network includes:

[0011] By introducing the edge weight distribution function into the double-layer SAR propagation network model, the node propagation information reception probability between the information propagation nodes in the model is obtained according to the edge weight distribution function;

[0012] According to the psychological change state of the hesitant group among individuals in the network information dissemination, the information adoption function of the hesitant group is constructed; at the same time, according to the psychological change state of the normal group among individuals in the network information dissemination, the information adoption function of the normal group is constructed;

[0013] According to the group heterogeneity, the difference of group hesitant general HECO characteristics, the probability of receiving node propagation information, the information adoption function of the hesitant group and the information adoption function of the general group, the information propagation node reception probability model of the cavity state node that does not propagate information, the information propagation probability model that does not adopt the propagated information in the social network for a certain period of time and is still in the susceptible state, and the probability model of no information propagation in the social network are constructed in the double-layer SAR propagation network model.

[0014] According to the three probability models, the changing states of information with different propagation ranges and different propagation probabilities are obtained respectively, and the information propagation mechanism of complex networks is quantified according to the changing states.

[0015] Preferably, the acquisition of the node propagation information reception probability between the information propagation nodes includes:

[0016] The number of network nodes in the double-layer SAR propagation network model is N = 10000, and the degree distribution is P (k). At any time, individuals in the network model will be in one of the susceptible state S, the adopting state A and the recovery state R. The susceptible state S can receive information from adjacent nodes and can transmit information to adjacent nodes.

[0017] Introducing edge weight distribution function into the network model to characterize the degree of interaction between individuals in the network model;

[0018] For the edge weight between adjacent nodes i and j, use ε ij Indicates that the weight distribution function is f(ε); when node i in state A sends information to node j in state S, the probability of node j receiving the information is:

[0019]

[0020] Where: λ is the probability of information unit transmission, as ε ij The increase of β ε gradually increases monotonically; when ε ij =1,β ε =λ.

[0021] Preferably, the construction of the information adoption function of the hesitant group includes:

[0022] For the hesitant group, the hesitant group is in a state of active adoption at the beginning of information dissemination, and the active adoption state increases with the increase of information obtained, and then enters the passive adoption stage. When the hesitant group obtains information that reaches the information adoption decision-making ability value, it returns to a positive mentality and stabilizes;

[0023] Then the information adoption function of the hesitant group is:

[0024]

[0025] Among them: e represents the hesitant group; a is the hesitation parameter; x represents the ratio of the total number of messages successfully received by the node m to the degree.

[0026] Preferably, the construction of the information adoption function of the normal state group includes:

[0027] For the normal group, they will normally receive and adopt information during the information dissemination process, and the more information they get, the more they will be encouraged to adopt information.

[0028] Then the information adoption function of the normal group is:

[0029]

[0030] Where: q represents the normal state group; a is the hesitation parameter; x is the ratio of the total number of messages successfully received by the node m to the degree.

[0031] Preferably, the construction of the information propagation node reception probability model of the cavity state node that does not propagate information includes:

[0032] Nodes in the cavity state can receive information from neighboring nodes but cannot transmit information to neighboring nodes. Based on the principle of random distribution of edge weights, the node reception probability model of nodes in the cavity state is:

[0033]

[0034] Where: θ ε (t) is the probability that at time t, the A-state node does not propagate information to the adjacent node in the S-state through the edge with weight ε; f(ε) represents the weight distribution function.

[0035] Preferably, the construction of the information propagation probability model for information that has not been adopted in the social network for a certain period of time and is still in a susceptible state includes:

[0036] According to the group heterogeneity, the difference in the characteristics of group hesitation and normal HECO, the probability of receiving node propagation information, the information adoption function of the hesitant group and the information adoption function of the normal group, the nodes are considered to be in the hesitant state and the normal state respectively; the probability model of the hesitant node that has not adopted the information and is still in the S state after time t is constructed as follows:

[0037]

[0038] Where: e represents the hesitant group; P(k i ) is the degree distribution of hesitant node i; k i represents the degree of hesitant node i; φ e (k i ,m,t,a) indicates the degree is k i The probability that the hesitant node i is still in the S state after receiving m pieces of information after time t; a is the hesitation parameter;

[0039] The probability model of a normal state node that has not adopted information and is still in the S state after time t is constructed as follows:

[0040]

[0041] Where: q represents the normal population; P(k i ) is the degree distribution of normal node i; k i represents the degree of normal node i; φ e (k i ,m,t,a) indicates the degree is k i The probability that the normal state node i is still in the S state after receiving m pieces of information after time t; a is the hesitation parameter.

