CN104834392B - A kind of Chinese character input method of stroke dynamic group word - Google Patents

Abstract

The invention discloses a kind of Chinese character input method of stroke dynamic group word, methods described is during using input through keyboard keystroke sequence, pass through the searching algorithm based on Controlling UEP, matching word fitting algorithm and component sequence fitting algorithm, it will it is expected that Chinese character includes in viewing area by order of writing strokes, from part to entirety, immediately, so as to simulate stroke dynamic combined into the process of Chinese character, and form visual feedback.The present invention realize it is a kind of needed for code element is few, be easy to portable application, efficient Chinese character input method.

Description

A kind of Chinese character input method of stroke dynamic group word

Technical field

The present invention relates to a kind of information processing Chinese character input method, more particularly to a kind of Chinese character of stroke dynamic group word are defeated Enter method.

Background technology

(1) Chinese character entering technique Brief on Development

Since being come out from electronic computer technology, Chinese character entering technique experienced multiple innovation.

【1】Hanzi coding input method

Input generation of computers code is easy in so-called encoding of chinese characters, exactly one kind for Chinese character design, general to use coding rule Then describe, and be specifically reflected as keyboard layout and code table (Chinese character and the corresponding table of encoding of chinese characters).Hanzi coding input method, Searching and the method for inputting Chinese character are exactly carried out to code table using the keyboard layout.

Encoding of chinese characters is divided into shape code, tone code, shape tone code, phonetic-stroke code etc..The representative of shape code has：

1) Cangjie's code：First it is born in invention in 1976 again by Zhu nation, it is rear to turn into the main flow shape code in complex form of Chinese characters region；

2) king's code (the Five-stroke Method)：Nineteen eighty-three invention is first born in by Wang Yongmin, turns into simplified Chinese character main flow shape code afterwards, pole is propagated To be extensive；

3) Zheng's code：Invention in 1989 is first born in Zheng Yi, is also used widely；

4) the fast codes of Wang Lin：Invention in 1997 is born in by Wang Linxian, phonetic-stroke code or shape tone code is also also served as, now still has fan to make With.

5) stroke input method：By the U.S., special prompt communication (Tegic Communications) proposed in 1999, was earliest For the T9 input methods of mobile phone, using 5 kinds of strokes, propagate extremely wide (hereinafter referred to as common C5 stroke coding).Separately there is Chinese Normalizing input method, is born in invention in 2006 by Li Tiexian, uses 26 kinds of strokes；Six code stroke input methods, it is first by Hong Kong Bu Liwen Invention in 2007 is born in, using 5 kinds of strokes, totally 6 yards of 3 ends 3 headed by coding is fixed.

Note：According to State Language Work Committee's issue in 1997《Information processing GB13000.1 character set Hanzi component specifications》(with Lower abbreviation " part specification "), stroke is defined as：Constitute a minimum even unit for regular script Chinese character pattern.The form of a stroke or a combination of strokes is defined in addition Concept：The form of a stroke or a combination of strokes is the concrete shape of stroke, and the form of a stroke or a combination of strokes of regular script Chinese character can be divided into tens of kinds, but its most base according to the style of brushwork and trend Five classes such as this classification is one, Shu, Pie, Dian, Ya.Stroke described in this specification, has stroke and the form of a stroke or a combination of strokes defined in " part specification " concurrently Concept.

Almost contemporaneity tone code occurs with shape code, and representative has：

1) phonetic code：Include the schemes such as spelling, simplicity, Two bors d's oeuveres again, be mainly used in simplified Chinese character region；

2) phonetic notation code：It is mainly used in complex form of Chinese characters region.

The contemporaneity occurred in above-mentioned encoding of chinese characters, other also substantial amounts of miscellaneous encodings of chinese characters or software layer Go out not poor, the situation of formation a kind of " ten thousand yards of Pentium ".However as the popularization of the equipment such as computer and mobile phone, five, phonetic, pen Draw this several coding method and be increasingly becoming main flow, wherein being extensive especially using phonetic.

【2】Chinese Card and dynamic group word, word-base technology

At the beginning of computer technology comes out, due to computer memory space, the limitation of operational capability, Chinese character entering technique to Two different directions are developed：

1) Chinese Card technology：Chinese Card is that Chinese character input method, Hanzi font library and its driver are cured as one only by one kind Read the computer extender card of memory.Earliest Chinese Card is Legend Company in association's Chinese Card of exploitation in 1986.After nineteen ninety-five, With the lifting of computer capacity, the technology such as UCDOS and Windows operating system makes Chinese Card be gradually backed out the arena of history.

