Naslov
Developing a Lexicon for Calculating Emotions based on Social Media Reactions

Autori
Šimić, Andrej

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

Fakultet
Fakultet elektrotehnike i računarstva

Mjesto
Zagreb

Datum
29.06

Godina
2022

Stranica
31

Mentor
Bagić Babac, Marina

Ključne riječi
natural language processing ; sentiment lexicons ; emotion lexicons

Sažetak
With the constant rise of social media networks during recent years, the amount of user-generated content is increasing substantially. Recent research has shown that an immense amount of this social media content is not neutral and has a tendency of emotionally affecting its readers. As some of the social media platforms include a feature of denoting posted content with labels representing different emotional reactions, it has become possible to identify and analyze how exactly and to what extent the content affects users it has reached. This thesis covers the process of developing specific lexicons from user reactions on social media that are based on labelled datasets from different application domains. An algorithm for applying these lexicons is designed, implemented, and compared to existing lexicon-based algorithms in the field of emotion analysis.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA