Graph-structured data is everywhere: social networks, the web, and even mobile phone records. Viewing data as graphs can reveal valuable insights for targeting ads, recommending products, and predicting behavior. GraphX is the graph processing library included in Spark. GraphX comes with a range of graph algorithms and makes it easy to write your own using a simple API that can intermix graphs and RDDs. This talk will cover graph algorithms, the GraphX API and internals, and the future of the project.
Ankur is a PhD student in the UC Berkeley AMPLab and a maintainer for GraphX.