Identifying influential spreaders in social networks: A two-stage quantum-behaved particle swarm optimization with Lévy flight

The influence maximization problem aims to select a small set of influential nodes, termed a seed set, to maximize their influence coverage in social networks. Although the methods that are based on a greedy strategy can obtain good accuracy, they come at the cost of enormous computational time, and are therefore not applicable to practical scenarios in large-scale networks. In addition, the centrality heuristic algorithms that are based on network topology can be completed in relatively less time. However, they tend to fail to achieve satisfactory results because of drawbacks such as overlapped influence spread. In this work, we propose a discrete two-stage metaheuristic optimization combining quantum-behaved particle swarm optimization with Lévy flight to identify a set of the most influential spreaders. According to the framework,first, the particles in the population are tasked to conduct an exploration in the global solution space to eventually converge to an acceptable solution through the crossover and replacement operations. Second, the Lévy flight mechanism is used to perform a wandering walk on the optimal candidate solution in the population to exploit the potentially unidentified influential nodes in the network. Experiments on six real-world social networks show that the proposed algorithm achieves more satisfactory results when compared to other well-known algorithms.

CRB:A new rumor blocking algorithm in online social networks based on competitive spreading model and influence maximization

The virtuality and openness of online social platforms make networks a hotbed for the rapid propagation of various rumors.In order to block the outbreak of rumor,one of the most effective containment measures is spreading positive information to counterbalance the diffusion of rumor.The spreading mechanism of rumors and effective suppression strategies are significant and challenging research issues.Firstly,in order to simulate the dissemination of multiple types of information,we propose a competitive linear threshold model with state transition(CLTST)to describe the spreading process of rumor and anti-rumor in the same network.Subsequently,we put forward a community-based rumor blocking(CRB)algorithm based on influence maximization theory in social networks.Its crucial step is to identify a set of influential seeds that propagate anti-rumor information to other nodes,which includes community detection,selection of candidate anti-rumor seeds and generation of anti-rumor seed set.Under the CLTST model,the CRB algorithm has been compared with six state-of-the-art algorithms on nine online social networks to verify the performance.Experimental results show that the proposed model can better reflect the process of rumor propagation,and review the propagation mechanism of rumor and anti-rumor in online social networks.Moreover,the proposed CRB algorithm has better performance in weakening the rumor dissemination ability,which can select anti-rumor seeds in networks more accurately and achieve better performance in influence spread,sensitivity analysis,seeds distribution and running time.

Research on Fine-Grained Recognition Method for Sensitive Information in Social Networks Based on CLIP

With the emergence and development of social networks,people can stay in touch with friends,family,and colleagues more quickly and conveniently,regardless of their location.This ubiquitous digital internet environment has also led to large-scale disclosure of personal privacy.Due to the complexity and subtlety of sensitive information,traditional sensitive information identification technologies cannot thoroughly address the characteristics of each piece of data,thus weakening the deep connections between text and images.In this context,this paper adopts the CLIP model as a modality discriminator.By using comparative learning between sensitive image descriptions and images,the similarity between the images and the sensitive descriptions is obtained to determine whether the images contain sensitive information.This provides the basis for identifying sensitive information using different modalities.Specifically,if the original data does not contain sensitive information,only single-modality text-sensitive information identification is performed;if the original data contains sensitive information,multimodality sensitive information identification is conducted.This approach allows for differentiated processing of each piece of data,thereby achieving more accurate sensitive information identification.The aforementioned modality discriminator can address the limitations of existing sensitive information identification technologies,making the identification of sensitive information from the original data more appropriate and precise.

Habits on Social Networks at Workplace: A Survey of Motivations and Behaviour

This article explores the use of social networks by workers in Abidjan, Côte d’Ivoire, with particular emphasis on a descriptive or quantitative analysis aimed at understanding motivations and methods of use. More than five hundred and fifty questionnaires were distributed, highlighting workers’ preferred digital channels and platforms. The results indicate that the majority use social media through their mobile phones, with WhatsApp being the most popular app, followed by Facebook and LinkedIn. The study reveals that workers use social media for entertainment purposes and to develop professional and social relationships, with 55% unable to live without social media at work for recreational activities. In addition, 35% spend on average 1 to 2 hours on social networks, mainly between 12 p.m. and 2 p.m. It also appears that 46% believe that social networks moderately improve their productivity. These findings can guide marketing strategies, training, technology development and government policies related to the use of social media in the workplace.

