Context-aware re-formating of an input

• US 9,672,202 B2
• Filed: 03/20/2014
• Issued: 06/06/2017
• Est. Priority Date: 03/20/2014
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Alphonso, Issac, Levit, Michael, Parthasarathy, Sarangarajan, Kibre, Nick
• Primary Examiner(s): ROBERTS, SHAUN A
• Application Number:
  US14/220,916
• Publication Number:
  US 20150269136A1
• Time in Patent Office: 1,174 Days
• Field of Search: 704 9
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs
• CPC Class Codes: G10L 15/24 Speech recognition using non-acoustical features
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

Various components provide options to re-format an input based on one or more contexts. The input is received that has been submitted to an application (e.g., messaging application, mobile application, word-processing application, web browser, search tool, etc.), and one or more outputs are identified that are possibilities to be provided as options for re-formatting. A respective score of each output is determined by applying a statistical model to a respective combination of the input and each output, the respective score comprising a plurality of context scores that quantify a plurality of contexts of the respective combination. Exemplary contexts include historical-user contexts, domain contexts, and general contexts. One or more suggested outputs are selected from among the one or more outputs based on the respective scores and are provided as options to re-format the input.

Claims

1. A method of re-formatting an input based on one or more contexts comprising:

receiving the input that has been submitted to an application;

identifying a plurality of outputs comprising possible suggestions for re-formatting the input;

calculating a respective score of each output of the plurality of outputs by applying a statistical model to a respective combination of the input and each output,

wherein a respective score of each output comprises a plurality of context scores that quantify a plurality of contexts of the respective combination of the input and each output; and

wherein a context score of a context is calculated by applying a customizable weight assigned to the context to a frequency with which the input was previously re-formatted to the output when the context was applicable;

selecting one or more suggested outputs from among the one or more outputs based on the respective scores; and

providing the one or more suggested outputs as options to re-format the input.

2. The method of claim 1, wherein the plurality of contexts include a general-population context, a user-specific context, a user-group context, and a message-recipient context.

3. The method of claim 1, wherein the plurality of contexts includes vertical domains.

4. The method of claim 1, wherein one context of the plurality of contexts includes the application.

5. The method of claim 4, wherein the application includes a messaging application.

6. The method of claim 1, wherein the plurality of context scores represent a general context, a historical-user-selection context, and a domain context, each of which is factored by a respective weight when combined in the statistical model, and wherein the respective weight of each term is customizable to tune the statistical model.

7. The method of claim 1, wherein selecting the one or more suggested outputs includes comparing the respective scores to each other and selecting the one or more suggested outputs that correspond to a group of respective scores that are the highest among the respective scores, and wherein the group includes a pre-established quantity.

8. The method of claim 1, wherein selecting the one or more suggested outputs includes determining that the respective scores exceed a minimum score threshold, and wherein respective scores that do not exceed the minimum score threshold are deemed ineligible to be provided as an option to reformat.

9. The method of claim 1, wherein the context score includes a factor of the customizable weight and the frequency.

10. The method of claim 1, wherein the frequency is multiplied by the customizable weight to calculate the context score.

11. A computer storage device storing computer-readable instructions that, when executed, perform operations for re-formatting an input based on one or more contexts comprising:

receiving the input that has been submitted to an application, wherein the input is associated with a user profile of a specific user;

identifying one or more outputs to be used as suggestions for re-formatting the input by looking up the input in an index stored in a memory device;

determining a respective score of each output of the one or more outputs by applying a statistical model to a respective combination of the input and each output, wherein the respective score comprises;

a historical score of the respective combination quantifying a frequency with which the input has been reformatted to a respective output in a context of the specific user, wherein a first customizable weight is applied to the frequency to selectively adjust an influence of the historical score on the statistical model, and

a domain score of the respective combination quantifying a frequency with which the input has been reformatted to the respective output in a context of the application, wherein a second customizable weight is applied to the frequency to selectively adjust an influence of the domain score on the statistical model;

selecting one or more suggested outputs from among the one or more outputs based on the respective scores; and

providing the one or more suggested outputs as options to re-format the input.

12. The computer storage device of claim 11, wherein the application includes a messaging application and wherein the input includes a body of a message drafted using the messaging application.

13. The computer storage device of claim 12, wherein the respective score further comprises another history score of the respective combination quantifying a frequency with which the input has been reformatted to a respective output in a context of a specified recipient to whom the message is designated to be sent.

14. The computer storage device of claim 11, wherein the one or more suggested outputs provide an auto-correction option, an auto-completion option, an inverse-text-normalization option, a speech-to-text translation, or a combination thereof.

15. The computer storage device of claim 11, wherein the statistical model mathematically combines a plurality of contexts in a linear equation to calculate the respective score.

16. A computing device having a processor and a computer storage device, the processor for executing instructions stored on the computer storage device to perform operations comprising:

receiving an input that has been submitted to an application and that is associated with a user profile of a specific user;

communicating the input to an input re-formatter, which determines a respective score of each output of a plurality of outputs by applying a statistical model to a respective combination of the input and each output,

wherein a respective score comprises a plurality of terms that quantify a plurality of context-specific scores of the respective combination, and

wherein the statistical model applies a linear equation that includes dividing the respective score of each output by a summation of scores of the plurality of outputs;

receiving one or more suggested outputs from the input re-formatter; and

presenting the one or more suggested outputs as options to re-format the input using an output device of the computing device.

17. The computing device of claim 16, wherein the input re-formatter is a client-side input re-formatter.

18. The computing device of claim 16, wherein the input re-formatter is a server-side re-formatter.