I'm building a Japanese study app on anime and drama subtitles,and having the subtitle library made this cheap to count.
Every line is split into dictionary words. The 2,000 most-common-word frequency list is created by looking at 6,000 titles, each weighted equally. A show's score is
the share of its words that fall in that set. Character names are removed using
each season's cast list.
If you want to talk about language learning let me know at will@<the domain I linked to>! I love talking about this stuff.
zahlman•3m ago
Interesting, but what do you propose to do with the result? This is at best a quite crude proxy for the difficulty of understanding the text.
willmorrison•11m ago
Every line is split into dictionary words. The 2,000 most-common-word frequency list is created by looking at 6,000 titles, each weighted equally. A show's score is the share of its words that fall in that set. Character names are removed using each season's cast list.
If you want to talk about language learning let me know at will@<the domain I linked to>! I love talking about this stuff.
zahlman•3m ago