Human Evaluation of Online Machine Translation Services for English/Russian-Croatian

Seljan, Sanja; Tucakovic, M; Đunđer, Ivan

doi:10.1007/978-3-319-16486-1_108

Cited by 11 publications

(9 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the research of Seljan, Tucaković, and Dunđer (2015) devoted to the assessment of online translators quality on the examples of the English, Russian, and Croatian languages it is asserted according to the criteria of translation accuracy and speech naturalness some optimistic results were achieved, and the inner entirety of the text is described as high.…”

Section: Online Translators and Their Peculiaritiesmentioning

confidence: 99%

Use Of Online Translators To Develop Translator’s Competence Of ‘Digital Natives’ Generation

Ivanova¹

2018

The European Proceedings of Social and Behavioural Sciences

View full text Add to dashboard Cite

show abstract

Section: Online Translators and Their Peculiaritiesmentioning

confidence: 99%

Use Of Online Translators To Develop Translator’s Competence Of ‘Digital Natives’ Generation

Ivanova¹

2018

The European Proceedings of Social and Behavioural Sciences

View full text Add to dashboard Cite

show abstract

“…A new language-pair, i.e. Russian-Croatian, is introduced into the analysis in (Seljan, Tucaković, & Dunđer, 2015) with an additional online translation service, i.e. Yandex.Translate.…”

Section: Croatian Language Error Analysismentioning

confidence: 99%

A comparative error analysis of English and German MT from and into Croatian

Bakarić¹,

Babić²,

Dajak³

et al. 2017

Integrating ICT in Society

View full text Add to dashboard Cite

This paper first gives an insight into the problem of evaluating translations and lists existing approaches. After introducing error taxonomies, a comparative error analysis of Google Translate results is conducted based on the selected taxonomy. The analysis covers translations from English and German into Croatian, and from Croatian into English and German. The intra and inter-annotator agreements, which are rarely accounted for, are reported where applicable. The aim of the paper is better understanding of different notions of error analysis and detecting weak points of such analysis. A preliminary list of guidelines for error analysis is suggested.

show abstract

“…As the human evaluation requires more costs and time, automatic evaluation is requisite in the process of machine translation. Evaluation of machine translated text for Croatian, still obtain lower scores by various online tools when using different methods -1) human evaluation with one or more reference translations or 2) various automatic metrics (Seljan, 2015a;Seljan and Dunđer, 2015b). In this paper, human evaluation was used.…”

Section: Online Translation Toolsmentioning

confidence: 99%

“…Online summarization tool for Croatian -CroWebSum was presented by Mikelic Preradovic et al (2010) yet not integrated in the broader process for information transfer. Free online machine translation services for Croatian still obtain lower grades than for other languages (Seljan et al 2015a(Seljan et al , 2015b, due to scare language resources, differences between language pairs, specific domains, etc. and depending on user's expectations.…”

Section: Introductionmentioning

confidence: 99%