IMDB has looots of tags.

Just scraping them (even with limit) is quite senseless. You end up with lots of tags only having one movie.

Here is a list of all tags including their movie numbers:

https://contribute.imdb.com/updates/field/keywords/value

EMM could scrape that and let the user decide which tags to whitelist for scraping. Or, let the user enter a low bottom value (e.g. 100) about how much movies have to be provided with that tag.