Method and system for implementing social network information dissemination based on threshold adaptive change
By introducing adaptive threshold changes and time difference mechanisms into the threshold model, the problem of threshold remains unchanged in the prior art is solved, the impact of negative comments on information dissemination is simulated, and more accurate understanding and research tools are provided for the information dissemination process.
Patent Information
- Application Number
- CN202111009821.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-08-31
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-08-31
AI Technical Summary
The existing threshold model assumes that the node threshold remains unchanged in the process of describing the information dissemination of social networks, and fails to effectively capture the adaptive change of individual thresholds during information dissemination, especially under the influence of factors such as negative comments.
A social network information dissemination model based on threshold adaptive changes is proposed. By introducing four node states and three threshold adaptive changes strategies, combined with the time difference mechanism, the impact of negative comments on information dissemination is simulated, including linearity, exponential and logarithmic and distinguish the propagation mode of the existence and absence of time difference.
This model can more accurately simulate the information dissemination process, show the impact of negative comments on communication, provide a better understanding of the network information dissemination mechanism, and provide new insights and help for the study of human collective behavior.
Smart Images

Figure CN113807976B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of network information dissemination, and in particular to a method for implementing social network information dissemination based on adaptive threshold changes and a system for implementing the method. Background Art
[0002] Modeling the information diffusion process (such as fashion, innovation, viral memes, opinions, and rumors) is of great significance for analyzing the propagation pattern of information on social networks and then controlling its propagation process. So far, scientists have proposed a variety of models to describe this process, among which the most classic models include the independent cascade propagation model and the threshold model.
[0003] The threshold model was first proposed by sociologist Granovetter to describe collective behavior, and has been widely used to describe a series of binary decision-making phenomena in economics and sociology. The main idea of the model is that individuals in the network exhibit group behavior, that is, they make decisions based on the behavior of their neighbors. In 2002, Watts thoroughly studied the physical properties of the threshold model and proved the conditions for when global information cascades will occur on random networks. Since then, the model has attracted widespread attention from network scientists. It mainly focuses on the following aspects: 1. The influence of the underlying network structure. Centola et al. first studied the case of small-world networks and pointed out that random edges in small-world networks would hinder the spread of complexity. Galstyan studied the influence of weakly coupled community network structure, and the results showed that the speed of information propagation would show a double-layer phenomenon. Payne et al. considered the influence of degree-correlated networks, and they found that positive correlation between nodes (that is, nodes with large degrees are connected to nodes with large degrees) can expand the cascade window. Yagan et al. examined the case of multiple networks and showed that the propagation weights of different types of edges (bias towards the spread of a product or information) would have a great impact on the propagation dynamics behavior. 2. The influence of the initial seed size. Gleeson et al. developed a mean field method to study the impact of multiple seeds on information cascade propagation. 3. The impact of different threshold distributions. Karampourniotis et al. studied the case where the node threshold is Gaussian distributed, and found that the proportion of people in the final active state will produce non-monotonic changes with the increase of the standard deviation (the parameter of the Gaussian distribution function, indicating the degree of dispersion).
[0004] Although there have been so many studies on threshold models, most of the current studies assume that the node threshold remains unchanged throughout the entire propagation. However, with the rapid development of computer science and big data in recent years, people have found in some online empirical data that real social systems show some key factors that are ignored by traditional threshold models (and their extended models). For example, when people adopt a new product or watch a new movie, they will evaluate the product, and some negative feedback will have a certain degree of impact on other users, causing their "threshold" to change. Analysis of the Yelp data set shows that the bad reviews received by a restaurant will have a certain impact on the subsequent turnover of the restaurant. This proves to some extent that the individual's adoption threshold will change adaptively with the feedback of neighboring nodes. Summary of the invention
[0005] In order to overcome the deficiency of considering the node threshold to be a constant value (i.e., not changing over time) in the existing threshold model research, the present invention provides a social network information propagation implementation method based on threshold adaptive change, and proposes a threshold adaptive change propagation model combined with a time difference mechanism and different threshold adaptive change strategies.
