What if social media algorithms tried not to maximize only engagement, but reputational engagement? What if content was promoted not only on the basis of how many people engaged with that content, but also the reputation of those who engage with that content?
This program is a first hack at evaluating an "endorsement score" of tweets or Mastodon statuses, which attempts to gauge the average reputation of those who have engaged with it. If you input a list of tweet ids, it will determine the endorsement score for each tweet, and display each tweet as ranked by its endorsement score.
Important: In order to use this program, you'll need a Twitter API key. You can get one at the Twitter developer portal.
To install, first make sure you have the build tools cmake and ninja installed on your machine.
Secondly, I recommend installing dependencies cpr and nlohmann-json from your package manager (if available). If you do not have these installed, cmake should compile them from source, but it will take significantly longer to build.
Homebrew:
brew install cpr nlohmann-json
Apt (does not include cpr):
apt install nlohmann-json3-dev
Yay:
yay -S cpr nlohmann-json
For the Debian-based distros I tried, it seems you also have to apt install libssl-dev in order to build cpr.
Clone the repository, recursing submodules:
git clone https://github.com/dcourv/endorsement_score --recurse-submodules
And build:
cd endorsement_score
cmake -B build -S . -G Ninja && cmake --build build
The endorsement_score executable will be placed in the build directory.
Lastly, put your Twitter API Bearer Token (see above) into a one-line file called twitter_bearer_token.txt in the directory that you cloned the repsoitory into. The program will not function without a valid Twitter API Bearer Token.
./build/endorsement_score -t will prompt you to sign in to Twitter, and will fetch 50 recent Tweets from your timeline and rank them using a reputational engagement algorithm. (Note: I do not store of your data, it is all locally processed on your machine).
You may also anonymously run the program (without signing into Twitter) with ./build/endorsement_score -t [ids] where [ids] is a space-delimited string of Twitter status ids.
Example: ./build/endorsement_score -t 1492120137205526528
Note that for either option, you will still need to have valid Twitter API keys.
You may also pass -m [ids] instead of -t [ids] and use Mastodon status ids.