Talking handheld dictionaries

By Galina Vítková

We have already discussed the importance of the good dictionary in studying foreign languages, in our case Technical English (see Dictionary – your best helper in mastering English words and Electronic dictionaries). Modern technologies have brought new possibilities in application of dictionaries making it more comfortable and effective. First of all, it is connected with occurrence of talking mobile dictionaries. Talking dictionaries are produced to translate in 2 ways individual words, phrases or entire sentence, full text (in this case besides they pursue a function of translators).

Talking handheld dictionaries use several technologies. They are as follows:

TTS (Text-to-Speech or speech synthesis) technology enables the conversion of the computer readable text into the audible speech. The speech synthesis technology comprises a) taking any ordinary text, b) analysing it by means of a speech synthesis engine, c) processing the information, and finally d) transforming it into human speech () in a voice stream form. The transforming is performed by a computer, which renders the voice of a native speaker of a target language. The voice stream form can be stored in several audio file formats.

Speech Recognition technology allows applications to “understand” voice commands, i.e. the user may talk to the device, which provides a translation based on this input. The quality of speech recognition is chiefly influenced by a) speed of speaking, b) correctness of pronunciation, c) male, female, child voices. Hence software affords initial voice recognition training. Then it is able to choose a translation that matches the input from a set of possible meanings of words or phrases. There are systems that don´t need voice recognition training.

Machine Translation tools apply computational linguistic analysis of source documents to prepare draft translations of the input. Even though the prepared translation is illegible sometimes. Some phrases require more accurate human translation.

The technology of image translation enables the user take a picture (using the device camera) of some printed text (a secret code, a restaurant menu, a page of a document etc.), and then translate it into the chosen language.

These technologies are used in:

• Dictionaries – talking and non-talking applications provide translation from pre-installed databases
• Phrasebooks – include talking and non-talking translation of travel or professional related phrases
• Picture Dictionaries – provide translation of words based on picture selection
• Flashcards – collections of learning games to improve vocabulary and memorization
• Language learning

Supported languages. All talking dictionaries support the English language. Most dictionaries support other European languages, mostly Spanish, French, German, sometimes Russian. Due to a remarkable increase of the number of modern mobile devices with advanced features, the number of language pairs presented by talking dictionaries grew too. Many producers propose translation from and into over 20 languages, or over 200 language pairs, including most Latin languages. Japanese and Chinese service providers commonly offer cross-translation for Japanese, Chinese, English and Korean.

There are a lot of producers of talking dictionaries. Many of them use advanced technologies, which is especially important for the speech synthesis and speech recognition. Ectaco, Inc., a New York-based company, designs and manufactures handheld talking devices. Its software division LingvoSoft develops and releases software for more than 40 languages and for most major operating systems and platforms including Palm OS, Windows, and Symbian. LingvoSoft applications usually contain a combination of the company’s own technology and technologies of other companies. Fonix Speech, Inc. offers very interesting “voice technology solutions”. These solutions are based upon the Company’s proprietary artificial neural network-based automatic speech recognition (ASR) and Fonix DECtalk, a text-to-speech system (TTS). Fonix speech technology is user-independent, i.e. no voice training is needed. The Company is much engaged on the emerging international market of electronic dictionaries and cell phones, especially in Asia. Its main partners in Asia are Casio, Canon, Seiko Instruments. Find more information about other producers in the list Electronic Dictionary from other manufacturers

References: http://en.wikipedia.org, http://www.ectacoinc.com/, Ectaco Inc., LingvoSoft , Fonix Speech Group – http://www.fonixspeech.com/