2) dynamic group word technology：It is coding theory and technology of a kind of Chinese character in fields such as computers, independent of occupancy The character library of much room, it is only necessary to a small amount of part, instant " two dimension group word " is done using programme-control, is shown in a recursive manner Show on display.The example of " dynamic group word " technology includes：" the Cangjie's system " of Mr. Zhu Bangfu, Unicode IDS language, The CDL fonts description language of company of galaxy of literary talent of the U.S., " dynamic group word " theory of Mr. Zhang Shizhao etc..With the extension of character library, temporarily When, partly solve the not enough difficulty of Chinese character pattern, but dynamic group word technology is still with the presence of its value and significance.Dynamic Group word technology the characteristics of be：The X-Y scheme of Chinese character is immediately formed, therefore is the skill for surmounting general Hanzi coding input method Art, and the latter is substantially the retrieval of Chinese character.

Character library is a set being made up of the Chinese character of certain amount, font inequality, and information processing experienced many with character library It is secondary to expand, usually mainly there is following level：

1) GB2312-80 Chinese Character Set Code for Informatis baseset：Include 6763 Chinese characters (wherein one-level Chinese Word 3755, the Chinese characters of level 2 3008) (hereinafter referred to as 6763 Chinese Character Sets).

2) GB 13000.1-93 information technologies universal multiple-octet coded character set (UCS) Part I：Architecture and base This multi-lingual plane：Include 20902 Chinese characters (containing GB2312) (hereinafter referred to as 20902 Chinese Character Sets).

3) GB 18030-2005 information technologies Chinese character code character set：Include 70244 Chinese characters (containing GB 13000.1, Extend A6582 words, extension B42711 words, extension C) (hereinafter referred to as 70244 Chinese Character Sets).

【3】Other new technologies

In recent years since, with the development of computer technology, some advanced Chinese character entering techniques have been occurred in that, such as：

1) handwriting input：Refer to the orderly trace information produced when writing on handwriting equipment is converted into the Chinese by computer The technology of word ISN；

2) phonetic entry：Refer to that (also known as acoustic control is defeated into the input technology of Chinese character the speech recognition of user by computer Enter).

These new technologies normally only just have advantage under specific occasion, thus not as good as Hanzi coding input method application Extensively.Also some new technologies, such as cloud storage technology, search technique, whole sentence input technology, enhance encoding of chinese characters input The ease for use of method.

【4】Brief summary and prospect

The development of Chinese character entering technique is taken an overall view of, Hanzi coding input method has the advantages that other technologies are incomparable, from Leading position is occupied at the very start.But even as main flow encoding of chinese characters also all also exist some it is unsatisfactory it Place, such as phonetic has the shortcomings that the repetition rate of coding is high, can cause to forget what to write when actually writing, regional difference substantially (such as Guangdong language, the Wu dialect)；And five, Either the high-leveled and difficult memories of shape code efficiency such as picture, or good memory effect is low.

Because Chinese character is the ideograph that carries thousands of years of civilizations, in order to continue this tradition, it is to avoid Chinese character quilt Alphabetic writing replaces, and the effort of the coding method of more easy-to-learn high-efficiency of people's exploitation based on shape code does not stop yet so far.

(2) general principle of Hanzi coding input method

As shown in figure 1, the general principle of Hanzi coding input method is summarized as follows：

1) inscape of Hanzi coding input method：

A) coding rule：Code element and code fetch mode are defined, and is specifically reflected as keyboard layout and encoding of chinese characters table；

B) man-machine interface：Including keyboard, candidate word region, it is text filed (usually, keyboard and it is text filed be independently to deposit , it is not belonging to Hanzi coding input method, but necessary external environment condition)；

C) searching algorithm：It is keystroke sequence that it, which is inputted, and encoding of chinese characters table is inquired about, candidate list of words is output as.

2) general flow of Hanzi coding input method is as follows：

A) user carries out button operation to keyboard, generates keystroke sequence；

B) searching algorithm inquiry button sequence in encoding of chinese characters table, draws candidate list of words, and be shown in candidate block Domain；

C) user retrieves to candidate list of words again, it is determined that expecting Chinese character and to be shown in it text filed；

D) user obtains text filed visual feedback, completes to expect the input of Chinese character.