Mitigation Strategy against Cascading Failures on Social Networks

Cascading failures are common phenomena in many of real-world networks,such as power grids,Internet,transportation networks and social networks.It's worth noting that once one or a few users on a social network are unavailable for some reasons,they are more likely to influence a large portion of social network.Therefore,an effective mitigation strategy is very critical for avoiding or reducing the impact of cascading failures.In this paper,we firstly quantify the user loads and construct the processes of cascading dynamics,then elaborate the more reasonable mechanism of sharing the extra user loads with considering the features of social networks,and further propose a novel mitigation strategy on social networks against cascading failures.Based on the realworld social network datasets,we evaluate the effectiveness and efficiency of the novel mitigation strategy.The experimental results show that this mitigation strategy can reduce the impact of cascading failures effectively and maintain the network connectivity better with lower cost.These findings are very useful for rationally advertising and may be helpful for avoiding various disasters of cascading failures on many real-world networks.

Optimal Deep Learning-based Cyberattack Detection and Classification Technique on Social Networks

Cyberbullying(CB)is a distressing online behavior that disturbs mental health significantly.Earlier studies have employed statistical and Machine Learning(ML)techniques for CB detection.With this motivation,the current paper presents an Optimal Deep Learning-based Cyberbullying Detection and Classification(ODL-CDC)technique for CB detection in social networks.The proposed ODL-CDC technique involves different processes such as pre-processing,prediction,and hyperparameter optimization.In addition,GloVe approach is employed in the generation of word embedding.Besides,the pre-processed data is fed into BidirectionalGated Recurrent Neural Network(BiGRNN)model for prediction.Moreover,hyperparameter tuning of BiGRNN model is carried out with the help of Search and Rescue Optimization(SRO)algorithm.In order to validate the improved classification performance of ODL-CDC technique,a comprehensive experimental analysis was carried out upon benchmark dataset and the results were inspected under varying aspects.A detailed comparative study portrayed the superiority of the proposed ODL-CDC technique over recent techniques,in terms of performance,with the maximum accuracy of 92.45%.

Deep Learning Empowered Cybersecurity Spam Bot Detection for Online Social Networks

Cybersecurity encompasses various elements such as strategies,policies,processes,and techniques to accomplish availability,confidentiality,and integrity of resource processing,network,software,and data from attacks.In this scenario,the rising popularity of Online Social Networks(OSN)is under threat from spammers for which effective spam bot detection approaches should be developed.Earlier studies have developed different approaches for the detection of spam bots in OSN.But those techniques primarily concentrated on hand-crafted features to capture the features of malicious users while the application of Deep Learning(DL)models needs to be explored.With this motivation,the current research article proposes a Spam Bot Detection technique using Hybrid DL model abbreviated as SBDHDL.The proposed SBD-HDL technique focuses on the detection of spam bots that exist in OSNs.The technique has different stages of operations such as pre-processing,classification,and parameter optimization.Besides,SBD-HDL technique hybridizes Graph Convolutional Network(GCN)with Recurrent Neural Network(RNN)model for spam bot classification process.In order to enhance the detection performance of GCN-RNN model,hyperparameters are tuned using Lion Optimization Algorithm(LOA).Both hybridization of GCN-RNN and LOA-based hyperparameter tuning process make the current work,a first-of-its-kind in this domain.The experimental validation of the proposed SBD-HDL technique,conducted upon benchmark dataset,established the supremacy of the technique since it was validated under different measures.

Graph Transformer for Communities Detection in Social Networks

Graphs are used in various disciplines such as telecommunication,biological networks,as well as social networks.In large-scale networks,it is challenging to detect the communities by learning the distinct properties of the graph.As deep learning hasmade contributions in a variety of domains,we try to use deep learning techniques to mine the knowledge from large-scale graph networks.In this paper,we aim to provide a strategy for detecting communities using deep autoencoders and obtain generic neural attention to graphs.The advantages of neural attention are widely seen in the field of NLP and computer vision,which has low computational complexity for large-scale graphs.The contributions of the paper are summarized as follows.Firstly,a transformer is utilized to downsample the first-order proximities of the graph into a latent space,which can result in the structural properties and eventually assist in detecting the communities.Secondly,the fine-tuning task is conducted by tuning variant hyperparameters cautiously,which is applied to multiple social networks(Facebook and Twitch).Furthermore,the objective function(crossentropy)is tuned by L0 regularization.Lastly,the reconstructed model forms communities that present the relationship between the groups.The proposed robust model provides good generalization and is applicable to obtaining not only the community structures in social networks but also the node classification.The proposed graph-transformer shows advanced performance on the social networks with the average NMIs of 0.67±0.04,0.198±0.02,0.228±0.02,and 0.68±0.03 on Wikipedia crocodiles,Github Developers,Twitch England,and Facebook Page-Page networks,respectively.

