IP Library Granted Patent US 8,990,064
Granted Patent B2
US 8,990,064 · App. 12/510,913 · Granted Mar 24, 2015

Translating documents based on content

Inventors: Daniel Marcu (Hermosa Beach, CA); Radu Soricut (Manhattan Beach, CA); Narayanaswamy Viswanathan (Palo Alto, CA)
Assignee: Language Weaver, Inc.
G06F17/289
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,990,064
App. No.
12/510,913
Granted
Mar 24, 2015
Kind
B2
Abstract

A document containing text in a source language may be translated into a target language based on content associated with that document, in conjunction with the present technology. An indication to perform an optimal translation of a document into a target language may be received via a user interface. The document may then be accessed by a computing device. The optimal translation is executed by a preferred translation engine of a plurality of available translation engines. The preferred translation engine is the most likely to produce the most accurate translation of the document among the plurality of available translation engines. Additionally, the preferred translation engine may be identified based on content associated with the document. The document is translated into the target language using the preferred translation engine to obtain a translated document, which may then be outputted by a computing device.

Claims (50)

1. A method using a computing system for translating documents based on content, the method comprising:

receiving an request via a user interface of the computing system to perform an optimal translation of a document into a target language, the document comprising text in a source language;

identifying keywords included in a plurality of training datasets and in the document using a keyword module of the computing system;

selecting a preferred translation engine associated with a training data set having identified keywords that are related to identified keywords included in the document; and

directing the preferred translation engine to generate a translated document comprising text in the target language from the document.

2. The method of claim 1 , further comprising:

tagging the identified keywords; and

categorizing the document and training datasets based on the tagging.

3. The method of claim 1 , wherein each of a plurality of available translation engines is associated with a training dataset having a different subject matter.

4. The method of claim 3 , wherein the subject matter associated with the preferred translation engine is related to content associated with the document.

5. The method of claim 1 , further comprising accessing the document based on location information received through the user interface.

6. The method of claim 1 , further comprising providing a graphical user interface that enables a user to select the optimal translation or an alternate translation, the alternate translation associated with a user-selected translation engine among a plurality of available translation engines.

7. A method for translating documents based on content, using a computing device that comprises a processor and memory for storing executable instructions, the processor executing the instructions to perform the method, the method comprising:

accessing a document comprising text in a source language;

predicting a translation quality associated with each of a plurality of translation engines using a predictor module;

measuring a degree of alignment between the content associated with the document and content included in each of a plurality of training datasets, each of the plurality of training datasets associated with a different available translation engine;

selecting a preferred translation engine based on the predicted translation quality;

selecting the translation engine associated with the training dataset having the closest degree of alignment as the preferred translation engine;

directing the preferred translation engine to translate the document into a target language to obtain a translated document; and

outputting the translated document.

8. The method of claim 7 , wherein the preferred translation engine is most likely to produce the most accurate translation of the document relative to the rest of a plurality of available translation engines.

9. The method of claim 8 , wherein each of the plurality of available translation engines is associated with different subject matter.

10. The method of claim 7 , further comprising determining the preferred translation engine from the plurality of available translation engines.

11. The method of claim 10 , wherein the determining comprises:

evaluating, using a translator evaluation module, previous translations performed by each of the plurality of translation engines;

predicting a translation quality associated with each of the plurality of available translation engines based on the previous translations; and

selecting the translation engine with the highest translation quality prediction as the preferred translation engine.

12. A system for translating documents based on an alignment of content, the system comprising:

a computing device to receive an indication via a user interface to perform an optimal translation of a document into a target language, the document comprising text in a source language, the optimal translation to be executed by a preferred translation engine;

a plurality of available translation engines each including a training dataset for a different subject matter;

an alignment module to measure, using cross correlation, a degree of alignment between content associated with the document and content included in each of the training datasets;

a recommendation engine stored in memory and executable by a processor to identify a preferred translation engine based on the degree of alignment of the training dataset included in the selected translation engine; and

a computing device to output a translated document obtained via the optimal translation executed using the preferred translation engine, the translated document comprising text in the target language.

13. The system of claim 12 , wherein the recommendation engine is further configured to identify the preferred translation engine based on content associated with the document.

14. The system of claim 12 , wherein each of the plurality of available translation engines is associated with different subject matter.

15. The system of claim 12 , wherein the subject matter associated with the preferred translation engine is related to content associated with the document.

16. The system of claim 12 , further comprising a communications module stored in memory and executable by a processor to access the document based on location information associated with the document and received through the user interface.

17. The system of claim 12 , further comprising an interface module stored in memory and executable by a processor to provide a graphical user interface that enables a user to select the optimal translation or an alternate translation, the alternate translation associated with a user-selected translation engine of the plurality of available translation engines.

18. A non-transitory computer-readable storage medium having a program embodied thereon, the program being executable by a processor to perform a method for translating documents based on content, the method comprising:

receiving an indication via a user interface to perform an optimal translation of a document into a target language, the document comprising text in a source language;

measuring a degree of alignment between the text associated with the document and content included in each of a plurality of training datasets, each of the plurality of training datasets associated with a different available translation engine;

selecting the translation engine associated with the training dataset having the closest degree of alignment as a preferred translation engine, the optimal translation to be executed by the preferred translation engine;

requesting the preferred translation engine to generate a translated document comprising text in the target language from the document; and

requesting to output of the translated document.

19. A non-transitory computer-readable storage medium having a program embodied thereon, the program being executable by a processor to perform a method for translating documents based on content, the method comprising:

accessing a document comprising text in a source language;

identifying keywords included in a plurality of training datasets and in the document;

selecting a preferred translation engine associated with a training data set having identified keywords that are related to identified keywords included in the document;

translating the document into a target language using the preferred translation engine to obtain a translated document; and

outputting the translated document.

Assignments (3)
MERGER Recorded Feb 16, 2016
From: LANGUAGE WEAVER, INC.
To: SDL INC.
Reel/Frame 037745/0391 →
CORRECTIVE ASSIGNMENT TO CORRECT THE RECEIVING PARTY ADDRESS PREVIOUSLY RECORDED AT REEL: 023023 FRAME: 0522. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT. Recorded Dec 31, 2014
From: LANGUAGE WEAVER, INC.
To: LANGUAGE WEAVER, INC.
Reel/Frame 034741/0731 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 29, 2009
From: MARCU, DANIEL; SORICUT, RADU; VISWANATHAN, NARAYANASWAMY
To: LANGUAGE WEAVER, INC.
Reel/Frame 023023/0522 →
Continuity (1)
Related Publication 20110029300A1 · Feb 3, 2011