Probability-based approach to recognition of user-entered data

• US 8,589,149 B2
• Filed: 08/05/2008
• Issued: 11/19/2013
• Est. Priority Date: 08/05/2008
• Status: Active Grant

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Cecil, Matthew, Sharan, Santosh, LaChapelle, Jason
• Primary Examiner(s): Borsetti, Greg
• Application Number:
  US12/186,425
• Publication Number:
  US 20100036655A1
• Time in Patent Office: 1,932 Days
• Field of Search: 704 9- 10, 715/256, 715/257, 715/259, 715/260, 715/261, 715/263, 345/168
• US Class Current: 704/9
• CPC Class Codes: G06F 3/0237 using prediction or retrieval techniques
• CPC Class Codes: G06F 3/04886 by partitioning the display area of the touch-screen or the surface of the digitising tablet into independently controllable areas, e.g. virtual keyboards or menus
• CPC Class Codes: G06F 40/189 Automatic justification
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs

Abstract

A method for entering keys in a small key pad is provided. The method comprising the steps of: providing at least a part of keyboard having a plurality of keys; and predetermining a first probability of a user striking a key among the plurality of keys. The method further uses a dictionary of selected words associated with the key pad and/or a user.

Claims

1. A computer-implemented method comprising:

receiving input indicative of a user striking one or more keys in a keyboard having multiple keys;

predetermining a probability distribution,

wherein the probability distribution indicates, for each pair of keys in the keyboard comprised of at least a first key and a second key, a probability that a user unintentionally strikes the first key in the pair when the user intends to strike the second key in the pair;

wherein the probability distribution includes a probability that the user will miss the keyboard entirely when a keystrike was intended; and

;

wherein the keyboard is a touchscreen keyboard and

in response to receiving input indicative of the user striking one or more keys in the keyboard, using a processor and the predetermined probability distribution to select at least one candidate word.

2. The method of claim 1 further comprising using a dictionary to select the at least one candidate word.

3. A non-transitory computer-readable medium storing instructions that, when executed by a data processing device, performs an operation for identifying candidate words from user input, comprising:

receiving input indicative of a user striking one or more keys in a keyboard having multiple keys;

associating a probability distribution to manual user selection of individual keys of the keyboard,

wherein the probability distribution indicates, for each set of keys in the keyboard comprised of at least a first key and a second key, a probability that a user unintentionally strikes the first key in the set when the user intends to strike the second key in the set; and

in response to receiving input indicative of the user striking one or more keys in the keyboard, using the probability distribution to select at least one candidate word;

wherein the operation further comprises selecting one of several customized and predetermined probability matrices, and

wherein one matrix is associated with left-handed users and another matrix is associated with right-handed users.

4. The computer-readable medium of claim 3 wherein the operation further comprises using a dictionary of words to select the at least one candidate word.

5. The computer-readable medium of claim 4 wherein the operation further comprises associating every word in the dictionary with a probability, wherein the probability is a probability that the word is intended based on word frequency, grammatical rules, or proportion of an item in an inventory.

6. The computer-readable medium of claim 4 wherein the operation further comprises associating a set of n (n being a natural number) letters entered to a set of words in the dictionary.

7. The computer-readable medium of claim 6 wherein using the probability distribution to select at least one candidate word comprises selecting words in the set of words having the highest probability.

8. The computer-readable medium of claim 3 wherein the operation further comprises assigning a probability to a likelihood that a user will miss the keyboard entirely when a keystrike was intended, and wherein the keyboard is a touchscreen keyboard.

9. The computer-readable medium of claim 3 wherein the operation further comprises providing a learning phase where values of a probability matrix are seeded.

10. The computer-readable medium of claim 3 wherein the keyboard is a touchscreen keyboard, wherein errors occur from text input because keys of the touchscreen keyboard are smaller than the user'"'"'s finger, and wherein the operation further includes generating an N by N matrix where N is a number of keys on the touchscreen keyboard and wherein a matrix entry at position (I, J) is a probability that the user presses a J^th key when the user intended to press an I^th key.

11. A non-transitory computer-readable medium storing instructions that, when executed by a data processing device, performs an operation for selecting candidate words based on a user'"'"'s manual input to a keyboard, the operation comprising:

associating a probability distribution to a user'"'"'s manual input via a keyboard having multiple keys,

wherein the probability distribution indicates, for each set of keys in the keyboard comprised of at least a first key and a second key, a probability that the user will unintentionally select the first key in the set when the user intended to select the second key in the set instead;

providing a dictionary, wherein the dictionary includes multiple words;

in response to receiving a particular sequence of selected keys entered by a user, using the probability distribution to associate each of at least two words in the dictionary with a probability,

wherein the probability for each word reflects a likelihood that the user intended to enter the word;

selecting as a candidate word, at least one word having the highest probability among the at least two words in the dictionary; and

wherein associating a probability distribution to a user'"'"'s manual input via a keyboard further comprises selecting one of several customized and predetermined probability distributions, wherein one probability distribution is associated with left-handed users and another probability distribution is associated with right-handed users.

12. The computer-readable medium of claim 11, wherein associating each of at least two words in the dictionary with a probability comprises multiplying multiple probabilities, wherein each of the multiple probabilities is associated with a particular key in the received particular sequence of selected keys.

13. The computer-readable medium of claim 12, wherein associating each of at least two words in the dictionary with a probability further comprises adjusting the probability for a word to reflect a frequency of use of the word.

14. The computer-readable medium of claim 12, wherein the operation further comprises utilizing a predetermined probability that a user will miss the keyboard entirely when a keystrike was intended by the user, and wherein the keyboard is a touchscreen keyboard.

15. The computer-readable medium of claim 12, wherein the operation further comprises providing a learning phase where values of the probability distribution are seeded.

16. The computer-readable medium of claim 12 wherein associating a probability distribution to a user'"'"'s manual input via a keyboard further comprises selecting one of several customized and predetermined probability distributions, wherein one probability distribution is associated with a left-handed user and another probability distribution is associated with a right-handed user.

17. The computer-readable medium of claim 12 wherein the keyboard is a touchscreen keyboard, wherein errors occur from text input because keys of the touchscreen keyboard are smaller than the user'"'"'s finger, and wherein the operation further comprises generating an N by N matrix where N is a number of keys on the touchscreen keyboard and wherein a matrix entry at position (I, J) is a probability that the user presses a J^th key when the user intended to press an I^th key.

18. A computer-implementable method to perform an operation for selecting candidate words based on a user'"'"'s manual input to a keyboard, the method comprising:

associating a probability distribution to a user'"'"'s manual input via a keyboard having multiple keys,

wherein the probability distribution indicates, for each set of keys in the keyboard comprised of at least a first key and a second key, a probability that the user will unintentionally select the first key in the set when the user intended to select the second key in the set instead;

providing a dictionary, wherein the dictionary includes multiple words;

in response to receiving a particular sequence of selected keys entered by the user, using the probability distribution to associate each of at least two words in the dictionary with a probability,

wherein the probability for each word reflects a likelihood that the user intended to enter the word; and

selecting as a candidate word, at least one word having the highest probability among the at least two words in the dictionary;

wherein the operation further comprises utilizing a predetermined probability that a user will miss the keyboard entirely when a keystrike was intended by the user, and wherein the keyboard is a touchscreen keyboard.