System and method for interactive automatic translation

• US 9,613,026 B2
• Filed: 12/28/2012
• Issued: 04/04/2017
• Est. Priority Date: 12/28/2011
• Status: Active Grant

• Current Assignee: Bloomberg Finance LP (Bloomberg LP)
• Original Assignee: Bloomberg Finance LP (Bloomberg LP)
• Inventors: Hodson, James Alexander
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): Sharma, Neeraj
• Application Number:
  US13/729,838
• Publication Number:
  US 20130173247A1
• Time in Patent Office: 1,558 Days
• Field of Search: 704 2, 704 3, 704 4, 704 7, 704 9, 704275, 704254, 704255, 704 10, 715259, 715816, 707760, 709229
• 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

Abstract

Embodiments of the invention relate to computerized systems and methods for automatically translating text between natural languages. Text for translation may be broken into segments, which may themselves be templates including placeholders. If an exact match for a segment is found in a translation database, the segment may be matched with its translation from the database. Otherwise, alternate source language text, corresponding to the segment, for which an exact translation is available, may be identified and presented to the user for selection. Text selected by the user may replace the original segment and be matched with its translation.

Translation of a segment may be influenced by other translated segments. Additionally, translations may be modified as the user interactively enters and edits text.

Claims

1. A method of automatically translating text from one natural language to another using a computer system that comprises one or more processors, one or more interfaces operatively coupled to at least one of the processors, and one or more computer-readable storage media operatively coupled to at least one of the processors, the method comprising:

receiving text in a first language through at least one of the interfaces;

at least one of the processors executing instructions retrieved from at least one of the computer-readable storage media to identify within the text a plurality of segments, each segment comprising a plurality of words and a complete or partial phrase or a complete or partial sentence, with particular ones of the segments having a dependency upon particular other ones of the segments and providing a context for the translating of the text;

at least one of the processors executing instructions retrieved from at least one of the computer readable storage media to generate an intermediate representation of the text using one or more segments of the plurality of segments, and to generate a plurality of first candidate source segments based on the intermediate representation, each of the first candidate source segments differing from the one or more segments and having in a database an exact translation into a second language that is not the first language, at least one first candidate source segment of the plurality of first candidate source segments generated by identifying, using the intermediate representation, one or more invariant symbols over the translating of the text;

in response to receiving input through at least one of the interfaces that selects one of the first candidate source segments, designating the selected first candidate source segment as a first source segment; and

at least one of the processors executing instructions retrieved from at least one of the computer-readable storage media to identify in the database a first translation of the first source segment into the second language at least in part by mapping, using a parameterized template, an exact translation of words and phrases in the first source segment into the second language; and

wherein the parameterized template includes a templated phrase having at least two parameters, and the mapping identifies words in the first language from the first source segment in place of the at least two parameters and inserts translated counterparts in the second language therefore in the first translation of the first source segment.

2. The method of claim 1, comprising at least one of the processors executing instructions retrieved from at least one of the computer-readable storage media to identify in the database a second translation of the first source segment into a third language that is not the first language or the second language.

3. The method of claim 1, wherein the parameterized template includes a string of constants and constrained variables and each segment of the plurality of segments is examined and tagged for attributes associated with one of the constants or one of the constrained variables, and wherein the parameterized template is identified from a plurality of parameterized templates in the database by matching the tagged attributes of the segment with the plurality of parameterized templates.

10. The method of claim 1 wherein the intermediate representation includes a representation of the text as a parse tree.

11. The method of claim 10 wherein the parse tree indicates one or more relationship among the text of the intermediate representation.

12. The method of claim 10 further comprising:

transliterating one or more of portions of the intermediate representation.

