Links to other free translation service providersbabelfish.de
searches millions of translations of professional translators, websites and dictionaries. translate.google.com Google's free online translation service translates text and web pages in seconds. This translation tool supports: German, ... online-translator.com PROMT - a free text and website translator - English, German, French, Spanish, Portuguese (Brazilian) and Russian. systran.de Free online translation German-English. For your German-English translations you can use the SYSTRAN technology for free. worldlingo.com The free translator. Translations into and from: English, French, German, Italian, Portoguese, Spanish, Russian, Dutch, ... pons.com ... on PONS. Translate texts from eg English, German, Spanish, French, Russian, Polish and many more. ... your free translation.) http://www.freetranslation.com/en/translation-english-deutsch SDL Freetranslation.com offers you everything you need for German translations. Translation.babylon.com Free online translation from the house of Babylon. Translate German, English, Spanish, French, Chinese, Japanese and more ... How good are free translations?As correctly stated in a Wikipedia article, translators usually have several years of training (or study) - and over the course of professional practice, of course, over many years of experience - and know how to avoid many translation traps. If you need a translation, preferably free of charge, maybe you will remember your circle of friends and ask for a moment: "Can not you quickly translate this for me? It is also very little text. "And which translator has not yet experienced that, a demand for a translation, of course, without payment. The common reasons why a qualified translation service should not be paid typically include statements like these:
Beginnings of machine translationThe beginnings of machine translation can be found in the military field. There was one of the first projects a Russian English translation program for the US military. For the first time, without the help of a human translator or interpreter, one could directly get an idea of the contents of Russian documents. However, in 1966, a US Department of Defense report concluded that machine translation was ineffective, putting down all research in this area.It was not until the 1980s that Siemens AG started research and a Collaborative Research Center "Electronic Language Research" was founded at Saarland University in Saarbrücken. This is how the translation system SUSY was created. The toll-free translation tools include:
In the following I would like to show you with some examples the strengths and weaknesses of such tools - by means of Bing and Google Translator. Please note that these are examples, not a complete, comprehensive analysis. And with this list listed here, I make no claim to completeness. Since I am a translator for English, I concentrate on the german english language pair. StrengthenThe strengths include the translation of individual words and terms - including technical terms, even from the legal field. For example, the term "district court" of Bing is correctly reproduced with "local court". On the one hand, it is impressive that Bing has meanwhile used words that have several meanings (homonyms), such as: For example, the word "lock" that includes context. If you enter "Schloss" alone, Bing translates this first with "castle". However, Bing alters the translation in real time in accordance with your own text input. If you enter "The key stayed in the lock" Bing translates "The key remains in the castle". If one adds the word "stuck", ie the sentence reads "The key was stuck in the lock", Bing changes the translation completely to "The key remains stuck in the lock" and that is correct. In that sense, Bing has definitely evolved in recent years by now incorporating the context, which was not the case in the early days. For Google Translate, "lock" - whether large or small - is the same as "lock".However, Google Translate immediately displays a list of other possible translations, such as: B. castle as well as synonyms in German for "Schloss" Here is another example: If you enter "by mistake", Bing displays "Provided". But if you enter "I accidentally took it with you", Bing changes his translation correctly to "I've taken this by mistake" Why is that?According to a blog post by the Goethe Institute , a funny gibberish came out of machine translation programs in the 1990s because they were still working differently than they do today. They were fed by programmers with the rules of the structure of a language, that is, the rules of grammar and sentence structure and with the individual words. Nevertheless, the computer translated word for word and failed with ambiguous words. Today, machine translation is based on large amounts of data. As explained in a video on Google Translate , the computers are fed with large volumes of texts and their man-made translations (eg, from the United Nations, books, etc.), which have been analyzed for two pairs of languages , Through the so-calledneural machine translation , the translations have greatly improved. Here are translated whole sentences rather. Also Google Translate has been improved with the neural linkage. The computer virtually "discovers" the rules itself. According to the aforementioned Goethe Institut blog article, today's machine translation programs compare, among other things, the frequency of consecutive or nearby words. So, if the word "lock" is near or after or before the word "dwelling", "lock" is