Method and device for transliteration

1. A method for transliteration by a transliteration device including a receiver, a processor, and a display, the method comprising:

receiving, by the receiver, a text input in a source language from a user;

creating, by the processor, source language sub-phonetic units for each word in the text input;

converting, by the processor, the source language sub-phonetic units for the each word in the text input to target language sub-phonetic units;

retrieving, by the processor, ranking of each of the target language sub-phonetic units from a database;

creating, by the processor, target language words based on the target language sub-phonetic units and the ranking of each of the target language sub-phonetic units;

identifying, by the processor, candidate target language words by performing a reverse transliteration for the created target language words; and

displaying, by the display, the candidate target language words to the user.

2. The method of claim 1 , further comprising:

calculating a preference coefficient for each of the target language sub-phonetic units; and

annotating a preference with each of the target language sub-phonetic units, based on the preference coefficient,

wherein the preference coefficient indicates strength of relationship of the source language sub-phonetic units with the target language sub-phonetic units.

3. The method of claim 2 , wherein the preference coefficient is calculated by using Bayes estimation.

4. The method of claim 1 further comprising generating a correspondence between the each source language sub-phonetic unit and the each target language sub-phonetic unit.

5. The method of claim 4 , wherein the correspondence between the each source language sub-phonetic unit and the corresponding target language sub-phonetic unit is generated based on mapping of the each source language sub-phonetic unit with the each target language sub-phonetic unit.

6. The method of claim 4 , wherein the correspondence between the each source language sub-phonetic unit and the each target language sub-phonetic unit is formed by using reflective coefficients, wherein the reflective coefficients provide probabilities of bindings of source language sub-phonetic units with the target language sub-phonetic units.

7. The method of claim 6 , wherein the reflective coefficients are measured by using Bayesian factor.

8. The method of claim 1 , further comprising performing training instances of the sub-phonetic units.

9. The method of claim 8 , wherein the training instances comprises:

receiving feedback, based on the target language words selected by the user;

identifying sub-phonetic units of target language words selected by the user;

modifying preference coefficients for the identified sub-phonetic units; and

storing the identified sub-phonetic units and the modified preference coefficient.

10. The method of claim 8 , wherein the performing training instances comprises:

selecting learned words from a target language corpus;

forming learned sub-phonetic units based on frequently used phoneme sequences in the target language corpus;

analyzing learned sub-phonetic units to derive preferences; and

storing learned sub-phonetic units and corresponding preferences in the database.

11. The method of claim 1 , wherein the displayed candidate target language words are selected by the user.

12. The method of claim 1 further comprising truncating the candidate target language words based on preferences of the sub-phonetic units identified in candidate words,

wherein the preferences of the sub-phonetic units are identified based on preference coefficients and a threshold.

13. The method of claim 12 , wherein target language words are generated according to a reverse transliteration of truncated candidate words into the source language.

14. The method of claim 13 , wherein the truncated candidate words are reverse transliterated by using a correspondence between a source language sub-phonetic and a target language sub-phonetic units based on mapping of the each source language sub-phonetic unit with the each target language sub-phonetic unit, and reflective coefficients.

15. The method of claim 1 , further comprising disambiguating the candidate target language words.

16. The method of claim 1 , wherein identifying candidate target language words by performing the reverse transliteration comprises:

dividing the target language words to form transliterated sub-phonetic units;

retrieving ranking for each transliterated sub-phonetic unit from the database;

forming words in the source language from the each translated sub-phonetic units; and

identifying the candidate target language words according to a degree of matching between the formed source language words and the target language words.

17. The method of claim 1 further comprising validating the candidate target language words based on a target language corpus.

18. The method of claim 1 , wherein the text input in the target language is displayed as an ordered list.

19. An electronic device comprising:

a receiver configured to receive text input in a source language from a user;

a processor configured to create source language sub-phonetic units for each word in the text input, convert the source language sub-phonetic units for the each word in the text input to target language sub-phonetic units, retrieve ranking of each of the target language sub-phonetic units from a database, create target language words based on the target language sub-phonetic units and the ranking of the each of the target language sub-phonetic units, and identify candidate target language words performing a reverse transliteration for the created target language words; and

a display configured to display the candidate target language words to the user.

20. The electronic device of claim 19 , further comprising a disambiguation module configured to disambiguate the candidate words in the target language.

21. The electronic device of claim 19 , wherein the processor is further configured to calculate a preference coefficient for the each of the target language sub-phonetic units, and annotate a preference with each of the target language sub-phonetic units, based on the preference coefficient,

wherein the preference coefficient indicates strength of relationship of the source language sub-phonetic units with the target language sub-phonetic units.

22. The electronic device of claim 19 , wherein the processor generates a correspondence between the each source language sub-phonetic unit and the each target language sub-phonetic unit,

wherein the correspondence is generated based on mapping of the each source language sub-phonetic unit with the each target language sub-phonetic unit or by using reflective coefficients, and

the reflective coefficients provide probabilities of bindings of source language sub-phonetic units with the target language sub-phonetic units.

23. The electronic device of claim 19 , wherein, in order to train instances of the sub-phonetic units, the processor receives feedback based on the target language words selected by the user, identifies sub-phonetic units of target language words selected by the user, modifies preference coefficients for the identified sub-phonetic units, and stores the identified sub-phonetic units and the modified preference coefficient.

24. The electronic device of claim 19 , wherein, in order to train instances of the sub-phonetic units, the processor selects learned words from a target language corpus, forms learned sub-phonetic units based on frequently used phoneme sequences in the target language corpus, analyzes learned sub-phonetic units to derive preferences, and stores learned sub-phonetic units and corresponding preference in the database.

25. The electronic device of claim 19 , wherein the processor truncates the candidate target language words based on preferences of the sub-phonetic units identified in candidate words,

wherein the preferences of the sub-phonetic units are identified based on preference coefficient and a threshold.

26. The electronic device of claim 19 , wherein the processor generates target language words according to a reverse transliteration of truncated candidate words into the source language,

wherein the truncated candidate words are reverse transliterated by using a correspondence between a source language sub-phonetic and a target language sub-phonetic units based on mapping of the each source language sub-phonetic unit with the each target language sub-phonetic unit, and reflective coefficients.

27. The electronic device of claim 19 , wherein, in order to identify candidate target language words by performing the reverse transliteration, the processor divides the target language words to form transliterated sub-phonetic units, retrieves ranking for each transliterated sub-phonetic unit from the database, forms words in the source language from the each translated sub-phonetic units, and identifies the candidate target language words according to a degree of matching between the formed source language words and the target language words.

28. The electronic device of claim 19 , wherein the processor validates the candidate target language words based on a target language corpus.