Primjena i vrednovanje metoda za detekciju sarkazma u engleskome jeziku

undergraduate thesis

• Title (english): Application and evaluation of methods for detecting sarcasm in the English language
• Author: Josip Pavlić
• Mentor: Jan Šnajder (mentor)
• Committee member: Jan Šnajder (predsjednik povjerenstva)
• Committee member: David Dukić (član povjerenstva)
• Committee member: Ana Barić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-08, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj završni rad bavi se primjenom i vrednovanjem metoda za detekciju sarkazma u engleskom jeziku. Cilj je bio ispitati učinkovitost različitih modela strojnog učenja, uključujući stroj potpornih vektora (SVM), BERT i višezadaćno učenje, na tri podzadatka: identifikacija sarkazma, određivanje podkategorija sarkastičnih tekstova i prepoznavanje sarkazma u rečenicama i njihovim nesarkastičnim inačicama. Rezultati pokazuju da modeli BERT daju bolje rezultate od SVM-a, dok višezadaćno učenje nije pokazalo poboljšanja zbog velike razlike u prirodi podzadataka i ograničenosti skupa podataka.

Abstract (english)

This thesis explores the application and evaluation of methods for detecting sarcasm in the English language. The aim was to examine the effectiveness of various machine learning models, including Support Vector Machines (SVM), BERT, and multitask learning, on three subtasks: identifying sarcasm, determining subcategories of sarcastic texts, and recognizing sarcasm in sentences and their non-sarcastic paraphrases. The results show that BERT models perform better than SVM, while multitask learning did not show improvements due to the significant difference in the nature of the subtasks and the limitations of the dataset.

• Keywords: obrada prirodnog jezika; detekcija sarkazma; transformatori; višezadaćno učenje; strojno učenje
• Keywords (english): natural language processing; sarcasm detection; transformers; multitask learning; machine learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:220942
• 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: 2024-10-12 21:47:42