Detekcija tipova ličnosti korištenjem učenja prijenosom i jezičnih modela

master's thesis

• Title (english): Personality Type Detection using Transfer Learning and Language Models
• Author: Mihaela Bošnjak
• Mentor: Jan Šnajder (mentor)
• Committee member: Jan Šnajder (predsjednik povjerenstva)
• Committee member: Bojana Dalbelo-Bašić (član povjerenstva)
• Committee member: Domagoj Jakobović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-10, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Detekcija ličnosti je problem koji pronalazi brojne primjene u marketingu i unaprijeđenu korisničkih usluga ali nailazi na probleme poput nedostatka podataka i modela koji još uvijek ne rade dovoljno dobro za uporabu u produkciji. Ovaj rad se primarno bavi problemom nedostatka podataka te kako možemo iskoristiti metode prijenosnog da olakšamo taj problem. U ovom radu smo proveli pokuse koji testiraju moći modela BERT kao modela za reprezentaciju riječi te usporedili rezultate s ostalim poznatim rezultatima na istim datasetovima.

Abstract (english)

Personality Detection is a problem that finds numerous applications in marketing and in user services but faces problems such as the lack of data and models that are still not working well enough for use in production. This thesis deals primarily with the lack of data and how we can use transfer learning to facilitate this problem. In this thesis, we conducted experiments that test the power of BERT as a word representation model and several experiments on different interesting hypothesis we have laid out while exploring out dataset. In the end, we compare our results with other well-known results on the same datasets.

• Keywords: Detekcija ličnosti; BERT; prijenosno učenje
• Keywords (english): Personality Type Detection; BERT; Transfer Learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:582671
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 16:56:17