Reprezentativno učenje u obradi prirodnog jezika

undergraduate thesis

• Title (english): Representation learning for natural language processing
• Author: Ana Pribanić
• Mentor: Nives Mikelić Preradović (mentor)
• Committee member: Tomislava Lauc (predsjednik povjerenstva)
• Committee member: Sanja Kišiček (član povjerenstva)
• Committee member: Nives Mikelić Preradović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Humanities and Social Sciences; (Department of information and Communication sciences); Zagreb
• Defense date and country: 2021-09-17, Croatia
• Scientific / art field, discipline and subdiscipline: SOCIAL SCIENCES; Information and Communication Sciences

Abstract

Obrada prirodnog jezika postaje sve aktivnijim područjem istraživanja. Razvoj tehnika reprezentativnog učenja od lokalnih do distribuiranih reprezentacija omogućio je tehnologijama obrade prirodnog jezika bolje razumijevanje semantičkog značenja i konteksta riječi i rečenica na prirodnim jezicima. Veliki napredak ostvaren je dvosmjernim treniranjem modela transformera čija je izvedba prikazana u ovom radu vizualizacijom skupa podataka IMDB recenzija na hrvatskom i engleskom jeziku.

Abstract (english)

Natural language processing has become an increasingly active area of research. The development of representative learning techniques from local to distributed representations has enabled natural language processing technology to better understand the semantic meaning and context of words and sentences in natural languages. Great progress has been made in the bidirectional training of transformer models whose performance is presented in this paper by their visualization of a set of IMDB reviews in Croatian and English language.

• Keywords: obrada prirodnog jezika; reprezentativno učenje; lokalna reprezentacija; distribuirana reprezentacija; transformeri; BERT
• Keywords (english): natural language processing; representation learning; local representation; distributed representation; transformers; BERT
• Language: croatian
• URN:NBN: urn:nbn:hr:131:538908
• Study programme: Title: Information Sciences; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) informacijskih znanosti (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) informacijskih znanosti)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2021-09-17 15:29:01