Multi-Mode Input Method Editor

• US 20100217581A1
• Filed: 04/09/2008
• Published: 08/26/2010
• Est. Priority Date: 04/10/2007
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Hong, Feng
• Granted Patent: US 8,543,375 B2
• Time in Patent Office: Days
• US Class Current: 704/3
• CPC Class Codes: G06F 40/129 Handling non-Latin characters, e.g. kana-to-kanji conversion
• 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

Abstract

Methods, systems, and apparatus, including computer program products, in which an input method editor receives composition inputs and determines language context values based on the composition inputs. Candidate selections based on the language context values and the composition inputs are identified.

Claims

1. A method, comprising:

receiving composition inputs;

determining language context values based on the composition inputs; and

identifying candidate selections for presentation based on the language context values and the composition inputs.

2. The method of claim 1, wherein determining language context values based on the composition inputs comprises:

determining an English context value; and

determining a Chinese context value.

3. The method of claim 2, wherein determining an English context value comprises:

applying the composition inputs in an English context; and

identifying English words that correspond to the composition inputs.

4. The method of claim 3, wherein determining an English context value comprises:

comparing the composition inputs to prefixes of English words; and

setting the English context value based on the comparison.

5. The method of claim 1, wherein identifying candidate selections for presentation based on the language context values and the composition inputs comprises ranking the candidate selections for two or more different languages according to the language context values, wherein the candidate selections are determined for the two or more different languages based on the composition inputs.

6. The method of claim 2, wherein determining a Chinese context value comprises:

applying the composition inputs in a Chinese context; and

identifying Chinese characters that correspond to the composition inputs.

7. The method of claim 6, wherein determining Chinese context value comprises:

evaluating the Chinese characters against a Chinese language model; and

setting the Chinese context value based on the evaluation.

8. The method of claim 7, wherein evaluating the characters against a Chinese language model comprises evaluating the Chinese characters against a Chinese grammar model.

9. The method of claim 8, wherein the Chinese grammar model is based on serial verb construction.

10. The method of claim 8, wherein the Chinese grammar model is based on aspects of perfectives and imperfectives.

11. The method of claim 6, wherein identifying characters that correspond to the composition inputs comprises identifying Hanzi characters based on Pinyin inputs.

12. The method of claim 1, wherein the composition inputs comprise keyboard inputs.

13. The method of claim 1, wherein a language context value is representative of a likelihood of a user'"'"'s intent to input the compositions inputs in a corresponding language.

14. A system, comprising:

an input device configured to receive composition inputs; and

means for determining language context values based on the composition inputs and for identifying candidate selections for presentation to a user based on the language context values and the composition inputs.

15. A computer-implemented method, comprising:

receiving composition inputs;

determining candidate selections for the two or more different languages based on the composition inputs;

evaluating the candidate selections for the two or more different languages against language models for the two or more different languages, wherein each language model includes a rule set for a language, and wherein the language models collectively include rule sets for the two or more different languages;

determining a language context value for each of the two or more different languages based on the evaluation; and

identifying candidate selections for presentation based on the language context values.

16. The method of claim 15, wherein determining a language context value for each of the two or more different languages comprises:

determining an English language context value; and

determining a Chinese language context value.

17. The method of claim 16, wherein:

determining an English language context value comprises;

applying the composition inputs in an English context;

comparing the composition inputs to prefixes of English words; and

setting the English context value based on the comparison; and

determining a Chinese language context value comprises;

applying the composition inputs in a Chinese context;

identifying Chinese characters that correspond to the composition inputs;

evaluating the Chinese characters against a Chinese language model; and

setting the Chinese context value based on the evaluation.

18. The method of claim 17, wherein the language models include a Chinese language model that includes a grammar rule set for the Chinese language.

19. The method of claim 18, wherein the Chinese language model is based one or more of serial verb construction or aspects of perfectives and imperfectives.

20. A system, comprising:

a data store configured to store input method editor language models, wherein each language model includes a rule set for a language, and wherein the language models collectively include rule sets for two or more different languages; and

an input method editor engine configured to receive composition inputs and determine candidate selections for the two or more different languages based on the received composition inputs, to evaluate the candidate selections against the language models and determine language context values for each of the two or more different languages, and to identify candidate selections for presentation based on the language context values.

21. The system of claim 20, wherein the language models include a Chinese language model that includes a grammar rule set for the Chinese language.