Would be great if the trainng tool was able to train to like or dislike specific words that are inside the article.
Ohh I wish! It’s just way too much work to go searching through full text on top of title, tags, and author.
So this is a impossible to exist feature in newsblur?