Akrati Saxena

Understanding Spreading Patterns on Social Networks Based on Network Topology

Ever since the introduction of the first epidemic model, scientists have tried extrapolating the damage caused by a contagious disease, given its spreading pattern in the premature stage. However, understanding epidemiology remains an elusive mystery to researchers specifically because of the unavailability of large amount of data. We utilise the study of diffusion of memes in a social networking website to solve this problem. In this paper, we analyse the impact of specific meso-scale properties of a network on a meme traversing over it. We have employed SCCP (Scale free, Communities, Core Periphery structure) networks for analysis purpose. We propose a new meme propagation model for real world social networks and observe the cause of virality of a meme. We have tested and validated our model with the real world information spreading pattern.

Modeling Memetics Using Edge Diversity

The study of meme propagation and the prediction of meme trajectory are emerging areas of interest in the field of complex networks research. In addition to the properties of the meme itself, the structural properties of the underlying network decides the speed and the trajectory of the propagating meme. In this paper, we provide an artificial framework for studying the meme propagation patterns. Firstly, the framework includes a synthetic network which simulates a real world network and acts as a testbed for meme simulation. Secondly, we propose a meme spreading model based on the diversity of edges in the network. Through the experiments conducted, we show that the generated synthetic network combined with the proposed spreading model is able to simulate a real world meme spread. Our proposed model is validated by the propagation of the Higgs boson meme on Twitter as well as many real world social networks.

Three is The Answer: Combining Relationships to Analyze Multilayered Terrorist Networks

In this paper we introduce a methodology to create multilayered terrorist networks, taking into account that the main challenges of the data behind the networks are incompleteness, fuzzy boundaries, and dynamic behavior. To account for these dark networks’ characteristics, we use knowledge sharing communities in determining the methodology to create 3-layered networks from each of our datasets. We analyze the resulting layers of three terrorist datasets and present explanations of why three layers should be used for these models. We also use the information of just one layer, to identify the Bali 2005 attack community.

Observe Locally Rank Globally

Most real world dynamic networks are evolving very fast with time. It is not feasible to collect the entire network at any given time to study its characteristics. This creates the need to propose local algorithms to study various properties of the network. In the present work, we estimate degree rank of a node without having the entire network. The proposed methods are based on the power law degree distribution characteristic or sampling techniques. We further study the efficiency and feasibility of these approaches in different contexts. The proposed methods are simulated on synthetic networks, as well as on real world social networks. Results show that the degree rank of a node can be estimated with high accuracy using only 1% samples of the network size. The accuracy of the estimation decreases from high ranked to low ranked nodes.

Evolving models for meso-scale structures

Real world complex networks are scale free and possess meso-scale properties like core-periphery and community structure. We study evolution of the core over time in real world networks. This paper proposes evolving models for both unweighted and weighted scale free networks having local and global core-periphery as well as community structure. Network evolves using topological growth, self growth, and weight distribution function. To validate the correctness of proposed models, we use K-shell and S-shell decomposition methods. Simulation results show that the generated unweighted networks follow power law degree distribution with droop head and heavy tail. Similarly, generated weighted networks follow degree, strength, and edge-weight power law distributions. We further study other properties of complex networks, such as clustering coefficient, nearest neighbor degree, and strength degree correlation.

Estimating the degree centrality ranking

Complex networks have gained more attention from the last few years. The size of the real world complex networks, such as online social networks, WWW networks, collaboration networks, is exponentially increasing with time. It is not feasible to completely collect, store and process these networks. In the present work, we propose a method to estimate degree centrality ranking of a node without having complete structure of the graph. The proposed method uses degree of a node and power law exponent of the degree distribution to calculate the ranking. We also study simulation results on Barabasi-Albert model. Simulation results show that average error in the calculated ranking is approximately 5% of total number of nodes.

Estimating degree rank in complex networks

Identifying top-ranked nodes can be performed using different centrality measures, based on their characteristics and influential power. The most basic of all the ranking techniques is based on nodes degree. While finding the degree of a node requires local information, ranking the node based on its degree requires global information, namely the degrees of all the nodes of the network. It is infeasible to collect the global information for some graphs such as (i) the ones emerging from big data, (ii) dynamic networks, and (iii) distributed networks in which the whole graph is not known. In this work, we propose methods to estimate the degree rank of a node, that are faster than the classical method of computing the centrality value of all nodes and then rank a node. The proposed methods are modeled based on the network characteristics and sampling techniques, thus not requiring the entire network. We show that approximately

Fast Estimation of Closeness Centrality Ranking

1\%