Online Social Networks Research Articles

Analyzing user migration in blockchain online social networks through network structure and discussion topics of communities on multilayer networks

User migration, i.e. the movement of large sets of users from one online social platform to another one, is one of the main phenomena occurring in modern online social networks and even involves the most recent alternative paradigms of online social networks, such as blockchain online social networks (BOSNs). In these platforms, user migration mainly occurs through hard forks of the supporting blockchain, i.e. a split of the original blockchain and the creation of an alternative blockchain, to which users may decide to migrate. However, our understanding of user migration and its mechanisms is still limited, particularly regarding the role of densely connected user groups (communities) during migration and fork events. Are there differences between users who stay and those who decide to leave, in terms of network structure and discussion topics? In this work, we show, through network-based analysis centered on the identification of communities on multilayer networks and text mining that a) the “position” of a group within the network of social and economic interactions is connected to the likelihood of a group to migrate, i.e. marginal groups are more likely to leave; b) group network structure is also important, as users in densely connected groups interacting through monetary transactions are more likely to stay; c) users who leave are characterized by different discussion topics; and d) user groups interacting through monetary transactions show interest in migration-related content if they are going to leave. These findings highlight the importance of social and economic relationships between users during a user migration caused by fork events In general, in the larger context of online social media, it motivates the need to investigate user migration through a network-inspired approach based on groups and specific subgraphs while leveraging user-generated content, at the same time.

Integrating GIN-based multimodal feature transformation and multi-feature combination voting for irony-aware cyberbullying detection

With the increasing diversity of expressions, irony-aware cyberbullying has emerged as a significant issue in online social networks. However, detecting irony-aware cyberbullying is challenging, as it requires a comprehensive understanding of context and external factors beyond literal meanings. To take full advantage of multiple features of multimodal data to detect challenging irony-aware cyberbullying, we propose an integration framework (GINBV_MFCV). The multimodal feature construction method with Graph Isomorphism Network (GIN) feature transformation (GINBV) leverages the message passing and aggregation operations of GIN to extract the potential representations of text-image features, which enriches the structural information of multimodal data. In addition, the multi-feature combination voting strategy (MFCV) soft-votes the prediction results of constructed multimodal features and multiple combinations of GIN, Bidirectional Encoder Representations from Transformers (BERT), and Vision Transformers (ViT) embedded features to reduce the data structure information bias, which has a positive effect on irony-aware cyberbullying detection. Experimental results on a real-world dataset from Weibo demonstrate that GINBV_MFCV achieves an F1-score of 83.29% and an AUC of 91.21% in irony-aware cyberbullying detection, improving 8.65% and 15.73% over the baseline algorithm, respectively. These promising results confirm the potential of GINBV_MFCV for detecting irony-aware cyberbullying.

Strengths and Challenges of Using iNaturalist in Plant Research with Focus on Data Quality

iNaturalist defines itself as an “online social network of people sharing biodiversity information to help each other learn about nature” and it is likely one of the largest citizen science web portals in the world, as every year millions of observations across thousands of species are gathered and collectively compiled by an engaged community of nearly 3 million users (November 2023). The strengths and potentialities that explain the success of the platform are reviewed and include, among others, its usability and low technical requirements, immediacy, open-access, the possibility of interacting with other users, artificial-intelligence-aided identification, versatility and automatic incorporation of the validated records to GBIF. iNaturalist has, however, features that scientists need to carefully consider when using it for their research, making sure that the quality of observations does not limit or hinder its usefulness in plant research. While these are identified (e.g., the lack of representative photographs for many observations or the relatively frequent identification errors), we provide some suggestions to overcome them and, by doing so, improve the use and add value to iNaturalist for plant research.

Community-enhanced Link Prediction in Dynamic Networks

The growing popularity of online social networks is quite evident nowadays and provides an opportunity to allow researchers in finding solutions for various practical applications. Link prediction is the technique of understanding network structure and identifying missing and future links in social networks. One of the well-known classes of methods in link prediction is a similarity-based method, which uses local and global topological information of the network to predict missing links. Some methods also exist based on quasi-local features to achieve a trade-off between local and global information on static networks. These quasi-local similarity-based methods are not best suited for considering community information in dynamic networks, failing to balance accuracy and efficiency. Therefore, a community-enhanced framework is presented in this article to predict missing links on dynamic social networks. First, a link prediction framework is presented to predict missing links using parameterized influence regions of nodes and their contribution in community partitions. Then, a unique feature set is generated using local, global, and quasi-local similarity-based as well as community information-based features. This feature set is further optimized using scoring-based feature selection methods to select only the most relevant features. Finally, four machine learning-based classification models are used for link prediction. The experiments are performed on six well-known dynamic networks and three performance metrics, and the results demonstrate that the proposed method outperforms the state-of-the-art methods.

