The rest of this article will focus on electronic translators, but that doesn't end the winnowing process. As a result, translators are ideal for people who want to travel to a country without learning the native tongue, while dictionaries appeal to students or travelers with a good working knowledge of the language(s) of their destination. They organize content around thematic subjects - hotels, restaurants or airports, for example - making it possible to locate commonly used expressions or phrases that obey proper rules of syntax and grammar. They offer no insights into the structure or rules of a language, so while they can tell you the French word for "ketchup" - it's ketchup! - they won't help you ask, "May I please have some ketchup with my french fries?" Electronic dictionaries enable you to look up a word and find its equivalent in another language. Search the sites of a few leading brands, and you'll quickly see two basic kinds of devices - dictionaries and translators. Before you rush to get yours, with dreams of speaking fluent Spanish as you stroll the streets of Madrid, keep reading.įor a device that's meant to simplify the lives of travelers or foreign language students, it's amazing how confusing the product category can be. Now companies like Franklin, ECTACO and Lingo feature a stunning - and stupefying - array of electronic translators.
Ectaco itranslate voice translator software#
The hardware and software born from this R&D eventually trickled down into consumer electronics. Defense Advanced Research Projects Agency began throwing money around to develop translators for military use, the momentum finally shifted, and hand-held translators moved from the realm of science fiction to reality. Suddenly, huge amounts of bilingual text became available, making statistical machine translation both feasible and, because of the huge reservoir of data available to it, more accurate than traditional, rules-based methods of translation. That final information barrier collapsed when the Internet was commercialized in the 1990s. The problem? It was difficult to get enough raw material to make the calculations possible, like trying to base probabilities on five coin tosses versus 5 million tosses. They directed computers to analyze translated texts to determine the probability that a word or phrase in one language matched a word or phrase in another. In the 1980s, computer scientists devised a way to make translations using statistical probability instead of complex rules based on syntax, grammar and semantics. In 1966, in a symbolic shrug of defeat, the Automatic Language Processing Advisory Committee reported that humans could perform faster, more accurate translation at half the cost. Developing algorithms that could accurately translate foreign languages proved to be enormously challenging. Here's a quick recap: After Georgetown University and IBM developed a machine capable of translating 60 sentences from Russian to English in 1954, scientists predicted that computers would perform near-perfect translation within five years.
0 kommentar(er)
