Computer Network Information System Security Prevention Methods under the Background of Big Data

,


Introduction
e degree of social informatization continues to improve, forming a globally integrated information society, which leads to the rapid increase in people's dependence on computer network [1]. e computer network has been worked in people's daily life and become one of the important tools in the process of learning, given the convenience and the open sharing of computer network. Computer information network has been not only worked for people's life and great learning convenient but also brought a greater threat to the society, especially with the new era of computer network hacker wood horse disease and arti cial destruction. e use of network information brings great security threat to computer, and the information technology is gradually changing our way of thinking and way of life Internet-based information globalization provides human beings with a huge information-based communication platform, which has become an indispensable part of people's daily production and life [2,3]. At the same time, a series of operation management service system based on the Internet has emerged at the right moment and shows a strong prospect for development.
Since the arrival of the era of big data, dig data connect countries all over the world into an inseparable network [4]. Countries around the world are also competing to expand the value extension of big data, no matter military or political, information, as the most valuable resource of human society is the key to economic and social development at present but also the winning factor of international social competition. If information resources are grasped, it will be the artery of economic development. Government departments have the most comprehensive control. With the help of big data, the most authoritative information resources and the construction of e-government information resource sharing are accelerated. Meanwhile, the administrative level and capacity of the Chinese government are promoted [5].
e government now are paying more and more resources sharing in the face of the present situation of the high-speed growth for egovernment information resources. ey are constantly exploring the big data technologies, that is, how to improve the level of e-government information resources and sharing large data e-government information resource construction. It is the fundamental basis of information sharing and efficient; the lack of effective management of information resource sharing will be a serious impediment to government departments, and between regions of the communication and collaboration. It is unfavorable to improve administrative efficiency and it is also unable to highlight the advantages of big data technology. To this end, the State Council issued the interim, that is, the measures for the management of government information resource sharing in 2016, aiming to accelerate the development and risk prevention of the interconnection [6].
At the same time, the big data under the background of the construction of government affair information shares first article only in the first half of the year of 2019. e national computer network emergency coordination center as well as the detection and coordination of the disposal of the government website has been tampered, including the universal security vulnerabilities [7]. As government websites not only represent the external image of the government, but also contain a large amount of important information and data, it will pose a great threat to the Chinese government and citizens under the background of data to strengthen risk prevention that has become one of the important issues faced by governments to speed up the construction of the electronic office.
In the face of the increasingly severe information security situation, how to protect information safely and effectively has become the focus of information security research [8]. e solution to information security problems involves all aspects, and the formulation of relevant policies and regulations, the improvement of management standards, and the improvement of information security technology are indispensable. In order to solve this problem, it is necessary to take relevant measures to prevent information security risks from the perspective of system engineering. Based on the information system, which can identify the risk assessment and control, to reduce unnecessary risks at the same time the risk of objective existence should be cut to an acceptable level. Information security risk assessment, as the core of information security management, is to use reasonable methods and scientific analysis to assess the risks faced by the system, so as to have an objective and comprehensive cognition of the possible harm degree of security events [9].
For the potential risk of the information system of uncertainty and random problems, from asset value, the potential threat to information security risk assessment, and a comprehensive evaluation of information system analysis, and to judge the probability of the system security incidents and the damage they may cause, the corresponding risk management measures are put forward. rough information security risk assessment, on the one hand, the security status of the system and the main security risks can be defined, to provide effective security for information system security [10]. On the other hand, the evaluation results can provide a reference for the construction of a safety technology system and a management system. However, traditional methods are difficult to carry out effective and comprehensive improvement of information processing speed requirements and the increasingly complicated and complex business processes that need to be dealt with, resulting in the continuous expansion of the system scale and the complexity and dynamic characteristics of the system architecture [11]. In addition to the potential systems mentioned above, information systems are also susceptible to the following risks. e main significance of information security risk assessment lies in the following: (1) comprehensive cognition and confirmation of security risks, and risk assessment is a process of confirming risks and analyzing risks as a whole, which is the key to solve information security problems.
rough information security risk assessment, we can effectively recognize the risks faced by the evaluated system and the severity of these risks. Managers should be helped to fully understand the causes of risks, so as to choose targeted measures to reduce and control risks; (2) the focus should be provided on information security, and information security is faced with a variety of risks, only on the basis of risk assessment to establish a sound information security management mechanism, in order to effectively manage to avoid the possible risks [12][13][14].