[0042] Preferably, the construction of the probability model of no information propagation in the social network includes:

[0043] Since the probability that node i in state S can successfully adopt the information from neighboring node j in state A through the corresponding edge with weight is β ε ; Then the probability model of a node losing interest in information dissemination from the adoption state to the recovery state with probability γ is constructed, and the probability model of no information dissemination on the edge in the network is constructed as follows:

[0044]

[0045] in: Represents node i with degree k j The probability of node j making contact; <k>is the average degree of the network; ψ(k j , n,∞,a) represents the probability that node j in S state still remains in S state after receiving n pieces of information cumulatively after infinite ∞ time.

[0046] Preferably, the changing states of different propagation ranges of the information with different propagation probabilities include:

[0047] When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.8, the variation law of information dissemination range with information dissemination probability λ is obtained;

[0048] When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.5, the variation law of information dissemination range with information dissemination probability λ is obtained;

[0049] When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.2, the variation law of information dissemination range with information dissemination probability λ is obtained.

[0050] The embodiment of the present invention provides a complex network information propagation method based on the general characteristics of individual hesitation. Compared with the prior art, the method has the following beneficial effects:

[0051] The present invention explores the hesitant ordinary HECO characteristics of groups in weighted social networks by constructing a propagation network model based on uncorrelated configuration, and quantifies the information propagation mechanism of individuals using the edge partitioning theory based on edge weight and hesitant ordinary HECO characteristics. In terms of group adoption heterogeneity, the present invention divides the group into hesitant state group and normal state group, and quantifies the propagation mechanisms of individuals in the hesitant state group and the normal state group respectively, so as to form an accurate propagation mechanism for information propagation in complex networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic diagram of the overall process of a complex network information dissemination method based on individual hesitation common characteristics provided by an embodiment of the present invention;

[0053] Figure 2 A schematic diagram of a weighted social network propagation model for a complex network information propagation method based on the general characteristics of individual hesitation provided by an embodiment of the present invention;

[0054] Figure 3 A schematic diagram of information adoption functions for different groups of people in a complex network information dissemination method based on individual hesitant general characteristics provided by an embodiment of the present invention;

[0055] Figure 4 A schematic diagram of the variation of the propagation range with the propagation probability λ when the proportion of the hesitant population is 0.8 for a complex network information propagation method and device based on the general characteristics of individual hesitation provided by an embodiment of the present invention;

[0056] Figure 5 A schematic diagram of the variation of the propagation range with the propagation probability λ when the proportion of the hesitant population is 0.5 for a complex network information propagation method based on the general characteristics of individual hesitation provided by an embodiment of the present invention;

[0057] Figure 6 A schematic diagram of the variation of the propagation range with the propagation probability λ when the proportion of the hesitant population is 0.2 for a complex network information propagation method based on the general characteristics of individual hesitation provided by an embodiment of the present invention;

[0058] Figure 7 A phase transition diagram of the final adoption effect of individuals in a network of a complex network information propagation method based on the general characteristics of individual hesitation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific embodiments disclosed below.

[0060] See also Figure 1 to Figure 3 The embodiment of the present invention provides a complex network information propagation method based on individual hesitant common characteristics, comprising the following steps:

[0061] Step S1: Network model generation and information reception probability.

[0062] An ER complex network with N nodes (N=10000) is constructed with a degree distribution of P(k). The edges between nodes are randomly generated according to the set network model, and all nodes are set to the susceptible state. Nodes with a proportion of q0 in the network are randomly selected as ordinary nodes, and the remaining nodes with a proportion of e0 are hesitant nodes (from which it can be concluded that e0+q0=1). Then, nodes with a proportion of ρ0 are randomly selected from the global nodes as the adopted state, and the remaining nodes are in the default susceptible state.

