Closed ninpnin closed 1 year ago
Great!
I suggest a paragraph on what we want to do as a usecase. My example:
I want to eftract all speeches from 1994-2020 by speakers party and type of debate (partiledardebatt, interpellationsdebatt etc.). I will run a seeded perspective topic model on the corpus to study the saliency and framing of immigration during the period.
Johan and I want to extract all individual speeches from 1920–2020 by speaker’s gender and party belonging. We want to run a LDA topic model on the material to see which topics men and women tend to discuss in the parliament.
Erik wants to extract all individual speeches from 1920–2020 by speaker’s gender and party belonging. He wants study how the concept “internationell” (and alike) is discussed by party belonging and gender. Regarding methods he will probably want to start with some simpler co-occurrence methods that Roger has implemented in Jupyter. But he would probably also be interested in looking into a topic model on the same corpora.
Miriam want to follow a set of words (nedsättande and vardagliga) over time from say 1995 ish by party and year.
Let's gather some straightforward use cases for the corpus here. welfare-state-analytics/riksdagen-corpus#3 is a good example.