What is this?
This repository contains Python and R code that can be useful for data scientists to download, modify and analyze fbTREX data (and in the near future ytTREX and other platforms supported by tracking.exposed). The tool has some scripts to download and produce/modify datasets, and other scripts to visualize data. It includes a R shiny app and Python Notebooks. The tools aim at understanding more about algorithms and their impact on peoples lives and on society, and are not written or designed to allow research on users. Note that you can only use your own token(s) with the dashboard, so you should install the facebook.tracking.exposed first.
Who is this for?
Data scientists with some coding skills in Python or R can use this repository as a starting point to develop more complex visualizations with their own tracking.exposed datasets. For example, they can make comparisons between users or between what a user sees and what specific sources produce. Many features can be implemented in the library, just write your code and make a PR or open an issue.
This tools are free to use and modify and can be employed to conduct research in an academic context. Social scientists as well as researchers with a different background can use the tools to answer questions about the inner mechanics of the algorithm. The contributors of this repository are open to any suggestion and willing to develop further following indications by researchers.