[0063] Step S2: Classification of crowd decision-making capabilities.

[0064] In order to describe the decision-making ability of the crowd, the present invention introduces the hesitation parameter a. The larger a is, the stronger the hesitation is, that is, the weaker the decision-making ability of the crowd is, and vice versa. In addition, in order to characterize the impact of group adoption heterogeneity on information dissemination, two functions are introduced to illustrate the individual's information adoption decision-making ability. For the hesitant crowd, in the early stage of information dissemination, it first manifests as an active adoption stage. The active adoption mentality increases with the increase of information obtained. Due to the existence of hesitation, it then enters the passive adoption stage. When the number of information obtained reaches the information adoption decision-making ability value, they return to the positive mentality again and no longer change. For the general population, they will normally receive and adopt information, and as more information is obtained, the more this part of the population is promoted to adopt information.

[0065] Step S3: Modeling the propagation process considering individual characteristics.

[0066] The present invention considers the information propagation of non-redundant information memory with group adoption heterogeneity on a weighted network, proposes an edge partitioning theory based on edge weight and HECO characteristics, and then constructs an information propagation model; at the same time, the present invention also introduces nodes in a cavity state, which can receive information from neighbors but cannot transmit information to other nodes; on this basis, combined with the differences in HECO characteristics of the population and the information adoption function, an individual reception and propagation information probability model and a critical propagation probability model are constructed.

[0067] Step S4: simulation verification analysis.

[0068] The present invention mainly discusses the propagation of information on a weighted ER network, in which the nodes obey a Poisson distribution. Matlab is used for simulation modeling, and by setting different proportions of hesitant people (three cases of hesitant proportions of 0.2, 0.5, and 0.8 are set respectively), the joint effect of the unit propagation probability and the hesitation degree parameter on the final adoption range of individuals in the weighted ER network is analyzed.

[0069] Specifically:

[0070] Step S1: Network model generation and information reception probability.

[0071] ① The present invention constructs a two-layer propagation network model based on uncorrelated configuration to explore the impact of differences in HECO characteristics of people in weighted social networks on information dissemination; wherein, the number of network nodes is N (10,000 nodes), the degree distribution is P (k), the propagation model is the SAR model, and at any time each individual will be in one of the three states of susceptible state (S), adopted state (A) and recovered state (R). Individuals in the susceptible state have not adopted information yet and can receive information from adjacent nodes. Individuals in the adopted state have adopted information and will transmit the information to adjacent nodes. Individuals in the recovered state have lost interest in the information and will no longer participate in the subsequent information dissemination process (that is, they will not adopt the information and will not disseminate the information again). The dissemination process is as follows: Figure 1 shown.

[0072] ② The present invention introduces an edge weight distribution function to characterize the degree of interaction between individuals. Different weight distributions reflect the heterogeneity of information transmission and reception between nodes. For the edge weight between adjacent nodes i and j, ε ij Indicates that the weight distribution function is f(ε); when node i in state A sends information to node j in state S, the probability of node j receiving the information is:

[0073]

[0074] Where: λ is the probability of information unit transmission, as ε ij The increase of β ε gradually increases monotonically; when ε ij =1,β ε =λ, which means that the weight value has no effect on information propagation.

[0075] Let m be the total number of messages successfully received by the S-state node. There is no information propagation in the initial weighted social network, that is, m of the S-state node j j = 0, and then at each propagation time point, if node j successfully receives the information transmitted by the neighbor node i in state A through the edge, the number of adopted information of node j increases by 1, that is, m j →m j +1.

[0076] Step S2: Classification of crowd decision-making capabilities.

