Skip to content

Auxiliary part for the diploma, on the topic Language features of verbalization of emotions in the crisis media discourse of 2020-2021

Notifications You must be signed in to change notification settings

AIGUL2/tonality_analysig

Repository files navigation

Analysig tonality

Auxiliary part for the diploma, on the topic Language features of verbalization of emotions in the crisis media discourse of 2020-2021

The empirical material for this study was the headlines of news texts posted on such information resources as the Russian News Agency (hereinafter RIA Novosti) and Rosbusinessconsult (hereinafter RBC) in Russian.

In total, the empirical material amounted to 8,052 text headings of articles, of which 142 were taken from RBC, and 7910 from RIA Novosti.

The following marked-up lexicons were used.

  1. Productcenters, 2012 (https://www.semanticscholar.org/paper/Extraction-of-Russian-Sentiment-Lexicon-for-Product-Chetviorkin-Loukachevitch/e23b1430b3c4c4850db1336c5ba9c51c2084f29b?p2df)
  2. Emolex, 2013 (https://paperswithcode.com/paper/crowdsourcing-a-word-emotion-association)
  3. Chan-Skiena's Lexicon, 2014 (https://sites.google.com/site/datascienceslab/projects/multilingualsentiment)
  4. LinisCrowd, 2016 (http://www.linis-crowd.org/)
  5. RuSentiLex, 2017 (http://www.labinform.ru/pub/rusentilex/rusentilex_2017.txt)
  6. SentiRusColl, 2019 (https://github.com/kotelnikov-ev/SentiRusColl)

I hope for your attention and understanding. The code serves informational purposes and gives an idea of the overall picture of the study.

About

Auxiliary part for the diploma, on the topic Language features of verbalization of emotions in the crisis media discourse of 2020-2021

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published