Group viewpoint guiding method, system and device based on individual fixity

Abstract

The invention belongs to the technical field of predictive optimization, and particularly relates to a group viewpoint guiding method, system and device based on individual fixity, which deeply considers individual fixity difference characteristics on the basis of viewpoint evolution cooperatively driven by an information network and a social network. Guided individuals are divided into a fixed person group and a non-fixed person group according to individual fixed degree differences, a targeted guiding method is designed for the two groups, and for non-fixed extreme individuals, the fixed degree of the individuals is intervened by using an intelligent information recommendation technology and a'hinting 'psychology; for the fixed extreme individuals, the interest change rate is used for describing the process that the interests of the fixed extreme individuals change along with the proportion of the positive viewpoints in the viewpoint environment, then the viewpoint variation is calculated, and finally the effect of guiding the group viewpoints to develop towards the preset expected viewpoints is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of prediction and optimization technology, and specifically relates to a method, system and device for guiding group opinions based on individual stubbornness. Background Art

[0002] The Opinion Evolution Model is an important tool for studying the evolution of group opinions. For example, the publication number CN117290433A discloses an opinion evolution model driven by the synergy of information networks and social networks. When the opinion evolution trend has a restrictive effect on social stability, business development, and the direction of public opinion, a key scientific opinion guidance method is urgently needed to effectively guide it to ensure that social activities develop in a healthy direction. At present, relevant research on opinion guidance strategies includes: guidance strategies that make some nodes ineffective, guidance strategies based on informed agents, degree guidance strategies, and guidance strategies based on opinion leaders.

[0003] The above guidance strategies usually change the opinions of individuals in the model too rigidly, that is, directly impose the target opinions on individuals without considering the individual's acceptance of the target opinions. However, the evolution of individual opinions and opinion guidance is a complex process involving multiple factors. Directly changing individual opinions or forcibly adding target opinion values ​​to individuals will lose research significance. Compared with directly pushing a very different guiding opinion to the guided individual at the beginning of the guidance, the guidance strategy based on moderate opinions calculates the distance between the opinions of the two interacting parties. If the individual can accept the guiding target opinion, the target opinion is directly pushed. Otherwise, a moderate opinion that the individual can accept and is biased towards the target opinion is pushed.

[0004] However, the above guidance strategy does not take into account the impact of individual stubbornness differences caused by different personalities on the guidance of group opinions, nor does it consider the use of different strategies to guide the group. Summary of the invention

[0005] The present invention provides a method, system and device for guiding group opinions based on individual stubbornness.

[0006] The technical solution of the present invention is as follows: The present invention provides a method for guiding group opinions based on individual stubbornness, comprising the following steps: S1: Based on the interaction between information, an information network is constructed; according to the reliability of the information content in the information network, the reliability of the information source and the external recognition of the information, the influence weight of the information on the current individual opinion is calculated, which is recorded as the first weight; According to the first weight and the current individual opinion, the opinion is evolved to obtain the evolution of the individual opinion in the information network, which is recorded as the first opinion; S2: Based on the interaction between individuals, a social network is constructed; according to the cognitive ability of individuals in the social network and the intimacy between individuals, the influence weight of neighbor individuals on the current individual's opinion is calculated, which is recorded as the second weight; The opinion is evolved according to the second weight and the first opinion, and the evolution of the individual opinion in the social network is obtained, which is recorded as the second opinion; S3: Based on the preset expected viewpoint, the difference between the second viewpoint and the preset expected viewpoint is calculated, and individuals whose difference is not less than the first preset difference are recorded as extreme individuals. The proportion of the number of extreme individuals in the social network is calculated. If the proportion is greater than or equal to the guidance threshold, S4 is executed; S4: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

[0007] In step S4, based on the dynamic stubbornness, the second weight and the first opinion, the opinion of the non-stubborn extreme individual is updated, specifically: According to the formula: , update the views of non-stubborn extreme individuals; In the formula, A newer perspective for non-stubborn extreme individuals; Indicates the current round; is the dynamic stubbornness; The current view of non-stubborn extreme individuals; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

[0008] In step S4, the dynamic stubbornness of the non-stubborn extreme individual is calculated according to the opinion similarity threshold, specifically: Obtain the number of opinions received by non-stubborn extreme individuals, compare the similarity between the opinions of non-stubborn extreme individuals and the received opinions, record opinions with similarity not less than the opinion similarity threshold as similar opinions, and obtain the number of similar opinions; record opinions with similarity less than the opinion similarity threshold as repulsive opinions, and obtain the number of repulsive opinions; If the number of similar opinions is greater than the number of repulsive opinions, the individual stubbornness is linearly increased and used as the dynamic stubbornness of the non-stubborn extreme individual; If the number of similar opinions is equal to the number of repulsive opinions, the individual stubbornness is taken as the dynamic stubbornness of the non-stubborn extreme individual; If the number of repelling opinions is greater than the number of similar opinions, the individual stubbornness will be linearly reduced and used as the dynamic stubbornness of the non-stubborn extreme individual.

