Twitter infuence metrics

@NCSR Demokritos

‘REVEAL-REVEALing hidden concepts in Social Media’ (Ε-11834) program, research and development in knowledge extraction from graphs representing Social Networks.

Exploring the concept of influence in a social media platform (Twitter). Investigation of simple metrics of influence, comparisons in terms of correlation. Presentation slides can be found here.

Instructions/How-to run

1. edit Main.py file, replace paths and parameters as desired.
   Params: save : saves results into pickle files
   resultsFolder: folder path to save results, if desired
   dataset_path : location of dataset
   type :

   • simple : simple graph and centrality measures
   • local_networks: graph and local network subgraphs, running centralities on overall graph as well as local networks
   • weighted: creating weighted graph and running weighted centralities

2. python Main.py to run