[0006] The technical concept of the present invention is: the present invention theoretically explains the impact of negative comments on information dissemination during product marketing or message dissemination through simulation experiments, helping researchers to better understand the impact of negative comments in the marketing process.
[0007] The technical solution adopted by the present invention to solve the technical problem is as follows:
[0008] The network information propagation model based on threshold adaptive change includes four node states: uncontacted state S, contacted but no comment state E, contacted and positive comment state I P , contacted and posted negative comments status I N The method for implementing network information propagation based on adaptive threshold changes includes the following steps:
[0009] S1: Construct a network G = (V, E), where V = {v1, v2, ..., v N} is the network node set, N is the total number of nodes, is the network edge set;
[0010] S2: Initialization: assign the same initial threshold to all nodes in the network Randomly select a node as the information dissemination source;
[0011] S3: Consider three threshold adaptive change strategies: (1) The threshold of the S-state node changes with the number of neighbors. N The number of S-state nodes increases linearly; (2) The threshold of S-state nodes increases with the number of neighbors.N The number of S-state nodes increases exponentially; (3) The threshold of S-state nodes increases with the number of neighbors. N The number of dynamic nodes grows logarithmically;
[0012] S4: Network information cascade propagation process, considering two propagation modes, namely "with time difference" and "without time difference", the difference lies in whether the S state node immediately changes to I with probability p P The nodes propagate information according to the threshold model mechanism. For the S-state nodes that meet the threshold conditions, the node state changes with probability p according to whether there is a "time difference" until no new non-S-state nodes appear in the network, and the propagation ends.
[0013] S5: Calculate the final cascade range of information propagation, conduct multiple experiments for each set of experimental indicators, and calculate the proportion of non-S-state nodes in the total number of nodes in the network after the propagation is completed.
[0014] Furthermore, the step S1 specifically includes:
[0015] Consider a random network with average degree z, N=1000 nodes, and edges satisfying
[0016] Furthermore, the step S2 specifically includes:
[0017] Each node v in the network i All have node thresholds, denoted as For all nodes in the network, the same initial threshold is given Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P state.
[0018] Furthermore, the step S3 specifically includes:
[0019] The present invention designs three threshold adaptive change strategies. Assuming that the S-state node v i Neighbors I N The number of nodes in the state is n neg :The first is the linear threshold adaptive change strategy, node v i The threshold value is Growth based on The second is the exponential change strategy, where node v i The threshold value is Growth based on The third is the logarithmic change strategy, where node v i The threshold value is Growth based on The variable δ is the node threshold growth rate indicator, and its value is set to 0.01; the range of the node threshold is [0, 1].
[0020] Furthermore, the step S4 specifically includes:
[0021] In the information propagation process of step S4, there are two situations: "time difference exists" and "time difference does not exist". "Time difference" means that the S-state node v i After satisfying the threshold judgment formula, it immediately changes to I with probability p P state node, or first transform to the E state and then transform to the I state with probability p after a period of time Δt P state;
[0022] The specific propagation process of “no time difference” is as follows:
[0023] S4.1: Randomly select a node in the network as the source of the comment, publish a positive comment, and set its state to I P state;
[0024] S4.2: At the beginning of each step of information propagation, for the S-state node v in the current network i , according to its neighbors I N The number of state nodes n neg And randomly select a threshold adaptive change strategy to update its node threshold
[0025] S4.3: Randomly select an S-state node v i , assuming that the number of positive comments posted by neighbor nodes is n pos , using the following threshold determination formula:
[0026]
[0027] Where k is the node v i The number of neighbors; for the S-state node v that satisfies the threshold judgment formula (1) i , generate a random number r and compare it with the probability value p. When the random number r is less than or equal to p, node v i The state changes to I P state, otherwise it will become I N state;
[0028] S4.4: Repeat steps S4.1 to S4.3 until no new non-S-state nodes appear in the network;
[0029] The specific process of propagation of “time difference” is as follows:
[0030] T4.1: Set all node clocks to 0; randomly select a node in the network as the source of the comment, publish a positive comment, and set its status to I P state;
[0031] T4.2: At the beginning of each step of information propagation, the clock of each E-state node is updated by adding 1. For E-state nodes that reach a certain "time difference", they are transformed into I-state nodes with probability p. P state node, transforms to I with probability 1-p N state node; and for the S state node v in the current network i , according to its neighbors I N The number of state nodes n neg Update its node threshold
[0032] T4.3: Randomly select an S-state node v i , assuming that among its neighbor nodes I P The number of state nodes is n pos , I N The number of state nodes is n neg , the number of E-state nodes is n mid , using the following threshold determination formula:
[0033]
[0034] For an S-state node v that satisfies the threshold judgment formula (2), i , the node state immediately changes to E state, and the node clock is set to 1;
[0035] T4.4: Repeat steps T4.1 to T4.4 until no new non-S-state nodes appear in the network.
