Sykole

The problem with a crosslink database is that the best ones are based on user information, thus requiring a lot of people to actively use the program to work well. Obviously it's also possible to do it via genres and tags but that will either require a lot of work (as in manually registering similarities between bands or assigning subgenres) or an amazing algorithm able to automatically create an amount of precise tags for each and every song.
I suppose it's possible to use YT's suggestion bar to find similarities between bands but even that isn't amazing.

Look at this, he calls himself the "Admin".

Did you also remove 2 of your ribs recently ?

@sylvix95 i created it so that makes me the admin, feels good

@Kaliber84 yeah I thought about that, however, it is possible for me to have the user apply tags to the video manually when adding it to their list, or simply retrieve the video's tags myself using Youtube's API and store it in a database.

It would require me to have a DB structure based on genres, this way the criteria would run faster & more efficiently.

Say for example the following is how it will look like:

electronic:[

https://www.youtube.com/watch?v=gS4domW0_yQ

{

      users:<list of users with this song in any of their playlists>

      songgenres:<list of other genres this song belongs to>

      familiartags:<tags/genres similar to electronic to show in suggestion if wanted>

      familiarvideos<id of similar videos>

}]

So yeah, it's either manually done, or will take some time to fully automate it. Will require an algorithm either way to decide whether a video suggestion fits in or not