[0009] In step S4, based on the change in opinion, the individual stubbornness, the second weight and the first opinion, the opinion of the extremely stubborn individual is updated, specifically: According to the formula: , update the views of stubborn and extreme individuals; In the formula, A newer perspective for stubborn extreme individuals; for individual stubbornness; For the current views of stubborn extreme individuals; is the amount of change in opinion; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

[0010] In step S4, based on the interest change rate, the change amount of the opinion of the stubborn extreme individual is calculated, specifically: According to the formula: , , calculate the change in opinions of extremely stubborn individuals; In the formula, is the amount of change in opinion; The rate of change of interest.

[0011] In step S4, based on the viewpoint environment, the interest change rate of the stubborn extreme individual is calculated, specifically: According to the formula: , , calculate the rate of change of interests of extremely stubborn individuals; In the formula, The rate of change for the benefit; Represents the view environment.

[0012] In step S4, the opinion environment is calculated according to the individual opinion of each individual in the social network, specifically: According to the formula: ,Computational Viewpoint Environment; In the formula, Indicates the viewpoint environment; represents the number of individuals with positive opinions in the social network; Represents the total number of individuals in the social network.

[0013] After step S4, the method further includes calculating the difference between the guided opinion and the preset expected opinion. If the proportion of individuals whose difference is not less than the second preset difference in the social network group is greater than or equal to the preset proportion, step S4 is repeated.

[0014] The present invention also provides a group opinion guidance system based on individual stubbornness, comprising: Information network opinion evolution module: Based on the interaction between information, an information network is constructed; according to the reliability of information content, the reliability of information sources and the external recognition of information in the information network, the influence weight of information on the current individual opinion is calculated, which is recorded as the first weight; According to the first weight and the current individual opinion, the opinion is evolved to obtain the evolution of the individual opinion in the information network, which is recorded as the first opinion; Social network opinion evolution module: Based on the interaction between individuals, a social network is constructed; according to the cognitive ability of individuals in the social network and the intimacy between individuals, the influence weight of neighbor individuals on the current individual's opinion is calculated, which is recorded as the second weight; The opinion is evolved according to the second weight and the first opinion, and the evolution of the individual opinion in the social network is obtained, which is recorded as the second opinion; Judgment module: based on the preset expected opinion, calculate the difference between the second opinion and the preset expected opinion, record the individuals whose difference is not less than the first preset difference as extreme individuals, calculate the proportion of the number of extreme individuals in the social network, and if the proportion is greater than or equal to the guidance threshold, enter the opinion guidance module; Opinion guidance module: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

[0015] The present invention also provides a group opinion guiding device based on individual stubbornness, comprising a processor and a memory, wherein the processor implements the group opinion guiding method based on individual stubbornness when executing a computer program stored in the memory.

[0016] Beneficial Effects Based on the evolution of opinions driven by the synergy of information networks and social networks, the present invention deeply considers the characteristics of individual stubbornness differences, divides the guided individuals into stubborn groups and non-stubborn groups according to the differences in individual stubbornness, and designs targeted guidance methods for these two groups. For non-stubborn extreme individuals, intelligent information recommendation technology and "suggestion" psychology are used to intervene in the stubbornness of individuals; for stubborn extreme individuals, the interest change rate is used to describe the process of how the interests of stubborn extreme individuals change with the proportion of positive opinions in the opinion environment, and then calculates the change in opinions, which ultimately guides the group's opinions to develop towards the preset expected opinions. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of a method for guiding group opinions based on individual stubbornness is provided in an embodiment.

[0018] Figure 2 (a) is the simulation result of the ISOE method without a guiding strategy, and (b) is the simulation result of the PDGM guided method of this application.

[0019] Figure 3 This is a trend chart of the change in group average opinions under the ISOE method and the PDGM method.

[0020] Figure 4 This is a trend chart of the evolution steps when the viewpoint guidance method of the present application reaches a stable state.

[0021] Figure 5 This is a trend chart of changes in the average opinions of a group using the opinion guidance method of this application under different initial data distributions.

[0022] Figure 6 A graph showing the changing trend of the proportion of different extremely stubborn individuals versus the number of evolutionary steps when the viewpoint guidance method of the present application is used to reach a stable state.

[0023] Figure 7This is a trend chart of the changes in the average opinions of the group using the opinion guidance method of this application under different proportions of the number of stubborn and extreme individuals.

[0024] Figure 8 This is a trend diagram of the change in the number of evolution steps when the viewpoint guidance method of the present application reaches a stable state for different guidance ratios. DETAILED DESCRIPTION

[0025] The following examples are intended to illustrate the present invention rather than to further limit the present invention.

[0026] The present invention provides a method for guiding group opinions based on individual stubbornness, such as Figure 1 As shown, the following steps are included: S1: Based on the interaction between information, an information network is constructed; according to the reliability of the information content in the information network, the reliability of the information source and the external recognition of the information, the influence weight of the information on the current individual opinion is calculated, which is recorded as the first weight; The opinion is evolved according to the first weight and the current individual opinion, and the evolutionary opinion of the individual opinion in the information network is obtained, which is recorded as the first opinion.

[0027] First, in the information network, as disclosed in publication number CN117290433A, the network structure in the opinion evolution model driven collaboratively by information networks and social networks is introduced. Each piece of information has a unique information ID, nodes represent information, and edges represent interactions between information.

[0028] Then, according to the reliability of information content in the information network, the reliability of information sources and the external recognition of information, the influence weight of information on the current individual opinion is calculated and recorded as the first weight, which is: According to the formula: , calculate the first weight; In the formula, is the first weight; Indicates the reliability of information content; The reliability of the information source; External recognition of information; , and The larger it is, the more authentic and reliable the information is; are the coefficients of the reliability of information content, the reliability of information source and the external recognition of information, respectively, and , .

[0029] Among them, the method for judging the reliability of information content and the reliability of information source refers to patent CN117290433A.

[0030] For the external recognition of information, first use the formula: , quantify the external recognition of information; In the formula, Indicates the external recognition of quantitative information; The like rate of the information. is the comment rate of the information, is the forwarding rate of information; are the coefficients of the information’s like rate, information’s comment rate, and information’s forwarding rate, respectively. .

[0031] According to the formula: , normalize it and get the external recognition of information; In the formula, It represents the minimum value of the external recognition of the quantitative information; Represents the maximum value of the external recognition of quantitative information.

[0032] Finally, the opinion is evolved according to the first weight and the current individual opinion, and the evolution of the individual opinion in the information network is obtained, which is recorded as the first opinion. Specifically: According to the formula: , get the first view; In the formula, For the first point of view; Indicates the current round; For individuals Stubbornness; For individuals Initial view; represents the comprehensive impact of information in the information network on the current individual's opinion, where is the first weight; Represents an individual Views from social networks.

[0033] Regarding individual stubbornness, stubbornness refers to a person's insistence on his or her own views and unwillingness to change or compromise, while stubbornness reflects the degree of a person's stubbornness. This application assumes that the degree of stubbornness of individuals to their original views is heterogeneous, and the individual stubbornness is recorded as .in, The closer it is to 1, the more individual The more stubborn a person is, the harder it is to change his or her opinion. The closer it is to 0, the more individual The less stubborn you are, the easier it is to change your views.

[0034] S2: Based on the interaction between individuals, a social network is constructed; according to the cognitive ability of individuals in the social network and the intimacy between individuals, the influence weight of neighbor individuals on the current individual's opinion is calculated, which is recorded as the second weight; The opinion is evolved according to the second weight and the first opinion, and the evolution of the individual opinion in the social network is obtained, which is recorded as the second opinion.

[0035] First, in a social network, nodes represent individuals (users), each of whom has an opinion, and edges represent connections between individuals. The edges between social networks and information networks represent the interaction between individuals and information, that is, individuals can both publish information on information networks and obtain information from information networks. The two-layer network structure based on information networks and social networks has been disclosed in the "Information Network and Social Network Collaboratively Driven Opinion Output Method and System" with publication number CN117290433A, which will not be repeated here.

[0036] Secondly, according to the cognitive ability of individuals in the social network and the intimacy between individuals, the influence weight of neighbor individuals on the current individual's opinion is calculated, which is recorded as the second weight. Specifically, According to the formula: ; , calculate the second weight; In the formula, is the second weight; A quantitative value that represents the cognitive ability of individuals in a social network and the closeness between individuals; Neighbor cognitive abilities; For individuals With neighbor individuals The intimacy between , are the coefficients of individual cognitive ability and inter-individual intimacy, respectively.

[0037] Finally, the opinion is evolved according to the second weight and the first opinion, and the evolution of the individual opinion in the social network is obtained, which is recorded as the second opinion. Specifically: According to the formula: = , get a second opinion; In the formula, For the second point of view; For individuals Stubbornness; For the first point of view; Represents neighbor individuals in a social network The combined impact on current views, including is the second weight.

[0038] Individual opinions exist in the form of information in the information network, that is, individuals (users) can obtain information from various online media platforms and publish their own opinions on various media platforms. Therefore, this application assumes that the change of individual opinions will eventually form their own opinions through alternating evolution on information networks and social networks. In addition, in the collaborative process between social networks and information networks, after each round of opinion evolution, the opinions in the social network and the information in the information network will be synchronously updated through the inter-layer information transmission mechanism. Specifically, in the information network, once the evolution process is over, due to the connection between layers, the updated information of the individual will be promptly transmitted to the social network. Similarly, in the social network, once the evolution is over, due to the connection between layers, the updated opinions of the individual will also be promptly input into the information network. Therefore, the inter-layer information transmission mechanism ensures the synchronous update between the two layers of the network.