Practical and privacy-preserving geo-social-based POI recommendation

With the rapid development of location-based services and online social networks, POI recommendation services considering geographic and social factors have received extensive attention. Meanwhile, the vigorous development of cloud computing has prompted service providers to outsource data to the cloud to provide POI recommendation services. However, there is a degree of distrust of the cloud by service providers. To protect digital assets, service providers encrypt data before outsourcing it. However, encryption reduces data availability, making it more challenging to provide POI recommendation services in outsourcing scenarios. Some privacy-preserving schemes for geo-social-based POI recommendation have been presented, but they have some limitations in supporting group query, considering both geographic and social factors, and query accuracy, making these schemes impractical. To solve this issue, we propose two practical and privacy-preserving geo-social-based POI recommendation schemes for single user and group users, which are named GSPR-S and GSPR-G. Specifically, we first utilize the quad tree to organize geographic data and the MinHash method to index social data. Then, we apply BGV fully homomorphic encryption to design some private algorithms, including a private max/min operation algorithm, a private rectangular set operation algorithm, and a private rectangular overlapping detection algorithm. After that, we use these algorithms as building blocks in our schemes for efficiency improvement. According to security analysis, our schemes are proven to be secure against the honest-but-curious cloud servers, and experimental results show that our schemes have good performance.

Gender-specific homophily on Instagram and implications on information spread

More and more social interactions happen online. On online social networks such as Instagram, millions of users share, like, and comment on photos and videos every day, interacting with other users world wide, at large scale and at a high rate. These networks do not only introduce new user experiences, but they also enable new insights into human behavior. Here, we use these new possibilities to study homophilic behavior—the tendency of individuals to bond with people similar to themselves. While homophilic behavior has been observed in many contexts, little is known about gender-specific differences and the extent of homophilic behavior of female and male users in online social networks. Based on a unique and extensive data set, covering over 800,000 (directed) Instagram interactions and a time span of three years, we shed light on differences between genders and uncover an intriguing asymmetry of homophily. In particular, we show that female users exhibit homophily to a larger extent than male users. The magnitude of this asymmetry depends on the type of interaction, as differences are more pronounced for ‘comment’-interactions than for ‘like’-interactions. Given these empirical observations, we further study the implications of such gender differences on the spread of information in social networks in a basic model. We find that on average, a piece of information that originates from a female group reaches significantly more female users than male users.

Game-theoretic modeling and analysis of cyberbullying spreading on OSNs

The rapid development of online social networks (OSNs) has enabled seamless cyberbullying activities, which not only cause psychological harm to victims but also bring disharmony to society. From the perspective of OSN platform administrators, strategic decision making towards when/how to launch an anti-cyberbullying campaign is the key to mitigating its negative impact. Hence, this paper develops a dynamic anti-cyberbullying investment solution to mitigate the adverse impact of strategic cyberbullying phenomena cost-effectively. Specifically, this paper first quantifies the payoffs of the attacker/bullying instigator and the defender/platform administrator based on a newly proposed propagation model. Then, considering the budget-constrained factor for both sides, a differential game-theoretical model is proposed to simulate each party's optimal trade-off between cost and benefit. Next, by utilizing Pontryagin Maximum/Minimum Principle, a set of necessary conditions for a Nash equilibrium strategy profile have been proposed. Based on the forward-backward sweep method, an iterative algorithm (SCR optimization algorithm) is tailored to solve the conditions numerically. Through experiments conducted under Facebook, Youtube and Twitter sub-networks, the algorithm's performance has been evaluated by inspecting its convergence speed and payoff compared with randomized profiles. Consequently, the resulting strategies are verified to exhibit superior performance in dealing with cyberbullying spreading.