[0036] Furthermore, the step S5 specifically includes:
[0037] Change the network average degree value z and transform the S-state nodes into I P The probability p of the state node and the selection of different threshold adaptive change strategies are carried out for 10*N times. The proportion of the number of non-S state nodes in the network after the propagation is calculated as the final information propagation cascade range, and the average value of multiple experiments is calculated.
[0038] A system for implementing the method for implementing social network information propagation based on threshold adaptive change includes: a random network construction module, an initialization module, a threshold adaptive change module, an information cascade propagation module and a cascade range calculation module;
[0039] The random network construction module constructs a random network including N nodes and random edges between nodes, and the average degree of the network is denoted as z;
[0040] The initialization module assigns the same initial threshold to all nodes in the network. Each node in the network v iAll have node thresholds, denoted as Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P state;
[0041] The threshold adaptive change module includes two types of threshold adaptive change strategies: the first type is a linear threshold adaptive change strategy, where the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg Linear growth; the second category is the nonlinear threshold adaptive change strategy, which includes two nonlinear strategies, one is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg The other is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg Logarithmic growth;
[0042] The information cascade propagation module includes two propagation modes, namely, "with time difference" and "without time difference", the difference lies in whether the S-state node immediately changes to I with probability p P state node; randomly select a node as the information dissemination source, and carry out information dissemination according to the threshold model mechanism. For the S-state node that meets the threshold condition, the node state is changed with probability p according to the actual situation of whether there is a "time difference", until no new non-S-state nodes appear in the network, and the dissemination ends;
[0043] The cascade range calculation module performs 10*N experiments for each set of experimental indicators z, p and different threshold adaptive change strategies, calculates the proportion of the number of non-S-state nodes in the network after the propagation is completed in the total number of network nodes, as the final information propagation cascade range, and calculates the average value of multiple experiments;
[0044] The random network construction module, initialization module, threshold adaptive change module, information cascade propagation module and cascade range calculation module are connected in sequence.
[0045] The beneficial effects of the present invention are:
[0046] (1) This invention proposes a method for implementing information propagation in social networks based on threshold adaptive changes, which intuitively displays and explains the propagation process of positive and negative comments on the network, providing new insights for researchers to better understand the information propagation process in the network;
[0047] (2) At the same time, the present invention adds a time difference mechanism and designs different threshold adaptive strategies to simulate the impact of negative comments on information dissemination in real marketing processes or information dissemination processes, which provides great help for further research on human collective behavior. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1 It is the algorithm flow chart of the present invention;
[0049] Figure 2 is a schematic diagram of a random network according to an embodiment of the present invention;
[0050] Figure 3 When "no time difference exists", the trend graph of network information cascade range changing over time;
[0051] Figure 4 When "time difference exists", the trend chart of network information cascade range changing over time;
[0052] Figure 5 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION
[0053] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings.
[0054] Reference Figure 1 to Figure 4 , a method and system for implementing social network information dissemination based on threshold adaptive change, comprising the following steps:
[0055] The present invention uses an artificial random network for modeling analysis, the total number of nodes N = 1000, the network average degree z ranges from 0 to 8, and the node is transformed into I P The probability p of the state node ranges from 0 to 1, and the number of experiments for each group of indicators is 10*N times;
[0056] S1: Construction Figure 2 The random network G = (V, E) shown, where V = {v1, v2, ..., v N} is a random network node set, N is the total number of nodes, It is a random network edge set, where nodes are randomly connected to each other, including:
[0057] Consider a random network with average degree z, N=1000 nodes, and edges satisfying
[0058] S2: Assign the same initial threshold to all nodes in the network Randomly select a node as the information dissemination source; specifically:
[0059] Each node v in the network iAll have node thresholds, denoted as For all nodes in the network, the same initial threshold is given Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P state.