Related Work
Compared with Europe and the United States and other developed countries, China's information security risk assessment development lags behind, after the 1980s. With the promotion and application of computers, computer security issues began to attract attention, but due to the lack of relevant experience, the information security protection effect is not ideal in the 1990s. e Internet application is wide. China also began to increase the information security risk assessment research, based on which it developed the corresponding information security technical standards and information management norms [15]. With the deepening of the understanding and research of information security issues, China's attention to information security risk assessment is also increasing in February 1994, and China promulgated the regulations that clearly put forward the implementation of security-level protection of the information system. After a period of time to increase the importance of information security, since February 2005, the Information Office of the State Council has carried out risk assessment pilot projects on the relevant information systems of important infrastructure facilities such as the People's Bank of China, the State Grid, and the State Administration of Taxation in many places to explore the methods of risk assessment in order to improve the relevant standards [16]. In 2007, GB/T 20984-2007 was officially released, which specifically deployed the content and operation steps of risk assessment, pointing out the direction for the further development of China's information security risk assessment, and risk assessment has become an important part of China's information security protection work [17][18][19].
Experts and scholars have paid attention to the influencing factors of government affairs information sharing, measures to improve the level of sharing, and mechanisms and strategies to establish government affairs information sharing and prevention of risks in government affairs information sharing. Literature [20][21][22][23] studied how to better realize the sharing of government information among various government departments and conducted empirical analysis and research by constructing a conceptual model of influencing factors of government information sharing. e model mainly started from technical means, organization, management, laws and regulations, and other factors. Literature [23][24][25] studied the current situation of government affairs information sharing in China and believed that factors affecting the regulation construction planning and the government's lack of understanding of the importance of information resource sharing and the lack of financial investment focused on the influence of government affairs in their research. rough empirical research, they concluded that individual organization and environment are the three elements of government information sharing. Literature [26] emphasized government in their research. e sharing of government information resources is a complicated project.
is study analyzes some main obstacles in the sharing of government information in China and puts forward some corresponding solutions.
Foreign scholars and research institutions have been studied information security risk assessment for a long time, and many remarkable achievements have been made in the field of risk assessment. At present, developed countries led by Britain and Germany have built relatively complete information risk defense architecture and information security guarantee architecture on the basis of early research and application. On the whole, the development process of foreign information security risk assessment represented by the United States has roughly experienced the following three stages: the first stage, computer-based information security stage. e concept and measures of communication security have existed for a long time, but it was not until the 1950s that systematic theoretical research on information security began. In the 1960s and 1970s of last century, the computer began to be gradually applied to the government and the army; at this point, the confidentiality of the information security risk assessment is focused on information authentication, and access control and audit are the most basic link in the information security control, at the same time, such as the path analysis and integrity detection. e second stage is the information security protection based on computer and network. After the 1980s, computer networks began to appear and be used on a large scale, and confidentiality alone could not meet the needs of information security protection. At this time, risk assessment began to consider more information security attributes represented by integrity and availability. With the "General Assessment Criteria for Information Technology Security" (CC) as the representative and the formulation of various standards, information security risk assessment began to override a single technology, from the perspective of overall protection to protect information security. e third stage is the information security guarantee stage aimed at ensuring the critical infrastructure of the information system. With the advent of the twenty-first century, the application of computer network is becoming more and more widely, the information system as a core infrastructure in many areas, represented by the United States, developed countries began to ascend to the level of the national strategic security information security, information security attributes and further admittedly expands controllability. e object of information security and safeguard measures has become clearer.
Foreign experts and scholars have carried out extensive and in-depth research in the field of information security risk assessment and put forward many risk assessment models and methods. Literature [24] puts forward a quantitative assessment method based on investigation and analyzes security risks by considering current demands. Literature [27] proposed an evaluation method based on the combination of fuzzy set theory and D-S evidence theory and finally determined the risks of the whole system by calculating the risks of each component. Literature [28] puts forward an asset evaluation model considering asset correlation, which represents the dependency relationship between assets and asset types, and then uses the security risk analysis method to detect and determine the priority of security risks, and then a method for comprehensive evaluation of information security risk level, which combines decision laboratory analysis (DEMATEL) with network analysis (ANP), so as to reduce the subjectivity of evaluation results, in order to identify the causal relationship between risk factors. e risk analysis model uses a Bayesian network to simultaneously define risk factors and their causal relationship, so as to obtain the maximum probability risk value of the information system. Literature [29] proposed a fuzzy decision theory, which effectively improved the reliability and accuracy of assessment results. Literature [30] proposed a risk assessment method for information security based on Bayesian theory to deduce and calculate the probability of risk occurrence in an information system. Literature [31] used a Bayesian risk graph to model risks, so as to quantitatively calculate the risks faced by the system. Literature [32] proposed an information security risk assessment method based on the combination of a genetic algorithm and BP neural network. e contributions of this study are summarized as follows: (1) is study analyzes the information security status of small Internet of ings system based on the characteristics of Internet of ings information security (2)

