Techniques for performing language detection and translation for multi-language content feeds

• US 8,812,295 B1
• Filed: 10/24/2011
• Issued: 08/19/2014
• Est. Priority Date: 07/26/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Swerdlow, Andrew, Jagpal, Navdeep Singh
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): ESTES, ERNEST LANGSTON
• Application Number:
  US13/279,568
• Time in Patent Office: 1,030 Days
• Field of Search: 704 1- 10
• US Class Current: 704/8
• CPC Class Codes: G06F 3/04817 using icons graphical or visual programming using iconic symbols G06F8/34
• CPC Class Codes: G06F 3/04842 Selection of displayed objects or displayed text elements G06F3/0482 takes precedence
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/263 Language identification
• 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
• CPC Class Codes: H04L 51/046 Interoperability with other network applications or services

Abstract

A technique for translating a portion of a website includes receiving a language set of a user indicating a primary language of the user is received. A content feed to be displayed to the user is received and parsed to identify a text portion of a user generated content. The original language of the text portion is determined and compared with the one or more languages in the language set. When the original language of the text portion does not match the languages in the language set: (i) the text portion, the original language, and the primary language are provided to a translation engine, (ii) a translated version of the text portion is received from the translation engine, (iii) the translated version of the text portion is inserted into the content feed to obtain a modified content feed, and (iv) the modified content feed is displayed to the user.

Claims

1. A computer implemented method for translating a portion of a website comprising:

determining whether the website comprises a parsable content feed including user generated content having at least one text portion; and

when the website is determined to comprise the parsable content feed;

receiving a language set of a viewing user, the language set including one or more languages and a primary language indicator that identifies one of the one or more languages as a primary language of the viewing user;

receiving an aggressiveness setting of the viewing user;

parsing a document object model representing the parsable content feed;

determining a user generated content element in the document object model, the user generated content element representing the user generated content; and

determining a text portion element of the user generated content element, the text portion element containing the at least one text portion;

determining an original language of the at least one text portion by;

(i) providing the at least one text portion to a language determination engine and receiving a potential language classification for the at least one text portion and a confidence score associated therewith,

(ii) comparing the confidence score to a threshold, wherein the threshold is based on the received aggressiveness setting,

(iii) when the confidence score is greater than the threshold, adopting the potential language classification of the at least one text portion as the original language, and

(iv) when the confidence score is less than the threshold, adopting one of the one or more languages in the language set as the original language such that original language of the at least one text portion matches one of the one or more languages in the language set;

comparing the original language of the at least one text portion with the one or more languages in the language set; and

when the original language of the at least one text portion does not match one of the one or more languages in the language set;

(i) providing, to a translation engine, the at least one text portion, the original language of the at least one text portion and the primary language indicator,

(ii) receiving, from the translation engine, a translated version of the least one text portion, the translated version corresponding to a translation of the least one text portion from the original language to the primary language of the viewing user,

(iii) inserting the translated version of the least one text portion into the parsable content feed to obtain a modified content feed, and

(iv) providing the modified content feed for display to the viewing user.

2. A computer implemented method for translating a parsable content feed comprising:

determining whether a website comprises the parsable content feed including user generated content having at least one text portion; and

when the website is determined to comprise the parsable content feed;

receiving a language set of a viewing user, the language set including one or more languages and a primary language indicator that identifies one of the one or more languages as a primary language of the viewing user;

parsing the parsable content feed to identify the user generated content and the at least one text portion;

determining an original language of the at least one text portion, wherein determining the original language of the least one text portion comprises;

receiving an aggressiveness setting of the viewing user,

receiving a potential language classification and a confidence score associated therewith,

comparing the confidence score to a threshold specific to the viewing user, wherein the threshold is based on the aggressiveness setting,

when the confidence score is greater than the threshold, adopting the potential language classification as the original language, and

when the confidence score is less than the threshold, adopting one of the one or more languages in the language set as the original language such that original language of the at least one text portion matches one of the one or more languages in the language set;

comparing the original language of the at least one text portion with the one or more languages in the language set; and

when the original language of the at least one text portion does not match one of the one or more languages in the language set;

(i) providing, to a translation engine, the at least one text portion, the original language of the at least one text portion and the primary language indicator,

(ii) receiving, from the translation engine, a translated version of the least one text portion, the translated version corresponding to a translation of the least one text portion from the original language to the primary language of the viewing user,

(iii) inserting the translated version of the least one text portion into the parsable content feed to obtain a modified content feed, and

(iv) providing the modified content feed for display to the viewing user.

3. The computer-implemented method of claim 2, wherein parsing the parsable content feed comprises:

parsing a document object model representing the parsable content feed;

determining a user generated content element in the document object model, the user generated content element representing the user generated content; and

determining a text portion element of the user generated content element, the text portion element containing the at least one text portion.

4. The computer-implemented method of claim 3, wherein inserting the translated version of the least one text portion into the parsable content feed to obtain the modified content feed comprises replacing the at least one text portion with the translated version of the at least one text portion.

5. The computer-implemented method of claim 4, wherein inserting the translated version of the least one text portion into the parsable content feed to obtain the modified content feed further comprises:

(i) creating an original text portion element in the document object model corresponding to the user generated content element, and (ii) inserting the at least one text portion in the original text portion element.

6. The computer-implemented method of claim 5, wherein the at least one text portion is selectively displayed to the viewing user.

7. The computer-implemented method of claim 2, wherein the computer-implemented method is executed by a web browser associated with the viewing user, the web browser being embodied as computer readable instructions executed by a processor.

8. The computer-implemented method of claim 2, wherein the computer-implemented method is executed by a web server hosting the website.

9. A system for translating a portion of a website comprising:

a user settings datastore that stores a language set of a viewing user, the language set including one or more languages and a primary language indicator that identifies one of the one or more languages as a primary language of the viewing user; and

one or more processors configured to;

determine whether the website comprises a parsable content feed including user generated content having at least one text portion; and

when the website is determined to comprise the parsable content feed;

parse the content feed to identify the user generated content and the at least one text portion;

determine an original language of the at least one text portion by;

retrieving an aggressiveness setting of the viewing user from the user settings datastore,

receiving a potential language classification and a confidence score associated therewith,

comparing the confidence score to a threshold specific to the viewing user, wherein the threshold is based on the aggressiveness setting,

when the confidence score is greater than the threshold, adopting the potential language classification as the original language, and

when the confidence score is less than the threshold, adopting one of the one or more languages in the language set as the original language such that original language of the at least one text portion matches one of the one or more languages in the language set;

compare the original language of the at least one text portion with the one or more languages in the language set; and

when the original language of the at least one text portion does not match one of the one or more languages in the language set;

(i) provide the at least one text portion, the original language of the at least one text portion and the primary language indicator to a translation engine,

(ii) receive a translated version of the least one text portion from the translation engine, the translated version corresponding to a translation of the least one text portion from the original language to the primary language of the viewing user,

(iii) insert the translated version of the least one text portion into the parsable content feed to obtain a modified content feed, and

(iv) provide the modified content feed to a display module for displaying the modified content feed to the viewing user.

10. The system of claim 9, wherein the one or more processors are configured to parse the parsable content feed by:

parsing a document object model representing the parsable content feed;

determining a user generated content element in the document object model, the user generated content element representing the user generated content; and

determining a text portion element of the user generated content element, the text portion element containing the at least one text portion.

11. The system of claim 10, wherein the one or more processors are configured to insert the translated version of the least one text portion into the parsable content feed to obtain the modified content feed by replacing the at least one text portion with the translated version of the at least one text portion.

12. The system of claim 11, wherein the one or more processors are configured to insert the translated version of the least one text portion into the content feed to obtain the modified content feed by:

(i) creating an original text portion element in the document object model corresponding to the user generated content element, and (ii) inserting the at least one text portion in the original text portion element.

13. The system of claim 9 further comprising a web browser associated with the viewing user, the web browser being embodied as computer readable instructions executed by the one or more processors.

14. The system of claim 9 further comprising a web server hosting the website.