Dynamic generation of auto-suggest dictionary for natural language translation

• US 9,262,403 B2
• Filed: 01/14/2011
• Issued: 02/16/2016
• Est. Priority Date: 03/02/2009
• Status: Active Grant

• Current Assignee: SDL Ltd.
• Original Assignee: SDL PLC (RWS Holdings Plc)
• Inventors: Christ, Oliver
• Primary Examiner(s): Neway, Samuel G
• Application Number:
  US13/007,445
• Publication Number:
  US 20110184719A1
• Time in Patent Office: 1,859 Days
• Field of Search: 704 1- 10
• US Class Current: 1/1
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 40/45 Example-based machine translation; Alignment
• CPC Class Codes: G06F 40/47 Machine-assisted translation, e.g. using translation memory
• CPC Class Codes: G06F 40/49 using very large corpora, e.g. the web
• CPC Class Codes: G06F 40/51 Translation evaluation
• 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

Abstract

The present technology dynamically generates auto-suggest dictionary data from translation data stored in memory at a server. The auto-suggest dictionary data may be transmitted to a remote device by the server for use in language translation. The auto-suggest dictionary data may be transferred as part of a package which includes content to be translated, translation meta-data, and various other data. The auto-suggest dictionary data may be generated at a first computing device, periodically or in response to an event, from translation data stored in memory. The auto-suggest dictionary may be transferred to a remote device along with content to be translated and other data, as part of a package, for use in translation of the content at the remote device.

Claims

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

extracting auto-suggest dictionary data including a plurality of sentence sub-segment pairs, each pair comprising a source sentence sub-segment extracted from a source sentence in a source language and a target sentence sub-segment extracted from a translation of the source sentence in a target language, the source and target sentences stored in translation data;

generating a package including translation content and the extracted auto-suggest dictionary data;

transmitting the package from a server to a remote device configured to;

display a plurality of target sentence sub-segments as predictive translations based on correspondence between data input to the remote device by a human translator and at least a portion of the target sentence sub-segments,

highlight a suggested best predictive translation in the plurality of predictive translations,

receive a selection of one of the plurality of predictive translations from the human translator, and

provide the received selection to the server; and

updating the extracted auto-suggest dictionary data based on the received selection.

2. The non-transitory computer readable storage medium of claim 1, the method further comprising storing parameters for the translation in the form of meta-data in the package.

3. The non-transitory computer readable storage medium of claim 1, wherein the auto-suggest dictionary data is configured to be accessed by a translation application on the remote device.

4. The non-transitory computer readable storage medium of claim 1, the method further comprising providing the updated auto-suggest dictionary data to a second remote device.

5. The non-transitory computer readable storage medium of claim 1, the method further comprising updating an auto-suggest dictionary stored in the package based at least in part on detection of an auto-suggest dictionary update event.

6. The non-transitory computer readable storage medium of claim 1, the method further comprising updating the auto-suggest dictionary data stored in the package based on a change in size of the auto-suggest dictionary data.

15. The non-transitory computer readable storage medium of claim 1, wherein the predictive translations provided from the auto-suggest dictionary data are further based on correspondence between the content being translated by the human translator and the sentence segment in the source language.

16. The non-transitory computer readable storage medium of claim 1, wherein a predictive translation of the plurality of predictive translations is highlighted based on an amount of correspondence between the data input by the human translator and the plurality of predictive translations, and the plurality of predictive translations are ranked based on ranking factors, the ranking factors comprising:

a likelihood of selection by the human translator; and

a length of one of the predictive translations.

21. The non-transitory computer readable storage medium of claim 1, wherein the stored translation data comprises translated text in a plurality of target languages.

22. The non-transitory computer readable storage medium of claim 1, further comprising determining a highest ranked predictive translation, wherein the highest ranked predictive translation is visually emphasized.

7. A method comprising:

generating auto-suggest dictionary data including a plurality of sub-segment pairs, each sub-segment pair including a source sub-segment in a source language and a target sub-segment corresponding to a translation of the source sub-segment into a target language, each sub-segment pair extracted from a stored translation unit including a source segment and a corresponding target segment corresponding to a translation of the source segment;

combining translation units, content to be translated, and the plurality of sub-segment pairs in a translation package;

transmitting the translation package from a server to a remote device configured to;

access the translation package,

present content in the package to a human translator,

receive data input in the target language from the human translator,

display a plurality of predictive translations to the human translator in response to the received data input, each predictive translation being a translation of a source sub-segment from the source language to a corresponding target sub-segment of the target language in the package, each predictive translation based on correspondence between the data input from the human translator and at least a portion of the target sub-segment in the package,

rank the predictive translations based on an amount of displayed elements,

indicate a best predictive translation based on a number of text characters in the predictive translation in common with the data input, and

receive a selection of the one of the ranked plurality of predictive translations from the human translator;

receiving new translation units from the remote device, the new translation units based on selections of the predictive translations by the human translator;

updating the auto-suggest dictionary data based on the received new translation units; and

storing the updated auto-suggest dictionary in the package.

8. The method of claim 7, wherein the package further includes metadata for the translation project, placeable identification, and conversion data.

9. The method of claim 7, wherein updating the auto-suggest dictionary is triggered by a change in size of the translation memory greater than a threshold over an interval of time.

10. The method of claim 7, the method further comprising:

updating the auto-suggest dictionary data; and

providing the updated auto-suggest dictionary data to a second remote device.

11. The method of claim 7, the method further comprising updating an auto-suggest dictionary based at least in part on detection of an auto-suggest dictionary update event.

12. The method of claim 7, the method further comprising updating the auto-suggest dictionary data based on a change in size of the auto-suggest dictionary data.

17. The method of claim 7, wherein the predictive translations provided from the auto-suggest dictionary data are further based on correspondence between the content being translated by the human translator and the sentence segment in the source language.

18. The method of claim 7, wherein the plurality of predictive translations are ranked based on ranking factors and a predictive translation is highlighted based on an amount of correspondence between the data input by the human translator and the plurality of predictive translations, the ranking factors comprising:

a likelihood of selection by the human translator; and

a length of one of the predictive translations.

13. A system for managing translation of content, the system comprising:

a memory;

a dictionary generation module stored in the memory and executable by a processor to extract auto-suggest dictionary data comprising source sentence sub-segments and corresponding target sentence sub-segments from stored translation data including source sentence segments and corresponding translated target sentence segments;

a translation module to display a plurality of predictive translations to a human translator, each predictive translation received as a target sentence sub-segment from the auto-suggest dictionary data based on correspondence between data input by the human translator and at least a portion of the received target sentence sub-segment;

a ranking module to rank the plurality of displayed predictive translations;

an input/output module to receive a selection from the human translator of a predictive translation from the plurality of ranked predictive translations; and

a package management module stored in the memory and executable by the processor to;

generate a package including content to be translated, the extracted auto-suggest dictionary data which corresponds to the source language and target language for the translation job to be performed, parameters for translation in the form of metadata, and placeable identification and conversion data,

provide the generated package to the translation module, and

update the auto-suggest dictionary data in the package based on the received selection.

14. The system of claim 13, the dictionary generation module configured to generate second auto-suggest dictionary data based on updated stored translation data.

19. The system of claim 13, wherein the predictive translations provided from the auto-suggest dictionary data are further based on correspondence between the content being translated by the human translator and the source sentence segments.

20. The system of claim 13, wherein the plurality of predictive translations are ranked based on ranking factors, the ranking factors comprising:

a likelihood of selection by the human translator;

an amount of words in the source sub-segments to which words in the respective predictive translations correspond; and

a length of one of the predictive translations.