[0060] S3: Consider three threshold adaptive change strategies: (1) The threshold of the S-state node changes with the number of neighbors. N The number of S-state nodes increases linearly; (2) The threshold of S-state nodes increases with the number of neighbors. N The number of S-state nodes increases exponentially; (3) The threshold of S-state nodes increases with the number of neighbors. N The number of dynamic nodes grows logarithmically; specifically:
[0061] The invention designs three threshold adaptive change strategies. Assuming that the S-state node v i Neighbors I N The number of nodes in the state is n neg :The first is the linear threshold adaptive change strategy, node v i The threshold value is Growth based on The second is the exponential change strategy, where node v i The threshold value is Growth based on The third is the logarithmic change strategy, where node v i The threshold value is Growth based on The variable δ is the node threshold growth rate indicator, and its value is set to 0.01; the range of the node threshold is [0, 1].
[0062] S4: Network information cascade propagation process, considering two propagation modes, namely "with time difference" and "without time difference", the difference lies in whether the S state node immediately changes to I with probability p P The nodes propagate information according to the threshold model mechanism. For the S-state nodes that meet the threshold conditions, the node state changes with probability p according to whether there is a "time difference" until no new non-S-state nodes appear in the network and the propagation ends. Specifically, it includes:
[0063] The method for implementing information dissemination in a social network based on adaptive threshold change is characterized in that: in the information dissemination process of step S4, there are two situations: "there is a time difference" and "there is no time difference"; "time difference" means that the S-state node v i After satisfying the threshold judgment formula, it will immediately change to I with probability p P state node, or first transform to the E state and then transform to the I state with probability p after a period of time. P state;
[0064] The specific propagation process of “no time difference” is as follows:
[0065] S4.1: Randomly select a node in the network as the source of the comment, publish a positive comment, and set its state to I P state;
[0066] S4.2: At the beginning of each step of information propagation, for the S-state node v in the current network i , according to its neighbors I N The number of state nodes n neg And randomly select a threshold adaptive change strategy to update its node threshold
[0067] S4.3: Randomly select an S-state node v i , assuming that the number of positive comments posted by neighbor nodes is n pos , using the following threshold determination formula:
[0068]
[0069] Where k is the node v i The number of neighbors; for the S-state node v that satisfies the threshold judgment formula (1) i , generate a random number r and compare it with the probability value p. When the random number r is less than or equal to p, node v i The state changes to I P state, otherwise it will become I N state;
[0070] S4.4: Repeat steps S4.1 to S4.3 until no new non-S-state nodes appear in the network;
[0071] The specific propagation process of "time difference" is as follows:
[0072] T4.1: Set all node clocks to 0; randomly select a node in the network as the source of the comment, publish a positive comment, and set its status to I P state;
[0073] T4.2: At the beginning of each step of information propagation, the clock of each E-state node is updated by adding 1. For E-state nodes that reach a certain "time difference", they are transformed into I-state nodes with probability p. P state node, transforms to I with probability 1-p N state node; and for the S state node v in the current network i , according to its neighbors I N The number of state nodes n neg Update its node threshold
[0074] T4.3: Randomly select an S-state node v i , assuming that among its neighbor nodes I P The number of state nodes is n pos , I N The number of state nodes is n neg , the number of E-state nodes is n mid , using the following threshold determination formula:
[0075]
[0076] For an S-state node v that satisfies the threshold judgment formula (2), i , the node state immediately changes to E state, and the node clock is set to 1;
[0077] T4.4: Repeat steps T4.1 to T4.4 until no new non-S-state nodes appear in the network;
[0078] S5: Calculate the final cascade range of information propagation. Perform multiple experiments for each set of experimental indicators and calculate the proportion of non-S-state nodes in the total number of nodes in the network after the propagation is completed. Specifically, it includes:
[0079] Change the network average degree value z and transform the S-state nodes into I P The probability p of the state node and the selection of different threshold adaptive change strategies are carried out for 10*N times. The proportion of the number of non-S state nodes in the network after the propagation is calculated as the final information propagation cascade range, and the average value of multiple experiments is calculated.
