System and method for understanding text using a translation of the text

• US 10,191,899 B2
• Filed: 01/26/2017
• Issued: 01/29/2019
• Est. Priority Date: 06/06/2016
• Status: Active Grant

• Current Assignee: Comigo Ltd.
• Original Assignee: Comigo Ltd.
• Inventors: Lasser, Menahem
• Primary Examiner(s): Spooner, Lamont M
• Application Number:
  US15/415,952
• Publication Number:
  US 20170351661A1
• Time in Patent Office: 733 Days
• Field of Search: 704 2- 8
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/47 Machine-assisted translation, e.g. using translation memory
• CPC Class Codes: G06F 40/51 Translation evaluation
• CPC Class Codes: G06F 40/53 Processing of non-Latin text kana-to-kanji conversion G06F40/129; vowelisation G06F40/232

Abstract

Devices and methods for determining the content of a first segment of text in a first language, using a second segment of text in a second language. The second segment of text is a translation of the first segment of text.

Claims

1. A method for automatically determining content of a text associated with video media content and for generating an output based on the content, the method comprising:

a) obtaining, by a processor associated with a media playing device or with a set top box feeding said media playing device, a first segment of text in a first language, said first segment of text being part of subtitles in said first language of video media content played by said media playing device, said subtitles in said first language being generated prior to initiating playing of said video media content by said media playing device;

b) obtaining, by said processor, a second segment of text in a second language, said second segment of text being part of subtitles in said second language of said video media content played by said media playing device, said second segment of text being a translation of said first segment of text, said subtitles in said second language being generated prior to initiating playing of said video media content by said media playing device;

c) using said processor, automatically determining content of said first segment of text, in accordance with a predefined analysis rule, said automatically determining of said content of said first segment of text being done while said media playing device is playing said video media content, said rule including;

i) initially automatically analyzing said first segment of text without using said second segment of text, thereby attempting to unambiguously determine said content of said first segment of text;

ii) in response to determining, during said analyzing of said first segment of text, that said first segment of text contains no ambiguous words which have multiple meanings in said first language, automatically and unambiguously determining said content of said first segment of text;

iii) in response to identifying, during said analyzing of said first segment of text, one or more ambiguous words in said first segment of text which have multiple meanings in said first language, for at least one identified ambiguous word having multiple meanings in said first language;

A) automatically identifying a corresponding word of said second language that appears in said second segment of text and is a translation of said identified ambiguous word;

B) automatically determining whether said corresponding word has an unambiguous meaning in said second language; and

C) in response to determining that said corresponding word has an unambiguous meaning in said second language, automatically assigning said unambiguous meaning of said corresponding word to said identified ambiguous word, and automatically and unambiguously determining said content of said first segment of text based on said assigned unambiguous meaning;

d) generating a recommendation based on said determined content of said first segment of text, said recommendation being for content associated with said video media content or with said first segment of text; and

e) outputting, by said processor, said recommendation.

2. The method of claim 1, wherein said outputting said recommendation comprises displaying said recommendation visually on a display associated with said media playing device.

3. The method of claim 1, further comprising, subsequent to said automatically determining whether said corresponding word has an unambiguous meaning in said second language and in response to determining that said corresponding word in said second language is an ambiguous word having multiple meanings in said second language:

automatically identifying a corresponding word of a third language that appears in a third segment of text and is a translation of said identified ambiguous word of said first language;

automatically determining whether said corresponding word of said third language has an unambiguous meaning in said third language; and

in response to determining that said corresponding word of said third language has an unambiguous meaning in said third language, automatically assigning said unambiguous meaning of said corresponding word of said third language to said identified ambiguous word of said first language;

wherein (i) said third segment of text is part of subtitles in said third language of said video media content played by said media playing device, (ii) said third segment of text is a translation of said first segment of text, and (iii) said subtitles in said third language are generated prior to initiating playing of said video media content by said media playing device.

4. The method of claim 1, wherein said identifying one or more ambiguous words comprises looking up words of said first segment of text in a pre-compiled list of ambiguous words for said first language.

5. A method for automatically determining content of a text and for generating an output based on the content, the method comprising:

a) obtaining, by a processor associated with a device operable to analyze text for determining content of the text, a first segment of text in a first language;

b) obtaining, by said processor, a second segment of text in a second language, said second segment of text being a translation of said first segment of text;

c) subsequent to said obtaining said first segment of text and said second segment of text and using said processor, automatically determining content of said first segment of text, in accordance with a predefined analysis rule, said rule including;

i) initially automatically analyzing said first segment of text without using said second segment of text, thereby attempting to unambiguously determine said content of said first segment of text;

ii) in response to determining, during said analyzing of said first segment of text, that said first segment of text contains no ambiguous words which have multiple meanings in said first language, unambiguously and automatically determining said content of said first segment of text;

iii) in response to identifying, during said analyzing of said first segment of text, one or more ambiguous words in said first segment of text which have multiple meanings in said first language, for at least one identified ambiguous word having multiple meanings in said first language;

A) automatically identifying a corresponding word of said second language that appears in said second segment of text and is a translation of said identified ambiguous word;

B) automatically determining whether said corresponding word has an unambiguous meaning in said second language; and

C) in response to determining that said corresponding word has an unambiguous meaning in said second language, automatically assigning said unambiguous meaning of said corresponding word to said identified ambiguous word, and automatically and unambiguously determining said content of said first segment of text based on said assigned unambiguous meaning;

d) generating a message based on said determined content of said first segment of text; and

e) outputting, by said processor, said message.

