Reaction: Balancing Systematic and Flexible Exploration of Social Networks

This paper mainly addresses the complexity of a network and tries to visualize the data present in a network. Network analysis is very complex and involves various parameters to be considered and hence the network visualization becomes messy. Network analysis involves understanding the data represented by the nodes and the relationships between the nodes amongst themselves. The visualization of networks becomes difficult because, as the network grows large the labels of nodes become illegible and the links overlap.

Hence to solve this issue, the authors propose a tool for social network analysis which is called a SocialAction. This approach applies attribute ranking and coordinated views to identify the extreme-values nodes. This approach balances systematic and flexible exploration for the networks. The authors thus present a novel tool for network analysis and visualisation of the data in a network.

Some of the features provided by the tool include attribute rankings and coordinated views. The visualization of data is proposed in an ordered list view with appropriate color coding and dynamic querying for the users. Here the visualization is depicted by viewing the list and the network side by side. However as the network grows large, the list becomes quite long and network layouts become illegible. Hence to avoid this, the SocialAction tool provides a methodology to aggregate the nodes based on link structure. Thus the network can be compressed and examined well. Also analysts could examine a part of the section from the aggregated structure. To view the data efficiently, SocialAction also allows filtering of data in the network and the list using the rankings. This filtering increases the managing of the network in terms of legibility. Analysis across two rankings can also be carried out in a node structure using a Scatterplot. The scatterplot allows managing network across the two dimensions by plotting the points across the two axes selected. However at times, scatterplot becomes crowded and inefficient for interpretation. Hence SocialAction also supports aggregation of ranking for the subgroups. This allows isolation of a group of nodes based on their structural properties. SocialAction also provides a matrix summary for multiples link types or multiplex networks.

Thus the method proposed by the authors in this paper and their work seems to provide an efficient solution for network visualization. The SocialAction tool provides stability to the network and better visibility to the user for analysis.