Botometer and tweetbotornot--Bot Detection Algorithms

How to use Botometer:

The most recent (May 2023) Botometer script with Tutorial video And Github link to this work: https://github.com/Big-Data-Analytics-Lab-USF/ga_scripts 

• Warning: Twitter v.2 requires extra steps to receive an "elevated access (tier 2)." Please contact Dr. Hagen for this. 

(Old but useful) Video (some conversations on Botometer and brief description on what each chunk of the python script mean): https://usf.box.com/s/isljlbi3y30ctimdnahh4kbiyqxmb6q7 

(Old but useful) ppt file on how to (use the script below saved in our github): botometer tutorial.pptx

Botometer Script

Previous Botometer script: https://github.com/Big-Data-Analytics-Lab-USF/covid19-random_scripts/blob/master/python/populate_bot_files.py 

External link: https://github.com/IUNetSci/botometer-python 

Botometer Page

Botometer is a bot detection algorithm using feature-based machine learning techniques. The output of the algorithm is between 0 and 1. 1 means that an account is highly likely to be a bot. 0 means that an account is highly likely to be a human account. Please read their documentation very carefully before starting the work.

tweetborornot2

This R implementation includes feature contribution to bot score. It is relatively easy to implement. Let's try this too!