RBOTUE: Rumor Blocking Considering Outbreak Threshold and User Experience

When a rumor breaks out in online social network (OSN), it can lead to significant negative impact on human society, especially in the context of public emergencies, such as pandemic. Toward restraining rumor outbreak in OSN, one of the effective containment measures is to block influential users to minimize the spread of rumors. However, most of previous efforts ignore the imbalance between the cost and effect of rumor suppression. To fill this gap, from the perspective of public opinion crisis, a dynamic rumor spread model called PISIR model is established, which takes into account the overall popularity and individual tendency of rumors. Based on this model, two rumor blocking algorithms considering outbreak threshold and user experience, called 1-Hop and 2-Hop RBOTUE algorithms, are proposed, respectively. In the algorithms, a hyperbolic discount effect-based user experience mode is introduced as the constraint to ensure the user experience in OSN, then the blocking strategy is implemented on the selected subset of nodes to keep the rumor spread scale always below the outbreak warning line. The experimental results in two synthetic networks and four real OSNs indicate that both 1-Hop and 2-Hop RBOTUE algorithms have lower rumor infection rate and require less number of blocked nodes, which means that proposed algorithms can achieve better blocking performance with less restraining cost of rumors in mainstream social networks, and the two algorithms also have different adaptability for different OSNs.

Contextual Integrity and Personal Health Information Sharing in Online Social Networks

Many online social network (OSN) users exchange health-related information on OSNs to manage their health, obtain support, and offer assistance to others facing similar medical challenges. Nonetheless, OSN users are concerned about the privacy of their health information. Extant studies on online information privacy have largely resorted to a calculus perspective, focusing on the costs and benefits of information exchange with commercial vendors. Such a perspective, however, fails to capture the nuanced flows of personal health information (PHI) on OSNs, which involves more than two parties. To address this gap and further advance the research on information privacy, this paper embraces the theory of contextual integrity to explore the distinctive contextual idiosyncrasies of PHI flow on OSNs. In particular, we examine how context-specific values and norms affect OSN users’ personal privacy norms and their PHI disclosure intentions. Our theory-driven empirical results suggest that contextual values and normative expectations related to technology control, legislation, and peer users collectively influence OSN users’ personal privacy norms. In addition, perceived personal value and perceived altruistic value serve as motivators to promote PHI disclosure intentions, working in tandem with personal privacy norms. The key findings of the study also have important implications for policymakers and OSN providers in the context of OSN-mediated healthcare through (1) advancing our understanding of the complex dynamics of PHI flow on OSNs and (2) designing effective information privacy measures (i.e., policies, technical features, and programs) to uphold contextual integrity and nurture the development of patient-centered healthcare paradigms.

Designing a Testbed for Privacy-Preserving Voice-based Social Networks

Over the past decade, existing online social networks have emerged as vital platforms for fostering and nurturing social relationships. However, their reliance on text and multimedia-based interactions has posed limitations, particularly in terms of accessibility and hands-free operation, and in turn, impeded the full potential of online social networking. Voice-based social networks (VBSNs) offer a way to address these constraints by enabling multitasking through real-time audio communication, catering to a broader spectrum of users. To ensure the widespread adoption and practical utility of VBSNs, it is imperative that voice-based social networks and their applications offer secure and privacy-preserving services. In this paper, we outline the urgent requirements for voice-based social networks and evaluate existing open-source online social networking software and platforms to find the most suitable one for developing a testbed for VBSNs. Employing the chosen platform, we propose a design for VBSN testbed that facilitates the development and evaluation of VBSN applications while ensuring robust security and privacy measures through a blockchain-based decentralized identity and reputation mechanism. This proposed testbed serves as a valuable resource for researchers and developers to explore and address security and privacy concerns within the realm of voice-based social networking, enhancing the quality and reliability of their innovations.

When Six Degrees of Separation Meets Online Social Networks: How Low Can the Degree Be?

The proposition of Six Degrees of Separation prompts the realization that the world is not as vast as previously envisioned. Despite the current global population exceeding 8 billion, two strangers can still establish contact through limited intermediaries. When online social networks have garnered widespread influence, can people connect through shorter distances? This issue merits contemplation. This paper described Six Degrees of Separation and the limitations of this theory in practical applications. Combined with online social networks, the paper analyzed the development and change of interpersonal connection degrees. Finally, the paper considered the actual coverage of online social networks and other issues, while also providing a reasoned projection of potential future degrees within the Six Degrees of Separation paradigm.

