This project is a simulation of Twitter bots that runs on the multiagents framework MASON.
The objective of this simulation is to test a trained classifier in an original way. This trained classifier identifies what Twitter accounts are from real persons and which ones are not (companies, fansites, bots...). This is useful for example when coding a social network as Google+ that only wants real people with real profiles.
This project simulates a lot of Twitter bots running at the same time, each one of them running as an agent. Each agent on the simulations has a profile and a behaviour when tweeting, following other accounts, etc. Each 1000 steps everyone of them stops to check its status with the classifier (human or not human).
Because you don't have acces to all the code.
- The classifier: it exists, but you don't have access to the code (sorry). You can code your own classifier and put it in a server, that's how each agent communicates with it.
/data/names: List of real names, one at a line./data/sources: List of tweets sources, one at a line.