Skip to content

Using Natural Language Processing to understand topic prevalence in oral debates on AI held in the U.S. Congress and European Parliament.

Notifications You must be signed in to change notification settings

Char19lotte/Master-Thesis

Repository files navigation

MPP-Master-Thesis

This GitHub Repo documents my Master Thesis, submitted in May 2022 at the Hertie School

Thesis title

Understanding topic prevalence in oral debates on Artificial Intelligence: A comparative Natural Language Processing analysis of European Parliament and U.S. Congress debates held in 2021

Contact me

c.kaiser@mpp.hertie-school.org

About

Using Natural Language Processing to understand topic prevalence in oral debates on AI held in the U.S. Congress and European Parliament.

Topics

Resources

Stars

Watchers

Forks