GCN-LSTM: multi-label educational emotion prediction based on graph Convolutional network and long and short term memory network fusion label correlation in online social networks

Although there are a lot of methods for multi-label classification in the past research, there are still many problems. For example, in the real world, labels are not necessarily independent of each other, and there may be some connection between labels. Therefore, exploring and utilizing the interdependence between labels is a key issue in current research. For example, in the photo category, a picture that contains blue sky often also contains white clouds, and in the text category, a political story is less likely to be entertainment news. Therefore, the key to improve the accuracy of multi-label classification is to effectively learn the possible correlations between each label. Therefore, we propose a novel multi-label educational emotion prediction based on graph convolutional network and long and short term memory network fusion label correlation in online social networks. This model uses Word2Vec method to train word vectors and combines graph convolutional neural network (GCN) with long and short term memory network (LSTM). The GCN is used to dig deeper word features of text, the LSTM layer is used to learn the longterm dependence relationship between words, and the multi-granularity attention mechanism is used to assign higher weight to the affective word features. At the same time, label correlation matrix is used to complete the label feature vector and text features as the input of the classifier, and the correlation between labels is investigated. The experimental results on the open data set show that the proposed model has a good classification effect compared with other advanced methods. The research results promote the combination of deep learning and affective computing, and can promote the research of network user behavior analysis and prediction, which can be used in personalized recommendation, targeted advertising and other fields, and has wide academic significance and application prospects.

A novel deep fully convolutional encoder-decoder network and similarity analysis for English education text event clustering analysis

Education event clustering for social media aims to achieve short text clustering according to event characteristics in online social networks. Traditional text event clustering has the problem of poor classification results and large computation. Therefore, we propose a novel deep fully convolutional encoder-decoder network and similarity analysis for English education text event clustering analysis in online social networks. At the encoder end, the features of text events are extracted step by step through the convolution operation of the convolution layer. The background noise is suppressed layer by layer while the target feature representation is obtained. The decoder end and the encoder end are symmetrical in structure. In the decoder end, the high-level feature representation obtained by the encoder end is deconvolved and up-sampled to recover the target event layer by layer. Based on the linear model, text similarity is calculated and incremental clustering is performed. In order to verify the effectiveness of the English education text event analysis method based on the proposed approach, it is compared with other advanced methods. Experiments show that the performance of the proposed method is better than that of the benchmark model.

Cyber vaccinator for image tamper resilient and recovery using invertible neural network

People frequently interact with their families, friends, and colleagues through Online Social Networks (OSNs). People post and share their photos in online communities and content-sharing sites. The problem addressed in this paper is the susceptibility of digital images to tampering, which compromises security and privacy. Traditional image forgery detection methods face challenges in reproducing original content after manipulation. This paper introduces an advanced Image Immunization System leveraging Invertible Neural Networks. The system, which comprises the cyber vaccinator, vaccine validator, forward pass for tamper detection, and backward pass for image self-recovery, aims to proactively immunize images against various attacks. The run-length encoding in the backward pass transforms hidden perturbations into information, facilitating the recovery of the authentic image. The middleware's expansion to multimodal content analysis, including videos and audio, provides a more comprehensive defense against digital manipulation within OSNs. These advancements reflect a commitment to robust security and holistic content integrity. The Cyber Vaccinator, using Invertible Neural Networks (INNs) for image tamper resilience and recovery, demonstrates significant effectiveness in detecting tampering and restoring images, providing a robust solution for maintaining image integrity. The Cyber Vaccinator uses an Invertible Neural Network (INN) to safeguard image integrity. It detects tampering by analyzing invariant features and responds with precise recovery methods. By continuously monitoring images, it ensures real-time tamper detection and efficient restoration, maintaining image authenticity through advanced neural network resilience and recovery techniques.

Examination of Information Quality on Public Instagram Profiles Regarding Botulinum Toxin for Bruxism: A Study in São Paulo State, Brazil.

Currently, social media emerges as a swift and efficient channel for disseminating knowledge in dentistry; however, it is imperative to assess whether this information aligns with scientific evidence. This study aims to evaluate the quality of information found on public Instagram profiles in São Paulo State, Brazil, regarding the utilization of botulinum toxin (BTx) for bruxism treatment. The data were categorized into three qualitative groups: information pertaining to bruxism diagnosis, treatment options for bruxism, and the application of BTx for bruxism. Following the selection of pertinent publications, 50 public profiles were included in the analysis. The publications were assessed utilizing the Global Quality Scale (GQS). A total of 20,546 posts were tallied across the 50 profiles, with 230 relating to bruxism diagnosis, 166 discussing bruxism treatment options, and 78 mentioning the use of BTx for bruxism. Of these 78 posts addressing BTx for bruxism, 61% did not align with current scientific references, while 39% did. GQS analysis disclosed predominantly "poor quality" content (GQS = 2). It is concluded that the themes of bruxism and BTx are frequently broached on public Instagram profiles, yet the quality of the available information is generally subpar and often lacks scientific substantiation. Key words:Botulinum Toxins, Type A; Bruxism; Online Social Networking; Sleep Bruxism.

