Kavi Mehta

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2019-89

May 20, 2019

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-89.pdf

The language used in political discourse varies depending on the worldview of the speaker, and we describe specific instances of politically charged phrases as political frames. A frame is a group of words that appears consistently across political corpora, and is applied to evoke particular emotions or opinions regarding an issue. Such frames get used in all language, whether it is intentional or not. We propose a system that can automatically pull out frames from a corpus using statistical and linguistic analysis. The result of this system is a tool that ingests large amounts of texts and constructs frame matrices — an organizational structure to categorize and document political frames. Overall, the main contribution is in demonstrating the value of such a system and outlining the means through which it can be done.

Advisors: Jerome A. Feldman

---

BibTeX citation:

@mastersthesis{Mehta:EECS-2019-89,
    Author= {Mehta, Kavi},
    Title= {Underpinnings of Political Leaning: Using Collocation Extraction and Semantic Analysis to Categorize Ideological Frames},
    School= {EECS Department, University of California, Berkeley},
    Year= {2019},
    Month= {May},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-89.html},
    Number= {UCB/EECS-2019-89},
    Abstract= {The language used in political discourse varies depending on the worldview of the speaker, and we describe specific instances of politically charged phrases as political frames. A frame is a group of words that appears consistently across political corpora, and is applied to evoke particular emotions or opinions regarding an issue. Such frames get used in all language, whether it is intentional or not. We propose a system that can automatically pull out frames from a corpus using statistical and linguistic analysis. The result of this system is a tool that ingests large amounts of texts and constructs frame matrices — an organizational structure to categorize and document political frames. Overall, the main contribution is in demonstrating the value of such a system and outlining the means through which it can be done.},
}

EndNote citation:

%0 Thesis
%A Mehta, Kavi 
%T Underpinnings of Political Leaning: Using Collocation Extraction and Semantic Analysis to Categorize Ideological Frames
%I EECS Department, University of California, Berkeley
%D 2019
%8 May 20
%@ UCB/EECS-2019-89
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-89.html
%F Mehta:EECS-2019-89