From Bare-Bones to Interactive
Instructor: Dr. Filipi Silva
The advent of the information age has greatly changed the way science is made. Over the past few years, the deluge of data generated from experiments, digital reconstruction of real objects, the web, and social media; culminated into several scientific and technological advances. At the same time, processing, modeling, and interpreting this huge amount of data has become an increasingly challenging task. Network science emerged as a powerful framework to represent and model many of these systems. In addition to that, the increase of computational power and arrival of graphics processing units (GPUs) now allows the use of large-scale visualization techniques to better interpret the structure and dynamics of complex systems. Consequently, researchers can greatly benefit from the union of advanced visualization techniques and network science. Traditional static network visualization techniques can hardly be sufficient for the task of interpreting complex networks, which are generally large and display intricate relationships among their elements. Nevertheless, interactive data visualization allows the exploration of several relationships together without limiting a small set of isolated aspects of the system under analysis.
In this workshop, we will guide the participants through the traditional and some state-of-the-art techniques to visualize networks. This includes an introduction to several concepts used in network visualization, such as force-based layouts, neural network embedding methods (e.g., node2vec), dimensionality reduction, 2D and 3D rendering techniques, as well as approaches to effectively render such structures. The workshop will also cover some of the most used libraries and software that can be used for network visualization. This workshop is divided into two parts: a 90 minutes session dedicated to the introduction of the fundamentals in network visualization; and another 120 minutes session of hands-on tutorial in which some of these techniques will be implemented or imported from existing tools.
Throughout the hands-on session, we will briefly showcase some of the existing standalone tools, such as Cytoscape and Gephi. Next, the participants will be guided on constructing a network visualization pipeline programmatically using Python and other libraries such as igraph, networkx, graph-tool, matplotlib, Fury, and Helios. Then, we will briefly introduce a few web-based technologies and the usage of d3.js, WebGL, and Helios-Web to construct interactive network visualizations that can be deployed to the web.
This workshop will be held on-line, with a limited number of seats available in-person. Both attendance options require registration. Click here to register.
Return to the workshop series home page.