[0039] S3: Based on the preset expected viewpoint, the difference between the second viewpoint and the preset expected viewpoint is calculated, and individuals whose difference is not less than the first preset difference are recorded as extreme individuals. The proportion of the number of extreme individuals in the social network is calculated. If the proportion is greater than or equal to the guidance threshold, S4 is executed.

[0040] In the process of opinion evolution, when the proportion of extreme individuals in the social network group reaches a certain guidance threshold, the guidance activity will be started; when the proportion of extreme individuals is lower than the guidance threshold, the guidance activity will stop. This application denotes the guidance threshold as .

[0041] S4: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

[0042] For non-stubborn extreme individuals, they are good at interacting with others and will actively seek opinions and views from others. With the existence of information networks, they can be exposed to a large amount of suggestive information. Therefore, under the influence of the "suggestion" effect, the individual's stubbornness in his original views will be weakened, and then the acceptance of information with opposite views will continue to increase, and then gradually shorten the distance between the extreme individual's views and the preset expected views, and finally make the non-stubborn extreme individuals accept the expected views, achieving the effect of guiding the group's views.

[0043] Therefore, for non-extremely stubborn individuals, this application uses intelligent information recommendation technology and "suggestion" psychology to intervene in the stubbornness of the individual.

[0044] According to the formula: , , and obtain the dynamic stubbornness of non-stubborn extreme individuals.

[0045] In the formula, is the dynamic stubbornness of non-stubborn extreme individuals, indicating the stubbornness after intervention; is the individual stubbornness, indicating the stubbornness before intervention; Indicates the number of similar opinions; Indicates the number of opposing viewpoints; , , is a constant.

[0046] The specific operations are as follows: First, according to the opinion similarity threshold, the dynamic stubbornness of non-stubborn extreme individuals is calculated, specifically: Obtain the number of opinions received by non-stubborn extreme individuals, compare the similarity between the opinions of non-stubborn extreme individuals and the received opinions, record opinions with similarity not less than the opinion similarity threshold as similar opinions, and obtain the number of similar opinions; record opinions with similarity less than the opinion similarity threshold as repulsive opinions, and obtain the number of repulsive opinions; If the number of similar opinions is greater than the number of repulsive opinions, the individual stubbornness is linearly increased and used as the dynamic stubbornness of the non-stubborn extreme individual; If the number of similar opinions is equal to the number of repulsive opinions, the individual stubbornness is taken as the dynamic stubbornness of the non-stubborn extreme individual; If the number of repelling opinions is greater than the number of similar opinions, the individual stubbornness will be linearly reduced and used as the dynamic stubbornness of the non-stubborn extreme individual.

[0047] Among them, the opinion similarity threshold is used to determine whether the opinions of the two parties in the interaction are similar enough. This application assumes that the opinion similarity threshold is homogeneous and this threshold is recorded as , .

[0048] In other words, this application aims to describe the dynamic change process of individual stubbornness under the influence of intelligent information recommendation and "suggestion" psychology. Considering that the influence of "suggestion" effect on individual stubbornness is mainly reflected in two aspects, namely: (1) Reinforcement effect: Among all the opinions interacting with non-stubborn extreme individuals, when the number of similar opinions received by an individual exceeds the number of repulsive opinions, it will have a reinforcing effect on the individual's original opinions. This effect makes the individual more stubborn about the original opinion, making it more difficult for him to accept the opposite opinion.

[0049] (2) Inhibition effect: Among all the opinions interacting with non-stubborn extreme individuals, when the number of repulsive opinions received by an individual exceeds the number of similar opinions, an inhibitory effect will be produced on the individual's original opinions. This effect reduces the individual's stubbornness towards the original opinion, making it easier to accept the opposite opinion.

[0050] Then, based on the dynamic stubbornness, the second weight and the first opinion, the opinion of the non-stubborn extreme individual is updated, specifically: According to the formula: , update the views of non-stubborn extreme individuals; In the formula, A newer perspective for non-stubborn extreme individuals; is the dynamic stubbornness; The current view of non-stubborn extreme individuals; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

[0051] For extremely stubborn individuals, they tend to be stubborn and have difficulty accepting the opinions and views of others. Therefore, it is not ideal to achieve the purpose of guidance by adjusting their stubbornness. However, when the majority of the group is guided to hold positive opinions, the interests (credibility, authority, etc.) of stubborn people with negative opinions will be lost in the opinion environment, forcing them to change their opinions.

[0052] Therefore, for stubborn extreme individuals, this application first assumes that the stubbornness of stubborn extreme individuals is static and remains unchanged during the process of opinion guidance. However, under the pressure of the "opinion environment", when most group members hold positive opinions, stubborn extreme individuals holding negative opinions will face loss of interests, prompting them to change their existing opinions to cater to the "dominant" opinions.