[0077] ① For the hesitant group, they show a positive adoption state at the beginning of information dissemination. The positive adoption mentality increases with the increase of information obtained. Due to the hesitation, they enter the negative adoption stage. When the amount of information obtained reaches the information adoption decision-making ability value, they return to a positive mentality and no longer change. Therefore, for the hesitant group, their information adoption function can be expressed as:

[0078]

[0079] ② For the general population, they will normally accept and adopt information during the information dissemination process, and the more information they obtain, the more they will be encouraged to adopt information; therefore, the information adoption function of the general population can be expressed as:

[0080]

[0081] Among them: e and q represent the hesitant population and the general population respectively; x is the ratio of the total number of messages successfully received by the node m to the degree.

[0082] Step S3: Modeling the propagation process considering individual characteristics.

[0083] ① The board invention introduces nodes in the cavity state. Such nodes can receive information from neighbors but cannot transmit information to other nodes. Based on the principle of random distribution of edge weights, the probability that a node has not received information from its neighbors at time t is:

[0084]

[0085] Where: θ ε (t) is the probability that the A-state node does not propagate information to the adjacent node in the S-state through the edge with weight ε at time t; i The cumulative number of m messages received by node i in state S from its neighbors can be expressed as:

[0086]

[0087] ② Based on group heterogeneity, differences in HECO characteristics of the population and information adoption function; if node i is a hesitant node in state S, after time t, the cumulative reception After the information is received, the information is not adopted and the state is still in S state. The probability is expressed as:

[0088]

[0089] Similarly, when the received information m satisfies and When , the probability of not adopting the information and still being in the S state is expressed as:

[0090]

[0091] Then for any hesitant node in state S, the probability of not adopting information by time t is:

[0092]

[0093] Similarly, if node i is an ordinary node in the S state, after time t, it has received m pieces of information but has not adopted any information and is still in the S state, the probability is expressed as:

[0094]

[0095]

[0096] Then for any ordinary node in state S, the probability of not adopting information by time t is:

[0097]

[0098] Therefore, the probability that node i in S state still remains in S state after receiving m pieces of information after time t is:

[0099] φ(k i ,m,t,a)=(1-ρ0)[e0φ e (k i ,m,t,a)+q0φ q (k i ,m,t,a)]

[0100] Then in this weighted network, the proportion of nodes in S state at time t is:

[0101]

[0102] Due to the three states of the SAR model, the present invention introduces θ ε (t) can be calculated by first expressing it as:

[0103] θ ε (t) = η S,ε (t)+η A,ε (t)+η R,ε (t)

[0104] Where: η A,ε (t) is the probability that at time t, node i in S state contacts the adjacent node j in A state through the corresponding edge with weight ε, but node i fails to successfully adopt the information; η S,ε (t) and η R,ε (t) is the probability that a node i in state S contacts its neighboring node j in state S (or R) through a corresponding edge with weight ε.

[0105] ③Calculate η S,ε (t) Introducing the cavity node theory, a node i in the cavity state cannot transmit information to other nodes, so its degree is k j The node j in the S state can be obtained from other k j -1 neighboring nodes, so the probability that node j receives a total of n pieces of information from neighboring nodes at time t is:

[0106]

[0107] According to the differences in HECO characteristics of the crowd and the information adoption function; if node j is a hesitant node in state S, after time t, the cumulative reception After the information is received, the information is not adopted and the state is still in S state. The probability is expressed as:

[0108]

[0109] Similarly, when the received information m satisfies and When , the probability of not adopting the information and still being in the S state is expressed as:

[0110]

[0111] When node j is an ordinary node in S state, after time t, it has received n pieces of information but has not adopted any information and is still in S state. Its probability is expressed as:

[0112]

[0113]

[0114] Therefore, the probability that node j in S state still remains in S state after receiving n pieces of information after time t is:

[0115] ψ(k j ,n,t,a)=(1-ρ0)[e0ψ e (k j ,n,t,a)+q0ψ q (k j ,n,t,a)]

[0116] Then the probability that node i in S state can contact node j in S state through the corresponding edge with weight ε is:

[0117]

[0118] in: Represents node i with degree k j The probability of node j making contact; <k>is the average degree of the network.

