Naslov
Twitter sentiment Analysis before and after USA election 2020

Autori
Topić, Karla

Vrsta, podvrsta i kategorija rada
Ocjenski radovi, diplomski rad, preddiplomski

Fakultet
Fakultet elektrotehnike i računarstva

Mjesto
Zagreb

Datum
26.06

Godina
2023

Stranica
30

Mentor
Bagić Babac, Marina

Ključne riječi
emotion ; text analysis ; twitter ; US presidential elecions 2020 ; python ; nlp

Sažetak
The purpose of this study is to analyse sentiment before and after (more precisely during) USA presidential election 2020 and reveal how that sentiment varied. We sought to compare the emotional content of approximately million tweets containing #trump, #donaldtrump, #biden, #joebiden using NRC VAD Lexicon, NRC lexicon and VADER sentiment analysis lexicon. After that, our focus was on words that co-occur with words biden and trump, so we calculated PMI for that words and then perform series of linear regressions to see if there was relation between PMIs and the dimensions of the affect or four emotions labels (anger, fear, joy, sadness). Finally, we extract topics from tweets to see what are tweets about and which topic are dominant in tweets at that time. Keywords: emotion, text analysis, Twitter, USA Elections 2020, natural language processing.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA