Spreading Activation Connectivity Based Approach to Network Clustering

Techno-social systems generate data, which are rather different, than data, traditionally studied in social network analysis and other fields. In massive social networks agents simultaneously participate in several contexts, in different communities. Network models of many real data from techno-social systems reflect various dimensionalities and rationales of actor's actions and interactions. The data are inherently multidimensional, where “everything is deeply intertwingled”. The multidimensional nature of Big Data and the emergence of typical network characteristics in Big Data, makes it reasonable to address the challenges of structure detection in network models, including a) development of novel methods for local overlapping clustering with outliers, b) with near linear performance, c) preferably combined with the computation of the structural importance of nodes. In this chapter the spreading connectivity based clustering method is introduced. The viability of the approach and its advantages are demonstrated on the data from the largest European social network VK.

Community Detection in Large-Scale Social Networks

Community Structure is one of the most important properties of social networks. Detecting such structures is a challenging problem in the area of social network analysis. Community is a collection of nodes with dense connections than with the rest of the network. It is similar to clustering problem in which intra cluster edge density is more than the inter cluster edge density. Community detection algorithms are of two categories, one is disjoint community detection, in which a node can be a member of only one community at most, and the other is overlapping community detection, in which a node can be a member of more than one community. This chapter reviews the state-of-the-art disjoint and overlapping community detection algorithms. Also, the measures needed to evaluate a disjoint and overlapping community detection algorithms are discussed in detail.

Visualizing Co-Authorship Social Networks and Collaboration Recommendations With CNARe

Studies have analyzed social networks considering a plethora of metrics for different goals, from improving e-learning to recommend people and things. Here, we focus on large-scale social networks defined by researchers and their common published articles, which form co-authorship social networks. Then, we introduce CNARe, an online tool that analyzes the networks and present recommendations of collaborations based on three different algorithms (Affin, CORALS and MVCWalker). Through visualizations and social networks metrics, CNARe also allows to investigate how the recommendations affect the co-authorship social networks, how researchers' networks are in a central and eagle-eye context, and how the strength of ties behaves in large co-authorship social networks. Furthermore, users can upload their own network in CNARe and make their own recommendation and social network analysis.

Link Prediction in Social Networks

Social network is a topic of current research. Relations are broken and new relations are increased. This chapter will discuss the scope or predictions of new links in social networks. Here different approaches for link predictions are described. Among them friend recommendation model is latest. There are some other methods like common neighborhood method which is also analyzed here. The comparison among them to predict links in social networks is described. The significance of this research work is to find strong dense networks in future.

Can We Trust the Health Information We Find Online?

The Internet is becoming increasingly an important source of information for people who are seeking healthcare information. Users do so without professional guidance and may lack sufficient knowledge and training to evaluate the validity and quality of health web content. This is particularly problematic in the era of Web 2.0. Hence, the main goal of this research is to propose an approach to infer user reputation based on social interactions. Reputation is a social evaluation towards a person or a group of people. The results show that our rank methodology and the network topology succeeded in achieving user reputation. The results also show that centrality measures associated with the weighted ties approach suitably controls suitably the ranking of nodes.

Hadoop-Based Distributed K-Shell Decomposition for Social Networks

Complex network analysis comprises a popular set of tools for the analysis of online social networks. Among these techniques, k-shell decomposition of a network is a technique that has been used for centrality analysis, for communities' discovery, for the detection of influential spreaders, and so on. The huge volume of input graphs and the environments where the algorithm needs to run, i.e., large data centers, makes none of the existing algorithms appropriate for the decomposition of graphs into shells. In this article, we develop for a distributed algorithm based on MapReduce for the k-shell decomposition of a graph. We furthermore, provide an implementation and assessment of the algorithm using real social network datasets. We analyze the tradeoffs and speedup of the proposed algorithm and conclude for its virtues and shortcomings.

Undirected Bipartite Networks as an Alternative Methodology to Probabilistic Exploration

This chapter explores how graph analysis techniques are able to complement and speed up the process of learning analytics and probability theory. It uses a sample of 2,353 e-learners from six European countries (France, Germany, Greece, Poland, Portugal, and Spain), who were enrolled in their first year of open online courses offered by HarvardX and MITX. After controlling the variables for socio-demographics and online content interactions, the research reveals two main results relating student-content interactions and online behavior. First, a multiple binary logistic regression model tests that students who explore online chapters are more likely to be certified. Second, the authors propose an algorithm to generate an undirected bipartite network based on tabular data of student-content interactions (2,392 nodes, 25,883 edges, a visual representation based on modularity, degree and ForceAtlas2 layout); the graph shows a clear relationship between interactions with online chapters and chances of getting certified.

Parallelizing Large-Scale Graph Algorithms Using the Apache Spark-Distributed Memory System

The rapidly emerging area of Social Network Analysis is typically based on graph models. They include directed/undirected graphs, as well as a multitude of random graph representations that reflect the inherent randomness of social networks. A large number of parameters and metrics are derived from these graphs. Overall, this gives rise to two fundamental research/development directions: (1) advancements in models and algorithms, and (2) implementing the algorithms for huge real-life systems. The model and algorithm development part deals with finding the right graph models for various applications, along with algorithms to treat the associated tasks, as well as computing the appropriate parameters and metrics. In this chapter we would like to focus on the second area: on implementing the algorithms for very large graphs. The approach is based on the Spark framework and the GraphX API which runs on top of the Hadoop distributed file system.

Social Network Analysis of Different Parameters Derived From Real-Time Facebook Profiles

With the advent of online social media, it is possible to collect large information and extract information. One of the powerful roles that networks play is to bridge gap between the local and the global perspectives. For example, social network analysis could be used to offer explanations for how simple processes at the level of individual nodes and links can have a global effect. Social networks like Twitter, Facebook, LinkedIn are very large in size with millions of vertices and billions of edges. To collect meaningful information from these densely connected graphs and huge volume of data, it is important to find proper topology of the network as well as conduct analysis based on different network parameters. The main objective of this work is to study the network parameters commonly used to explain social structures. We extract data from the three real-time Facebook accounts using the Netvizz application, analyze and evaluate their network parameters on some widely recognized graph topology using Gephi, a free and open source software.

Graph Tools for Social Network Analysis

A social network can be defined as a complex graph, which is a collection of nodes connected via edges. Nodes represent individual actors or people in the network, whereas edges define relationships among those actors. Most popular social networks are Facebook, Twitter, and Google+. To analyze these social networks, one needs specialized tools for analysis. This chapter presents a comparative study of such tools based on the general graph aspects as well as the social network mining aspects. While considering the general graph aspects, this chapter presents a comparative study of four social network analysis tools—NetworkX, Gephi, Pajek, and IGraph—based on the platform, execution time, graph types, algorithm complexity, input file format, and graph features. On the basis of the social network mining aspects, the chapter provides a comparative study on five specialized tools—Weka, NetMiner 4, RapidMiner, KNIME, and R—with respect to the supported mining tasks, main functionality, acceptable input formats, output formats, and platform used.