/meta-aNEWSlysis

Exploring how certain issues are depicted in news coverage. The analysis hopes to capture differences in coverage between nation-wide, popular outlets and smaller more localized newspapers. The output would be a tool for understanding discourse within a country on particular issues and how the conversation is evolving.

Primary LanguagePython

NLP News: identifying how health issues are discussed in the news cycle around Canada

Project components

This project is aimed at developing toooling to do a few things:

  • Keep a track of major news coverage of specific health issues within Canada
  • Compare and contrast the sentiment associated with the particular issues globally across the country and locally within sub-regions
  • Identify trending sentiment and interest in particular issues to allow for earlier awareness or concentrate efforts
  • Also help to compare the coverage level for particualr issues as compared to their overall mortality impact in the country

Major scripts

  • getIssueArticleUrls.py: website -> list of URLs
  • scrapeArticle.py: url -> ArticleObject
  • analyzeArticle: ArticleObject -> None
  • analyzeWebsiteCoverage.py-> list of URLs -> data frame of article sentiment, statistics, metadata, date

Useful stats: num distinct journalists covering, time between articles, , number year to date for website, number ytd for all

  • reportFindings.py-> dataframe of article sentiment ->