ED-SWE:Event detection based on scoring and word embedding in online social networks for the internet of people

Online social media networks are gaining attention worldwide,with an increasing number of people relying on them to connect,communicate and share their daily pertinent event-related information.Event detection is now increasingly leveraging online social networks for highlighting events happening around the world via the Internet of People.In this paper,a novel Event Detection model based on Scoring and Word Embedding(ED-SWE)is proposed for discovering key events from a large volume of data streams of tweets and for generating an event summary using keywords and top-k tweets.The proposed ED-SWE model can distill high-quality tweets,reduce the negative impact of the advent of spam,and identify latent events in the data streams automatically.Moreover,a word embedding algorithm is used to learn a real-valued vector representation for a predefined fixed-sized vocabulary from a corpus of Twitter data.In order to further improve the performance of the Expectation-Maximization(EM)iteration algorithm,a novel initialization method based on the authority values of the tweets is also proposed in this paper to detect live events efficiently and precisely.Finally,a novel automatic identification method based on the cosine measure is used to automatically evaluate whether a given topic can form a live event.Experiments conducted on a real-world dataset demonstrate that the ED-SWE model exhibits better efficiency and accuracy than several state-of-art event detection models.

Fake Reviews Tell No Tales？ Dissecting Click Farming in Content-Generated Social Networks

Recently, there has been a radial shift from traditional online social networks to content-generated social networks(CGSNs). Contemporary CGSNs, such as Dianping and Trip Advisor, are often the targets of click farming in which fake reviews are posted in order to boost or diminish the ratings of listed products and services simply through clicking. Click farming often emanates from a collection of multiple fake or compromised accounts, which we call click farmers. In this paper, we conduct a three-phase methodology to detect click farming. We begin by clustering communities based on newly-defined collusion networks. We then apply the Louvain community detection method to detecting communities. We finally perform a binary classification on detected-communities. Our results of over a year-long study show that(1) the prevalence of click farming is different across CGSNs;(2) most click farmers are lowly-rated;(3) click-farming communities have relatively tight relations between users;(4) more highly-ranked stores have a greater portion of fake reviews.

Influence fast or later:Two types of influencers in social networks

In real-world networks,there usually exist a small set of nodes that play an important role in the structure and function of networks.Those vital nodes can influence most of other nodes in the network via a spreading process.While most of the existing works focused on vital nodes that can maximize the spreading size in the final stage,which we call final influencers,recent work proposed the idea of fast influencers,which emphasizes nodes’spreading capacity at the early stage.Despite the recent surge of efforts in identifying these two types of influencers in networks,there remained limited research on untangling the differences between the fast influencers and final influencers.In this paper,we firstly distinguish the two types of influencers:fast-only influencers and final-only influencers.The former is defined as individuals who can achieve a high spreading effect at the early stage but lose their superiority in the final stage,and the latter are those individuals that fail to exhibit a prominent spreading performance at the early stage but influence a large fraction of nodes at the final stage.Further experiments are based on eight empirical datasets,and we reveal the key differences between the two types of influencers concerning their spreading capacity and the local structures.We also analyze how network degree assortativity influences the fraction of the proposed two types of influencers.The results demonstrate that with the increase of degree assortativity,the fraction of the fast-only influencers decreases,which indicates that more fast influencers tend to keep their superiority at the final stage.Our study provides insights into the differences and evolution of different types of influencers and has important implications for various empirical applications,such as advertisement marketing and epidemic suppressing.

Evolutionary prisoner’s dilemma on Newman-Watts social networks with an asymmetric payoff distribution mechanism

In this paper, we introduce an asymmetric payoff distribution mechanism into the evolutionary prisoner＇s dilemma game （PDG） on Newman Watts social networks, and study its effects on the evolution of cooperation. The asymmetric payoff distribution mechanism can be adjusted by the parameter α： if α〉 0, the rich will exploit the poor to get richer; if α 〈 0, the rich are forced to offer part of their income to the poor. Numerical results show that the cooperator frequency monotonously increases with c~ and is remarkably promoted when c~ 〉 0. The effects of updating order and self-interaction are also investigated. The co-action of random updating and self-interaction can induce the highest cooperation level. Moreover, we employ the Gini coefficient to investigate the effect of asymmetric payoff distribution on the the system＇s wealth distribution. This work may be helpful for understanding cooperative behaviour and wealth inequality in society.

A Parallel Approach for Sentiment Analysis on Social Networks Using Spark

The public is increasingly using social media platforms such as Twitter and Facebook to express their views on a variety of topics.As a result,social media has emerged as the most effective and largest open source for obtaining public opinion.Single node computational methods are inefficient for sentiment analysis on such large datasets.Supercomputers or parallel or distributed proces-sing are two options for dealing with such large amounts of data.Most parallel programming frameworks,such as MPI(Message Processing Interface),are dif-ficult to use and scale in environments where supercomputers are expensive.Using the Apache Spark Parallel Model,this proposed work presents a scalable system for sentiment analysis on Twitter.A Spark-based Naive Bayes training technique is suggested for this purpose;unlike prior research,this algorithm does not need any disk access.Millions of tweets have been classified using the trained model.Experiments with various-sized clusters reveal that the suggested strategy is extremely scalable and cost-effective for larger data sets.It is nearly 12 times quicker than the Map Reduce-based model and nearly 21 times faster than the Naive Bayes Classifier in Apache Mahout.To evaluate the framework’s scalabil-ity,we gathered a large training corpus from Twitter.The accuracy of the classi-fier trained with this new dataset was more than 80%.

A Combination Approach to Community Detection in Social Networks by Utilizing Structural and Attribute Data

Community detection is one of the important tasks of social network analysis. It has significant practical importance for achieving cost-effective solutions for problems in the area of search engine optimization, spam detection, viral marketing, counter-terrorism, epidemic modeling, etc. In recent years, there has been an exponential growth of online social platforms such as Twitter, Facebook, Google+, Pinterest and Tumblr, as people can easily connect to each other in the Internet era overcoming geographical barriers. This has brought about new forms of social interaction, dialogue, exchange and collaboration across diverse social networks of unprecedented scales. At the same time, it presents new challenges and demands more effective, as well as scalable, graphmining techniques because the extraction of novel and useful knowledge from massive amount of graph data holds the key to the analysis of social networks in a much larger scale. In this research paper, the problem to find communities within social networks is considered. Existing community detection techniques utilize the topological structure of the social network, but a proper combination of the available attribute data, which represents the properties of the participants or actors, and the structure data of the social network graph is promising for the detection of more accurate and meaningful communities.

New Media, Public Sphere, and Arab Spring： Demystifying the Power of Social Networks

This paper looks at the new media, communication, and political environment in both Tunisia and Egypt during and after the revolution. The new environment provided activists, politicians, civil society, and youth among others, who want to express their opinions and share their views, with various channels and means of corranunication to be part of the political action and to participate in the decision-making process. Social media played an important role in mobilizing youth to rally and protest. This is to say that a new model of communication has emerged with this new environment. The receiver has become the sender and the producer of the message. The process of communication, therefore, has been changed from one to many to from many to many, and everybody became sender and receiver at the same time. The main research question this paper aims to answer is： Are social networks enough to change the political and economic scene in the Arab World？ And is there a relationship between the new communication environment and Arab spring？ The year 2011 has been in the Arab world the year of social networks and radical changes in the political scene where a score of dictators were ousted. New political communication networks and mechanisms took place, and for the first time in Arab political communication, public opinion was a major political player. Social networks helped tremendously the formation of new public sphere where the public finds its way in the media and communication processes. At their best, new media can mobilize crowds and masses to rally and protest. They can give a social perspective to movements. However, they can＇t make change and implement democracy. After the collapse of the regimes in Tunisia and Egypt, things are not getting any better. There is no democratic transition, and both countries are experiencing complex economic, social, and political problems.

Social Networks and Citizen Participation in the Collaborative Community Policing ——A Case Study of S Community in Beijing

It is of great significance to enhance collaborative community policing for crime prevention and better community-police relationships. Understanding the relational structure of collaborative community policing is necessary to pinpoint the pattern of interactions among key actors involved in community policing and improve the effectiveness of network governance. Based on 234 surveys of citizens of S Community in Beijing from April 2017 to May 2017, this paper empirically examines the characteristics of formal network and informal network of citizen participation in the collaborative community policing. Beijing is widely known for its active involvement of neighborhood volunteers in different types of community policing. We focused on four different types of interpersonal work relationships in this study： workflow, problem solving, mentoring and friendship, among resident committees, neighborhood administrative offices, media, police station, business security personnel, neighborhood volunteers, and security activists. The nature of relationships between individuals in networks can be treated as from instrumental ties to expressive ties. Expressive ties cover relationships that involve the exchange of friendship, trust, and socio-emotional support. We extended this intra-organizational insight into a community policing inter-organizational context. The collaborative network showed the trend of the distributed network. The clustering analysis showed that in the workflow network, we should make thll use of the close interaction between the citizens and activists in the community. Meanwhile, in the problem-solving network, mentoring network and friendship network, interactions between citizens and neighborhood committee are weak.

