Document translation system

• US 7,877,251 B2
• Filed: 05/07/2007
• Issued: 01/25/2011
• Est. Priority Date: 05/07/2007
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Kumaran, Arumugam, Vanderwende, Lucretia H., Joy, Joseph, Blagsvedt, Sean Olin
• Primary Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US11/745,291
• Publication Number:
  US 20080281578A1
• Time in Patent Office: 1,359 Days
• Field of Search: None
• US Class Current: 704/2
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30

Abstract

A document translation system is described. In various embodiments, the document translation system improves translations of text or speech (“documents”) that are performed by computing devices (“machine translations”). Upon translating a document into a target language, the system can provide a user interface containing both the original document and a translated document. The translated document can include portions that are translated automatically by a machine translation service, and other portions that have been amended or modified by users. Upon providing the translated document, the system may receive input from a user describing portions of the translated document that should be modified. These modifications may then be stored in corpora that the document translation system can employ during future requests to translate the document. When multiple modifications are stored, the system may select one of the stored modifications during translation, such as based on the document type, reputation of the user who provided the modification, and so forth. The stored corpora can then be employed to improve future machine translations.

Claims

1. A method performed by a computer system having a processor and memory to improve document translations, comprising:

receiving a request for a translated document in a preferred natural language;

retrieving an identified document, the identified document storing text in a first natural language;

translating using the processor the retrieved document to create the translated document containing text translated into the preferred natural language wherein the preferred natural language is a natural language that is not the first natural language;

providing the translated document;

receiving a modification to a word in the translated document;

storing the received modification to the translated word to improve a subsequent request for the translated document;

receiving a subsequent modification to the translated word in the translated document; and

storing in the storage the received subsequent modification to the translated word;

upon receiving the subsequent request for the translated document,

retrieving the identified document;

translating the retrieved document into the preferred natural language;

selecting either the modification or the subsequent modification based at least on an identification of users who made the modifications and a relative trustworthiness of the users, wherein the trustworthiness of a user is based on how many times other users modify a modification made b the user;

retrieving the selected modification;

modifying the translated document based on the retrieved modification; and

providing the modified translated document.

2. The method of claim 1 further comprising:

upon receiving the subsequent request for the translated document, retrieving the identified document;

translating the retrieved document into the preferred natural language or retrieving a stored version of the translated document;

retrieving the received modification to the translated word;

modifying the translated document based on the retrieved modification; and

providing the modified translated document.

3. The method of claim 2 further comprising identifying the modification as a user-provided modification.

4. The method of claim 1 further comprising providing the identified document in addition to the translated document wherein a sequence of words in the provided identified document is correlated with a sequence of words in the provided translated document.

5. The method of claim 4 wherein the correlation is a link that causes the sequence of words in the provided identified document and the provided translated document both to be highlighted when the sequence of words in the provided document is selected.

6. The method of claim 4 wherein the correlation is a link that causes the sequence of words in the provided identified document and the provided translated document both to be highlighted when the sequence of words in the provided translated document is selected.

7. The method of claim 1 further comprising:

receiving a subsequent modification to the translated word in the translated document; and

storing in the storage the received subsequent modification to the translated word;

upon receiving the subsequent request for the translated document, retrieving the identified document;

translating the retrieved document into the preferred natural language;

selecting either the modification or the subsequent modification;

retrieving the selected modification;

modifying the translated document based on the retrieved modification; and

providing the modified translated document.

8. The method of claim 1 wherein the trustworthiness of users changes over time.

9. The method of claim 1 further comprising providing a stored modification to a machine translation learning system to improve future machine translations.

10. A computer-readable medium storing computer-executable instructions that, when executed, cause a computer system to perform a method for improving document translations, the method comprising:

providing to a server an identification of a document and a preferred natural language; and

receiving from the server the document in the identified preferred natural language wherein the document in the identified preferred natural language contains a user-supplied modification to a translated word wherein the user-supplied modification was supplied to correct a prior translation, the user-supplied modification having been selected from multiple user-supplied modifications based at least on comparing a number of times the user-supplied modifications have been modified by users other than a user who initially provided the modification.

11. The computer-readable medium of claim 10 wherein the method further comprises:

receiving from the server the document in an original natural language;

displaying in a first region the document in the original natural language;

displaying in a second region the document in the identified preferred natural language;

receiving a selection of a sequence of words in one of the regions; and

identifying a corresponding set of words in the other of the regions.

12. The computer-readable medium of claim 11 wherein the method further comprises identifying in the second region the modified translated word.

13. The computer-readable medium of claim 12 wherein the method further comprises identifying in the first region a word corresponding to the modified translated word.

14. The computer-readable medium of claim 10 wherein the method further comprises:

receiving from a user a modification to a word in the document received in the identified preferred natural language; and

providing to the server the modification received from the user so that the modification can be provided by the server when it receives a subsequent request to translate the document.

15. The computer-readable medium of claim 10 wherein the method further comprises:

receiving from a user a correction to the modification; and

providing to the server the correction received from the user so that the correction can be provided by the server when it receives a subsequent request to translate the document.

16. A system for improving document translations, comprising:

a component configured to receive an identification of an original document that contains text in a first natural language, retrieve the original document, translate the text from the first natural language into a second natural language using a machine translation service to create a translated document, add editing code to the created translated document to enable a user to supply a translation modification, and provide the translated document; and

a component configured to (a) select a user-supplied modification from multiple user-supplied modifications based at least on comparing a number of times the user-supplied modifications have been modified by users other than a user who initially provided the modification; and

(b) modify the translated document based on the selected user-supplied modification.

17. The system of claim 16 wherein the translation server component additionally supplies editing code and the translation communicator component employs the editing code to receive a translation modification from a user and supplies the translation modification to the translation server component.

18. The system of claim 17 wherein the translation server component stores in a translation edits database the translation modification it receives from the translation communicator component and provides the translation modification when it subsequently receives a request to translate the original document into the second natural language.