Find high FREX and high lift words in Stranger Things dialogue

This screencast walks through how to use topic modeling and tidytext to learn topics and identify important words, with #TidyTuesday Stranger Things dialogue. Check out the code on my blog: juliasilge.com...

[@hris2570]

Thank you for all you do, I have learnt alot from you in R. Thank you.

[@cheninitayeb6590]

Thank you Julia for this session even WE miss you this few month bcz you did not lot of feature ingenuerie thanks lot for your efforts

[@CaribouDataScience]

Very nice! Thanks

[@olexiypukhov-KT]

Is there going to be a new video? I can't wait for the next one!

[@houssineabaali7882]

Still working as of today, ty!

[@saraoliver2557]

Dear Julia,
thank you very much for your content! Learning about FREX and high lift words will definitely improve my workflow. On a general note, I got the impression that there is currently a kind of hype regarding „BERTopic“. Do you know about an option to use this approach in R? Kind regards, Sara 😊

[@JuliaSilge]

I have not played around with this in R, no! It would be really interesting to see.

[@angvl8793]

Thx u !

[@sangwonchu7662]

is there any advantage of using logodds rather than tf-idf?

[@JuliaSilge]

It can have advantages because it is weighted in a different way. Check out this section (especially the last paragraph):
juliasilge.github.io/tidylo/articles/tidylo.html#jane-austen-and-bigrams
And this post by Tyler (one of the authors of tidylo with me):
medium.com/@TSchnoebelen/i-dare-say-you-will-never-use-tf-idf-again-4918408b2310