Contextualized Analysis of Social Networks:Collaboration in Scientific Communities

Currently, the collaboration in scientific communities has been studied in order to explain, among other things, the knowledge diffusion. The quality of Graduate Programmes is often associated with the scientific collaboration. This paper discusses how scientific collaboration processes can be identified and characterized through social and complex networks. For this purpose, collaboration networks of bibliographic production, research projects, and committees of PhD theses and Masters’ dissertations by researchers from a graduate program in computational modeling were studied. The data were obtained from CAPES’ reports of the period from 2001 to 2009. Among the studied indices, centrality indices indicate the presence of prominent researchers who influence others and promptly interact with other researchers in the network. The indices of complex networks reveal the presence of the small-world (i.e. these networks are favorable to increase coordination between researchers) phenomenon and indicate a behavior of scale-free degree distribution (i.e. some researchers promote clustering more than others) for one of the studied networks.

Fractal Analysis of Mobile Social Networks

Fractal and self similarity of complex networks have attracted much attention in recent years. The fractal dimension is a useful method to describe the fractal property of networks. However, the fractal features of mobile social networks （MSNs） are inadequately investigated. In this work, a box-covering method based on the ratio of excluded mass to closeness centrality is presented to investigate the fractal feature of MSNs. Using this method, we find that some MSNs are fractal at different time intervals. Our simulation results indicate that the proposed method is available for analyzing the fractal property of MSNs.

Model for Generating Scale-Free Artificial Social Networks Using Small-World Networks

The Internet of Things(IoT)has the potential to be applied to social networks due to innovative characteristics and sophisticated solutions that challenge traditional uses.Social network analysis(SNA)is a good example that has recently gained a lot of scientific attention.It has its roots in social and economic research,as well as the evaluation of network science,such as graph theory.Scientists in this area have subverted predefined theories,offering revolutionary ones regarding interconnected networks,and they have highlighted the mystery of six degrees of separation with confirmation of the small-world phenomenon.The motivation of this study is to understand and capture the clustering properties of large networks and social networks.We present a network growth model in this paper and build a scale-free artificial social network with controllable clustering coefficients.The random walk technique is paired with a triangle generating scheme in our proposed model.As a result,the clustering controlmechanism and preferential attachment(PA)have been realized.This research builds on the present random walk model.We took numerous measurements for validation,including degree behavior and the measure of clustering decay in terms of node degree,among other things.Finally,we conclude that our suggested random walk model is more efficient and accurate than previous state-of-the-art methods,and hence it could be a viable alternative for societal evolution.

Maximizing Influence in Temporal Social Networks:A Node Feature-Aware Voting Algorithm

Influence Maximization(IM)aims to select a seed set of size k in a social network so that information can be spread most widely under a specific information propagation model through this set of nodes.However,most existing studies on the IM problem focus on static social network features,while neglecting the features of temporal social networks.To bridge this gap,we focus on node features reflected by their historical interaction behavior in temporal social networks,i.e.,interaction attributes and self-similarity,and incorporate them into the influence maximization algorithm and information propagation model.Firstly,we propose a node feature-aware voting algorithm,called ISVoteRank,for seed nodes selection.Specifically,before voting,the algorithm sets the initial voting ability of nodes in a personalized manner by combining their features.During the voting process,voting weights are set based on the interaction strength between nodes,allowing nodes to vote at different extents and subsequently weakening their voting ability accordingly.The process concludes by selecting the top k nodes with the highest voting scores as seeds,avoiding the inefficiency of iterative seed selection in traditional voting-based algorithms.Secondly,we extend the Independent Cascade(IC)model and propose the Dynamic Independent Cascade(DIC)model,which aims to capture the dynamic features in the information propagation process by combining node features.Finally,experiments demonstrate that the ISVoteRank algorithm has been improved in both effectiveness and efficiency compared to baseline methods,and the influence spread through the DIC model is improved compared to the IC model.