[0080] The present invention also includes a social network information dissemination implementation system based on threshold adaptive change, such as Figure 5 As shown, it includes: a random network construction module, an initialization module, a threshold adaptive change module, an information cascade propagation module and a cascade range calculation module connected in sequence;
[0081] The random network construction module constructs a random network including N nodes and random edges between nodes, and the average degree of the network is recorded as z; specifically, it includes:
[0082] Consider a random network with average degree z, N=1000 nodes, and edges satisfying
[0083] The initialization module assigns the same initial threshold to all nodes in the network. Each node in the network v i The threshold value is recorded as Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P State; specifically including:
[0084] Each node v in the network i All have node thresholds, denoted as For all nodes in the network, the same initial threshold is given Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state P state as the source of transmission.
[0085] The threshold adaptive change module includes two types of threshold adaptive change strategies: the first type is a linear threshold adaptive change strategy, where the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg Linear growth; the second category is the nonlinear threshold adaptive change strategy, which includes two nonlinear strategies, one is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg The other is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg It grows logarithmically. When used, three threshold adaptive change strategies are randomly selected, including:
[0086] The invention designs three threshold adaptive change strategies. Assuming that the S-state node v i Neighbors I N The number of nodes in the state is n neg :The first is the linear threshold adaptive change strategy, node v i The threshold value is Growth based on The second is the exponential change strategy, where node v i The threshold value is Growth based on The third is the logarithmic change strategy, where node v i The threshold value is Growth based on The variable δ is the node threshold growth rate indicator, and its value is set to 0.01; the range of the node threshold is [0, 1]; and the three threshold adaptive change strategies are randomly selected.
[0087] The information cascade propagation module includes two propagation modes, namely, "with time difference" and "without time difference", the difference lies in whether the S-state node immediately changes to I with probability p Pstate node; randomly select a node as the information dissemination source, and carry out information dissemination according to the threshold model mechanism. For the S-state node that meets the threshold condition, the node state is changed with probability p according to the actual situation of whether there is a "time difference", until no new non-S-state nodes appear in the network, and the dissemination ends; specifically, it includes:
[0088] The specific propagation process of “no time difference” is as follows:
[0089] S4.1: Randomly select a node in the network as the source of the comment, publish a positive comment, and set its state to I P state;
[0090] S4.2: At the beginning of each step of information propagation, for the S-state node v in the current network i , according to its neighbors I N The number of state nodes n neg And randomly select a threshold adaptive change strategy to update its node threshold
[0091] S4.3: Randomly select an S-state node v i , assuming that the number of positive comments posted by neighbor nodes is n pos , according to the threshold judgment formula:
[0092]
[0093] Where k is the node v i The number of neighbors; for the S-state node v that satisfies the threshold judgment formula (1) i , generate a random number r and compare it with the probability value p. When the random number r is less than p, node v i The state changes to I P state, otherwise it will become I N state;
[0094] S4.4: Repeat steps S4.1 to S4.3 until no new non-S-state nodes appear in the network;
[0095] The specific process of propagation of “time difference” is as follows:
[0096] T4.1: Set all node clocks to 0; randomly select a node in the network as the source of the comment, publish a positive comment, and set its status to I P state;
[0097] T4.2: At the beginning of each step of information propagation, the clock of each E-state node is updated by adding 1. For the E-state node that meets the "time difference" condition, it is transformed into I with probability p. P state node, transforms to I with probability 1-pN state node; and for the S state node v in the current network i , according to its neighbors I N The number of state nodes n neg And randomly select a threshold adaptive change strategy to update its node threshold
[0098] T4.3: Randomly select an S-state node v i , assuming that among its neighbor nodes I P The number of state nodes is n pos , I N The number of state nodes is n neg , the number of E-state nodes is n mid , according to the threshold judgment formula:
[0099]
[0100] For an S-state node v that satisfies the threshold judgment formula (2), i , the node state immediately changes to E state, and the node clock is set to 1;
[0101] T4.4: Repeat steps T4.1 to T4.4 until no new non-S-state nodes appear in the network.
[0102] The cascade range calculation module conducts 10*N experiments for each set of experimental indicators z, p and different threshold adaptive change strategies, calculates the proportion of the number of non-S-state nodes in the network after the propagation is completed in the total number of network nodes, as the final information propagation cascade range, and calculates the average value of multiple experiments, specifically including:
[0103] Change the network average degree value z and transform the S-state nodes into I P The probability p of the S-state node and the selection of different threshold adaptive change strategies are carried out for each group of experiments for 10*N=10*1000=10000 times. The ratio of the number of non-S-state nodes in the network to the total number of network nodes after the propagation is completed under each group of indicators is calculated as the final information propagation cascade range, and the average value of multiple experiments is calculated.
[0104] The random network construction module, the threshold initialization module, the threshold adaptive change module, the information cascade propagation module and the cascade range calculation module are connected in sequence.
[0105] The contents described in the embodiments of this specification are merely an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms described in the embodiments. The protection scope of the present invention also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.
Claims
1. A method for implementing social network information dissemination based on threshold adaptive change, characterized by: The social network information propagation model based on threshold adaptive change includes four node states: uncontacted state S, contacted state without comment E, contacted state with positive comment I P , contacted and posted negative comments status I N ; comprising the following steps: S1: Construct a network G = (V, E), where V = {v1, v2, ..., v N } is the network node set, N is the total number of nodes, is the network edge set; S2: Initialization: assign the same initial threshold to all nodes in the network Randomly select a node as the information dissemination source; S3: Consider three threshold adaptive change strategies: (1) The threshold of the S-state node changes with the number of neighbors. N The number of S-state nodes increases linearly; (2) The threshold of S-state nodes increases with the number of neighbors. N The number of S-state nodes increases exponentially; (3) The threshold of S-state nodes increases with the number of neighbors. N The number of dynamic nodes grows logarithmically; S4: Network information cascade propagation process, considering two propagation modes, namely "with time difference" and "without time difference", the difference is whether the S state node immediately changes to I with probability p P The nodes propagate information according to the threshold model mechanism. For the S-state nodes that meet the threshold conditions, the node state changes with probability p according to whether there is a "time difference" until no new non-S-state nodes appear in the network, and the propagation ends. In the process of information propagation, there are two situations: "there is a time difference" and "there is no time difference". "Time difference" means that the S-state node v i After satisfying the threshold judgment formula, it immediately changes to I with probability p P state node, or first transform to the E state and then transform to the I state with probability p after a period of time Δt P The specific propagation process of "no time difference" is as follows: S4.1: Randomly select a node in the network as the source of the comment, publish a positive comment, and set its state to I P state; S4.2: At the beginning of each step of information propagation, for the S-state node v in the current network i , according to its neighbors I N The number of state nodes n neg And randomly select a threshold adaptive change strategy to update its node threshold S4.3: Randomly select an S-state node v i , the number of positive comments posted by neighbor nodes is n pos , using the following threshold determination formula: Where k is the node v i The number of neighbors; for the S-state node v that satisfies the threshold judgment formula (1) i , generate a random number r and compare it with the probability value p. When the random number r is less than or equal to p, node v i The state changes to I P state, otherwise it will become I N state; S4.4: Repeat steps S4.1 to S4.3 until no new non-S-state nodes appear in the network; In the process of information dissemination, the dissemination process of "time difference" is as follows: T4.1: Set all node clocks to 0; randomly select a node in the network as the source of the comment, publish a positive comment, and set its status to I P state; T4.2: At the beginning of each step of information propagation, the clock of each E-state node is updated by adding 1. For E-state nodes that reach a certain "time difference", they are transformed into I-state nodes with probability p. P state node, transforms to I with probability 1-p N state node; and for the S state node v in the current network i , according to its neighbors I N The number of state nodes n neg Update its node threshold T4.3: Randomly select an S-state node v i , among its neighbor nodes I P The number of state nodes is n pos , I N The number of state nodes is n neg , the number of E-state nodes is n mid , using the following threshold determination formula: For an S-state node v that satisfies the threshold judgment formula (2), i , the node state immediately changes to E state, and the node clock is set to 1; T4.4: Repeat steps T4.1 to T4.3 until no new non-S-state nodes appear in the network; S5: Calculate the final cascade range of information propagation, conduct multiple experiments for each set of experimental indicators, and calculate the proportion of non-S-state nodes in the total number of nodes in the network after the propagation is completed.