[0119] ④Analysis η A,ε (t) and η R,ε (t), since the probability that node i in state S can successfully adopt the information from neighboring node j in state A through the corresponding edge with weight is β ε , then θ ε (t) can continue to evolve into:

[0120]

[0121] In addition, the node in state A may lose interest in information dissemination with probability γ and change to state R, then η R,ε (t) can be evolved into:

[0122]

[0123] Using the initial condition θ ε (0) = 1 and η R,ε (0) = 0, we can get:

[0124]

[0125] Then we get θ ε (t) can be evolved into:

[0126]

[0127] In the entire network, the density change of each state can be expressed as:

[0128]

[0129] Finally, the state densities S(t), A(t), and R(t) of each state can be obtained at any time step.

[0130] When t→∞, the state of the nodes in the network no longer changes. At this time, there are only S-state nodes and R-state nodes in the network; that is, R(∞) is the final information adoption size; at this time, the probability that no information propagation occurs on the edge with weight ε is:

[0131]

[0132] Step S4: simulation verification analysis.

[0133] ① Discuss the variation law of the propagation range with the propagation probability λ when the initial adoption state node ratio is 0.001 and the hesitant state population ratio is 0.8; Figure 4 As shown, it indicates that when the hesitant population has strong decision-making ability, the smaller the proportion of the normal population, the faster the information propagation occurs and the earlier it breaks out, but the rate at which the propagation range increases with the unit propagation probability is relatively slow; on the contrary, when the proportion of the normal population is large, information propagation begins to break out later, but the rate at which the propagation range grows with the unit propagation probability is faster, and even a discontinuous phase transition phenomenon occurs, which can reach the global adoption state faster.

[0134] ② Discuss the variation of the propagation range with the propagation probability λ when the initial adoption state node ratio is 0.001 and the hesitant state population ratio is 0.5; Figure 5 As shown, it shows that when the decision-making ability of the hesitant population is moderate, the outbreak threshold of information dissemination is the same regardless of the proportion of the normal population. However, the larger the proportion of the normal population, the greater the jump in the adoption range when information dissemination breaks out, and it is easier to reach the global adoption state.

[0135] ③ Discuss the variation law of the propagation range with the propagation probability λ when the initial adoption state node ratio is 0.001 and the hesitant state population ratio is 0.2; Figure 6 As shown in the figure, when the decision-making ability of the hesitant population is low, the larger the proportion of the normal population, the faster the information adoption outbreak point can be reached at a smaller unit propagation probability λ, and the larger the final adoption range of information propagation in the steady state.

[0136] ④ Discuss the combined effects of different communication modes; e.g. Figure 7 As shown in the figure, the joint effect plane is divided into four regions according to different information propagation modes. In region I, when the hesitation degree parameter is very small, that is, when the hesitant crowd has extremely strong decision-making ability, there is no information propagation explosion. This is because the hesitant crowd has an inhibitory effect on information propagation when the decision-making ability is extremely strong. In region II, as the hesitation degree parameter increases, that is, the decision-making ability gradually weakens, the final adoption range growth presents a second-order continuous phase transition. The reason is that when the crowd has strong decision-making ability, the hesitant crowd dominates the initial stage of information propagation and guides the outbreak of information propagation. Later, with the increase of λ, the ordinary crowd continues to spread on the basis of the hesitant crowd. In region In region III, the hesitation degree parameter increases, the crowd's decision-making ability continues to weaken, and the final growth of the adoption range presents a first-order discontinuous phase transition. In this area, the decision-making ability of the hesitant crowd is moderate, and the adoption abilities of the hesitant and ordinary crowds are similar. There is no dominant party, and the outbreak thresholds of both parties are similar. They simultaneously break out during the information dissemination process, resulting in a first-order discontinuous phase transition. In region IV, the hesitation degree parameter is large, the crowd's decision-making ability is extremely low, and the final growth of the adoption range presents a second-order continuous phase transition. This is the result of the ordinary crowd dominating the dissemination when the hesitant crowd has a low decision-making ability. Therefore, the outbreak threshold of the second-order continuous dissemination will be smaller than the outbreak threshold in region II.

