Skip to content

zeromh/proj4_refugee_narratives

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Media Narratives on the Refugee Crisis

Frequencies of Moral Foundations Words in Refugee Articles

Objective

I wanted to see if the New York Times and Fox News had different narratives on the 2015 refugee crisis.

Data

I pulled 3000+ articles, letters, and transcripts from the NYT and Fox APIs. All articles were related to "refugees" and were dated to 2015.

I also used the "Moral Foundations Dictionary" provided here. Moral Foundations Theory is a sociological theory which argues that humans' moral concerns focus on 5 or so axes: harm, fairness, ingroup, authority, and purity. (Read more about the theory at the website.) The Moral Foundations dictionary contains a list of words shown to be relevant to each axis. I slightly edited the dictionary to contain multiple endings of each word (e.g. terrorist, terrorists, terrorism, etc.)

Analysis

In the second Jupyter notebook I perform several analyses:

  1. Topic modeling using non-negative matrix factorization. This didn't provide convincing, coherent categories for the articles. I ran NMF on just the NYT articles and also on the entire set from both sources.
  2. Comparison of frequencies of Moral Foundations words in each news source. This analysis provided the most insight to the project. The graphic above shows the most relevant differences!
  3. K-means clustering of articles based on their proportional concern with each Moral Foundations axis. Again the clusters here were not very coherent or insightful.

Slides from my project presentation are in the media folder.

About

Media Narratives on the Refugee Crisis

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages