Social Network Security Research Articles

Enhancing Security in Social Networks through Machine Learning: Detecting and Mitigating Sybil Attacks with SybilSocNet

This study contributes to the Sybil node-detecting algorithm in online social networks (OSNs). As major communication platforms, online social networks are significantly guarded from malicious activity. A thorough literature review identified various detection and prevention Sybil attack algorithms. An additional exploration of distinct reputation systems and their practical applications led to this study’s discovery of machine learning algorithms, i.e., the KNN, support vector machine, and random forest algorithms, as part of our SybilSocNet. This study details the data-cleansing process for the employed dataset for optimizing the computational demands required to train machine learning algorithms, achieved through dataset partitioning. Such a process led to an explanation and analysis of our conducted experiments and comparing their results. The experiments demonstrated the algorithm’s ability to detect Sybil nodes in OSNs (99.9% accuracy in SVM, 99.6% in random forest, and 97% in KNN algorithms), and we propose future research opportunities.

The role of agroforestry systems for addressing climate change livelihood vulnerability of farmers of Northwestern Ethiopia

Farmers in Ethiopia have been vulnerable to climate change in recent decades. In the face of this change, farmers have managed agroforestry systems to maintain their livelihoods. However, studies exploring the role of agroforestry in reducing household vulnerability are lacking in Northwestern Ethiopia. The objectives of the study were to (i) investigate households’ livelihoods vulnerability to climate change in Northwestern Ethiopia; (iii) assess the role of agroforestry in mitigating the negative impacts of climate change on farmers livelihoods. Key informant interviews, in-depth interviews, household surveys (387), and focus group discussions were used to collect the data. Descriptive statistics, principal component analysis, the X2-test, and the t-test were run to analyze the data. The findings revealed that households are vulnerable to rising temperatures, rainfall variability, frost, disease and pests, erosion, hailstorms, price hikes, wildlife damage to crops, and health stress. Agroforestry non-practitioners had a higher livelihood vulnerability index (LVI) (0.42 ± 0.081) than practitioners (0.46 ± 0.079). The Livelihood Vulnerability-Intergovernmental Panel for Climate Change Index (LVI-IPCC) showed that AF non-practitioners had a higher exposure (0.58), sensitivity (0.54) index, and a lower adaptive capacity index (0.44) than the exposure (0.34), sensitivity (0.38), and adaptive capacity index (0.51) of practitioners. Plant diversity, income level and diversity, livelihood activities, social network, and food security status of farmers were improved by agroforestry. Farmers were therefore less susceptible to adverse climate shocks. Thus, the AF system could be part of future adaptation and resilience programs that provide dependable tools to minimize households’ vulnerability to climate shocks. However, management guidelines, such as understanding local ecosystems, setting clear objectives, choosing suitable species, planning for diversity, considering the market, and regular maintenance and monitoring, are needed for agroforestry to improve its contribution.

DarkMor: A framework for darknet traffic detection that integrates local and spatial features

The dark web, as an integral part of the multi-layered structure of the internet, provides anonymity and high levels of concealment absent on the surface web. Users can engage in various online activities without leaving any trace, making the dark web a hub for illicit activities, such as drug and weapon trafficking. Consequently, this poses a significant threat to social order and network security. However, due to the high concealment of the dark web, traditional detection methods suffer with insufficient extraction of characteristic information from darknet traffic and inadequate consideration of feature correlations. As a result, the accuracy of these conventional detection methods in detecting the dark web in real network environments is suboptimal. This paper proposes a darknet traffic detection framework called DarkMor to address these challenges. It integrates local and spatial features, automates feature mining and fusion, and models spatial relationship between features to fully exploit their potential. The core components of DarkMor consist of the feature fusion module and the traffic perception module. Using an improved feature tokenizer transformer architecture, the feature fusion module enhances the extraction of local features within high-dimensional feature clusters, effectively combining local feature information with global context. Additionally, the traffic perception module leverages a temporal model that incorporates self-attention mechanisms to learn the spatiotemporal characteristics of fused features, thereby further enhancing the model’s detection. Experimental results demonstrated that DarkMor achieved an accuracy of 97.78% on real network datasets, surpassing the latest cross-modal darknet traffic detection models. Furthermore, DarkMor maintained an accuracy of 97.57% even in network environments with reduced training samples, confirming the feasibility and robustness of the proposed detection framework.