Machine learning for detecting fake accounts and genetic algorithm-based feature selection

Abstract People rely extensively on online social networks (OSNs) in Africa, which aroused cyber attackers’ attention for various nefarious actions. This global trend has not spared African online communities, where the proliferation of OSNs has provided new opportunities and challenges. In Africa, as in many other regions, a burgeoning black-market industry has emerged, specializing in the creation and sale of fake accounts to serve various purposes, both malicious and deceptive. This paper aims to build a set of machine-learning models through feature selection algorithms to predict the fake account, increase performance, and reduce costs. The suggested approach is based on input data made up of features that describe the profiles being investigated. Our findings offer a thorough comparison of various algorithms. Furthermore, compared to machine learning without feature selection and Boruta, machine learning employing the suggested genetic algorithm-based feature selection offers a clear runtime advantage. The final prediction model achieves AUC values between 90% and 99.6%. The findings showed that the model based on the features chosen by the GA algorithm provides a reasonable prediction quality with a small number of input variables, less than 31% of the entire feature space, and therefore permits the accurate separation of fake from real users. Our results demonstrate exceptional predictive accuracy with a significant reduction in input variables using the genetic algorithm, reaffirming the effectiveness of our approach.

Investigating the Formation Phase of a New Online Social Network

Investigating the Formation Phase of a New Online Social Network

Machine learning applied to social science research: The Nexus of Internet Access, Economic Freedom, and Corruption

This study employed a quantile regression model as a machine learning tool in Python to assess the nuanced effects and relationships within a dataset examining the combined impact of Internet connectivity and degrees of economic freedom on global corruption. The study uses a dataset covering 161 nations from 2010 to 2019, incorporating various sources of information for a range of variables rooted in theory. The findings suggest an influence of the combination of greater freedom and internet access in countering corruption. Consequently, policies that focus on improving the availability of open information, online social networking, and other elements linked to Internet accessibility should be considered as a practical approach to curbing corruption within the countries examined.

Covid-19, identity, and piety online: ultra-Orthodox discussions in WhatsApp and Telegram groups under social distancing regulations

ABSTRACT The Covid-19 pandemic has disrupted relationships between religious communities and the state. Churches, religious authorities, and believers have been obligated to modify key activities (prayer, ceremonies) and weigh their religious identity against state prohibitions. Accordingly, we ask how members of a reclusive religious community negotiated and performed their identity under a prolonged emergency. We analyse intra-communal discourse concerning the state’s social distancing regulations during Covid-19 outbreaks and lockdowns that occurred in closed groups in messaging apps. Specifically, this research case-studies Jewish ultra-Orthodox communities in Israel (Haredim). Extensive analysis of over 2,000 WhatsApp and Telegram posts in 35 ultra-Orthodox groups over a two-year period indicates that participants constantly sought to affirm their primary, dominant, communal identity. While efforts to integrate adherence to state prohibitions into this primary identity were evident, these efforts were well accepted only when justifying them via affirmation of the communal identity – superseding medical, political, and practical concerns. The findings advance understanding of minority groups’ attitudes towards state directives and illuminate contemporary pathways of identity dynamics in online social networks. The findings may also be applicable in a broader sense to the study of discourse dynamics in echo chambers and filter bubbles.

A Pre-Trained Model for Aspect-based Sentiment Analysis Task: using Online Social Networking

Aspect-based sentiment analysis (ABSA) is a crucial part of Natural Language Processing (NLP) that focuses on identifying emotions related to specific elements in written material. ABSA has gained widespread interest due to its ability to provide precise insights into sentiment expressions across different domains. Social media provides a valuable resource for ABSA, containing user-created content with viewpoints and feedback. However, the informal nature of social media text poses challenges for ABSA. This study investigates the performance enhancement of baseline and proposed models in the ABSA task context. Both baseline and proposed models were evaluated for accuracy and F1 score improvements. The results showed that the suggested model performs better than other baseline models, with an improvement of 0.52% F1 score and an overall accuracy increase of 1.16%. However, specific analyses for laptops indicated limitations in the proposed model's performance, with accuracy scores ranging from 72.65% to 84.98%.