Systems and methods for language detection

• US 10,162,811 B2
• Filed: 10/03/2016
• Issued: 12/25/2018
• Est. Priority Date: 10/17/2014
• Status: Active Grant

• Current Assignee: Mz Ip Holdings, LLC (AppLovin Corporation)
• Original Assignee: Mz Ip Holdings, LLC (AppLovin Corporation)
• Inventors: Bojja, Nikhil, Wang, Pidong, Guo, Shiman
• Primary Examiner(s): Spooner, Lamont
• Application Number:
  US15/283,646
• Publication Number:
  US 20170024372A1
• Time in Patent Office: 813 Days
• Field of Search: 704 2- 8
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/263 Language identification

Abstract

Implementations of the present disclosure are directed to a method, a system, and a computer program storage device for identifying a language in a message. Non-language characters are removed from a text message to generate a sanitized text message. An alphabet and/or a script are detected in the sanitized text message by performing at least one of (i) an alphabet-based language detection test to determine a first set of scores and (ii) a script-based language detection test to determine a second set of scores. Each score in the first set of scores represents a likelihood that the sanitized text message includes the alphabet for one of a plurality of different languages. Each score in the second set of scores represents a likelihood that the sanitized text message includes the script for one of the plurality of different languages. The language in the sanitized text message is identified based on at least one of the first set of scores, the second set of scores, and a combination of the first and second sets of scores.

Claims

1. A computer-implemented method of identifying a language in a message, the method comprising:

obtaining a text message generated by a user;

removing non-language characters from the text message to generate a sanitized text message;

detecting an alphabet and a script present in the sanitized text message, wherein

(i) detecting the alphabet comprises performing an alphabet-based language detection test to determine a first set of scores, and wherein each score in the first set of scores represents a likelihood that the sanitized text message comprises the alphabet for one of a plurality of different languages, and

(ii) detecting the script comprises performing a script-based language detection test to determine a second set of scores, and wherein each score in the second set of scores represents a likelihood that the sanitized text message comprises the script for one of the plurality of different languages;

providing one or more combinations of the first and second sets of scores as input to one or more classifiers including a first classifier and a second classifier, wherein the first classifier was trained using outputs from a first combination of language detection tests and the second classifier was trained using outputs from a second combination of language detection tests;

obtaining as output from at least one of the one or more classifiers a respective confidence score that the sanitized text message is in one of a plurality of different languages; and

identifying the language in the sanitized text message based on the confidence score from at least one of the one or more classifiers.

2. The method of claim 1, wherein the non-language characters comprise at least one of an emoji, a punctuation mark, an extra space, a carriage return, and a numerical character.

3. The method of claim 1, wherein the one or more combinations comprise an interpolation between the first and second sets of scores.

4. The method of claim 1, wherein providing one or more combinations of the first and second sets of scores comprises:

performing a language detection test on the sanitized text message to generate a third set of scores, wherein each score in the third set of scores represents a likelihood that the sanitized text message comprises one of a plurality of different languages.

5. The method of claim 4, wherein the language detection test is selected from a plurality of language detection tests based on at least one of the first set of scores, the second set of scores, and a combination of the first and second sets of scores.

6. The method of claim 4, wherein the language detection test comprises a language detection method.

7. The method of claim 6, wherein the language detection method comprises at least one of a dictionary-based language detection test, an n-gram language detection test, and a user language profile language detection test.

8. The method of claim 4, comprising:

processing the third set of scores using the one or more classifiers to identify the language in the sanitized text message.

9. The method of claim 1, wherein the one or more classifiers comprise at least one of a supervised learning model, a partially supervised learning model, an unsupervised learning model, and an interpolation.

10. A computer-implemented system for identifying a language in a message, comprising:

one or more computer processors programmed to implement a sanitizer module, a grouper module, and a language detector module,

wherein the sanitizer module obtains a text message generated by a user and removes non-language characters from the text message to generate a sanitized text message,

wherein the grouper module detects an alphabet and a script present in the sanitized text message, and wherein

(i) detecting the alphabet comprises performing an alphabet-based language detection test to determine a first set of scores, and wherein each score in the first set of scores represents a likelihood that the sanitized text message comprises the alphabet for one of a plurality of different languages, and

(ii) detecting the script comprises performing a script-based language detection test to determine a second set of scores, and wherein each score in the second set of scores represents a likelihood that the sanitized text message comprises the script for one of the plurality of different languages, and

wherein the language detector module is operable to perform operations comprising;

providing one or more combinations of the first and second sets of scores as input to one or more classifiers including a first classifier and a second classifier, wherein the first classifier was trained using outputs from a first combination of language detection tests and the second classifier was trained using outputs from a second combination of language detection tests;

obtaining as output from at least one of the one or more classifiers a respective confidence score that the sanitized text message is in one of a plurality of different languages; and

identifying the language in the sanitized text message based on the confidence score from at least one of the one or more classifiers.

11. The system of claim 10, wherein the non-language characters comprise at least one of an emoji, a punctuation mark, an extra space, a carriage return, and a numerical character.

12. The system of claim 10, wherein the one or more combinations comprise an interpolation between the first and second sets of scores.

13. The system of claim 10, wherein the grouper module is operable to perform operations comprising:

selecting the language detector module from a plurality of language detector modules based on at least one of the first set of scores, the second set of scores, and a combination of the first and second sets of scores.

14. The system of claim 10, wherein the language detector module comprises:

a language detection methods module operable to perform operations comprising;

performing a language detection test on the sanitized text message to generate a third set of scores, wherein each score in the third set of scores represents a likelihood that the sanitized text message comprises one of a plurality of different languages.

15. The system of claim 14, wherein the language detection test comprises at least one of a dictionary-based language detection test, an n-gram language detection test, and a user language profile language detection test.

16. The system of claim 14, wherein the language detector module comprises:

a classifier module operable to perform operations comprising;

processing the third set of scores using the one or more classifiers to identify the language in the sanitized text message.

17. The system of claim 16, wherein the classifier module is operable to perform operations comprising:

outputting an indication that the sanitized text message is in the identified language, wherein the indication comprises a confidence score.

18. The system of claim 10, wherein the one or more classifiers comprise at least one of a supervised learning model, a partially supervised learning model, an unsupervised learning model, and an interpolation.

19. An article, comprising:

a non-transitory computer-readable medium having instructions stored thereon that, when executed by one or more computers, cause the computers to perform operations comprising;

obtaining a text message generated by a user;

removing non-language characters from the text message to generate a sanitized text message;

detecting an alphabet and a script present in the sanitized text message, wherein

(i) detecting the alphabet comprises performing an alphabet-based language detection test to determine a first set of scores, and wherein each score in the first set of scores represents a likelihood that the sanitized text message comprises the alphabet for one of a plurality of different languages, and

(ii) detecting the script comprises performing a script-based language detection test to determine a second set of scores, and wherein each score in the second set of scores represents a likelihood that the sanitized text message comprises the script for one of the plurality of different languages;

providing one or more combinations of the first and second sets of scores as input to one or more classifiers including a first classifier and a second classifier, wherein the first classifier was trained using outputs from a first combination of language detection tests and the second classifier was trained using outputs from a second combination of language detection tests;

obtaining as output from at least one of the one or more classifiers a respective confidence score that the sanitized text message is in one of a plurality of different languages; and

identifying the language in the sanitized text message based on the confidence score from at least one of the one or more classifiers.