Analiza sentimenta na Twitteru prije i poslije američkih izbora 2020

undergraduate thesis

• Title (english): Twitter sentiment Analysis before and after USA election 2020
• Author: Karla Topić
• Mentor: Marina Bagić Babac (mentor)
• Committee member: Marina Bagić Babac (predsjednik povjerenstva)
• Committee member: Melita Fertalj (član povjerenstva)
• Committee member: Kristijan Poje (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-07, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Analiza sentimenta na Twitteru prije i poslije američkih izbora 2020
Cilj ovog istraživanja je analizirati sentiment prije i poslije (točnije tijekom) predsjedničkih izbora SAD-a 2020. godine i otkriti kako se taj sentiment mijenjao. Cilj nam je analizirati emocionalni sadržaj otprilike milijun tweet-ova koji sadrže #trump, #donaldtrump, #biden i #joebiden koristeći NRC VAD leksikon, NRC leksikon i VADER leksikon za analizu sentimenta. Nakon toga, usredočili smo se na riječi koje se pojavljuju zajedno sa riječima biden i trump. Izračunali smo PMI za te riječi, a zatim izveli niz linearnih regresija kako bismo vidjeli postoji li veza između PMI-eva i dimenzija efekta ili četiriju emocija (ljutnja, tuga, sreća, strah). Konačno, izvukli smo teme iz tweet-ova kako bismo vidjeli o čemu tweet-ovi zapravo govore i koje su teme dominantne u tweet-ovima u to vrijeme.
Ključne riječi: emocija, tekst analiza, Twitter, SAD predsjednički izbori 2020

Abstract (english)

Summary
Twitter sentiment Analysis before and After USA election 2020
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.

• Keywords: emocije; tekst analiza; twitter; SAD predsjednički izbori2020; python; nlp
• Keywords (english): emotion; text analysis; twitter; US presidential elecions 2020; python; nlp
• Language: english
• URN:NBN: urn:nbn:hr:168:783984
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 10:48:38