Cross-Social-Network User Identification Based on Bidirectional GCN and MNF-UI Models

Due to the distinct functionalities of various social network platforms, users often register accounts on different platforms, posing significant challenges for unified user management. However, current multi-social-network user identification algorithms heavily rely on user attributes and cannot perform user identification across multiple social networks. To address these issues, this paper proposes two identity recognition models. The first model is a cross-social-network user identification model based on bidirectional GCN. It calculates user intimacy using the Jaccard similarity coefficient and constructs an adjacency matrix to accurately represent user relationships in the social network. It then extracts cross-social-network user information to accomplish user identification tasks. The second model is the multi-network feature user identification (MNF-UI) model, which introduces the concept of network feature vectors. It effectively maps the structural features of different social networks and performs user identification based on the common features of seed nodes in the cross-network environment. Experimental results demonstrate that the bidirectional GCN model significantly outperforms baseline algorithms in cross-social-network user identification tasks. The MNF-UI (multi-network feature user identification) model can operate in situations with two or more networks with inconsistent structures, resulting in improved identification accuracy. These two user identification algorithms provide technical and theoretical support for in-depth research on social network information integration and network security maintenance.

Analysis of Information Security Under the Conditions of Hybrid War in Ukraine: Social Aspects

Abstract The development of the post-industrial society requires the acceleration of the integration of the national economy into the globalized economic space. This stage is characterized by active informatization of all spheres of life in society, which requires information security and cyber protection for high-quality information provision of the country’s population, intellectualization of economic processes and prevention of destructive informational influence on the social status of the individual. It is also necessary to consider that for Ukraine the specified stage is complicated by the hybrid war with Russia, which requires strengthening the protection of information from cyber-attacks and the formation of new approaches in preserving the quality of information on social networks. Therefore, the purpose of this article is the development of scientific and methodological approaches in information security management, strengthening its social significance. This requires solving a certain range of tasks of identifying and preventing socially dangerous information based on the use of economic and mathematical methods and models. The article highlights the main directions of socialization of modern technological development and theoretically substantiates its significant impact on human consciousness and behavior, puts a person in front of serious challenges and, under conditions of hybrid warfare, requires strengthening of information security in social networks. In the work, the components of information security are supplemented, its impact on social aspects of society’s life is highlighted. As a result of the research, the authors proposed a scientific and methodical approach to the construction of a system for countering the spread of socially dangerous information in social networks. Besides, its functional elements are highlighted. A method of combating the spread of harmful information in social networks has been developed, which solves the problem of information support of the decision-making process and includes providing the chosen and alternative options to the person making the decision, with the justification of the choice. This creates prerequisites for information support for making a well-founded management decision.

FedKG: A Knowledge Distillation-Based Federated Graph Method for Social Bot Detection.

Malicious social bots pose a serious threat to social network security by spreading false information and guiding bad opinions in social networks. The singularity and scarcity of single organization data and the high cost of labeling social bots have given rise to the construction of federated models that combine federated learning with social bot detection. In this paper, we first combine the federated learning framework with the Relational Graph Convolutional Neural Network (RGCN) model to achieve federated social bot detection. A class-level cross entropy loss function is applied in the local model training to mitigate the effects of the class imbalance problem in local data. To address the data heterogeneity issue from multiple participants, we optimize the classical federated learning algorithm by applying knowledge distillation methods. Specifically, we adjust the client-side and server-side models separately: training a global generator to generate pseudo-samples based on the local data distribution knowledge to correct the optimization direction of client-side classification models, and integrating client-side classification models' knowledge on the server side to guide the training of the global classification model. We conduct extensive experiments on widely used datasets, and the results demonstrate the effectiveness of our approach in social bot detection in heterogeneous data scenarios. Compared to baseline methods, our approach achieves a nearly 3-10% improvement in detection accuracy when the data heterogeneity is larger. Additionally, our method achieves the specified accuracy with minimal communication rounds.

Unsupervised twitter social bot detection using deep contrastive graph clustering

The emergence of malicious Twitter social bots poses a considerable threat to the security of social networks, and the detection of evolving social bots has become challenging. State-of-the-art detection methods are usually supervised, but the label acquisition process suffers from time-consuming and inaccurate problems in addition to its inability to cope with the challenge of the continuous evolution of social bots. Moreover, structural relationships within social networks are under-explored using current unsupervised methods. To address current challenges, we propose an unsupervised social bot detection method based on deep contrastive graph clustering (BotDCGC). This approach utilises a graph attentional encoder and an inner product decoder to acquire user node embeddings. By incorporating information from both user account features and topological structure, the model employs a contrastive learning technique based on structure to discern node embeddings of distinct classes within the feature space. Subsequently, confident cluster assignments are used as soft labels to guide the embedding process by calculating the similarity between each node and the clustering centre, enabling the joint optimisation of user node embeddings and clustering results. Experiments show that BotDCGC is more effective than the state-of-the-art baselines with an accuracy of 0.8095 in the Twibot-20 dataset and 0.9334 in the Cresci-2015 dataset, and the experimental results indicate the effectiveness of the graph autoencoder module, contrastive learning and deep clustering module.

Blockchain solutions for trustworthy decentralization in social networks

The influence of blockchain on technology, especially in the field of finance, has brought about the age of cryptocurrencies and decentralized control, spreading its impact across various industries to ensure precision. As the trend moves towards decentralized social networks, diverse approaches are explored, becoming crucial tools for defining trust and privacy in individuals’ personal lives during uncertain environments. Blockchain’s transformative role reaches beyond finance, revolutionizing how we establish trust and privacy in our interconnected digital societies. Concurrently, researchers recognize the efficacy of trust models in enhancing the security and reliability of social networks applications. However, a comprehensive review of recent advancements in trust models based on blockchain for social networks is lacking. To address this gap, our survey takes a comprehensive perspective on the theme of trust within decentralized networks. We examine various concepts such as the trust concept itself, trust models, trust prediction methodologies, decentralization in trust-based social networks, and the role of blockchain in fortifying trust and security in social networks. We investigate the intersection of trust and blockchain technology, encompassing concepts, types of blockchain, and their applications in social networks. These models are categorized based on different parameters, and we examine both their advantages and limitations. Finally, we propose prospective paths for researchers interested in exploring this area.

A Fake Profile Detection Model Using Multistage Stacked Ensemble Classification

Fake profile identification on social media platforms is essential for preserving a reliable online community. Previous studies have primarily used conventional classifiers for fake account identification on social networking sites, neglecting feature selection and class balancing to enhance performance. This study introduces a novel multistage stacked ensemble classification model to enhance fake profile detection accuracy, especially in imbalanced datasets. The model comprises three phases: feature selection, base learning, and meta-learning for classification. The novelty of the work lies in utilizing chi-squared feature-class association-based feature selection, combining stacked ensemble and cost-sensitive learning. The research findings indicate that the proposed model significantly enhances fake profile detection efficiency. Employing cost-sensitive learning enhances accuracy on the Facebook, Instagram, and Twitter spam datasets with 95%, 98.20%, and 81% precision, outperforming conventional and advanced classifiers. It is demonstrated that the proposed model has the potential to enhance the security and reliability of online social networks, compared with existing models.

Design of a Trusted Content Authorization Security Framework for Social Media

Artificial intelligence-generated content (AIGC) technology has had disruptive results in AI, representing a new trend in research and application and promoting a new era of AI. The potential benefits of this technology are both profound and diverse. However, the benefits of generative tools are accompanied by a series of significant challenges, the most critical of which is that it may cause AI information pollution on social media and mislead the public. Traditional network security models have shown their limitations in dealing with today’s complex network threats, so ensuring that generated content published on social media accurately reflects the true intentions of content creators has become particularly important. This paper proposes a security framework called “secToken”. The framework adopts multi-level security and privacy protection measures. It combines deep learning and network security technology to ensure users’ data integrity and confidentiality while ensuring credibility of the published content. In addition, the framework introduces the concept of zero trust security, integrates OAuth2.0 ideas, and provides advanced identity authentication, fine-grained access control, continuous identity verification, and other functions, to comprehensively guarantee the published content’s reliability on social media. This paper considers the main issues of generative content management in social media and offers some feasible solutions. Applying the security framework proposed in this paper, the credibility of generated content published on social media can be effectively ensured and can help detect and audit published content on social media. At the operational level, when extracting key information summaries from user-generated multimodal artificial intelligence-generated content and binding them to user identity information as a new token to identify user uniqueness, it can effectively associate user identity information with the current network status and the generated content to be published on the platform. This method significantly enhances system security and effectively prevents information pollution caused by generative artificial intelligence on social media platforms. This innovative method provides a powerful solution for addressing social and ethical challenges and network security issues.