[0053] In the process of opinion guidance, most individuals in the group gradually move closer to the preset expected opinions. As the proportion of positive opinions increases, the attention, credibility, and authority of stubborn and extreme individuals with negative opinions will also decrease. This application refers to this phenomenon as "decline in interests". For this reason, this application uses the interest change rate to describe the process in which the interests of stubborn and extreme individuals change with the proportion of positive opinions in the opinion environment.

[0054] Furthermore, in the process of reducing benefits, the stubborn extreme individuals have to change their views to cater to the existing "dominant" views. In this application, the views changed due to the reduction of benefits are referred to as the change of views.

[0055] The specific operations are as follows: First, considering that the change of interests is closely related to the proportion of positive opinions in the group, it is preferred to calculate the opinion environment based on the individual opinions of each individual in the social network, specifically: According to the formula: ,Computational Viewpoint Environment; In the formula, Indicates the viewpoint environment; represents the number of individuals with positive opinions in the social network; Represents the total number of individuals in the social network.

[0056] Secondly, consider that the interests of stubborn extreme individuals will vary with the change in the number of people with positive views. When the number of individuals with positive views is low, the interests of stubborn people with negative views will decline slowly; and when the number of individuals with positive views increases, the interests of stubborn people with negative views will decline rapidly. Therefore, the rate of interest decline shows a trend of first slow and then fast. Preferably, based on the view environment, the rate of change of the interests of stubborn extreme individuals is calculated, specifically: According to the formula: , , calculate the rate of change of interests of extremely stubborn individuals; In the formula, The rate of change for the benefit; Represents the view environment.

[0057] Thirdly, based on the rate of interest change, the change in opinion of the stubborn extreme individual is calculated as follows: According to the formula: , , calculate the change in opinions of extremely stubborn individuals; In the formula, is the amount of change in opinion; The rate of change of interest.

[0058] Finally, based on the change in opinion, individual stubbornness, the second weight, and the first opinion, the opinion of the extremely stubborn individual is updated as follows: According to the formula: , update the views of stubborn and extreme individuals; In the formula, Updated perspectives for stubborn extreme individuals; for individual stubbornness; For the current views of stubborn extreme individuals; is the amount of change in opinion; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

[0059] In addition, after completing a round of guidance using the above guidance method, it is determined whether the opinions of the social network group have reached a stable state, where the opinions of the social network group after guidance are similar to the preset expected opinions. The difference between the opinions after guidance and the preset expected opinions is calculated. If the proportion of individuals whose difference is not less than the second preset difference in the social network group is greater than or equal to the preset proportion, step S4 is repeated; otherwise, the guidance is terminated.

[0060] Based on the evolution of opinions driven by the synergy of information networks and social networks, the present invention deeply considers the characteristics of individual stubbornness differences, divides the guided individuals into stubborn groups and non-stubborn groups according to the differences in individual stubbornness, and designs targeted guidance methods for these two groups to guide the group opinions to develop towards preset expected opinions.

[0061] The present invention also provides a group opinion guidance system based on individual stubbornness, comprising: Information network opinion evolution module: Based on the interaction between information, an information network is constructed; according to the reliability of information content, the reliability of information sources and the external recognition of information in the information network, the influence weight of information on the current individual opinion is calculated, which is recorded as the first weight; According to the first weight and the current individual opinion, the opinion is evolved to obtain the evolution of the individual opinion in the information network, which is recorded as the first opinion; Social network opinion evolution module: Based on the interaction between individuals, a social network is constructed; according to the cognitive ability of individuals in the social network and the intimacy between individuals, the influence weight of neighbor individuals on the current individual's opinion is calculated, which is recorded as the second weight; The opinion is evolved according to the second weight and the first opinion, and the evolution of the individual opinion in the social network is obtained, which is recorded as the second opinion; Judgment module: based on the preset expected opinion, calculate the difference between the second opinion and the preset expected opinion, record the individuals whose difference is not less than the first preset difference as extreme individuals, calculate the proportion of the number of extreme individuals in the social network, and if the proportion is greater than or equal to the guidance threshold, enter the opinion guidance module; Opinion guidance module: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

[0062] The present invention also provides a group opinion guiding device based on individual stubbornness, comprising a processor and a memory, wherein the processor implements the group opinion guiding method based on individual stubbornness when executing a computer program stored in the memory.

[0063] Experimental simulation analysis To ensure the accuracy and reliability of the experiment, this application sets the default values ​​of various parameters. The specific parameter settings are shown in Table 1.