e Big Data Era.
With the rapid development of modern society, the progress of science and technology, information quantity speed growth, and an increasingly important role of information flow of communication between people more and more close, big data are in this context of arrogance, since the data are read in from all areas of social attention, but for how to define the concept of big data, at present. ere is no unified opinion in the academic circle, which usually refers to the dataset that cannot be captured, managed, and processed within a certain period of time by using conventional software tools. Big data generally have five characteristics, such as scale, high speed, diversity, value, and authenticity scale that refers to that the value and potential information of data depending on the size of data. A certain data scale is a basis for big data to play its role. High speed means that large-scale data determine the high speed of data acquisition. Diversity refers to the variety of data types and wide source channels; value refers to the use of big data that can create high value at a low cost. True real sex should be a real big index according to the data analysis that can be used to predict user behavior analysis. Otherwise, some other advanced data analysis in today's society can be said to be a digital society. Along with its application in the field of increasingly widespread, the big data can be applied to many aspects of our lives which is deeply changing our way of life. At present, the application of big data has been attached to great importance by governments all over the world. Various types of large databases of local governments in China are gradually being built, and government information is also one of the important aspects of big data application. Figure 1 gives the visualizing impact of big data on computing performance.

Improvement of Analytic Hierarchy Process.
Generally speaking, the AHP method in the process of evaluating hierarchy weighting accuracy that depends on the construction of the judgment matrix, the traditional layer, and at the same time the analysis solution of different assessment results, often using the simple averaging method on the whole, while ignoring the differences behind the implicit information, therefore, the introduction of the concept of information entropy and entropy value method, with the aid of entropy. e value method adjusts the weights according to the estimated differences. erefore, information entropy can be used as the standard to measure the value of information to some extent: e greater the uncertainty of a variable, the greater the entropy, and the more information it takes to make sense of it. erefore, it is a better method to use information entropy to measure the relationship among attributes when evaluating different decision-making processes. When the evaluation of the genus is beneficial (the larger the value is, the better), When the evaluation attribute is cost (the smaller the better), en, the entropy of the evaluated attribute can be calculated by the following formula: (4) e entropy weight calculation of attributes is evaluated: e output error of neuron of the kth output layer is as follows: e sum of error energy of the system output layer is as follows: en, according to the AHP scale theory, pairwise comparative judgment matrices are constructed for different schemes, where K represents the number of schemes: e constructed judgment matrix is normalized e ownership weight vector is combined into a weight matrix, where k j represents the weight of the j index in the kth scheme: Accordingly, there are Based on the equations (1)-(12), Figure 2 gives the security domain division of the information system based on AHP.

Introduction to Experimental
Dataset. Current e-government information shares in government in the field of academic research, but the academic research and practical operation are put in bigger difference. erefore, China's egovernment information sharing in the current process remains in the exploratory stage and the relevant government affair information sharing system of laws and regulations formulates technical operation personnel. e operation mode and so on are in the rising and developing stage, and the academic research on government affairs information sharing in this stage has no accurate and mature questionnaire for reference compared with the side stress theory and data analysis. In the process of questionnaire design, it is fully based on the relevant provisions and policies of the current government affairs information sharing state, combined with the relevant research of the academic circle, and deeply discusses its development status.
A total of 387 questionnaires on the status quo of government information sharing were distributed. After  Mathematical Problems in Engineering manual and machine screening, 296 valid questionnaires were finally obtained, with an efficiency of 76.5%. As 80% of the questionnaires were distributed online, the quality of the questionnaires could not be guaranteed. According to the set of questions and the control of the basic information of the respondents, and in order to ensure the validity of the data, a relatively strict screening method is adopted, which is also the reason for the low efficiency of the questionnaire. However, the efficiency of the questionnaire is not directly proportional to the validity of the questionnaire data, so the validity of the questionnaire data and survey results will not be affected.

