Rosette® Name Translator (RNT) accurately translates names from many languages to a standard English spelling suitable for integrating into software used by people and information systems in finance, law enforcement, intelligence, border control, and litigation support.
Governments and financial institutions use Rosette Name Translator to standardize the spelling of names of people and places to ensure that information retrieval systems will find all occurrences of a person in their data. For over 20 years, Basis Technology has provided software tools for analyzing and extracting information from multilingual text.
Whether your application just needs a way to import data not written in Latin script, or your users need to get an idea of the names in a document in a foreign script, Rosette Name Translator easily integrates into applications via a Java API.
Rosette Name Translator delivers English translations of names from written Arabic, Persian (Farsi and Dari), Pashto, Urdu, Korean, Chinese, Japanese, and Russian.
These capabilities can be used to:
Annotate names in search results for users not proficient in the language of the documents found
Support name analysis systems that only accept Latin characters as input
Standardize the English spellings of foreign names for data consistency and more complete information retrieval
Rosette Name Translator can also take names transliterated from Arabic, Pashto, and Dari and translate them back to the original language. The example below shows how seemingly different names in English actually map back to the same name in the original Arabic. The name translator understands and correctly handles these variations and delivers accurate and consistent results.
| Transliterated Arabic names | Original Arabic |
|---|---|
| Abul-qassem El-Chabby | ابو القاسم الشابي |
| Abo Al-Qassim Al-Shabbi |
Data in the world is frequently inconsistent, incomplete or otherwise “messy.” Rosette Name Translator handles spelling ambiguities caused by typos and missing spaces. For example, names written in “unvocalized” Arabic (i.e., without short vowels) are automatically vocalized to enable an accurate phonetic translation.
If the name’s language is unknown, Rosette Name Translator will take a “best guess” at detecting the language and then translate accordingly. The translated names on output adhere to the user-chosen transliteration standard, from major standards such as BGN, IC, or FBIS for Arabic names.
Since some translations have to be a “best guess” due to ambiguities of the language, Rosette Name Translator always returns a confidence score with each translated name, so that when accuracy is a very high priority, translations can be sent to a human being to review when necessary.
Transliteration is conversion of a name between scripts, based on the spelling of the original name. The Japanese girl’s name あずさ is transliterated to Latin script substituting “a” for あ, “zu” for ず, and “sa” for さ to produce “Azusa.”
In Arabic and Russian, Rosette Name Translator will convert names foreign to that language to their conventional English spelling. Take the name جورج بوش written in Arabic. Its transliteration is “Jwrj Bwsh,” but the application will return its true translation of “George Bush.”
Rosette Name Translator combines dictionaries, linguistic algorithms, and statistical inference to derive accurate translations. For ease of integration with analysis systems, Rosette Name Translator provides multilingual name translation via a single API. Since a name can produce more than one possible translation due to linguistic ambiguities, RNT provides a confidence score with each translation it returns.
For more information about Rosette Name Translator, download the product datasheet, request a product evaluation, or read a presentation about “Building Applications with Rosette Name Indexer and Rosette Name Translator” by Basis Technology CTO Benson Margulies.