Transliteration
From Nlpwiki
Definition
Transliteration is the systematic conversion of one script into another. Transliteration is not translation. The output of transliteration is not a word in the target language, but rather a loose phonetic representation of a source-language word in the target-language script.
Romanization is a type of transliteration, that of a foreign script into the Roman alphabet. For example, the Greek word συντάγμα can be Romanized as {syndagma}.
Transliteration schema may differ from one another as defined by a few key properties—whether they are predictable, pronounceable, unambiguous, and complete. One transliteration scheme is not unconditionally better than another for having or not having one of these properties, but is rather less or more suited for a certain task. In short, there is not a single, most desirable transliteration mapping for any pair of scripts, only a scheme which best fits its intended use.
Transliteration Tools
| Tool Name | Company | Website | Languages | License | Interface | Platform | Price | Romanization |
|---|---|---|---|---|---|---|---|---|
| ICU Transform | ICU Project | ICU Project | 23 languages; configurable for any script pair | Open source. ROTS | C, C++, Java APIs | Any | Free | Yes |
| English NOW | The MITRE Corp. | n/a | 11 languages | None. ROTS | Command line | Any | Free to Government | Yes |
| Arabic Editor | Basis Technology Corp. | Basis Tech | Arabic | COTS. Single workstation and more | PC, Linux | Varies, $99.99 for single workstation | Yes | |
| Microsoft Transliteration Utility | Microsoft | Microsoft Downloads | Bosnian Cyrillic | Serbian Cyrillic | Malayalam | Inuktitut | Korean-Hangul | User-generated modules can address any language pair | None. ROTS | GUI, command line | PC | Free | Yes |
| NameSphere | Apptek | Apptek | many | Single or multi-user licenses | SDK and API available | Cross-platform integration possible | Price quote needed | Yes, for proper nouns |
| Google Translate | Google Translate | 22 languages | None | Web GUI | Any | Free | OOV items | |
| Chinese Script Converter | Basis Technology Corp. | Basis Tech | Chinese - Simplified | Chinese - Traditional | Server, SDK available | PC, Linux | $32,000-$100,000 | No | |
| Azhagi | SRV Consultants | Azhagi.com | Tamil | COTS. Single, multi-user licenses available with online registration. | GUI | PC | Free or Professional Edition at 1000 Indian rupees ($20) + Rs.500 ($10) per additional user | Yes |
| Translit Online for Microsoft Office | LatKey | LatKey | Arabic | Greek | Hebrew | Hindi | Russian | None | MS Office embedded | PC | Free | Yes |
| Shahmukhi-Gurmukhi Transliteration solution | Punjabi University at Patiala | Advanced Center for Punjabi Language | Punjabi | Gurmukhi | Shahmukhi | TBD | Standalone or Web GUI | Standalone for PC; any browser | TBD; Web version free | No, Punjabi only. |
| Sangam | Punjabi University at Patiala | Punjabi Univ. | Punjabi | Shahmukhi | Gurmukhi | English | Single-user and Organizational licenses | Standalone | PC | $100 single/$400 Organization | Yes |
| Russian Decoder | SmartLink Corporation | SmartLink | Cyrillic | Russian | Server | Standalone or Web GUI | PC | Online version free. Quote needed for Professional edition. | Yes |
| Freeware transliterators | n/a | n/a | Many, with an emphasis on Cyrillic and Indic scripts | Numerous freeware tools offer transliteration capabilities for one specific script. Most of these operate via a web-based GUI where the user types or pastes text into a field for processing. | Free | Often | ||
| IME/Scripting Tools | n/a | n/a | Many | Many of the professional and freeware programs which are said to perform "Transliteration of English to language x" do not technically do "transliteration". Rather, these tools are typed-text converters, converting ASCII characters typed on a QWERTY keyboard into a foreign script.
Romanization is not possible with such a tool. | Varies, some free | No | ||
