Common errors produced by machine translation systems in translations from English into Czech

4.2 Common errors produced by machine translation systems in translations from English into Czech

The Czech word order presents another very peculiar problem. MT systems maintain the English word order, exceptions are only the cases where the system already knows the pattern (translated the sentence many times).

The translations of short sentences tend to be more accurate. The shorter the original text is the more accurate machine translation is produced. On the contrary, the longer the sentence in the source language, the poorer the translation provided by automated translation.

Translators can often discover also trivial mistakes, such as misspelling, incorrect punctuation, and incorrectly capitalized words.

Conclusion

Statistical machine translation systems dominate the field of technical translation at the moment and provide satisfactory results for major languages (spoken by tens of millions speakers), however, Czech language counts as a smaller language with a complex grammatical system and the parallel language corpora does not allow the production of consistently good translations.

The practical part of this thesis attempted to analyze the results of technical translations produced by two statistical machine translation systems – Google Translate and Bing Translator. Analyzed texts consisted of user manual, instruction manual, and technical documentation, such texts are generally supposed to be most suitable for the use of machine translation. The findings confirmed that machine translation from English to Czech faces many problems and requires a lot of attention on the part of translators and also many corrections.

Sentences translated by machine translation systems often convey the correct meaning, yet require edits to enhance the fluency and comprehensibility. Translations of shorter sentences seem to be less problematic and also the simple and clear often used instructions do not produce many errors or mistakes. Problematic parts of sentences are formed by verbs and predicates and also the subject verb agreement tends to be a troublesome task for MT systems.

The use of machine translation systems actually provides an advantage of lower cost to customers who demand and order translations, however, for translators the contemporary machine translation still poses a challenge.

Primary Sources

Secondary Sources

Alexander, L. G., and R. A. Close. Longman English Grammar. London: Longman, 1988. Print.

Apertium. Web. 6 Apr. 2016. .

Baisa, Vít. Strojový Překlad. Brno: Masaryk University, 2013. PDF.

Berneking, Steve, and Scott S. Elliott. Translation and the Machine: Technology, Meaning, Praxis. Roma: Edizioni Di Storia E Letteratura, 2008. Print.

Bojar, Ondřej, and Daniel Zeman. "Czech Machine Translation In The Project Czechmate." Prague Bulletin Of Mathematical Linguistics 101.1 (2014): 71-96. Academic Search Complete. Web. 21 Apr. 2016.

Bojar, Ondřej, Georg Rehm, and Hans Uszkoreit. The Czech Language in the Digital Age = Čeština v Digitálním Věku. Heidelberg: Springer, 2012. PDF.

Byrne, Jody. Scientific and Technical Translation Explained: A Nuts and Bolts Guide for Beginners. Manchester, UK: St. Jerome Pub., 2012. Print.

"Česílko." Cesilko. Web. 6 Apr. 2016. .

De Roeck, Anne. Anatomy of Eurotra: A Multilingual Machine Translation System (Rep.). Université de Liege, 1981. PDF.

Gobbo, Federico. "Machine Translation as a Complex System, And the Phenomenon of Esperanto." Interdisciplinary Description of Complex Systems 13.2 (2015): 264-274. Academic Search Complete. Web. 20 Apr. 2016.

Goutte, Cyril. Learning Machine Translation. Cambridge, Mass: The MIT Press, 2009. eBook Academic Collection (EBSCOhost). Web. 15 Apr. 2016

Hutchins, John. "The History of Machine Translation in a Nutshell." N.p., 2014. Web. 5 Apr. 2016.

Hutchins, John, and Evgenii Lovtskii. "Petr Petrovich Troyanskii (1894–1950): A forgotten pioneer of mechanical translation." Machine Translation 15.3 (2000): 187-221. Web. 5 Apr. 2016

Kelly, Louis G. "History of Translation." Concise History of the Language Sciences: From the Sumerians to the Cognitivists (2014): 419. PDF.

Lagarda, A-L., et al. "Statistical post-editing of a rule-based machine translation system." Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics, Companion Volume: Short Papers. Association for Computational Linguistics, 2009. Web. 5 Apr. 2016

Macháček, Matouš, and Ondřej Bojar. "Evaluating Machine Translation Quality Using Short Segments Annotations." Prague Bulletin Of Mathematical Linguistics 103.1 (2015): 85-110. Academic Search Complete. Web. 20 Apr. 2016.

Newmark, Peter. A textbook of translation. Vol. 1. New York: Prentice hall, 1988. PDF.

"Official Website of the European Union." EUROPA. Web. 6 Apr. 2016. .

Pavlovič, Radek. "David Čaněk: V Některých Jazycích Se Strojový Překlad Přiblíží Tomu Lidskému." www.euro.cz. N.p., 24 Dec. 2014. Web. 10 Apr. 2016.

Richardson, Stephen D. Machine Translation: From Research to Real Users 5th Conference of the Association for Machine Translation in the Americas, AMTA 2002, Tiburon, CA. USA, October 8-12, 2002: Proceedings. Berlin: SpringerLink, 2002. Print.

"SDL Trados Studio - Translation Software." SDL. Web. 6 Apr. 2016. .

Syea, Anand. "The EUROTRA Machine Translation System." Manchester, UK: Centre for Computational Linguistics, 1990. Web. 5 Apr. 2016

"THE TRANSLATION PLATFORM." Memsource: Translation Software That Includes Translation Memory, Machine Translation, Terminology Management, and a Translator's Workbench. Web. 28 Apr. 2016. .

Trujillo, Arturo. Translation Engines: Techniques for Machine Translation. London: Springer, 1999. Print.

Wright, Sue Ellen, and Leland D. Wright Jr, eds. Scientific and technical translation. John Benjamins Publishing, 1993. PDF.

Yang, Jin, and Elke D. Lange. "SYSTRAN on AltaVista a user study on real-time machine translation on the Internet." Machine Translation and the Information Soup. Springer Berlin Heidelberg, 1998. 275-285. Web. 6 Apr. 2016.

Zetzsche, Jost Oliver. "Tool Kit * What Makes a Translation Environment Tool a Good TEnT?" Translators Café.com. N.p., 20 Nov. 2007. Web. 22 Apr. 2016.

English resume

This bachelor’s thesis, titled Machine Translation and Its Use in Technical Translation from English into Czech, focuses on machine translation and provides a concise overview of history and development of machine translation systems. It also describes machine translation systems and translation tools that use MT systems. An important part of this thesis is formed by practical examples of machine translation use in technical translations from English into Czech.

The first chapter deals with technical translation, describes the differences between literary, scientific, and technical translation, and highlights the reasons why technical texts are more suitable for the use of machine translation. The topic of the second chapter is the theory of machine translation, its history and development. The third chapter focuses on the Czech environment. It includes information regarding the Czech language and explains possible errors and mistakes that can occur in machine translated texts.

Practical examples of machine translation use in technical translations from English into Czech are described in Chapter 4. Several scenarios (user guide, user instructions, and technical documentation) are used in these examples. The output of machine translation services (Google Translate and Bing Translator) is analyzed and compared with human translated texts.

The last part is devoted to the overview of the findings and contains a summary of specific errors and mistakes that often occur in machine translated texts.

Czech resume

Tato bakalářská práce s názvem Machine Translation and Its Use in Technical Translation from English into Czech se zabývá problematikou strojového překladu, obsahuje stručný přehled vývoje systémů strojového překladu a popisuje některé systémy strojového překladu a překladatelské nástroje, které strojový překlad využívají. Důležitou součástí této práce jsou i příklady použití strojového překladu při překladu technických textů, které ukazují, do jak velké míry je strojový překlad pro překladatele přínosem.

První část se věnuje technickému překladu a popisuje rozdíly mezi literárním, vědeckým a technickým překladem a vysvětluje, proč je výhodné a vhodné používat právě při technických překladech strojový překlad. Tématem druhé části je teorie strojového překladu, jeho historie a vývoj systémů strojového překladu. Třetí část se zaměřuje zejména na české prostředí, obsahuje informace týkající se českého jazyka a snaží se vysvětlit problémy, ke kterým může při překladu z angličtiny do češtiny s využitím strojového překladu docházet.

Čtvrtá část je věnována praktickým příkladům strojového překladu z angličtiny do češtiny s využitím návodů k použití, návodů k obsluze a další technické dokumentace. Výsledné překlady strojových překladačů (Google Translate a Bing Translator) jsou porovnány s překlady uvedenými v těchto dokumentech a jsou podrobeny jazykové analýze.

Závěrečná část obsahuje přehled zjištěných výsledků a shrnuje konkrétní problémy, které se při strojovém překladu technických textů často vyskytují.

1^ Machine translation is a field of computational linguistics that concerns with the draft, implementation and application of automated systems (programs) for text translations with minimized human input. (translated from Czech by the author of this thesis)

Download 168.63 Kb.

Share with your friends: