Finally, you should check out the Text-Analytics-as a Service from
GATE & @sheffielduni.bsky.social (@vera-ai.bsky.social partner). There are plenty of useful tools, including hyper-partisan scorers 🙃

Here is a useful crash course on Hugging Face transformers provided by
@journocoders.bsky.social
paper.dropbox.com/doc/Journoco...

For text classification, you can also consult Hugging Face Transformers, which offers multiple state-of-the-art NLP models. There, you can find the Offensive Language Identification Dataset (OLID) or the GoEmotions dataset, which contains labels for 27 emotion categories.

When using it to analyse JD Vance's speech from the Munich conference, sentences flagged as negative are the anecdotal stories about Europe's alleged issues with freedom of expression and religious liberties

Prompt Compass is available here github.com/ErikBorra/Pr... thanks to Erik Borra via
@digitalmethods.bsky.social

Prompt-Compass allows you to automatically sift through sentences to surface those with negative or positive sentiment. It also supports hate speech detection and identifies named entities around which the speech has been crafted