DISAMBIGUATION OF TELEPHONE STYLE KEY PRESSES TO YIELD CHINESE TEXT USING SEGMENTATION AND SELECTIVE SHIFTING

• US 20080235003A1
• Filed: 03/18/2008
• Published: 09/25/2008
• Est. Priority Date: 03/22/2007
• Status: Active Grant

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: Tegic Communications, Inc. (Microsoft Corporation)
• Inventors: Lai, Jenny Huang-Yu, Wu, Jianchao
• Granted Patent: US 8,103,499 B2
• Time in Patent Office: Days
• US Class Current: 704/8
• CPC Class Codes: G06F 3/018 Input/output arrangements for oriental characters
• CPC Class Codes: G06F 3/0237 using prediction or retrieval techniques

Abstract

To enter Chinese text, a user enters the corresponding phonetic spelling via telephone style keypad. Some or all keys represent multiple phonetic letters. In disambiguating entered key presses to yield a valid phonetic spelling, a computer divides the key presses into segments, while still preserving key press order. Each segment must correspond to an entry in a dictionary of Chinese characters, character phrases, and/or character components such as radicals or other predetermined stroke groupings. Upon arrival of a new key press that cannot form a valid entry when appended to the current segment, key presses are incrementally reallocated from the previous segment. As for already-resolved segments occurring prior to the previous and current segments, these are left intact. After each shifting attempt, the computer reinterprets key presses of the last two segments, and accepts the new segmentation if the segments form valid dictionary entries.

Claims

1. A process to facilitate user entry of Chinese logographic text via computer having a display and a multi-key keypad, the process comprising operations of:

the computer receiving user entry of a series of key presses representing a corresponding series of letters collectively spelling-out an intended series of one or more phonetic entries from predetermined dictionary, where each dictionary entry represents a phonetic representation of one or more Chinese characters;

where the key presses are inherently ambiguous because, according to a predetermined mapping, at least some of the specific keys represent multiple letters;

during the course of receiving the series of key presses, the computer performing operations comprising;

interpreting the series of key presses to identify a corresponding series of dictionary entries spelled out by the series of key presses;

maintaining and displaying a collection buffer containing the series of dictionary entries according to a current interpretation;

presenting Chinese characters corresponding to at least a first dictionary entry of the collection buffer;

maintaining and displaying an output buffer accumulating user selections of presented Chinese characters, in order of selection;

where the key presses whose interpretation formed a final dictionary entry in the collection buffer constitute a current segment, and key presses whose interpretation formed a next-to-last dictionary entry in the collection buffer constitute a previous segment;

where the interpreting operation includes, in response to each new key press, the computer responding by performing operations as follows;

the computer identifying all possible strings of letters spelled out by the key presses of the current segment appended by the new key press, and determining which if any identified string forms a recognized entry in the dictionary;

upon success of the determining operation, changing the final entry in the collection buffer to one of the recognized entries;

upon failure of the determining operation, shifting key presses one-by-one from the previous segment to the current segment appended by the new key press, and with each shift, performing operations comprising;

for each of the previous segment as shifted and the current segment appended by the new key press as shifted, performing a test to identify all possible strings of letters spelled out by respective segment, and determining which if any of the identified strings forms a recognized entry in the dictionary;

performing one of the following;

as soon as the test determines that letters of the previous segment and the current segment appended by the new key press respectively spell out recognized entries in the dictionary, discontinuing further shifting and revising the collection buffer to present said recognized entries;

continuing the shifting to exhaust all shifting possibilities, evaluating the recognized entries against a predetermined quality popularity criteria, and for each of the current and previous segments as shifted choosing a recognized entry preferred by the criteria, and revising the collection buffer to preferred entries for the segments.

2. A computer-driven process to facilitate user entry of logographic Chinese text, comprising operations of:

receiving user entry of an ordered sequence of key presses representing an intended series of letters collectively spelling-out a sequence of one or more entries occurring in a predetermined dictionary, where each dictionary entry represents the phonetic spelling of one or more of the following items according to a predetermined system for phonetically representing logographic Chinese text;

a Chinese character, a set of multiple Chinese characters, a radical or other grouping of multiple Character strokes;

where user intention as to the intended series of letters is ambiguous because some or all of the key presses represent multiple phonetic letters;

interpreting the sequence of key presses by allocating key presses in the sequence into contiguous segments such that letters of each segment'"'"'s key presses spell a entry occurring in the dictionary, where a segment being appended by new key presses is a current segment, and where a next-earlier segment to the current segment in the sequence forms a previous segment;

where the interpreting operation comprises;

responsive to each new key press, determining whether if the new key press were to be appended to a trailing end of the current segment to form a new segment, letters of key presses of the new segment would by spell out any entries recognized by the dictionary;

if the determining operation succeeds, presenting an output of one or more of the entries spelled out by the key presses of the new segment, and deeming the new segment to be the current segment;

if the determining operation fails, appending the new key press to a trailing end of the current segment and attempting to shift key presses from a trailing end of the previous segment to a leading end of the current segment until letters of the key presses of both the current segment and the previous segment spell out respective entries occurring in the dictionary;

limiting performance of said attempt to shift key presses to the current segment and the previous segment, and leaving intact all segments occurring earlier than the previous segment in the sequence of key presses;

if said attempt to shift key presses succeeds, presenting an output of one or more of said entries spelled out by the key presses of the previous segment as shifted and the current segment as shifted;

presenting an output of logographic Chinese text represented by one or more entries spelled by key presses of the segments into to which the sequence of key presses have been allocated.

