Vrednovanje strojnog prevoditelja Google Translate na primjeru jezičnog para turski i hrvatski

master's thesis

• Title (english): Evaluation of the machine translation service Google Translate for Turkish-Croatian language pair
• Author: Željka Leskovar
• Mentor: Ivana Simeon (mentor)
• Committee member: Matea Filko (predsjednik povjerenstva)
• Committee member: Marko Tadić (član povjerenstva)
• Committee member: Ivana Simeon (član povjerenstva)
• Granter: University of Zagreb; Faculty of Humanities and Social Sciences; (Department of Linguistics); Zagreb
• Defense date and country: 2021-03-03, Croatia
• Scientific / art field, discipline and subdiscipline: HUMANISTIC SCIENCES; Philology; General Linguistics

Abstract

Ovim radom vrednovan je alat za strojno prevođenje Google Prevoditelj na konkretnom jezičnom paru turski i hrvatski. Analiza polazi od kreiranja korpusa materijala koji se sastoje od originalnih tekstova izvornog jezika, njihova strojnog i konvencionalnog prijevoda kao referentne točke. Odabrani istraživački pristup koristi se kvalitativnom i kvantitativnom analizom pogrešaka koje su prepoznate u izlaznim podacima strojnog prijevoda s ciljem uvida u trenutačno stanje kvalitete prijevoda na manje istaknutim ili korištenim jezicima na globalnoj razini.
Strukturne razlike odabranog jezičnog para dodatno su naglašene razlikom u jezičnim stilovima odabranih tekstova koji su dovedeni u korelaciju s količinom i vrstom pogrešaka izlaznih podataka. Uz statistički prikaz prepoznatih pogrešaka provedena je i lingvistička obrada rezultata s ciljem dublje analize samog jezičnog para i pogrešaka koje su specifične za njega.
Na samom kraju rad donosi nedostatke predmetne analize i prijedlog za promjenu pristupa u budućim istraživanjima.

Abstract (english)

This thesis evaluates machine translation provided by online tool Google Translate, on a specific language pair Turkish and Croatian. Analysis started off by building a corpus consisted of source language texts as a base and then their machine translation using online tool Google translate and last but not least, conventional translation texts as a reference point. The research part of the thesis is mainly based on qualitative and quantitative analysis of errors identified in machine translation output for gaining insight into current state of translation quality in less prominent languages used on a global scale.
The structural differences between chosen languages are further accentuated by the difference in written style of the selected texts that can be put into correlation with the amount and type of output errors. In addition to the statistical presentation of the identified errors, a linguistic processing of the results was performed with the aim of a deeper analysis of the language pair itself and their characteristical errors.
At the very end, the thesis brings the shortcomings of the analysis and a potential proposal for changing the approach in future research.

• Keywords: strojno prevođenje; analiza pogrešaka; turski jezik; hrvatski jezik; Google Prevoditelj
• Keywords (english): machine translation; error analysis; Turkish; Croatian; Google Translate
• Language: croatian
• URN:NBN: urn:nbn:hr:131:503041
• Study programme: Title: Linguistics (double major); specializations in: General Linguistics, Comparative Linguistics, Computational Linguistics, Aplied Linguistics and Cognitive Linguistics, Cognitive; Course: Computational Linguistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra lingvistike (magistar/magistra lingvistike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2021-06-23 21:02:48