Different Hanzi coding input methods, it is general only to be distinguished on coding rule, keyboard layout, encoding of chinese characters table, Other are then similar.When being implemented as software, some Hanzi coding input methods can be spelled in text filed instant playback Write dotted portion in text (including coded sequence or matching word), i.e. Fig. 1.

(3) weak point that existing shape code is present

It is preceding to have addressed, it is easy to learn due to the defect natural in terms of Chinese-character writing tradition is passed on of the tone code using phonetic as representative Efficient shape code is the main target of encoding of chinese characters.So-called efficient, one refers to that input Chinese character quantity is more in the unit interval, and two refer to It can realize that (typically finger does not see keyboard always more, and by instructing button to grasp from the text filed visual feedback obtained for touch system Make), it is easy to learn and be efficiently the contradiction for being difficult to reconcile for a pair but for shape code.

Existing shape code has the following disadvantages：

【1】Code element is more, hardly possible memory

The code element (i.e. font unit) of shape code, by quantity from being more to roughly divided into following several levels less：

1) part：According to " part specification ", part is defined as：There is the structure word list for assembling Chinese word function by what stroke was constituted Position, wherein basis of Chinese character part are 560.Main flow shape code such as king's code, Cangjie, Zheng's code etc. are all using approximate number tuber of stemona part (or word Root, code element)；

2) stroke combination：About 25, typical case is that five kinds of basic strokes take 2 fully intermeshings such as Erbi input method；

3) stroke：According to " part specification ", stroke have 5 classes (one, Shu, Pie, Dian, Ya), as many as tens of kinds, actual use is from 5 To 26 (Chinese normalizing input method).

The shape code of main flow generally directed to alphabetic keypad (ISO/IEC 9995-1 include 26 alphabet keys) optimization design, Also it is at least in 26 or so, hundreds of symbol allocations to 26 keys using the quantity of button, the burden of memory can be brought to user, it is difficult To learn in a short time.User have first with for it is main the characteristics of, i.e., once learned certain shape code, be difficult to receive others again Shape code, even a kind of more excellent shape code, therefore a kind of new shape code is difficult often popularization and application.And if user is shape Code shelves a period of time after, it is easy to forgets, and once forgets, sense of frustration will be brought to user, makes user Abandon the wish relearned, and then leave shape code to find it difficult to learn the impression of difficult note.These are all the natural obstacle of shape code popularization and application.

【2】Code element is more, difficult portable application

Portable equipment typically carries numeric keypad (ISO/IEC 9995-4 include 10 digital keys) or telephone keypad (ISO/IEC 9995-8 include 10 digital keys and *, #), or with can Touch Screen, operated using hand or hands. It is difficult to for the code input method of standard alphabet keyboard optimization design more symbol allocation to digital keys, even if can touch Virtual alphabetic keypad is realized on control screen, also due to portable equipment size is limited, virtual key very little, and reduce easy-to-use Property：User can not the touch system as on alphabetic keypad, even if carefully touching virtual key, also very easily push the wrong, cause defeated Enter efficiency to decline to a great extent.

In fact, including spelling input method, the input method of every use alphabetic keypad, when for portable equipment, All there is problem above, thus the touch-typing operation as alphabetic keypad can not possibly be realized.

【3】The repetition rate of coding is high, and input efficiency is low

For the input of individual Chinese character, tone code repetition rate of coding highest, about 97.1%.The repetition rate of coding of all kinds of shape codes is below sound Yard, the main shape code such as the Five-stroke Method, Zheng's code, the repetition rate of coding about 5% in 6763 Chinese Character Sets, but Chinese character scope is expanded as After 20902 Chinese Character Sets, the repetition rate of coding can bring up to 30% or so.Because, this kind of shape code is in order to improve the usual pin of input efficiency Chinese characters in common use have been done with optimization design, but this optimization is not all pervasive to all Chinese characters.

The repetition rate of coding of common C5 stroke coding is 7.4% for 6763 Chinese Character Sets, is 9.3% for 20902 Chinese Character Sets, can See that the repetition rate of coding because being significantly improved the increasing of Chinese character quantity, does not have certain universality to all Chinese characters.But its input effect Rate is relatively low, mainly due to following reason：

1) coded sequence is long：The length of common C5 stroke coding is stroke number, and the average stroke number of 6763 Chinese characters is 10.6, The average stroke number of 20902 Chinese Character Sets is 12.8；

2) candidate word recall precision is low：Candidate word is typically simple by word frequency sequence, do not account for by the degree of correlation and With degree sequence, it is intended that to improve input efficiency, and the most frequently used word is preferentially shown, but can so reduce most of The recall precision of relatively low word frequency word.User is actually doing " retrieval " rather than " input " of Chinese character, thus also brings input not Smoothly experience.