3. The process of claim 2, where:

said operation of, if the determining operation succeeds, presenting an output of one or more of the entries spelled out by the key presses of the new segment, further comprises;

if multiple entries are spelled out, evaluating the multiple entries against predetermined criteria as to use history to identify a preferred entry, and limiting the presented output to the preferred entry;

said operation of, if said attempt to shift key presses succeeds, presenting an output of one or more of said entries spelled out by the key presses of the previous segment as shifted and the current segment as shifted, further comprises;

if multiple entries are spelled out, evaluating the multiple entries against predetermined criteria as to use history to identify a preferred entry, and limiting the presented output to the preferred entry.

4. The process of claim 2, where if said attempt to shift fails, rejecting the new key press and reverting the current segment to its contents before being appended by the new key press and undergoing said shifting.

5. The process of claim 2, where if said attempt to reallocate fails, performing operations comprising:

checking whether key presses of the current segment before being appended by the new key press and undergoing said shifting formed any valid entries from the dictionary;

if so, defining as the previous segment the contents of the current segment before being appended by the new key press and undergoing said shifting, and defining as the current segment the new key press alone;

if not, rejecting the new key press and reverting the current segment to its content before being appended by the new key press and undergoing said shifting.

6. The process of claim 2, the attempting shifting comprising:

considering all possible shifts of key presses, and choosing a resulting allocation as between previous segment and current segment that produces spelled out respective entries of highest ranking according to predetermined past criteria as to historical use.

7. The process of claim 2, the attempted shifting comprising:

shifting key presses one-at-a-time until conditions comprising the following are met;

the key presses of both the current segment and the previous segment spell out respective entries recognized by the dictionary; and

both of the respective entries satisfy predetermined criteria as to historical use.

8. The process of claim 2, where the operation of limiting performance comprises:

limiting said attempted shifting to the current segment, the previous segment, and one segment preceding the previous segment.

9. The process of claim 2, further comprising:

presenting a collection buffer including, for each segment into which the sequence of key presses has been allocated, a preferred one of said recognized entries spelled out by the key presses of said segment;

only for a first one of the segments in order of the sequence, performing operations comprising;

presenting a selection list including some alternative dictionary entries spelled out by key presses of the first segment;

presenting a list of some logographic Chinese text items phonetically represented by the preferred recognized entry corresponding to the first segment;

responsive to receiving user selection of one of the presented logographic Chinese text items, entering the selected item into a displayed output buffer and removing from the collection buffer the letters corresponding to the first segment;

responsive to receiving user selection of an entry from the selection list, revising the collection buffer to remove the preferred recognized entry for the first segment and to add the selected entry, and revising the list of logographic Chinese text items to present logographic Chinese text items representing the selected entry.

10. The method of claim 9, further comprising identifying the preferred entries by applying predetermined criteria as to historical use.

11. A computer-driven process to facilitate user entry of logographic Chinese text, comprising operations of:

receiving user entry of an ordered sequence of key presses representing an intended series of letters collectively spelling-out a sequence of one or more entries occurring in a predetermined dictionary, where each dictionary entry represents the phonetic spelling of one or more Chinese characters or character sets;

where user intention as to the intended series of letters is ambiguous because some or all of the key presses represent multiple phonetic letters;

interpreting the sequence of key presses by allocating key presses in the sequence into contiguous segments such that letters of each segment'"'"'s key presses spell a entry occurring in the dictionary, where a segment being appended by new key presses is a current segment, and where a next-earlier segment to the current segment in the sequence forms a previous segment;

where the interpreting operation comprises;

responsive to each new key press, determining the following;

if the new key press were to be appended to a trailing end of the current segment to form a new segment, whether letters of key presses of the new segment would by spell out any entries recognized by the dictionary;

if the determining operation succeeds, identifying a preferred one of all entries spelled out by the key presses of the new segment, presenting an output of the preferred entry, and deeming the new segment to be the current segment;

if the determining operation fails, appending the new key press to a trailing end of the current segment and attempting to shift key presses from a trailing end of the previous segment to a leading end of the current segment until letters of the key presses of both the current segment and the previous segment spell out respective entries occurring in the dictionary;

limiting performance of said attempt to shift key presses to the current segment and the previous segment, and leaving intact all segments occurring earlier than the previous segment in the sequence of key presses;

if said attempt to shift key presses succeeds, identifying a preferred one of all entries spelled out by the key presses of the previous segment as shifted, and identifying a preferred one of all entries spelled out by the key presses of the current segment as shifted, and presenting an output of the identified preferred entries;

continually presenting a collection buffer showing, for each segment into which the sequence of key presses have been allocated, a preferred entry identified for that segment.

12. The method of claim 11, further comprising:

only for a first one of the segments in order of the sequence, performing operations comprising;

presenting an alternatives-selection-list including some alternative dictionary entries spelled out by key presses of the first segment;

presenting a list of some logographic Chinese text items phonetically represented by the preferred entry corresponding to the first segment.

13. The method of claim 12, further comprising:

responsive to receiving user selection of one of the presented logographic Chinese text items, entering the selected item into a displayed output buffer and removing from the collection buffer the preferred entry spelled out by key presses of the first segment.

14. The method of claim 12, further comprising:

responsive to receiving user selection of an entry from the alternatives-selection-list, revising the collection buffer to remove the preferred entry spelled out by key presses of the first segment;

revising the list of logographic Chinese text items so as to present logographic Chinese text items representing the use selected entry.