prikaz prve stranice dokumenta Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima
No public access
undergraduate thesis
Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2018. urn:nbn:hr:168:371675
Ilakovac, Mihovil
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Ilakovac, M. (2018). Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima (Undergraduate thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:371675

Ilakovac, Mihovil. "Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2018. https://urn.nsk.hr/urn:nbn:hr:168:371675

Ilakovac, Mihovil. "Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2018. https://urn.nsk.hr/urn:nbn:hr:168:371675

Ilakovac, M. (2018). 'Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima', Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 18 November 2024, https://urn.nsk.hr/urn:nbn:hr:168:371675

Ilakovac M. Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima [Undergraduate thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2018 [cited 2024 November 18] Available at: https://urn.nsk.hr/urn:nbn:hr:168:371675

M. Ilakovac, "Modeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima", Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2018. Available at: https://urn.nsk.hr/urn:nbn:hr:168:371675

Metadata
TitleModeli strojnog učenja za prepoznavanje sarkazma u korisničkim komentarima
Title (english)Machine Learning Models for Sarcasm Detection in Users' Comments
AuthorMihovil Ilakovac
MentorJan Šnajder (mentor)
Committee memberJan Šnajder (predsjednik povjerenstva)
Committee memberMladen Karan (član povjerenstva)
Committee memberDomagoj Alagić (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2018-07-13, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Jedan od teških problema je zaključiti iz podataka nešto što zapravo nije tamo. Takav problem je problem prepoznavanja sarkazma unutar teksta. Sarkazam je govorna figura u kojoj osoba govori nešto suprotno od onoga što stvarno misli. Stvarali smo vlastiti skup podataka i razvili tipologiju sarkazma kako bismo što točnije označili dobivene podatke. Morali smo odustati od stvaranja vlastitog skupa podataka zbog krivo procjene potrebnog vremena za označavanje. Koristili smo Stanfordov skup podataka SARC koji sadrži samo-označene podatke s Reddita. Problem prepoznavanja sarkazma smo odlučili riješiti korištenjem strojnog učenja. Odabrali smo tri modela: stroj potpornih vektora, logistička regresija i slučajne šume. Ostvarili smo rezultate koji su bolji od osnovnog modela, ali ima još prostora za napredak.
Abstract (english)
One of hard problems is concluding something based on data that says explicitly one thing but the author meant something different. One of those problems is sarcasm detection inside of texts. Sarcasm is a speech figure that expresses something opposite to what would a sincere utterance express. We have worked on creating our dataset and developed a sarcasm typology in order to label the data precisely. We had to stop with the labeling process due to a wrong estimation of the time needed to complete the labeling task. Stanford had published a self-labeled Reddit dataset called SARC which we used. The problem of sarcasm classification we decided to solve with machine learning. Three models were selected: support vector machine, logistic regression and random forrest. We have achieved results that are better than the baseline but there is room for progress.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:371675
Study programmeTitle: 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 resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2019-03-13 20:56:13