3) visual feedback effect is poor：Occur the more degree of correlation and the relatively low information of matching degree in candidate list of words, input Strokes sequence is shown with one-dimensional linear array, it is impossible to effective visual feedback is constituted between the input of user and the output of display Loop.Especially when being generation input error in strokes sequence, it is impossible to soon recognize wrong stroke and correct.

Therefore, although the common C5 stroke coding repetition rate of coding is not comparatively high, input efficiency is still relatively low.

The content of the invention

In order to solve the above technical problems, it is an object of the invention to provide a kind of Chinese character input method of stroke dynamic group word, The advantage of common 5-stroke coding and " dynamic group word " technical concept are combined realization by this method.

The purpose of the present invention is realized by following technical scheme：

A kind of Chinese character input method of stroke dynamic group word, methods described is in the process using input through keyboard keystroke sequence In, pass through the searching algorithm based on Controlling UEP, matching word fitting algorithm and component sequence fitting algorithm, it would be desirable to which Chinese character is pressed Order of writing strokes, from part to it is overall, viewing area is immediately shown in, so as to simulate stroke dynamic combined into the mistake of Chinese character Journey, and form visual feedback.

Compared with prior art, one or more embodiments of the invention can have the following advantages that：

Solve many general code code elements, difficult memory, be difficult to the deficiency of portable application, while there is the low repetition rate of coding and efficient Rate.

【1】Code element is few, is easy to memory

Using only 12 kinds of code elements, in addition to five kinds of conventional basic strokes, user only needs seven kinds of extension strokes of memory, and all seven It is common typical, representative stroke to plant extension stroke, is easily discernible and remembers.

【2】Code element is few, is easy to portable application

General portable equipment carries digital keys or telephone keypad.12 kinds of code elements can be directly used for numeric keypad and electricity Talk about keyboard, for can Touch Screen portable equipment, 6 key dummy keyboards can be further simplified as.

【3】The repetition rate of coding is low, and input efficiency is high

According to statistics, 20902 Chinese characters are concentrated with about 700 words can not distinguish from stroke form completely, account for the 3.3% of sum, Li such as cen ridges, denier and mesh, first Shen, oneself in the sixth of the twelve Earthly Branches etc..Therefore 3.3% is the minimum repetition rate of coding that all stroke codings can reach. Because 12 code element stroke codings do not differentiate between horizontal and carry, put and press down, repeated code number of words is increased slightly, and is 858 words, and the repetition rate of coding is 4.1%, than the five kinds basic stroke coding repetition rates of coding have dropped 57%.

【4】Improve input efficiency

1) touch potential is reduced：By using Lian Bijian, substantially reduce the touch potential of long stroke Chinese character input.According to system Meter, the average stroke number of 6763 Chinese Character Sets is 10.6, after Lian Bijian, according to the preliminary statistics, and Chinese character average key press number is 6 keys, effect Rate improves 40%.By the Lian Bijian stroke positions replaced and quantity can flexibly change, therefore improve ease for use.

2) candidate's search words is automated：Algorithm is all given the retrieval work of most of candidate word to be automatically performed, only 4.1% repeat code Chinese character input, just needs user's manual retrieval's candidate list of words.Remaining about 21% Chinese character is inputted, it is only necessary to User confirms to complete rather than to candidate's search words, such as " spring " after the input of keystroke sequence (containing Lian Bijian) is completed, in input After the completion of, candidate word includes the word such as spring, stupid, and user is not required to retrieve candidate word, directly confirms to complete, you can input " spring ", without showing that the candidate list of words including " stupid " waits user's selection to user.Finally about 75% Chinese character is defeated Enter, after the input that user completes keystroke sequence (containing Lian Bijian), automatically confirm that completion.Therefore, conventional 100% candidate word inspection Rope process, be reduced to 4.1% repeated code search words, substantially increase ease for use.

