Of Bots and Trolls

We have captured many bots that manipulate Twitter trends for political gains. We are also aware of the data that Twitter shared which includes fake accounts, namely trolls, that try to influence politics in many countries. In this project, we will try to detect such accounts that could be malicious, using our ground truth. In the end, we will assess those accounts’ impact. Perhaps, those malicious accounts create the illusion that all goes wrong in the world?

The technical achievements could be introducing new features to bot and troll detection and producing better results in bot detection than state of the art.

Potential resources: bots, trolls

Deliverables: weekly analysis on jupyter notebooks and cool presentations

Requirements: Python, usual data science libraries (pandas, sklearn etc.)