It's something else, we got some inspiration from the QTS system proposed by nostr:npub1arkn0xxxll4llgy9qxkrncn3vc4l69s0dz8ef3zadykcwe7ax3dqrrh43w basically it's just social graph distance plus other validations like nip05, ln, 10002s, etc weighted to end up with a normalized score. This approach is much simpler and less sophisticated than any page rank algorithm. However, the beauty of it is that it all boils down to floating point operations, making it easy to tune as you like. You can add your own validators and use them to get a ranking based on what you care about. It's also an experiment ;)
There is more information at https://relatr.xyz/about
nostr:naddr1qvzqqqr4gupzp68dx7vvdlltl7sg2qdv8838ze3tl5tq76y0jnz966fdsana6dz6qq25gnzpveay5jns29z9xdrkdgehw5mvv46k6gd0088