The Influence of a Machine Translation System on Sentiment Levels
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3AS7XGR5GA" target="_blank" >RIV/00216208:11320/22:S7XGR5GA - isvavai.cz</a>
Result on the web
<a href="https://nlp.fi.muni.cz/raslan/raslan22.pdf#page=211" target="_blank" >https://nlp.fi.muni.cz/raslan/raslan22.pdf#page=211</a>
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
The Influence of a Machine Translation System on Sentiment Levels
Original language description
The aim of the paper is to verify the influence of the used machine translation system on the level of sentiment in the translated text from Slovak to English using the available systems Google Translate and DeepL. The experiment was carried out on a parallel corpus created from subtitles of movies of different styles. The raw parallel corpus contained subtitles in Slovak and English. IBM Watson Natural Language Understanding service was used to identify the sentiment in the subtitles of ten movies of different genres. The paper also describes the methodology of preparing the dataset suitable for sentiment analysis using the IBM NLU service. The research showed a high correlation between human text and machine translation of subtitles for both translation systems. The research results show a high level of onsistency of sentiment levels in both forms of translation. Based on the results obtained, the results of sentiment in machine translation can be generalized for the two most widely used translation systems.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
—
Others
Publication year
2022
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
RASLAN 2022: Proceedings of the 16th Workshop on Recent Advances in Slavonic Natural Languages Processing
ISBN
978-80-263-1752-4
ISSN
—
e-ISSN
—
Number of pages
9
Pages from-to
200-208
Publisher name
Tribun EU
Place of publication
—
Event location
Karlova Studánka, Czech Republic
Event date
Jan 1, 2022
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—