Typing Candidate Generating Method for Enhancing Typing Efficiency

• US 20080077396A1
• Filed: 09/27/2006
• Published: 03/27/2008
• Est. Priority Date: 09/27/2006
• Status: Active Grant

• Current Assignee: Academia Sinica (Government of The Republic of China)
• Original Assignee: Academia Sinica (Government of The Republic of China)
• Inventors: Hsu, Wen-Lian
• Granted Patent: US 8,364,468 B2
• Time in Patent Office: Days
• US Class Current: 704/10
• 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 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs

Abstract

For relieving typing burdens caused by incorrect spellings, typing errors, unknown spellings, and characters with diacritical marks, and for enhancing a typing efficiency of a typist with some simplified techniques, several candidate generating methods are provided for assisting the typist to pick a candidate word from a list of generated candidate words, or for selecting a candidate word from said list of generated candidate words in an automatic-selection manner. A proper-designed user interface may also be utilized for implementing the candidate generating methods.

Claims

1. A typing candidate generating method for enhancing typing efficiency, comprising:

providing a word dictionary;

performing an auto-selection method that generates at least one candidate word from the word dictionary for adapting a selected word in a user-typed sentence to match semantic properties of the user-typed sentence according to predetermined statistical data; and

replacing the selected word with the at least one candidate word.

2. The method of claim 1 wherein the auto-selection method further comprises:

calculating an edit distance between the selected word of the user-typed sentence and each of a plurality of candidate words in the word dictionary; and

if the calculated edit distances are all larger than zero, picking a candidate word having a shortest edit distance from the plurality of candidate words and replacing the selected word with the picked candidate word;

wherein the edit distance between the selected word and a corresponding candidate word is a least number of steps for transforming the selected word into the corresponding candidate word by operations including inserting, deleting, and substituting a character.

3. The method of claim 1 wherein the word dictionary is a system dictionary, a domain dictionary, or a user dictionary.

4. The method of claim 1 further comprising:

training and testing the word dictionary with a large plurality of sentences having a plurality of related words for establishing the predetermined statistical data with the large plurality of sentences and words of the large plurality of sentences, wherein the predetermined statistical data is associated with semantic properties of the large plurality of sentences.

5. The method of claim 1 wherein the auto-selection method is utilized according to an N-gram model method or to a linguistic template-matching system.

6. The method of claim 3 further comprising:

listing the plurality of candidate words by compiling user-typed documents; and

determining statistical attributes of the user-typed documents.

7. The method of claim 6 wherein the statistical attributes are a probability of a candidate word being chosen from the user-typed documents, or the probability that a specific word combination appears in the user-typed documents.

8. The method of claim 6 wherein typing errors are reported according to the user-typed documents.

9. The method of claim 6 wherein the typing errors are selected from a group consisting of a misspelled phrase, a misspelled sentence, a character-inserting error, a character-deleting error, and a character-permuting error.

10. The method of claim 3 further comprising:

providing a user interface for a user to select a candidate word from a list of candidate words generated by the auto-selection method.

11. A typing candidate generating method for enhancing typing efficiency, comprising:

(a) providing a phonetic word dictionary;

(b) segmenting an input word into a plurality of phonetic syllable sequences, each of the plurality of phonetic syllable sequences comprising a plurality of phonetic syllables;

(c) generating a plurality of candidate words, the plurality of candidate words comprising at least one candidate word for each of the plurality of phonetic syllable sequences according to a plurality of similarity matrices of the phonetic word dictionary; and

(d) picking a candidate word from the plurality of candidate words and replacing the input word with the picked candidate word, the picked candidate word having a smallest total substitution cost larger than zero of replacing the input word.

12. The method of claim 11 wherein step (a) and (c) further comprise:

storing a plurality of phonetic syllables in the phonetic word dictionary; and

calculating a plurality of similarity distances between each of the plurality of stored phonetic syllables; and

wherein a similarity distance is calculated according to a similarity between a first stored phonetic syllable and a second stored phonetic syllable in the phonetic word dictionary and a substitution cost of substituting the second stored phonetic syllable with the first stored phonetic syllable by substituting, inserting, deleting, and permutating some characters of the first stored phonetic syllable.

13. The method of claim 12 wherein the plurality of similarity distances is all larger or equal to 0, and is all smaller or equal to 1.

14. The method of claim 12 wherein a similarity matrix is a two-dimensional matrix generated for a candidate word and a phonetic syllable sequence;

the plurality of phonetic syllables of the candidate word are attributes of a first dimension of the similarity matrix, and the plurality of phonetic syllables of the phonetic syllable sequence are attributes of a second dimension of the similarity matrix;

elements of the similarity matrix are similarity distances between each of the plurality of phonetic syllables of the candidate word and each of the plurality of phonetic syllables of the phonetic syllable sequence.

15. The method of claim 14 wherein the total substitution cost of replacing the input word with the candidate word is determined by calculating a determinant of the similarity matrix relating to the input word and the phonetic syllable sequence.

16. The method of claim 14 wherein the total substitution cost of replacing the input word with the candidate word is determined by adding related similarity distances in the similarity matrix relating to the input word and to the phonetic syllable sequence.

18. The method of claim 14 further comprising:

providing a user interface;

wherein the user interface is utilized for performing steps (a), (b), (c), and (d).

19. The method of claim 14 further comprising:

performing steps (b), (c), and (d) in an auto-selection manner.

17. The method of claim 111 wherein the input word is inputted according to phonetic spelling.

20. A typing candidate generating method for enhancing typing efficiency, comprising:

providing a word dictionary;

determining if at least one candidate word exists, wherein the at least one candidate word is a word in the word dictionary that comprises an input word as a subsequence of the at least one candidate word and found anywhere within the at least one candidate word; and

selecting the at least one candidate word to replace the input word if it was determined that the at least one candidate word exists.

21. The method of claim 20 wherein the word dictionary comprises URLs or e-mail addresses or both URLs and e-mail addresses.

22. The method of claim 20 further comprising generating a list of candidate words if it is determined that a plurality of the at least one candidate word exists and selecting the at least one candidate word to replace the input word from the list.

23. The method of claim 22 wherein a user interface is utilized for selecting from the list.

24. The method of claim 22 wherein selecting from the list is performed in an auto-selection manner.

25. A typing candidate generating method for enhancing typing efficiency, comprising:

storing a word sequence; and

retrieving a word subsequence of the word sequence utilizing an acronym subsequence comprising a first character of one word and a first character of a subsequent word of the word sequence ordered in the order of corresponding words in the word sequence;

wherein the word subsequence comprises the one word, the subsequent word, and all intervening words between the one word and the subsequent word and ordered in the order of corresponding words in the word sequence.

26. The method of claim 25 wherein the acronym subsequence further comprises a first character of at least one of the intervening words wherein all characters of the acronym subsequence are ordered in the order of corresponding words in the word sequence.

27. The method of claim 25 further comprising generating a list of retrieved word subsequences to allow selection from the list utilizing a user interface if more than one word subsequence from one or more word sequences are retrieved utilizing a single acronym subsequence.

28. The method of claim 25 further comprising generating an acronym sequence corresponding to the word sequence, the acronym sequence comprising a first character of each word in the word sequence ordered in the order of corresponding words in the word sequence.

29. The method of 27 wherein generating the list of retrieved word subsequences to allow selection from the list comprises:

generating the list in an auto-selection manner.

30. A typing candidate generating method for enhancing typing efficiency, comprising:

(a) providing a word dictionary;

(b) generating at least one candidate word according to possible diacritical marks of a character in an input word; and

(c) picking a candidate word from the at least one candidate word;

wherein a plurality of possible diacritical marks of characters for forming a word is stored in the word dictionary.

31. The method of claim 30 wherein step (b) further comprises:

referring to semantic word combinations stored in the word dictionary and related to a plurality of possible diacritical marks.

32. The method of claim 30 further comprising:

providing a user interface;

wherein the user interface is utilized for performing steps (a) and (b).

33. The method of claim 30 further comprising:

performing steps (b) and (c) in an auto-selection manner.