System and method for improving text input in a shorthand-on-keyboard interface

• US 9,256,580 B2
• Filed: 03/12/2014
• Issued: 02/09/2016
• Est. Priority Date: 10/22/2005
• Status: Expired due to Fees

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Kristensson, Per-Ola, Zhai, Shumin
• Primary Examiner(s): PULLIAS, JESSE SCOTT
• Application Number:
  US14/206,920
• Publication Number:
  US 20140278374A1
• Time in Patent Office: 699 Days
• Field of Search: 704/1, 704/9, 704/10, 704231-257
• US Class Current: 1/1
• CPC Class Codes: G06F 3/0237 using prediction or retrieval techniques
• CPC Class Codes: G06F 3/04883 for inputting data by handwriting, e.g. gesture or text
• CPC Class Codes: G06F 40/10 Text processing natural language analysis G06F40/20; semantic analysis G06F40/30; processing or translation of natural language G06F40/40
• CPC Class Codes: G06F 40/237 Lexical tools

Abstract

A word pattern recognition system improves text input entered via a shorthand-on-keyboard interface. A core lexicon comprises commonly used words in a language; an extended lexicon comprises words not included in the core lexicon. The system only directly outputs words from the core lexicon. Candidate words from the extended lexicon can be outputted and simultaneously admitted to the core lexicon upon user selection. A concatenation module enables a user to input parts of a long word separately. A compound word module combines two common shorter words whose concatenation forms a long word.

Claims

1. A system for gesture recognition by a computing device, the system comprising:

one or more processors;

a memory coupled to the one or more processors;

a touch-sensitive display coupled to the one or more processors;

a virtual keyboard module configured to display a virtual keyboard on the touch-sensitive display and to receive a continuous stroke input by a user on the virtual keyboard,

wherein the continuous stroke comprises a start-point and an end-point;

a recognition module configured to obtain a most probable candidate word corresponding to the continuous stroke input by a user on the virtual keyboard; and

a candidate display module configured to display, over the virtual keyboard, one or more characters representing the most probable candidate word,

wherein a representation of the most probable candidate word is displayed at a particular location selected based at least in part on the continuous stroke end-point.

2. The system of claim 1 wherein the continuous stroke is entered by the user on the touch-sensitive display using a stylus, and wherein the most probable candidate word includes a suffix for a word matching the continuous stroke.

3. The system of claim 1 further comprising a concatenation module configured to:

automatically recognize that the most probable candidate word is part of a stem or compound word created from one or more previously entered words; and

replace the previously entered one or more words with the recognized stem or compound word.

4. The system of claim 1 further comprising:

a concatenation module configured to;

automatically recognize that the most probable candidate word is part of a stem or compound word created from one or more previously entered words; and

replace the previously entered one or more words with the recognized stem or compound word; and

a de-concatenation module configured to;

receive an input from the user; and

replace, based on the input, the recognized stem or compound word with two or more character sets separated by a SPACE symbol, wherein the two or more character sets are created by separating the characters comprising the recognized stem or compound word.

5. A computer-readable memory storing instructions that, when executed by a computing device, cause the computing device to perform gesture recognition operations, comprising:

receiving, via a virtual keyboard interface, a continuous stroke from a user, wherein the continuous stroke comprises a start-point and an end-point;

determining a most probable candidate word based at least in part on the received continuous stroke from the user; and

displaying, over the virtual keyboard interface, a representation of the determined most probable candidate word,

wherein the representation of the determined most probable candidate word is displayed to appear over the virtual keyboard and at a particular location selected based at least in part on the continuous stroke end-point.

6. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising comparing the received continuous stroke to a set of template patterns such that a n-best list of possible matching template patterns are selected.

7. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising:

comparing the received continuous stroke to a set of template patterns such that a n-best list of possible matching template patterns are selected;

wherein the determining a most probable candidate word is performed by determining a best-match from among the n-best list of possible matching template patterns.

8. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising:

comparing the received continuous stroke to a set of template patterns such that a n-best list of possible matching template patterns are selected;

wherein;

the n-best list of possible matching template patterns are selected by comparing the continuous stroke with a core lexicon and an extended lexicon;

each of the possible matching template patterns corresponds to a set of characters;

the sets of characters corresponding to the possible matching template patterns in the n-best list are displayed in a separate menu; and

character sets corresponding to the core lexicon are displayed with a first perceptual feature and character sets corresponding to the extended lexicon are displayed with a second perceptual feature not applied to the character sets corresponding to the first lexicon.

9. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising:

comparing the received continuous stroke to a set of template patterns such that a n-best list of possible matching template patterns are selected;

wherein;

the n-best list of possible matching template patterns are selected by comparing the continuous stroke with a core lexicon and an extended lexicon;

each of the possible matching template patterns corresponds to a set of characters;

the sets of characters corresponding to the possible matching template patterns in the n-best list are displayed in a separate menu; and

the n-best list of possible matching template patters has an order corresponding to a degree of similarity between the received continuous stroke and each template pattern, and wherein the separate menu first displays the sets of characters from the core lexicon in an order corresponding to the order of the n-best list, then displays the sets of characters from the extended lexicon in an order corresponding to the order of the n-best list.

10. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising:

comparing the received continuous stroke to a set of template patterns such that

a n-best list of possible matching template patterns are selected;

wherein;

the n-best list of possible matching template patterns are selected by comparing the continuous stroke with a core lexicon and an extended lexicon;

each of the possible matching template patterns corresponds to a set of characters; and

the sets of characters corresponding to the possible matching template patterns in the n-best list are displayed in a separate menu.

11. The computer-readable memory of claim 5 further storing instructions that, when executed by the computing device, cause the computing device to perform further operations comprising:

comparing the received continuous stroke to a set of template patterns such that a n-best list of possible matching template patterns are selected;

selecting, for each of the possible matching template patterns in the n-best list, a set of characters corresponding to the possible matching template pattern;

displaying the sets of characters corresponding to the possible matching template patterns in a separate menu;

receiving either a selection of the characters corresponding to the most probable candidate word or a selection of one of the sets of characters from the separate menu; and

entering, in a text entry area, the selected characters.

12. The computer-readable memory of claim 5 wherein determining a single most probable candidate word comprises a modified elastic matching algorithm using scale transformation and location transformation of the received gesture.

13. The computer-readable memory of claim 5 wherein the virtual keyboard is an ATOMIK virtual keyboard or a QWERTY virtual keyboard.

14. The computer-readable memory of claim 5 wherein the continuous stroke is entered using a stylus.

15. A gesture recognition method for a gesture entered via a virtual keyboard interface of a computing device, the method comprising:

receiving, via the virtual keyboard interface, a continuous stroke having a start-point and an end-point;

selecting a list of word candidates by comparing the received continuous stroke with a set of template patterns;

determining a best-match template pattern; and

displaying, over the virtual keyboard, and near a location defined by the endpoint of the continuous stroke, one or more characters that represent the best-match template pattern.

16. The method of claim 15 wherein:

the selecting comprises comparing the received continuous stroke with a first set of template patterns in a core lexicon and comparing the continuous stroke with a second set of template patterns in an extended lexicon; and

the determining a best-match template pattern comprises selecting a best-match template pattern from the core lexicon.

17. The method of claim 15 wherein the selecting comprises a modified elastic matching algorithm using scale transformation and location transformation of the received gesture.

18. The method of claim 15 further comprising:

displaying, in a menu separate from the characters that represent the best-match template pattern, one or more character sets, each character set corresponding to a template pattern in the list of word candidates,

wherein the separate menu is organized such that character sets corresponding to template patterns from the core lexicon appear before character sets corresponding to template patterns from the extended lexicon;

receiving a selection of one of the character sets from the separate menu, the selection corresponding to a template pattern from the extended lexicon; and

adding, to the core lexicon, the template pattern from the extended lexicon corresponding to the selection.

19. The method of claim 15 wherein determining a best-match template pattern comprises using context clues based on one or more previously received words.

20. The method of claim 15 wherein displaying, over the virtual keyboard, a line representing at least a portion of the continuous stroke comprises displaying, over a QWERTY virtual keyboard, a line representing at least a portion of the continuous stroke.