Visualization of evolving social networks using actor‐level and community‐level trajectories

Abstract

In recent years we witnessed an impressive advance in the social networks field, which became a "hot" topic and a focus of considerable attention. Also, the development of methods that focus on the analysis and understanding of the evolution of data are gaining momentum. In this paper we present an approach to visualize the evolution of dynamic social networks by using Tucker decomposition and the concept of temporal trajectory. Our visualization strategy is based on the definition of trajectories, both at the node-level and at the community-level, in a bidimensional space that preserves its structural properties. Furthermore, this approach can be used to identify similar actors, or communities of actors, by comparing the shape and position of trajectories. To illustrate the proposed approach we conduct a case study using a set of temporal friendship networks.

Tensor-based anomaly detection: An interdisciplinary survey

“…Tensors are used in analysis of Facebook data [20] , phone calls [63] , location-based social networks ( user × location × time ) [76] and analysis of physical social networks such as face-to-face contacts of individuals [77] . Apart from the traditional model, [78] proposed new tensor models such as nodes × measures × time and communities × measures × time for dynamic social networks where measures such as betweenness and degree closeness are computed from social network in each snapshot.…”

Quantitative analysis of a half-century of K-Pop songs: Association rule analysis of lyrics and social network analysis of singers and composers

“…A social network is defined as a set of actors and the relationships among them (Crespo & Antunes, ; Oliveira & Gama, ; Wasserman & Faust, ; Xu & Chen, ). SNA has been applied in various areas, including sociology, healthcare, business management, and intellectual property.…”

Analyzing the behavior dynamics of grain price indexes using Tucker tensor decomposition and spatio-temporal trajectories

“…However, the paper did not present a solution to deal with time series, or analyzed the time dimension, even though the evolution of such information may reveal important patterns. On the other hand, an approach discussed in Oliveira and Gama (2013), presents a methodology to track the evolution of dynamic social network, using advanced concepts of spatiotemporal trajectories.…”

“…In this research, our aim is to explore two of those techniques in order to provide a methodology for the analysis of agro economic data. The main techniques and framework used were Tucker decomposition (Tucker, 1966;Oliveira and Gama, 2012) and spatio-temporal trajectories (Oliveira and Gama, 2013). Some complementary statistical methods were used, namely, the correlation matrix (Kazmier, 2004), ANOVA and the sliding window model (Datar et al, 2002).…”