3) enhanced visual feedback, expectation Chinese character by order of writing strokes, from part to it is overall, to be immediately shown in needs defeated Enter the text filed of Chinese character, enhance visual feedback.User not exclusively ultimately forms from the text filed visual feedback obtained Expectation Chinese character, in addition to turn into the process of Chinese character from stroke combination, thus obtain the experience similar with writing Chinese characters.User The correctness of Chinese character can not only be confirmed to expect, additionally it is possible to the correctness of input is progressively confirmed in input process, thus can be with Text filed display is paid close attention to always, so as to realize touch system.

Brief description of the drawings

Fig. 1 is prior art Hanzi coding input method theory diagram；

Fig. 2 is stroke dynamic group character method for inputting Chinese characters theory diagram；

Fig. 3 is the searching algorithm based on Controlling UEP；

Fig. 4 is matching word fitting algorithm；

Fig. 5 is Controlling UEP algorithm；

Fig. 6 is matching degree parser；

Fig. 7 is component sequence fitting algorithm；

Fig. 8 is progressive expression formula algorithm；

Fig. 9 is alphabetic keypad layout；

Figure 10 is numeric keypad layout；

Figure 11 is telephone keypad layout；

Figure 12 is virtual keyboard layout.

Embodiment

To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with embodiment and accompanying drawing to this hair It is bright to be described in further detail.

As shown in Figure 2 there is provided a kind of Chinese character input method of stroke dynamic group word, this method is to use input through keyboard During keystroke sequence, calculated by the searching algorithm based on Controlling UEP, matching word fitting algorithm and component sequence fitting Method, it would be desirable to Chinese character by order of writing strokes, from part to it is overall, viewing area is immediately shown in, so as to simulate stroke dynamic The process of Chinese character is combined into, and forms visual feedback.

The above-mentioned searching algorithm based on Controlling UEP is by the 12 code elements pen in the keystroke sequence and 12 code element word tables Draw coding and do Controlling UEP, the character of the degree of correlation >=1 in 12 code element word tables is added to candidate list of words；When 12 code element word tables In be not present the degree of correlation >=1 character when, produce miscue.

Above-mentioned matching word fitting algorithm is that the keystroke sequence and candidate list of words done into matching degree analysis, and it is described can The character of matching degree >=3 in candidate list of words described in viewed area instant playback, automatically confirms that matching degree in the candidate list of words =4 character；When the matching degree maximum of character in the candidate list of words is 1, component sequence fitting algorithm is transferred to；Waiting The character of matching degree >=2 and total quantity more than 1 in candidate list of words described in the instant playback of word selection region.

(1) 12 code element stroke coding

12 code element stroke codings, refer to encode Chinese character by pen by order of writing strokes using 12 kinds of code elements.

【1】12 code elements are corresponding with accurate stroke

About 38 kinds of strokes used in Chinese character are classified as 12 kinds of code elements mutually distinguished by the present embodiment, and every kind of code element is with one Typical stroke is named, and correspondence several stroke, distributes a symbol code.12 kinds of code elements are divided into two classes, and a class is base This stroke, totally 5 kinds；One class is extension stroke, totally seven kinds.Code element is shown in Table 1 with 38 kinds of strokes, corresponding relations of symbol code：

Table 1

In above-mentioned table,For regular script " heart " distinctive stroke, Unicode 31C0 CJK Strokes are (hereinafter referred to as Unicode) include,《GF2001-2001 GB13000.1 character set Chinese character folded pen specifications》(hereinafter referred to as Chinese character folded pen specification) Do not include；

〇：Unicode is included, and Chinese character folded pen specification is not included；Stroke 〇 only sees a small number of Chinese characters, such as 〇,(being all not belonging to 20902 Chinese Character Sets).

Fly the 1st of word, Unicode is included, Chinese character folded pen specification is with " second ".

ㄣ：The 3rd of special word, Ze words it is the first sum of, Chinese character folded pen specification is included, and Unicode is not included；

I：Unicode, Chinese character folded pen specification are not included.

【2】Complete character set

The encoding of chinese characters table of the present embodiment corresponds to a complete character set, typically by a number of inequality element group Into so-called complete to refer to：The element that any one stroke number in the complete character set is more than 1 can be with the set Other elements be combined into come.

Because the quantity of Chinese character is not unlimited many, complete character set is clearly defined, it is necessary to define a base first This character set, it is generally the maximum character library that information processing system can be supported, such as 20902 Chinese Character Sets, or 70244 Chinese Character Sets.

