Language translation devices and methods

• US 20050038662A1
• Filed: 08/14/2003
• Published: 02/17/2005
• Est. Priority Date: 08/14/2003
• Status: Active Grant

• Current Assignee: Voxtec International, Inc.
• Original Assignee: Voxtec International, Inc.
• Inventors: Hall, John T., Sarich, Ace J.
• Granted Patent: US 7,369,998 B2
• Time in Patent Office: Days
• US Class Current: 704/277
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models

Abstract

The invention includes devices and methods for providing phrase-based translation. Devices of the invention have translation modules containing a set of phrases and their translations wherein the set of phrases is selected based on their suitability in the context in which the translation device is used. The methods of the invention determine the context in which a translation device is used and provide sets of phrases and their translations that are suitable in the context. The invention further includes devices containing software for executing the methods of the invention.

Claims

1. A method for providing a phrase vocabulary in an electronic translation device with a means for inputting a native language phrase and a means for outputting a translation of the phrase in a target language, the method comprising:

(a) evaluating a context in which the translation device will be used,

(b) selecting a set of native language phrases that would be useful in the context,

(c) providing translations of the native language phrases into a target language,

(d) associating each native language phrase with the corresponding target language translation, the set of associated phrases and translations comprising a translation module, and

(e) uploading the translation module onto the translation device;

wherein the translation module is stored in a memory means of the translation device.

2. The method of claim 1, wherein the translation device employs speaker independent voice recognition as the means for inputting phrases from the set of phrases that are translated by the translation device.

3. The method of claim 1, wherein the translation device employs audio output of the associated translation as the means for outputting a translation of the phrase in the target language, and the step of providing translations comprises the step of recording a spoken translation of the phrase.

4. The method of claim 3, further comprising the step of selecting the gender and tone of the spoken translation of the phrase based on the context.

5. The method of claim 2, further comprising adding alias phrases to the set of phrases, wherein each alias phrase has a similar meaning to a similar phrase in the set of phrases and is associated with the translation corresponding to the similar phrase.

6. The method of claim 2, wherein the electronic translation device is a handheld electronic translation device.

7. The method of claim 2, further comprising the step of deconflicting the selected set of phrases by comparing pairs of phrases in the set of phrases, determining whether the audio similarity of pairs of phrases in the set of phrases exceeds a predetermined threshold level of audio similarity, and eliminating one of the phrases in each pair of phrases that exceeds the predetermined threshold level of audio similarity.

8. The method of claim 7, further comprising the step of replacing a phrase that was eliminated from a pair of phrases that exceeds the predetermined threshold level of audio similarity.

9. The method of claim 2, further comprising the step of sorting phrases in the selected set of phrases or in the translation module into categories of phrases, wherein every phrase in a category is related to the same subject.

10. The method of claim 9, further comprising the steps of selecting a category of phrases on the electronic translation device, listing the phrases in the selected category of phrases, and selecting a phrase from the list of phrases in the selected category of phrases.

11. The method of claim 10, wherein the step of selecting a phrase from the list of phrases in the selected category of phrases comprises speaking the selected phrase.

12. The method of claim 2, further comprising the step of associating tag words with each phrase in the selected set of phrases or in the translation module, wherein the tag words associated with each phrase are related to the same subject as the associated phrase.

13. The method of claim 12, further comprising selecting a tag word, listing the phrases associated with the tag word, and selecting a phrase associated with the tag word.

14. The method of claim 13, wherein the step of selecting a tag word comprises the step of speaking the tag word.

15. The method of claim 1, wherein the context is selected from the group consisting of medical care, geriatric care, child care, tourism, and military operations.

16. An electronic translation device comprising:

(a) a means for inputting a native language input phrase,

(b) a means for outputting a translation of the native language input phrase in a target language, and

(c) a data storage device for storing a translation module containing the native language input phrase and the target language translation;

wherein the translation module is generated by a method comprising evaluating a context in which the translation device will be used, selecting a set of phrases including the native language input phrase, the selected set of phrases being useful in the context, providing translations of the phrases into the target language, providing a translation module by associating each native language phrase with the corresponding target language translation, and uploading the translation module into a memory of the translation device.