Experimental Result Analysis.
In order to verify the correctness of the improved AHP risk assessment model, this chapter takes different examples to study the risk assessment model proposed above. rough the above process, the weight vector of each index layer is obtained, and then, the fuzzy evaluation method is used to design the evaluation set for the importance of the index. Considering people's subjective evaluation results, it is more suitable to be expressed in the form of intervals. From the range of risk assessment results, the risk degree of the system is lower than the middle level, there are certain security risks, and appropriate control measures can be taken to enhance the    system security protection. Under this evaluation method, when setting risk response strategies, specific details can be evaluated by combining index evaluation value interval and hierarchy weight. For example, it can be seen from the evaluation value of the fourth-level index that there are loopholes in the authority management of the system. In the second-level index system, the weight of data communication risk is the highest, so it is necessary to increase investment in risk management. It is necessary to take appropriate information security risk control measures through the comprehensive index system of each layer index weight and risk value interval. Finally, the experimental results of different models are compared as shown in Figure 3.
Taking the small Internet of ings system introduced above as an example, the effect of the improved HAHP risk assessment model was tested through simulation experiments. e early risk assessment results of 70 such systems were used as simulation samples. e sample acquisition method was mainly a questionnaire. In combination with some risk assessment tools, 60 samples were used as training samples, and 10 samples were used as test samples. e samples were quantitatively processed by the fuzzy evaluation method to facilitate training and testing. Due to the limitations and inadequacy of the previous assessment samples, the risk assessment factors are adjusted and simplified in combination with the original risk assessment index system, and the risk assessment results are quantified into point value form to facilitate neural network training and detection. e risk assessment hierarchy model before and after improvement is shown in Figure 4. e validity of the method has been demonstrated in Figure 4, and the stability of the method will be verified below. In order to test the stability of the proposed method, several independent experiments were carried out to test the effectiveness of the algorithm in view of the randomness of the experimental results of the neural network. e trend chart of some test results is shown in Figure 5. It can be seen from the figure that the repeated experiment results are basically stable without much fluctuation. is suggests that the algorithm has good stability and repeatability, and several experimental results are obtained with the original risk value of the average contrast. It can be seen from the chart that the predictions of the simulation results and the actual risk assessment results are very close, the error between the two is less than 3%, and this shows that the algorithm can predict and prevent the risk assessment result. However, the results of the comparison algorithm are not very stable in the experimental process, and the results of several experiments are not ideal. Compared with this, the proposed method can better predict the risk results, which proves that the proposed algorithm has better applicability and stability in risk assessment. Among China's growing number of Internet users, some unwittingly violate Internet security by failing to pay attention to online activities according to law. e parameter of the computer crime model is set to population size. e scores of the search results are compared, as shown in Figure 6.
In this study, the JQC-3FF relay is designed to realize switching control of energy-saving socket power supply. PC0∼PC3 pin output control signal of MCU drives the control end of the relay. e corresponding relay module can also be configured according to the actual application situation to achieve multichannel control. In order to ensure that the control signal output by MCU can normally drive the relay module, the triode should be connected to amplify the control signal. At the same time, a continuing-current protection diode is also designed below the relay to prevent the reverse induction current generated by the relay coil from damaging the triode and other original devices, as shown in Figure 6.
Network in the learning process of the AHP may constantly adjust neurons and it make it step by step towards the optimal output of the party to study and optimize the network learning rate that will affect how fast and near it is.
at is, it is to determine the training sample when the learning rate is too low. e adjustment and correction progress of the matrix is slow, which leads to slow convergence speed and even discontinuation of training. When the learning rate is too high, the network tends to fall into local optimum and has poor generalization ability. When the learning rate is low, the accuracy of clustering results is low. It can be seen from the figure that the clustering accuracy of this network reaches 91.2%. If the network learning rate is improved, the clustering accuracy is not significantly improved, indicating that the network topology has been stable. If the network learning rate is changed, the network generalization capability may be reduced. All the above results are shown in Figure 7.
e test sample sets were, respectively, input into the trained network for testing. Both neural networks could effectively predict the risk values of the evaluation samples. e prediction results of risk assessment are shown in Figure 8. e network test error is obtained by risk error calculation of the expected output of the test sample set and the actual output of the network. e maximum error of the improved AHP method and the traditional BP neural network is 0.57 0.306, and the minimum error is 0.27 0.74, respectively. e average error and standard deviation of the two risk assessment methods are calculated. e results show that the proposed method has a smaller error and better risk prediction ability. e evaluation results are analyzed, and the validity and reliability of the method are verified.
rough the comparison experiment with the traditional BP neural network, the AHP has certain advantages in the convergence performance and risk prediction of network training.

Conclusions
With the deepening of information construction, the existing evaluation methods existing in complex mixed degree is higher, too subjective, and lack operability and many other limitations. erefore, to explore a kind of efficient and reliable risk assessment method has very important significance.
On the basis of fully understanding the research status at home and abroad, this study deeply studies the risk assessment methods based on the analytic hierarchy process (AHP) and neural network and triviality. An improved AHP risk assessment model and a risk prediction model based on an optimized neural network are proposed. However, this study is mainly focused on medium-sized data at present. How to solve the problem of big data is worth studying in the future.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.