[0137] The above experiments verify that the theoretical analysis and simulation results are consistent, indicating that an increase in the relative proportion of ordinary individuals will promote information dissemination and adoption; at the same time, a phase transition crossover phenomenon also occurs. As the proportion of the ordinary population increases, the growth pattern of the final adoption range changes from a first-order discontinuous phase transition to a second-order continuous phase transition.

[0138] The present invention firstly constructs a propagation network model based on uncorrelated configuration to explore the HECO characteristics of people in weighted social networks; secondly, the edge partitioning theory based on edge weight and HECO characteristics is used to quantitatively analyze the information propagation mechanism of individuals; the theoretical analysis and simulation results of the model consistently show that an increase in the relative proportion of ordinary individuals will promote information dissemination and adoption; at the same time, a phase transition crossover phenomenon also occurs. As the proportion of the ordinary population increases, the growth pattern of the final adoption range changes from a first-order discontinuous phase transition to a second-order continuous phase transition, indicating that an increase in the weight distribution index promotes information dissemination; in addition, a decrease in the heterogeneity of degree distribution will promote the spread of information; increasing the heterogeneity of degree distribution and weakening the decision-making ability of the population can both inhibit the spread of information.

[0139] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.< / k> < / k>

Claims

1. A complex network information propagation method based on the general characteristics of individual hesitation, characterized in that: The following steps are involved: According to the complex network theory, a double-layer SAR propagation network model is constructed in a complex network propagation environment; Based on the edge partitioning theory of edge weight and individual hesitant-ordinary HECO characteristics, the propagation of complex network information between individuals in the double-layer SAR propagation network model is separated into network information propagation between hesitant groups, network information propagation between ordinary groups, and network information propagation between hesitant ordinary groups, and the network information propagation mechanism between hesitant groups, ordinary groups, and hesitant ordinary groups is quantified. According to the differences in group hesitation and ordinary HECO characteristics and the degree of information adoption between the hesitant group and the ordinary group corresponding to the network information propagation mechanism, the information propagation mechanism corresponding to the complex network is quantified, and the information propagation of the complex network is carried out in the social network according to the information propagation mechanism corresponding to the complex network.

2. According to claim 1, a complex network information propagation method based on individual hesitant common characteristics is characterized by: The information propagation mechanism corresponding to the quantized complex network includes: By introducing the edge weight distribution function into the double-layer SAR propagation network model, the node propagation information reception probability between the information propagation nodes in the model is obtained according to the edge weight distribution function; According to the psychological change state of the hesitant group among individuals in the network information dissemination, the information adoption function of the hesitant group is constructed; at the same time, according to the psychological change state of the normal group among individuals in the network information dissemination, the information adoption function of the normal group is constructed; According to the group heterogeneity, the difference of group hesitant general HECO characteristics, the probability of receiving node propagation information, the information adoption function of the hesitant group and the information adoption function of the general group, the information propagation node reception probability model of the cavity state node that does not propagate information, the information propagation probability model that does not adopt the propagated information in the social network for a certain period of time and is still in the susceptible state, and the probability model of no information propagation in the social network are constructed in the double-layer SAR propagation network model. According to the three probability models, the changing states of information with different propagation ranges and different propagation probabilities are obtained respectively, and the information propagation mechanism of complex networks is quantified according to the changing states.

3. A complex network information propagation method based on individual hesitant common characteristics according to claim 2, characterized in that: The acquisition of the node propagation information reception probability between the information propagation nodes includes: The number of network nodes in the double-layer SAR propagation network model is N = 10000, and the degree distribution is P (k). At any time, individuals in the network model will be in one of the susceptible state S, the adopting state A and the recovery state R. The susceptible state S can receive information from adjacent nodes and can transmit information to adjacent nodes. Introducing edge weight distribution function into the network model to characterize the degree of interaction between individuals in the network model; For the edge weight between adjacent nodes i and j, use ε ij Indicates that the weight distribution function is f(ε); when node i in state A sends information to node j in state S, the probability of node j receiving the information is: Where: λ is the probability of information unit transmission, as ε ij The increase of β ε gradually increases monotonically; when ε ij =1,β ε =λ.

