Method for converting a phonetic character string into the text of an Asian language

• US 6,356,866 B1
• Filed: 10/07/1998
• Issued: 03/12/2002
• Est. Priority Date: 10/07/1998
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Urata, Kentaro, Oliver, David C., Rucker, Erik J., Pratley, Christopher H.
• Primary Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US09/167,914
• Time in Patent Office: 1,252 Days
• Field of Search: 704/1, 704/8-10, 707/530, 707/535, 707/532, 707/540, 345/150, 345/171
• US Class Current: 704/9
• CPC Class Codes: G06F 40/53 Processing of non-Latin text kana-to-kanji conversion G06F40/129; vowelisation G06F40/232

Abstract

Converting an incoming phonetically-coded string representing an East Asian language into its proper text characters. Conversion is triggered when dynamic IME analysis of the string indicates a threshold probability of correct conversion. Conversion of unconverted phonetic text is also triggered when the user relocates the insertion point. The user may select portions of text affected by conversion error and access a corresponding editable phonetic string and a prioritized list of alternate conversions for that string. The corresponding phonetic string is generated from the selected text if the original input string is unavailable.

Claims

1. A computer-implemented method for converting a phonetically-coded string into a text string comprising phonetic and nonphonetic characters for use by a program module running on a computer system, comprising the steps of:

(a) receiving phonetic character(s);

(b) generating a modified phonetically-coded string by adding the phonetic character(s) to the phonetically-coded string;

(c) converting a portion of the modified phonetically-coded string to the text string in response to the portion exceeding a threshold probability of accurate conversion; and

(d) providing the text string as an input to the program module.

2. The method of claim 1, wherein the program module comprises an Input Method Editor (IME), and steps (a)-(d) are completed as a background task by the IME.

3. The method of claim 1, wherein the portion comprises a longest continuous substring of the modified phonetically-coded string that exceeds the threshold probability of accurate conversion, wherein the threshold probability decreases as the size of the continuous substring decreases.

4. The method of claim 1 further comprising the step of correcting the text string by:

in response to an indication of selection of a section of the text string, identifying a phonetically-coded string corresponding to the text string section;

converting the identified phonetically-coded string to at least one corrected text string; and

displaying each corrected text string.

5. The method of claim 4, wherein the step of identifying a phonetically-coded string corresponding to the text string section comprises accessing the phonetically-coded string maintained in a memory storage device operatively coupled to the computer system.

6. The method of claim 4 wherein the step of identifying a phonetically-coded string corresponding to the text string section comprises converting the text string section to generate the corresponding phonetically-coded string.

7. The method of claim 4, wherein the step of identifying a phonetically-coded string corresponding to the text string section comprises completing a reverse-conversion by converting the text string section to generate the corresponding phonetically-coded string in the event that the corresponding phonetically-coded string is not resident in a memory storage device operatively coupled to the computer system.

8. The method of claim 1, wherein the portion of the modified phonetically-coded string includes the starting character of the phonetically-coded string.

9. A computer-implemented method for converting a string representing syllables of an Asian language into the written characters of the Asian language for use by a program module running on a computer system, comprising the step of:

(a) inputting a string of symbols representing the syllables;

(b) responsive to inputting the symbols, converting the symbol string into a language text string without manual intervention; and

(c) correcting a portion of the language text string corresponding to at least one symbol erroneously converted in step (b), by performing the steps of;

selecting the portion of the language text string;

identifying each symbol corresponding to the selected portion of the language text string;

generating a plurality of alternative conversions of each erroneously converted symbol;

displaying the plurality of alternative conversions; and

selecting one of the plurality of alternative conversions.

10. The method of claim 9, wherein each erroneously converted symbol is generated by a reverse-conversion of the selected portion of the language text string if the symbol is not resident on a memory storage device operatively coupled to the computer system.

11. The method of claim 9, wherein each step of displaying the plurality of alternative conversions further comprises displaying the erroneously converted symbol and the alternative conversions.

12. The method of claim 9, wherein the step of displaying the alternative conversions comprises displaying the alternative conversions in a list ordered by decreasing confidence of accurate conversion.

13. A computer-implemented method for converting a phonetically-coded string into a processed text string for use by a program module for processing an electronic document on a computer system, comprising the steps of:

detecting phonetic characters of the phonetically-coded string;

constructing an active portion comprising selected ones of the phonetic characters located behind an insertion point positioned in the phonetically-coded string;

identifying a substring of the active portion, the active portion substring comprising one of the selected phonetic characters positioned furthest from the insertion point, the active portion substring exceeding a threshold probability of accurate conversion;

converting the active portion substring to a converted substring comprising processed text characters; and

creating the processed text string by appending the converted substring to non-converted characters of the active portion.

14. The method of claim 13, wherein the selected phonetic characters of the active portion string are generated by reverse-conversion of corresponding text if the selected phonetic characters are not resident on a memory storage device operatively coupled to the computer system.

15. The method of claim 13, further comprising the steps of, in response to an insertion point relocation command for moving the insertion point from a prior position within the active portion to a new position within the active portion,

constructing an active portion comprising selected ones of the phonetic characters located behind the prior position of the insertion point;

converting the active portion to the converted string of processed text characters; and

providing the converted string of processed text characters as an input to the program module.

16. The method of claim 15, wherein converting the active portion substring to the converted string of processed text characters comprises:

identifying a conversion context defined by at least one of the phonetic characters positioned in front of and behind the active portion;

converting the active portion substring to the converted string of processed text characters based on the conversion context.

17. The method of claim 13, wherein the converting step comprises converting at least one of the selected phonetic characters just prior to excluding the one selected phonetic character from the active portion.

18. The method of claim 13, further comprising the step of correcting an error in the processed text string by performing the steps of:

selecting a portion of the processed text string corresponding to the error;

responsive to receiving a reconvert command, identifying a phonetically-coded portion corresponding to the portion of the processed text string;

generating alternate conversions by converting the phonetically-coded portion;

displaying the alternative conversions; and

in response to selection of one of the alternative conversion, replacing the portion of the processed text string with the selected conversion.

19. The method of claim 18, wherein the step of identifying a phonetically-coded portion corresponding to the selected portion of the processed text string comprises applying a reverse-conversion operation of the phonetically-coded portion string if the phonetically-coded portion is not available on a memory storage device coupled to the computer system.

20. The method of claim 18 wherein the step of displaying the alternative conversions comprises displaying the phonetically-coded portion and the alternative conversions.

21. The method of claim 18 further comprising the step of correcting the error in the processed text string in response to a user specification of correct characters for the conversion.

22. The method of claim 21 wherein the user specification of correct characters comprises a selection of the correct characters from a dictionary program module.

23. The method of claim 21 wherein the user specifies the correct characters comprises inputting the correct characters with a handwriting recognition program module.

24. A computer-readable medium having computer-executable instructions for performing steps comprising:

receiving a phonetically-coded input string representing text for an Asian language;

detecting a non-converted portion of the phonetically-coded input string;

responsive to detecting the non-converted portion of the phonetically-coded input string, converting the detected non-converted portion to a string comprising characters of the Asian language without manual intervention in the event that the detected non-converted portion exceeds a threshold probability of conversion accuracy;

generating a converted text string by replacing the detected non-converted portion with the string comprising Asian language characters; and

displaying the converted text string.

25. The computer-readable medium of 24, having further computer-executable instructions for performing the steps of:

recognizing a reconvert command in response to selection of at least one of the Asian language characters in the converted text string;

identifying a phonetic string corresponding to the selected characters;

generating alternative conversions for the phonetic string corresponding to the selected characters;

displaying the alternative conversions;

recognizing a selection of one of the alternative conversions; and

replacing the selected characters with the selected alternative conversion.

26. The computer-readable medium of 25, having further computer-executable instructions for the steps of:

recognizing an edit of the phonetic string;

generating an alternative conversion list for the edited phonetic string; and

replacing the alternative conversion list for the phonetic string with the alternative conversion list for the edited phonetic string.

27. The computer-readable medium of 25, having further computer-executable instructions for displaying the alternative conversion list in order of decreasing probability.

28. A computer-implemented method for converting a phonetically-coded string into a text string for use by a program module running on a computer system, comprising:

(a) generating a phonetically-coded string comprised of received phonetic characters;

(b) determining a probability of accurate conversion;

(c) if the probability of accurate conversion exceeds a threshold value, converting the string to a text string;

(d) if the probability of accurate conversion does not exceed the threshold value, removing the most recently entered character of the phonetically-coded string; and

(e) comparing the remaining substring to a lowered threshold value, where if the lowered threshold value is exceeded, the remaining substring is converted to a text string.

29. The method of claim 28, wherein if the lowered threshold value is not exceeded, removing the most recently entered character of the phonetically-coded string and repeating (b)-(e).

30. The method of claim 29, wherein if all characters of the remaining substring have been removed, receiving additional phonetic character(s).