[0064] Table 1 Experimental dataset

[0065] 1. Analysis of the evolution of viewpoints and the guiding process In order to verify the effectiveness of the group opinion guidance method (denoted as PDGM) proposed in this paper, this paper uses the opinion evolution only (denoted as ISOE) method as a comparison, which has no guidance strategy. This paper simulates the opinion evolution trend of ISOE and PDGM, and its maximum step size Set to 10000, social desirability view The initial opinion of the group is set to 1.0, and the initial opinion of the group is a random value between [0,1]. The rest of the parameters are tested using the default values ​​set in Table 1. Each case of the experimental results is the average result after 100 repeated experiments. The experiment stops when the group's opinion reaches a stable state or reaches the maximum step length, and the number of evolution steps is used as the speed of opinion evolution. The simulation results are shown in Figure 2 , Figure 3 As shown, Figure 2 The X-axis represents the number of evolutionary steps when the group's opinion evolution process reaches a stable state, and the Y-axis represents the individual's opinion value; Figure 3 The X-axis represents the number of evolutionary steps when the group opinion evolution process reaches a stable state, and the Y-axis represents the average opinion value of the group.

[0066] from Figure 2 It can be seen that compared with the ISOE method without guidance strategy ( Figure 2 Compared with (a), the proposed guided method PDGM ( Figure 2 (b) can make the group's views gradually approach the socially expected views within a certain period of time. This shows that in the PDGM method, individuals continue to receive information containing socially expected views recommended by intelligent information recommendation technology, and then under the influence of the "suggestion" psychology, the individual's stubbornness in his original views will gradually decrease, and then the trust in the information containing socially expected views will gradually increase. Therefore, in this process, the individual's views continue to approach the socially expected views, thereby achieving the effect of guiding views.

[0067] Figure 3 The trend of the average opinion of the group under the ISOE method and the PDGM method was simulated. The group's opinion is set to reach a stable state at 0.5. In the ISOE method, due to the absence of a guiding strategy, the final average opinion of the group is 0.5, while in the PDGM method, due to the existence of a guiding method, the average opinion of the group gradually increases between [0.5, 1] ​​and finally reaches the target opinion.

[0068] 2. The impact of different initial viewpoint distributions on guidance effects In order to verify the influence of different initial data distributions on the guiding effect of the bootstrapping method (PDGM) proposed in this application, this paper takes the initial data distribution as the independent variable, that is, sets the proportion of extreme individuals in all individuals in the group to 0, 20%, 40%, 60%, and 80%, to observe the experimental results. The experiment was carried out under an artificial network. This application sets the maximum step size to T=10000, and the initial opinions of individuals are randomly distributed between [0, 1]. In addition, in order to avoid the randomness of parameters, except for the initial opinions of the group, the remaining parameters use the default values ​​in Table 1, and each case of the experimental results is the average result obtained after 100 repeated experiments. The simulation results are shown in the figure. Figure 4 , Figure 5 As shown, Figure 4 The X-axis represents the proportion of extreme individuals in all individuals in the group, and the Y-axis represents the number of evolutionary steps when the opinion guidance process reaches a stable state; Figure 5 The X-axis represents the number of evolutionary steps when the opinion guidance process reaches a stable state, and the Y-axis represents the average opinion value of the group during the opinion guidance process.

[0069] Figure 4 The bar graph in the figure shows the specific value of the evolutionary steps when the opinion guidance process reaches a stable state, and the line graph shows the changing trend of the evolutionary steps. Figure 4 It can be seen that the greater the proportion of extreme opinions in the initial opinions of the group, the longer it takes for the opinion guidance process to reach a stable state. In other words, the time required for the opinion guidance process to reach a stable state is proportional to the number of individuals directly guided. This is because the proportion of positive opinions in the group's opinions is an important factor in the change of the interests of stubborn individuals. When the proportion of extreme individuals in a group is greater, the slower the interests of stubborn individuals decline, the longer it takes for their opinions to be changed, and therefore the time it takes for the group's opinions to be guided becomes longer. Therefore, the initial opinion distribution of the group, that is, the initial state of public opinion, is one of the important factors affecting the effect of opinion guidance. The more extreme the initial state of public opinion, the longer it takes for public opinion to be guided.

[0070] Figure 5 The line graph in shows the changes in the average opinion of the group under different initial data distributions. Figure 5 It can be seen that due to the different proportions of extreme individuals in the group in the initial state, the initial average views of the group are different, and the time and speed of guidance to reach a stable state are also different. Specifically, when the proportion of extreme individuals in the group is larger, the time and speed of the guidance process to reach a stable state are longer and slower; when the proportion of extreme individuals in the group is smaller, the time and speed of the guidance process to reach a stable state are shorter and faster. Therefore, in real-life public opinion events, the more people hold negative views, the more difficult it is to guide the event; conversely, the guidance is less difficult.

[0071] 3. The impact of the proportion of extremely stubborn individuals on the guidance effect In order to verify the influence of the proportion of different stubborn individuals on the guiding effect of the guiding method (PDGM) proposed in this application, this application takes the proportion of stubborn individuals as the independent variable, that is, sets the proportion of stubborn individuals in all individuals in the group to 0, 20%, 40%, 60%, and 80%, to observe the experimental results. The experiment was carried out under an artificial network. This application sets the maximum step size to T=10000, and the initial opinions of individuals are randomly distributed between [0,1]. In addition, in order to avoid the randomness of parameters, except for the stubbornness of the group as a variable, the remaining parameters use the default values ​​in Table 1 as constants in the experiment, and each case of the experimental results is the average result obtained after 100 repeated experiments. The simulation results are shown in the figure. Figure 6 ,in Figure 6 The X-axis represents the proportion of stubborn individuals in all individuals in the group, and the Y-axis represents the number of evolutionary steps when the opinion guidance process reaches a stable state.