4. According to claim 2, a complex network information propagation method based on individual hesitant common characteristics is characterized by: The construction of the information adoption function of the hesitant group includes: For the hesitant group, the hesitant group is in a state of active adoption at the beginning of information dissemination, and the active adoption state increases with the increase of information obtained, and then enters the passive adoption stage. When the hesitant group obtains information that reaches the information adoption decision-making ability value, it returns to a positive mentality and stabilizes; Then the information adoption function of the hesitant group is: Among them: e represents the hesitant group; a is the hesitation parameter; x represents the ratio of the total number of messages successfully received by the node m to the degree.

5. According to claim 2, a complex network information propagation method based on individual hesitant common characteristics is characterized by: The construction of the information adoption function of the normal state group includes: For the normal group, they will normally receive and adopt information during the information dissemination process, and the more information they get, the more they will be encouraged to adopt information. Then the information adoption function of the normal group is: Where: q represents the normal state group; a is the hesitation parameter; x is the ratio of the total number of messages successfully received by the node m to the degree.

6. A complex network information propagation method based on individual hesitant common characteristics according to claim 2, characterized in that: The construction of the information propagation node reception probability model of the cavity state node that does not propagate information includes: Nodes in the cavity state can receive information from neighboring nodes but cannot transmit information to neighboring nodes. Based on the principle of random distribution of edge weights, the node reception probability model of nodes in the cavity state is: Where: θ ε (t) is the probability that at time t, the A-state node does not propagate information to the adjacent node in the S-state through the edge with weight ε; f(ε) represents the weight distribution function.

7. The complex network information propagation method based on individual hesitant common characteristics according to claim 2 is characterized by: The construction of the information propagation probability model for information that has not been adopted in the social network for a certain period of time and is still in a susceptible state includes: According to the group heterogeneity, the difference in the characteristics of group hesitation and normal HECO, the probability of receiving node propagation information, the information adoption function of the hesitant group and the information adoption function of the normal group, the nodes are considered to be in the hesitant state and the normal state respectively; the probability model of the hesitant node that has not adopted the information and is still in the S state after time t is constructed as follows: Where: e represents the hesitant group; P(k i ) is the degree distribution of hesitant node i; k i represents the degree of hesitant node i; φ e (k i ,m,t,a) indicates the degree is k i The probability that the hesitant node i is still in the S state after receiving m pieces of information after time t; a is the hesitation parameter; The probability model of a normal state node that has not adopted information and is still in the S state after time t is constructed as follows: Where: q represents the normal population; P(k i ) is the degree distribution of normal node i; k i represents the degree of normal node i; φ e (k i ,m,t,a) indicates the degree is k i The probability that the normal state node i is still in the S state after receiving m pieces of information after time t; a is the hesitation parameter.

8. The complex network information propagation method based on individual hesitant common characteristics according to claim 2 is characterized by: The construction of a probability model that no information propagation occurs in the social network includes: Since the probability that node i in state S can successfully adopt the information from neighboring node j in state A through the corresponding edge with weight is β ε ; Then the probability model of a node losing interest in information dissemination from the adoption state to the recovery state with probability γ is constructed, and the probability model of no information dissemination on the edge in the network is constructed as follows: in: Represents node i with degree k j The probability of node j making contact; <k>is the average degree of the network; ψ(k j , n,∞,a) represents the probability that node j in S state still remains in S state after receiving n pieces of information cumulatively after infinite ∞ time.< / k> 9. The complex network information propagation method based on individual hesitant common characteristics according to claim 2 is characterized by: The changing states of different propagation ranges of the information with different propagation probabilities include: When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.8, the variation law of information dissemination range with information dissemination probability λ is obtained; When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.5, the variation law of information dissemination range with information dissemination probability λ is obtained; When the initial adoption node ratio is 0.001 and the hesitant group ratio is set to 0.2, the variation law of information dissemination range with information dissemination probability λ is obtained.

Images