2 ms·
I trained it by tokenizing my post history with nltk and using the number of likes (scraped via JS) as the target metric. Then I built a simple recommender that
by rcarmo 5y ago
I trained it by tokenizing my post history with nltk and using the number of likes (scraped via JS) as the target metric. Then I built a simple recommender that would take my draft and list the most interesting tags for it.