By Rob Mitchum // March 20, 2015
You may not realize it, but network analysis is a daily part of your life. Beyond the popular social networks of Twitter or Facebook, many companies use network data to provide some of the useful services we now take for granted: accurate web searches, book and movie recommendations, fraud detection. The art of transforming data on connections between people or entities into practical insights and predictions has made billions of dollars for companies such as Google or Amazon, creating high demand for network analysis skills.
The third installment of our Discovery Engines: Under The Hood workshops provided an introduction to these increasingly popular methods, as well as some of the software tools that make analysis and visualization tasks much easier. Feng “Bill” Shi, a post-doctoral researcher with the CI’s Knowledge Lab, uses network analysis in his own research to study scientific systems, but his tutorial demonstrated the broad flexibility of the approach, applicable to everything from the social dynamics of karate clubs to the prominence of Florence dynasties during the Renaissance to deciding which Oscar-nominated film a person would most enjoy watching.
The workshop, which you can watch below, used iPython notebook, the visualization software Gephi, and the toolkit iGraph to demonstrate the basic moves of network analysis. To follow along, install the tools above and download the datasets for the session at https://github.com/KnowledgeLab/CI-Workshop-Networks