2. The method for implementing social network information dissemination based on threshold adaptive change as claimed in claim 1, characterized in that: The step S1 specifically includes: Consider a random network with average degree z, N=1000 nodes, and edges satisfying 3. The method for implementing social network information dissemination based on threshold adaptive change as claimed in claim 1, characterized in that: The step S2 specifically includes: Each node v in the network i All have node thresholds, denoted as For all nodes in the network, the same initial threshold is given Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P state.
4. The method for implementing social network information dissemination based on threshold adaptive change as claimed in claim 1, characterized in that: The step S3 specifically includes: Design three threshold adaptive change strategies, S-state node v i Neighbors I N The number of nodes in the state is n neg :The first is the linear threshold adaptive change strategy, node v i The threshold value is Growth based on The second is the exponential change strategy, where node v i The threshold value is Growth based on The third is the logarithmic change strategy, where node v i The threshold value is Growth based on The variable δ is the node threshold growth rate indicator, and its value is set to 0.01; the range of the node threshold is [0, 1].
5. The method for implementing social network information dissemination based on threshold adaptive change as claimed in claim 1, characterized in that: The step S5 is specifically as follows: Change the network average degree value z and transform the S-state nodes into I P The probability p of the state node and the selection of different threshold adaptive change strategies are carried out for 10*N times. The proportion of the number of non-S state nodes in the network after the propagation is calculated as the final information propagation cascade range, and the average value of multiple experiments is calculated.
6. A system for implementing the method for social network information dissemination based on threshold adaptive change as claimed in claim 1, comprising: Random network construction module, initialization module, threshold adaptive change module, information cascade propagation module and cascade range calculation module; The random network construction module constructs a random network including N nodes and random edges between nodes, and the average degree of the network is recorded as z; The initialization module assigns the same initial threshold to all nodes in the network. Each node in the network i All have node thresholds, denoted as Set the initial state of all nodes in the network to S state, and randomly select a node to be in I state. P state; The threshold adaptive change module includes two types of threshold adaptive change strategies: the first type is a linear threshold adaptive change strategy, where the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg Linear growth; the second category is the nonlinear threshold adaptive change strategy, which includes two nonlinear strategies, one is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg The other is the S-state node v i The threshold is based on the number of neighbors I N The number of state nodes n neg Logarithmically increasing; three threshold adaptive change strategies are randomly selected when used; The information cascade propagation module includes two propagation modes, namely "with time difference" and "without time difference", the difference lies in whether the S-state node immediately changes to I with probability p P state node; randomly select a node as the information dissemination source, and carry out information dissemination according to the threshold model mechanism. For the S state node that meets the threshold condition, it changes to I state node with probability p according to the actual situation of whether it has "time difference". P state nodes until no new non-S state nodes appear in the network, and the propagation ends; The cascade range calculation module performs 10*N experiments for each set of experimental indicators z, p and different threshold adaptive change strategies, calculates the proportion of the number of non-S-state nodes in the network after the propagation is completed in the total number of network nodes, as the final information propagation cascade range, and calculates the average value of multiple experiments; The random network construction module, initialization module, threshold adaptive change module, information cascade propagation module and cascade range calculation module are connected in sequence.
Citation Information
Patent Citations
Method for realizing network information transmission for true and false messages
CN106992966A
Online social network information dissemination method based on status theory
CN113254719A
