Naslov
Application of affective lexicons in sports text mining: a case study of FIFA World Cup 2018

Autori
Avdić, Dino

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

Fakultet
Fakultet elektrotehnike i računarstva

Mjesto
Zagreb

Datum
12.07

Godina
2019

Stranica
36

Mentor
Bagić Babac, Marina

Ključne riječi
text mining ; sentiment analysis ; football ; Reddit ; supervised algorithms ; machine learning

Sažetak
Social media act as a living laboratory for monitoring human behaviour, since people write their comments, opinions, views and emotions on daily basis. For the purpose of emotion mining from text, it is especially interesting to investigate user behaviour in sports domain, where sports events usually evoke highly polarized emotions of fans. The goal of this thesis is to show an example of creating an affective lexicon from user-generated data. In particular, a dataset from social networks during the FIFA World Cup 2018 is to be exploited. Then, a comparison between the applications of general affective lexicons and domain specific lexicons is to be provided. The results are to be applied on user behaviour indicating what kind of affective dimensions can be recognized, and how they can be interpreted.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA