Method and apparatus for interactive source language expression recognition and alternative hypothesis presentation and selection
First Claim
1. In a language translation system, a method for providing a guide for a user, comprising:
- receiving an input that is representative of at least one word in a source language;
generating at least one recognition hypothesis in the source language in response to the input;
determining a best hypothesis from the at least one recognition hypothesis in the source language;
presenting the best hypothesis in the source language to a user;
presenting alternatives to a portion of the best hypothesis in the source language to the user;
receiving an indication of a choice of one of the alternatives from the user; and
presenting a revised version of the best hypothesis including the alternative chosen to the user.
1 Assignment
0 Petitions
Accused Products
Abstract
In a language translation system, a method for providing a guide for a user. In one embodiment, the method comprises the steps of receiving an input that is representative of at least one word in a source language, and generating at least one recognition hypothesis in the source language in response to the input. The method further comprise the steps of selecting a best hypothesis from the at least one recognition hypothesis in the source language, presenting the best hypothesis in the source language to a user, and presenting alternatives to a portion of the best hypothesis in the source language to the user. The method further includes receiving an indication of a choice of one of the alternatives from the user, and presenting a revised version of the best hypothesis including the alternative chosen to the user.
417 Citations
39 Claims
-
1. In a language translation system, a method for providing a guide for a user, comprising:
-
receiving an input that is representative of at least one word in a source language;
generating at least one recognition hypothesis in the source language in response to the input;
determining a best hypothesis from the at least one recognition hypothesis in the source language;
presenting the best hypothesis in the source language to a user;
presenting alternatives to a portion of the best hypothesis in the source language to the user;
receiving an indication of a choice of one of the alternatives from the user; and
presenting a revised version of the best hypothesis including the alternative chosen to the user. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12)
receiving an acceptance of the revised version of the best hypothesis in the source language; and
presenting an output that is representative of the revised version of the best hypothesis in a target language.
-
-
3. The method of claim 1, wherein the best hypothesis in the source language is a group of words in the source language, and wherein the portion of the best hypothesis in the source language is a subgroup of the group of words.
-
4. The method of claim 3, wherein, when multiple subgroups exist, a largest subgroup is used as the portion of the best hypothesis in the source language.
-
5. The method of claim 4, further comprising receiving a correction of the portion of the best hypothesis in the source language.
-
6. The method of claim 5, further comprising:
-
storing the correction; and
using the correction to generate a best hypothesis in the source language in response to a subsequent input that is similar to the input.
-
-
7. The method of claim 1, wherein the input received is spoken words.
-
8. The method of claim 1, wherein the input received is an output of a keyboard device.
-
9. The method of claim 1, wherein the input received is an output of a handwriting recognition process.
-
10. The method of claim 1, wherein the input received is an output of an optical character recognition (OCR) process.
-
11. The method of claim 1, wherein the input received is an output of one of a group comprising a handwriting recognition process and an optical character recognition (OCR) process.
-
12. The method of claim 1, wherein presenting comprises displaying words on a display device.
-
13. In a language translation system, a method for providing a guide for a user, comprising:
-
receiving an input that is representative of at least one word in a source language;
generating at least one recognition hypothesis in the source language in response to the input;
generating at least one recognition hypothesis in a target language in response to the input;
presenting the at least one recognition hypothesis in the source language and the at least one recognition hypothesis in a target language as a list of pairs ordered according to likelihood of being correct;
accepting a choice of one pair of the list of pairs from a user, wherein a chosen pair comprises a best hypothesis in the source language and a best hypothesis in the target language;
indicating a portion of the best hypothesis in the source language to the user, wherein the portion has at least one alternative expression;
presenting the at least one alternative expression to the user;
accepting a choice of one of the at least one alternative expressions; and
presenting the best hypothesis, including the chosen alternative, to the user in the source language; and
presenting the best hypothesis, including the chosen alternative, to the user in the target language. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20)
accepting a correction to the portion of the best hypothesis in the source language;
presenting the best hypothesis including the correction in the source language to the user; and
presenting the best hypothesis including the correction in the target language to the user.
-
-
16. The method of claim 15, wherein presenting comprises displaying on a display device.
-
17. The method of claim 15, wherein the input comprises spoken words.
-
18. The method of claim 15, wherein the input received is an output of a keyboard device.
-
19. The method of claim 15, wherein the input received is an output of a handwriting recognition process.
-
20. The method of claim 15, wherein the input received is an output of an optical character recognition (OCR) process.
-
21. In a language translation system, an interactive method for translating a group of words, comprising:
-
receiving an input that represents the group of words in a source language;
displaying at least one candidate hypothesis to the user, wherein the at least one candidate hypothesis comprises;
a source language representation of the group of words in the source language that is based upon a translation module'"'"'s understanding of the input;
a target language representation of the group of words in a target language that is based upon the translation module'"'"'s interpretation of the source language representation; and
a source language representation of a back-translation of the target language representation of the group of words. - View Dependent Claims (22, 23, 24, 25, 26)
indicating a subgroup of words in a source language representation portion of the best hypothesis, wherein there are alternative source language representations for the subgroup; and
displaying at least one alternative source language representation of the subgroup.
-
-
24. The method of claim 23, further comprising:
-
accepting an indication of a chosen alternative of the at least one alternatives from the user; and
displaying a revised best hypothesis, comprising;
a source language representation of the group of words, including the chosen alternative source language representation of the subgroup;
a target language representation of the group of words, including the translation module'"'"'s interpretation of the chosen alternative source language representation of the subgroup; and
a source language representation of a back-translation of the target language representation of the group of words.
-
-
25. The method of claim 24, further comprising receiving an indication from the user that the revised best hypothesis is acceptable.
-
26. The method of claim 23, further comprising accepting an edited version of the subgroup of words in the source language representation portion of the best hypothesis from the user.
-
27. A computer readable medium containing executable instructions which, when executed in a processing system, cause the system to perform a method for performing language translation, the method comprising:
-
receiving an input that is representative of at least one word in a source language;
generating at least one recognition hypothesis in the source language in response to the input;
selecting a best hypothesis from the at least one recognition hypothesis in the source language;
presenting the best hypothesis in the source language to a user;
presenting alternatives to a portion of the best hypothesis in the source language to the user;
receiving an indication of a choice of one of the alternatives from the user; and
presenting a revised version of the best hypothesis including the alternative chosen to the user. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39)
receiving an acceptance of the revised version of the best hypothesis in the source language; and
presenting an output that is representative of the best hypothesis in a target language.
-
-
29. The method of claim 27, wherein the best hypothesis in the source language is a group of words in the source language, and wherein the portion of the best hypothesis in the source language is a subgroup of the group of words.
-
30. The method of claim 29, wherein, when multiple subgroups exist, a largest subgroup is used as the portion of the best hypothesis in the source language.
-
31. The method of claim 30, further comprising receiving a correction of the portion of the best hypothesis in the source language.
-
32. The method of claim 31, further comprising:
-
storing the correction; and
using the correction to generate a best hypothesis in the source language in response to a subsequent input that is similar to the input.
-
-
33. The method of claim 27, wherein the input received is spoken words.
-
34. The method of claim 27, wherein the input received is an output of a keyboard device.
-
35. The method of claim 27, wherein the input received is an output of a handwriting recognition process.
-
36. The method of claim 27, wherein the input received is an output of an optical character recognition (OCR) process.
-
37. The method of claim 27, wherein presenting comprises displaying words on a display device.
-
38. The method of claim 27, wherein the display device is a touch-sensitive screen, and wherein receiving an indication of a choice of one of the alternatives from the user is a consequence of the user touching the touch screen.
-
39. The method of claim 37, wherein presenting alternatives to a portion of the best hypothesis in the source language to the user includes displaying a number associated with each alternative, and wherein receiving an indication of a choice of one of the alternatives from the user is a consequence of the user pressing a number on a keyboard device coupled to the display device.
Specification