A social network security user recommendation algorithm based on community user emotions

A social network security user recommendation algorithm based on community user emotions

Security analysis of the data of social networks using AI techniques

This proposed research explores the potential of AI techniques, particularly user engagement prediction, for analyzing social network data and identifying potential security threats. Utilizing a Random Forest classifier, we developed a highly accurate model achieving 100% accuracy and a 1.0 AUC-ROC score. This exceptional performance demonstrates the ability of engagement prediction to accurately flag suspicious accounts with unusually low engagement, often associated with bots or fake profiles. Based on these findings, we implemented mitigation strategies such as flagging low-engagement accounts for further investigation and analyzing engagement trends to inform proactive security measures. Furthermore, our work opens doors for future research in combining engagement prediction with other AI techniques like sentiment analysis for even more sophisticated threat detection, ultimately contributing to the development of robust solutions for enhanced social network security and user privacy protection.

Social media platform-oriented topic mining and information security analysis by big data and deep convolutional neural network

The aim of this work is to conduct topic mining and data analysis of social network security using social network-based big data. The deep convolution neural network (DCNN) is used to analyze social network security issues. Traditional neural network models cannot handle long sequence data when extracting information on Weibo security topics. Thus, the long short-term memory (LSTM) structure in the memory intelligence algorithm extracts Weibo topic information. Specifically, the social network security topics are mined through Big Data, and CNN searches Weibo security topics. CNN can learn the grammar and semantic information of Weibo topics to obtain in-depth data features. Afterward, the performance of the improved DCNN model is compared with the AlexNet, Convolutional Neural Network (CNN), and Deep Neural Network (DNN) by considering the model's accuracy, recall, and F1 value, respectively. The experimental results show that after 120 iterations, the recognition accuracy of the improved DCNN model peaks at 96.17 %, at least 5.4 % superior to the other three models. Additionally, the intrusion detection model's accuracy, recall, and F1 value are 88.57 %, 75.22 %, and 72.05 %, respectively. In the worst case, the constructed model's accuracy, recall, and F1 value are 3.1 % higher than those of the other methods. The training and testing time consumption of the improved DCNN security detection model stabilized at 65.86 s and 27.90 s, much shorter than similar literature approaches. The experimental conclusion is that the improved DCNN under deep learning has the characteristic of lower delay, and the model shows good network data security transmission.

Enhancing Social Network Security: A Cloud - Based Approach to Prevent Profile Cloning Attacks

Enhancing Social Network Security: A Cloud - Based Approach to Prevent Profile Cloning Attacks

Exploring the Convergence of Design, Security, and Human Dynamics in Social Networks in India

This Exploration dives into the complex exchange between social networks, zeroing in on their design, security measures, and the impact of human factors. As social networks keep on molding present day correspondence and connection, understanding their elements turns out to be progressively pivotal. The examination investigates the architecture and user-focused design rules that improve user experiences, encourage community associations, and empower self-expression. All the while, the review tends to the squeezing security challenges presented by these platforms, dissecting risks like identity theft and misinformation dispersal, and proposing techniques for powerful encryption and data protection. Moreover, the review digs into the domain of human behavior and brain research, clarifying how individual inspirations, cognitive biases, and social standards add to content creation, sharing, and commitment inside social networks. The bits of knowledge acquired give a complete perspective on what user activities mean for network elements and the development of online networks. By exploring these aspects, the review expects to illuminate platform developers and policymakers, offering important direction on upgrading design, carrying out powerful security measures, and figuring out user behavior. Eventually, this examination adds to an all-encompassing comprehension of social networks, introducing a diverse investigation of design, security, and human factors. The discoveries guide the advancement of more user-centric, secure, and socially significant social organization platforms that line up with user inclinations and contribute decidedly to the digital landscape.

Enhancing Predictive Expert Method for Link Prediction in Heterogeneous Information Social Networks

Link prediction is a critical prerequisite and foundation task for social network security that involves predicting the potential relationship between nodes within a network or graph. Although the existing methods show promising performance, they often ignore the unique attributes of each link type and the impact of diverse node differences on network topology when dealing with heterogeneous information networks (HINs), resulting in inaccurate predictions of unobserved links. To overcome this hurdle, we propose the Enhancing Predictive Expert Method (EPEM), a comprehensive framework that includes an individual feature projector, a predictive expert constructor, and a trustworthiness investor. The individual feature projector extracts the distinct characteristics associated with each link type, eliminating shared attributes that are common across all links. The predictive expert constructor then creates enhancing predictive experts, which improve predictive precision by incorporating the individual feature representations unique to each node category. Finally, the trustworthiness investor evaluates the reliability of each enhancing predictive expert and adjusts their contributions to the prediction outcomes accordingly. Our empirical evaluations on three diverse heterogeneous social network datasets demonstrate the effectiveness of EPEM in forecasting unobserved links, outperforming the state-of-the-art methods.

Multivariate time series anomaly detection by fusion of deep convolution residual autoencoding reconstruction model and ConvLstm forecasting model

Multivariate time-series anomaly detection is very essential to ensure the normal operation of physical equipment. Great achievements have been made in this field in recent years. However, two critical challenges limit the generalization ability of the model. Graph attention networks without prior knowledge are unable to extract the distribution features of dependencies among variables in a fine granularity. This reduces the accuracy of the dependence matrix to represent the correlations between variables. Beyond this challenge, the reconstruction model loses the fine-grained seasonal feature information in the spatial dimension during the reconstruction analysis process of the sample. In order to solve the above challenges, we propose a multivariate time-series anomaly detection model consisting of characterization network and forecasting network. In the characterization network, we first remove non-existent dependencies between variables by using prior knowledge. This approach can accurately capture the fine-grained dependency distribution features between variables and improve the accuracy of the dependence matrix representation of the correlations between variables. We then construct a deep convolutional residual autoencoder to reduce the loss of seasonal feature information in the spatial dimension. In forecasting network, we construct a temporal attention-based ConvLstm forecasting network to make fine-grained anomaly decisions on the output of the characterization network in the time dimension. We perform systematic experiments on six open-source datasets, in social network security scenario,the malicous acount behavior can be detected. Experimental results show that the proposed method outperforms the baseline anomaly detection methods with excellent detection performance and robustness. Specifically, our method achieves an average F1 score improvement of 0.23 over the baseline method at its highest level.

Enhancing security in online social networks: introducing the DeepSybil model for Sybil attack detection

Enhancing security in online social networks: introducing the DeepSybil model for Sybil attack detection

Information Security and Individual Privacy in Social Networks“Kurdish Users” as a Case Study

Information Security and Individual Privacy in Social Networks“Kurdish Users” as a Case Study

Livelihood Security of Livestock Farmer Producer Groups (Women Self Help Groups) in Telangana State

Livestock Farmer Producer Group is the primary organization of women members from existing SHGs involved in livestock rearing of sheep/goat, dairy animals or poultry. These groups shall be formed either in the same habitation or village and Livestock Farmer Producer Organisations at mandal/district/state level. The present aim is to identify the “Livelihood security of Livestock Farmer Producer Groups (women Self Help Groups) in Telangana state” was conducted in Sidddipet and Nagarkurnool districts 10 LFPGs were selected randomly from each district and from each group 6 members were selected thus a total of 120 members constituted the respondents of the study. The information was elucidated from respondents with help of a pre structured schedule through descriptive research design which was processed, analysed and interpreted. In the present study livelihood security was assessed by food security, health security, economic security, habitat security, educational security and social network security. In food security, the majority of the respondents depend upon the public distribution system, had three meals per day. Regarding health security, the majority of respondents fell ill only once or twice in a year and afford professional treatment with some difficulty. Regarding economic security, about sixty one percent of respondents have 1-2 lakh annual income and income sources are animal husbandry, agriculture and others. Educational security was concerned; the majority were educated up to metric and considered education as expensive. Regarding social network security, the majority have very strong relationships with family members and get help from them in times of distress.