translated as "castle", but if it is next to "fix" it will be translated as "lock". In this respect, the context is included statistically. Weaknesses:Reflexive Verbs In reflexive verbs, Bing can not differentiate and ignore the difference in meaning that a verb has in its nonreflective form over the reflexive form. Here is an example: promise - to promise promise (Bing: "promise"), promise (Bing stays with "promise"). Google says "To promise oneself" How about: "Sorry, I promised myself." Also, if you add a word to create even more clarity, such as: In: "Sorry, I just promised myself" Bing stays with his translation and simply expands it to: "Sorry, I just promised myself" Google Translate: "Excuse me, I just promised myself" sayings Example: It was raining cats and dogs. Bing makes it: "It was rain cats and dogs" But it would be correct: "It was raining heavily." Here, however, Google Translate performs better. The program correctly translates the phrase "It's pouring pails." dialect Example: Pfüa God, Pfüat Eich Although Bing recognizes "God", but can not do anything with "Pfüa" and "Eich" recognizes it as "oak and makes it" oak " Google Translate can not do anything with it and says: "Pfüa God, Pfüat Eich", translates only "God" Further weaknesses - in general:A free translation created by a computer program can not be translated to specific target groups. This can only be done by an experienced translator who has received a briefing from his client. Fundamental pros and cons of machine-made translations, as opposed to translations made by human translators, are often the subject of linguistic and translation studies, research, debate, and some discussion or lectures and presentations by academics on this topic - unfortunately only in English - to watch on Youtube, z. For example, the lecture Man vs. Machine: Translation in the Digital Age at the University of Arizona. From the presentation MORE THAN WORDS: Translating culture in language it becomes clear that the weaknesses of computer-generated translations lie here. In a lecture sponsored by Microsoft Research on Future (Present?) Of Machine Translation , the principle of neural translation is explained. At a sub-word level, first and last syllables are also taken into account here. Another lecture in English entitled " A Practical Guide to Neural Machine Translation " explains how neural networks take several words into account at once. So when does a free translation into English make sense?Even what's free, can be good. A machine translation can be useful if you want to inform yourself about the content of a foreign language website in an overview and takes into account that not everything is reproduced correctly. A machine translation can also be useful if you are traveling in a foreign country and want to quickly find out about a specific topic on the Internet. A machine translation can also be useful if you need a basic translation for professional purposes and realize that a rework on the translation is essential if you want to publish them. A machine translation can also be useful to understand a text in outline. Even if you have to translate large amounts of texts, at least a pre-translation by machine tools can be useful. Considering that a translator needs about an hour to translate one page of text (that's just an approximate value!), It immediately becomes apparent that machine translation can save a tremendous amount of time and productivity. In a video by Argos in English, these benefits are further illustrated and a comparison of the increase in productivity for different languages is shown. For large volumes of text, a machine translation can be agood basis for a later final editing by human hands . A blog post by Linguatec lists more uses of machine translation. This includes:
An embarrassing gibberish, referred to on a translator's blog in this context , arose during the visit of Canadian Prime Minister Justin Trudeau. During his visit to the White House, the television broadcaster responsible for translating the French speech of the Prime Minister decided to use a machine translation program associated with speech recognition. The resulting subtitles were often surprising and incomprehensible. Watch Justin Trudeau's speech at the White House in Washington DC on Youtube. In a webcast on neural machine translation by Google Translate, the speaker points out that translating longer passages of text through the computer is very slow and requires a lot of disk space. In addition, these programs have rare word problems and, in addition, computer programs often forget to translate words. As far as German word compounds, ie compounds, are concerned, today's programs can break them down into their components. Unwanted word joke due to incorrect syntax is another problem. Even with strongly inflecting languages such as Slavic languages, where one word changes according to number, case and gender, machine-controlled translation programs do not achieve satisfactory results.
The Future of Machine Translation - ConclusionThe future is bright for the machine-controlled translation of so-called "controlled" texts, ie technical texts. It is to be expected that machine translations will always be improved. Until then, it pays to invest in the editing of a qualified translator and have machine translations proofread and corrected by a trained translator. Translations from human hands are still more accurate,estimating that they are about 60% more accurate than machine translations. Düzenlemek için buraya tıklayın.
Yorumlar
|