Detect Tight Communities in a social Network
Using Spark GraphX to form graph and find cliques in the graph to detect smaller stronger sub-graphs representing tight communties.
Data Used : Facebook ( http://snap.stanford.edu/data/egonets-Facebook.html )
Execution :
- Download data and unzip jar into a folder ( upload folder to AWS S3 if running on EMR )
- Import as a sbt project into IntelliJ IDE
- Assembly build the project ( upload jar into S3 bucket )
- Run in IntelliJ or in AWS EMR cluster
Results : A file containing the node clusters representing tight communties