4. A computer system for automatically translating text from one natural language to another, the computer system comprising:

one or more processors;

one or more interfaces operatively coupled to at least one of the processors; and

one or more computer-readable storage media, operatively coupled to at least one of the processors and encoded with instructions that, when executed by at least one of the processors, cause the computer system at least to;

receive text in a first language through at least one of the interfaces;

identify within the text a plurality of segments, each segment comprising a plurality of words and a complete or partial phrase or a complete or partial sentence, with particular ones of the segments having a dependency upon particular other ones of the segments and providing a context for the translating of the text;

generate an intermediate representation of the text using one or more segments of the plurality of segments;

generate a plurality of first candidate source segments based on the intermediate representation, at least one of the first candidate source segments being generated by at least identifying, using the intermediate representation, one or more invariant symbols over the translating of the text, each of the first candidate source segments differing from the segments and having in a database an exact translation into a second language that is not the first language;

in response to receiving input through at least one of the interfaces that selects one of the first candidate source segments, designate the selected first candidate source segment as a first source segment; and

identify in the database a first translation of the first source segment into the second language at least in part by mapping, using a parameterized template, an exact translation of words and phrases in the first source segment into the second language;

and wherein the parameterized template includes a templated phrase having at least two parameters, and the mapping identifies words in the first language from the first source segment in place of the at least two parameters and inserts translated counterparts in the second language therefore in the first translation of the first source segment.

5. The computer system of claim 4, wherein the instructions comprise instructions that, when executed by at least one of the processors, cause the computer system at least to identify in the database a second translation of the first source segment into a third language that is not the first language or the second language.

6. The computer system of claim 5, wherein the parameterized template includes a string of constants and constrained variables and each segment of the plurality of segments is examined and tagged for attributes associated with one of the constants or one of the constrained variables, and wherein the parameterized template is identified from a plurality of parameterized templates in the database by matching the tagged attributes of the segment with the plurality of parameterized templates.

13. The computer system of claim 4 wherein the intermediate representation includes a representation of the text as a parse tree.

14. The computer system of claim 13 wherein the parse tree indicates one or more relationship among the text of the intermediate representation.

15. The computer system of claim 13 further comprising:

transliterating one or more of portions of the intermediate representation.

7. A non-transitory computer-readable storage medium encoded with instructions that, when executed by one or more processors within a computer system that comprises one or more interfaces operatively coupled to at least one of the processors, cause the computer system at least to:

receive text in a first language through at least one of the interfaces;

identify within the text a plurality of segments, each segment comprising a plurality of words and a complete or partial phrase or a complete or partial sentence, with particular ones of the segments having a dependency upon particular other ones of the segments and providing a context for translating the text;

generate an intermediate representation of the text using one or more segments of the plurality of segments;

generate a plurality of first candidate source segments using the intermediate representation, at least one of the first candidate source segments being generated by at least identifying, using the intermediate representation, one or more invariant symbols over the translating of the text, each of the first candidate source segments differing from the one or more segments and having in a database an exact translation into a second language that is not the first language;

in response to receiving input through at least one of the interfaces that selects one of the first candidate source segments, designate the selected first candidate source segment as a first source segment; and

identify in the database a first translation of the first source segment into the second language at least in part by mapping, using a parameterized template, an exact translation of words and phrases in the first source segment into the second language; and

wherein the parameterized template includes a templated phrase having at least two parameters, and the mapping identifies words in the first language from the first source segment in place of the at least two parameters and inserts translated counterparts in the second language therefore in the first translation of the first source segment.

8. The non-transitory computer-readable storage medium of claim 7, wherein the instructions comprise instructions that, when executed by at least one of the processors, cause the computer system at least to identify in the database a second translation of the first source segment into a third language that is not the first language or the second language.

9. The non-transitory computer-readable storage medium of claim 8, wherein parameterized template includes a string of constants and constrained variables and each segment of the plurality of segments is examined and tagged for attributes associated with one of the constants or one of the constrained variables, and wherein the parameterized template is identified from a plurality of parameterized templates in the database by matching the tagged attributes of the segment with the plurality of parameterized templates.

16. The non-transitory computer-readable storage medium of claim 7 wherein the intermediate representation includes a representation of the text as a parse tree.

17. The non-transitory computer-readable storage medium of claim 16 wherein the parse tree indicates one or more relationship among the text of the intermediate representation.