17. The translation device of claim 16, wherein the translation device employs speaker independent voice recognition as the means for inputting phrases from the set of phrases that are translated by the translation device.

18. The translation device of claim 16, wherein the translation device employs audio output of the translation through a speaker as the means for outputting a translation of the phrase in the target language, and providing translations of the phrases in the set of phrases is done by recording a spoken translation of the phrase.

19. The translation device of claim 18, wherein the gender and tone of the spoken translation of the phrase are selected based on the context.

20. The translation device of claim 17, wherein alias phrases are added to the set of phrases in generating the translation module, wherein each alias phrase has a similar meaning to a similar phrase in the set of phrases and is associated with the translation corresponding to the similar phrase.

21. The translation device of claim 17, wherein the selected set of phrases is deconflicted by comparing pairs of phrases in the set of phrases, determining whether the audio similarity of pairs of phrases in the set of phrases exceeds a predetermined threshold level of audio similarity, and eliminating one of the phrases in each pair of phrases that exceeds the predetermined threshold level of audio similarity.

22. The translation device of claim 21, wherein a phrase that was eliminated from a pair of phrases that exceeds the predetermined threshold level of audio similarity is replaced with a different phrase.

23. The translation device of claim 17, wherein phrases in the selected set of phrases or in the translation module are sorted into categories of phrases and every phrase in a category is related to the same subject.

24. The translation device of claim 23, wherein selecting a category on the translation device causes the translation device to list the phrases in the selected category.

25. The translation device of claim 17, wherein tag words are associated with each phrase in the selected set of phrases or in the translation module and the tag words associated with each phrase are related to the same subject as the associated phrase.

26. The translation device of claim 25, wherein selecting a tag word on the translation device causes the translation device to list the phrases associated with the selected tag word.

27. The translation device of claim 26, wherein phrases in the translation module are accessible by speaking associated tag words, and the translation device provides a list of phrases associated with a spoken tag word.

28. The translation device of claim 17, wherein the context is selected from the group consisting of medical care, geriatric care, child care, tourism, and military operations.

29. An device that stores or executes a software program, wherein the software program builds translation modules by executing a method comprising the steps of:

(a) receiving input comprising a set of native language input phrases,

(b) deconflicting the phrase set by comparing pairs of phrases in the set of phrases, determining whether the audio similarity of pairs of phrases in the set of phrases exceeds a predetermined threshold level of audio similarity, and eliminating one of the prhases in each pair of phrases that exceeds the predetermined threshold level of audio similarity,

(c) receiving input comprising a set of corresponding target language phrase translations, wherein each member of the set of native language input phrases has a corresponding target language phrase,

(d) compiling the phrase set and corresponding target language phrase translations into a translation module accessible by the translation software on a translation device, and

(e) uploading the translation module into the memory of a translation device.

30. The device of claim 29, wherein the corresponding target language phrase translations are recorded spoken translations of the native language input phrases in the target language.

31. The device of claim 30, wherein the gender and tone of the spoken translations of the phrases are selected based on the context.

32. The device of claim 30, wherein the software program further executes the step of receiving input comprising alias phrases, wherein each alias phrase has a similar meaning to a similar phrase in the set of native language input phrases and is associated with the translation corresponding to the similar phrase.

33. The device of claim 29, wherein the software program further executes the step of receiving input comprising a native language phrase with the same meaning as a native language phrase that was eliminated from a pair of phrases that exceeds the predetermined threshold level of audio similarity.

34. The device of claim 29, wherein the software program further executes the step of sorting phrases in the set of native language input phrases or in the translation module into categories of phrases, wherein every phrase in a category is related to the same subject.

35. The device of claim 29, wherein the software program further executes the step of associating tag words with each phrase in the set of native language input phrases or in the translation module, wherein the tag words associated with each phrase are related to the same subject as the associated phrase.

36. The device of claim 35, wherein phrases in the translation module are accessible by selecting associated tag words, and the translation device provides a list of phrases associated with a selected tag word.

37. The device of claim 36, wherein the step of selecting associated tag words comprises speaking the associated tag words.