6. The method of claim 5, wherein said generating said message comprises obtaining information that is related to said content of said first segment of text and defining said information to be said message.

7. The method of claim 5, wherein said generating said message comprises generating a translation of said first segment of text into a third segment of text in a third language, and defining said third segment of text to be said message.

8. The method of claim 5, wherein said first segment of text and said second segment of text are obtained from subtitles of video media content while said video media content is being played by a media playing device, and wherein said generating said message comprises generating a recommendation for content associated with said video media content or with said first segment of text, and defining said recommendation to be said message.

9. The method of claim 5, wherein said outputting said message comprises displaying said message visually on a display device functionally associated with said processor.

10. The method of claim 5, wherein said outputting said message comprises outputting said message to a non-transitory computer readable storage medium functionally associated with said processor.

11. The method of claim 5, further comprising, subsequent to said automatically determining whether said corresponding word has an unambiguous meaning in said second language and in response to determining that said corresponding word in said second language is an ambiguous word having multiple meanings in said second language:

automatically identifying a corresponding word of a third language that appears in a third segment of text and is a translation of said identified ambiguous word of said first language;

automatically determining whether said corresponding word of said third language has an unambiguous meaning in said third language; and

in response to determining that said corresponding word of said third language has an unambiguous meaning in said third language, automatically assigning said unambiguous meaning of said corresponding word of said third language to said identified ambiguous word of said first language,

wherein said third segment of text is a translation of said first segment of text that is generated prior to the analyzing of said first segment of text.

12. The method of claim 5, wherein said identifying one or more ambiguous words comprises looking up words of said first segment of text in a pre-compiled list of ambiguous words for said first language.

13. A method for automatically determining true content of a text associated with video media content and for generating an output based on the true content, the method comprising:

a) obtaining, by a processor associated with a media playing device or with a set top box feeding said media playing device, a first segment of text in a first language, said first segment of text being part of subtitles in said first language of video media content played by said media playing device, said subtitles in said first language being generated prior to initiating playing of said video media content by said media playing device;

b) obtaining, by said processor, a second segment of text in a second language, said second segment of text being part of subtitles in said second language of said video media content played by said media playing device, said second segment of text being a translation of said first segment of text, said subtitles in said second language being generated prior to initiating playing of said video media content by said media playing device;

c) using said processor, automatically determining true content of said first segment of text, in accordance with a predefined analysis rule, said automatically determining of said true content of said first segment of text being done while said media playing device is playing said video media content, said rule including;

i) initially automatically analyzing said first segment of text without using said second segment of text, thereby attempting to unambiguously determine content of said first segment of text;

ii) in response to said analyzing of said first segment of text succeeding to unambiguously determine said content of said first segment of text without using said second segment of text, accepting said determined content to be said true content of said first segment of text;

iii) in response to said analyzing of said first segment of text failing to unambiguously determine said content of said first segment of text;

A) automatically analyzing said second segment of text for determining content of said second segment of text; and

B) automatically using results of said analyzing of said second segment of text to determine said true content of said first segment of text;

d) generating a recommendation based on said true content of said first segment of text, said recommendation being for content associated with said video media content or with said first segment of text; and

e) outputting, by said processor, said recommendation.

14. The method of claim 13, wherein said outputting said recommendation comprises displaying said recommendation visually on a display associated with said media playing device.

15. A method for automatically determining true content of a text and for generating an output based on the true content, the method comprising:

a) obtaining, by a processor associated with a device operable to analyze text for determining content of the text, a first segment of text in a first language;

b) obtaining, by said processor, a second segment of text in a second language, said second segment of text being a translation of said first segment of text;

c) subsequent to said obtaining said first segment of text and said second segment of text and using said processor, automatically determining true content of said first segment of text, in accordance with a predefined analysis rule, said rule including;

i) initially analyzing said first segment of text without using said second segment of text, thereby attempting to unambiguously determine content of said first segment of text;

ii) in response to said analyzing of said first segment of text succeeding to unambiguously determine said content of said first segment of text without using said second segment of text, accepting said determined content to be said true content of said first segment of text; and

iii) in response to said analyzing of said first segment of text failing to unambiguously determine said content of said first segment of text;

A) automatically analyzing said second segment of text for determining content of said second segment of text; and

B) automatically using results of said analyzing of said second segment of text to determine said true content of said first segment of text;

d) generating a message based on said determined true content of said first segment of text; and

e) outputting, by said processor, said message.

16. The method of claim 15, wherein said generating said message comprises obtaining information that is related to said true content of said first segment of text and defining said information to be said message.

17. The method of claim 15, wherein said generating said message comprises generating a translation of said first segment of text into a third segment of text in a third language, and defining said third segment of text to be said message.

18. The method of claim 15, wherein said first segment of text and said second segment of text are obtained from subtitles of video media content while said video media content is being played by a media playing device, and wherein said generating said message comprises generating a recommendation for content associated with said video media content or with said first segment of text, and defining said recommendation to be said message.

19. The method of claim 15, wherein said outputting said message comprises displaying said message visually on a display device functionally associated with said processor.

20. The method of claim 15, wherein said outputting said message comprises outputting said message to a non-transitory computer readable storage medium functionally associated with said processor.