[0072] Figure 6 The bar graph in the figure shows the specific value of the evolutionary steps when the opinion guidance process reaches a stable state, and the line graph shows the changing trend of the evolutionary steps. Figure 6 It can be seen that the greater the proportion of stubborn extreme individuals in the group, the longer it takes for the opinion guidance process to reach a stable state. In other words, the time required for the opinion guidance process to reach a stable state is proportional to the number of stubborn extreme individuals in the group. This is because one of the factors affecting the opinion guidance effect of stubborn extreme individuals is the opinion environment. Therefore, when the number of non-stubborn extreme individuals is small and the number of stubborn extreme individuals is large, the opinion environment changes slowly. Therefore, the time it takes for the opinions of stubborn extreme individuals to be guided becomes longer. Therefore, the time required for the opinion guidance process to reach a stable state is proportional to the number of stubborn extreme individuals in the group.

[0073] Figure 7 The line graph in shows how the average opinion of a group changes when there are different proportions of stubborn individuals. Figure 7 It can be seen that due to the different proportions of stubborn individuals in the group, the initial average views of the group are different, and the time and speed of guidance to reach a stable state are also different. Specifically, when the proportion of stubborn individuals in the group is larger, the time it takes for the guidance process to reach a stable state is longer and the speed is slower; when the proportion of stubborn individuals in the group is smaller, the time it takes for the guidance process to reach a stable state is shorter and the speed is faster. Therefore, in real-life public opinion events, the more people hold negative views and are stubborn, the more difficult it is to guide the event; the fewer people hold positive views and are stubborn, the easier it is to guide.

[0074] 4. The influence of guidance ratio on guidance effect In order to verify the influence of the guidance ratio on the guidance effect of the guidance method (PDGM) proposed in this application, this application uses the guidance ratio as the independent variable of the experiment in this section, that is, the proportion of individuals directly guided in all individuals in the group is set to 40%, 50%, 60%, 70%, and 80%, to observe the experimental results. The simulation experiment is carried out under an artificial network. This application sets the maximum step size to T=10000, and the initial views of the individuals are randomly distributed between [0,1]. In addition, in order to avoid the randomness of the parameters, except that the guidance ratio value is dynamic, the other parameters use the default values ​​in Table 1, and each case of the experimental results is the average result obtained after 100 repeated experiments. The simulation results are shown in the figure. Figure 8 As shown, the X-axis represents the guidance ratio value, and the Y-axis represents the number of evolution steps when the opinion guidance process reaches a stable state.

[0075] Figure 8 The bar graph in the figure shows the specific value of the evolutionary steps when the opinion guidance process reaches a stable state, and the line graph shows the changing trend of the evolutionary steps. Figure 8 It can be seen that the greater the proportion of individuals directly guided in the group, the shorter the time required for the opinion guidance process to reach a stable state. In other words, the time required for the opinion guidance process to reach a stable state is inversely proportional to the number of individuals directly guided in the group. Experiments show that the guidance method proposed in this application can have a good guiding effect on individuals directly guided.

Claims

1. A method for guiding group opinions based on individual stubbornness, characterized in that: The following steps are involved: S1: Build information network based on the interaction between information; According to the reliability of information content in the information network, the reliability of information sources and the external recognition of information, the influence weight of information on the current individual's opinion is calculated and recorded as the first weight; According to the first weight and the current individual opinion, the opinion is evolved to obtain the evolution of the individual opinion in the information network, which is recorded as the first opinion; S2: Building social networks based on interactions between individuals; According to the cognitive ability and intimacy between individuals in the social network, the influence weight of neighbor individuals on the current individual's opinion is calculated and recorded as the second weight; According to the second weight and the first opinion, the opinion evolution is performed to obtain the evolution of the individual opinion in the social network, which is recorded as the second opinion; S3: Based on the preset expected viewpoint, the difference between the second viewpoint and the preset expected viewpoint is calculated, and individuals whose difference is not less than the first preset difference are recorded as extreme individuals. The proportion of the number of extreme individuals in the social network is calculated. If the proportion is greater than or equal to the guidance threshold, S4 is executed; S4: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

2. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, based on the dynamic stubbornness, the second weight and the first opinion, the opinion of the non-stubborn extreme individual is updated, specifically: According to the formula: , update the views of non-stubborn extreme individuals; In the formula, A newer perspective for non-stubborn extreme individuals; Indicates the current round; is the dynamic stubbornness; The current view of non-stubborn extreme individuals; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

3. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, the dynamic stubbornness of the non-stubborn extreme individual is calculated according to the opinion similarity threshold, specifically: Obtain the number of opinions received by non-stubborn extreme individuals, compare the similarity between the opinions of non-stubborn extreme individuals and the received opinions, record opinions with similarity not less than the opinion similarity threshold as similar opinions, and obtain the number of similar opinions; record opinions with similarity less than the opinion similarity threshold as repulsive opinions, and obtain the number of repulsive opinions; If the number of similar opinions is greater than the number of repulsive opinions, the individual stubbornness is linearly increased and used as the dynamic stubbornness of the non-stubborn extreme individual; If the number of similar opinions is equal to the number of repulsive opinions, the individual stubbornness is taken as the dynamic stubbornness of the non-stubborn extreme individual; If the number of repelling opinions is greater than the number of similar opinions, the individual stubbornness will be linearly reduced and used as the dynamic stubbornness of the non-stubborn extreme individual.

4. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, based on the change in opinion, the individual stubbornness, the second weight and the first opinion, the opinion of the extremely stubborn individual is updated, specifically: According to the formula: , update the views of stubborn and extreme individuals; In the formula, A newer perspective for stubborn extreme individuals; for individual stubbornness; For the current views of stubborn extreme individuals; is the amount of change in opinion; Represents neighbor individuals The combined impact on current views, including is the second weight; For the first point of view.

5. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, based on the interest change rate, the change amount of the opinion of the stubborn extreme individual is calculated, specifically: According to the formula: , , calculate the change in opinions of extremely stubborn individuals; In the formula, is the amount of change in opinion; The rate of change of interest.

6. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, based on the viewpoint environment, the interest change rate of the stubborn extreme individual is calculated, specifically: According to the formula: , , calculate the rate of change of interests of extremely stubborn individuals; In the formula, The rate of change for the benefit; Represents the view environment.

7. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: In step S4, the opinion environment is calculated according to the individual opinion of each individual in the social network, specifically: According to the formula: ,Computational Viewpoint Environment; In the formula, Indicates the viewpoint environment; represents the number of individuals with positive opinions in the social network; Represents the total number of individuals in the social network.

8. The method for guiding group opinions based on individual stubbornness according to claim 1, characterized in that: After step S4, the method further includes calculating the difference between the guided opinion and the preset expected opinion. If the proportion of individuals whose difference is not less than the second preset difference in the social network group is greater than or equal to the preset proportion, step S4 is repeated.

9. A group opinion guidance system based on individual stubbornness, characterized in that: include: Information network viewpoint evolution module: build information network based on the interaction between information; According to the reliability of information content in the information network, the reliability of information sources and the external recognition of information, the influence weight of information on the current individual's opinion is calculated and recorded as the first weight; According to the first weight and the current individual opinion, the opinion is evolved to obtain the evolution of the individual opinion in the information network, which is recorded as the first opinion; Social network opinion evolution module: build social networks based on interactions between individuals; According to the cognitive ability and intimacy between individuals in the social network, the influence weight of neighbor individuals on the current individual's opinion is calculated and recorded as the second weight; According to the second weight and the first opinion, the opinion evolution is performed to obtain the evolution of the individual opinion in the social network, which is recorded as the second opinion; Judgment module: based on the preset expected opinion, calculate the difference between the second opinion and the preset expected opinion, record the individuals whose difference is not less than the first preset difference as extreme individuals, calculate the proportion of the number of extreme individuals in the social network, and if the proportion is greater than or equal to the guidance threshold, enter the opinion guidance module; Opinion guidance module: Based on individual stubbornness, extreme individuals whose stubbornness is not greater than the preset stubbornness threshold are recorded as non-stubborn extreme individuals, and extreme individuals whose stubbornness is greater than the preset stubbornness threshold are recorded as stubborn extreme individuals. Opinion guidance is performed on non-stubborn extreme individuals and stubborn extreme individuals respectively. The specific opinion guidance operation is as follows: For non-stubborn extreme individuals, the dynamic stubbornness of non-stubborn extreme individuals is calculated based on the opinion similarity threshold; Based on the dynamic stubbornness, the second weight and the first opinion, the opinions of non-stubborn extreme individuals are updated; For extremely stubborn individuals, the opinion environment is calculated based on the individual opinions of each individual in the social network; Calculate the rate of change of interests of stubborn extreme individuals based on the viewpoint environment; Based on the rate of interest change, calculate the change in opinion of extremely stubborn individuals; Update the opinions of extremely stubborn individuals based on the amount of opinion change, individual stubbornness, the second weight, and the first opinion.

10. A group opinion guiding device based on individual stubbornness, characterized in that: The method comprises a processor and a memory, wherein when the processor executes the computer program stored in the memory, the method for guiding group opinions based on individual stubbornness as described in any one of claims 1 to 8 is implemented.

Images