Complete character set specifically includes following three classes subset：

1) Chinese Character Set is applied：It is the necessary Chinese character of information processing system application using Chinese character, generally refers to conventional written It is a nonvoid subset of basic Chinese characters collection, for example using involved Chinese character：The collection of 20902 Chinese Character Sets, all simplified Hanzis Close, the set of 3755 first-level Chinese characters, input required for certain information processing system are possible to the gathering of Chinese character, all place names Set of Chinese character etc..

2) part of application Chinese character：If having included some Chinese character using Chinese Character Set, the part of the Chinese character must be included, And the part of these parts is constituted, these parts may belong to basic Chinese characters collection, it is also possible to be not belonging to basic Chinese characters collection, example Such as：If using Chinese character " slender and delicate " has been included in Chinese Character Set, its part must be included" clothing ", and " clothing " is included simultaneously Part " Tou ",

3) stroke of application Chinese character：If having included some Chinese character using Chinese Character Set, all of the Chinese character must be included Stroke, these strokes are likely to belong to basic Chinese characters collection, it is also possible to be not belonging to basic Chinese characters collection, for example：If using Chinese character Concentration has included Chinese character " slender and delicate ", then must include its stroke " Pie ", "  ", " Dian ", " ㄅ ", " one ", "-",

As described above, being only that its complete character set also must at least include following for an application Chinese character " slender and delicate " Character：Slender and delicate Pie Bao spoonfulsClothing Tou Dian ㄅ mono--

Character outside the basic Chinese characters collection included for complete character set, should add to the word of information processing system In storehouse, so as to normal display.

【3】Expression formula

In the complete character set, any one stroke number be more than 1 element all can according to stroke order without interlock, font not Become principle to be split as at least including the sequence of two characters, these sequences are referred to as expression formula, wherein each character is also described The element of complete character set.Arrange simultaneously, in the complete character set stroke number be equal to the expression formula of 1 element for its from Body.

1) most simple expression formula

Same element may have multiple expression formulas, and wherein length is most short, can most save the expression formula of font style characteristic from damage, be referred to as Most simple expression formula.For example：" slender and delicate " can be split as "Clothing ", "Tou", " spoon ㄅ clothing ", " Bao Dian ㄅ Tou", " Pie  Dian ㄅ Dian One Pie-Pie" etc., designated length it is most short "Clothing " is the most simple expression formula of " slender and delicate ".In another example：" dry " can be split as " two Shu ", " 10 ", both length are all 2, and still " 10 " have saved the font style characteristic that stroke " one " " Shu " intersects in " dry " word from damage, Thus selected " 10 " are the most simple expression formula of " dry ".

Most simple expression formula be generally different from Chinese character have motivation split (original disassembly), for example, " name for ancient tribes in the east " Most simple expression formula be " one bow people ", rather than " longbow "；The most simple expression formula of " state " is " ㄇ jade one ", rather than " mouth is beautiful ".

2) progressive expression formula

For the element that some specific stroke number is M (M >=1), and given positive integer N (1≤N≤M), there is one Individual progressive expression formula, is defined as follows：Chosen in most simple expression formula since the 1st character, total stroke number is suitably larger than equal to N Subsequence, and more than situation when, constantly repeat " last character in the subsequence is replaced with the character Most simple expression formula, choose again since the 1st character, total stroke number is suitably larger than subsequence equal to N " operation, it is final The total stroke number gone out is equal to N subsequence.Progressive expression formula algorithm is as shown in specification Fig. 8.

As N=M, the progressive expression formula of the element is its own.

As N=1, the progressive expression formula of the element is its 1st stroke.

For example, the progressive expression formula (N=2) of " cherry " is " ten "；Progressive expression formula (N=6) is " wooden ㄇ "；Progressive expression formula (N=10) it is " wooden shellfish ㄇ ".

【4】Encoding of chinese characters list file form

The 12 code element word tables of the present embodiment are a files being made up of a number of data row, each data row correspondence An element in the complete character set.Each data row includes some fields, and its form is as shown in table 2：

Table 2

The implication of each field is as follows：

1) Chinese character：Some element i.e. in the complete character set corresponding to 12 code element word tables；

2) 12 code element stroke coding：Will the Chinese character according to order of writing strokes (《Modern Chinese general words normative stroke order》) tear open Solution into, the one-dimensional sequence that is attributed to 12 kinds of code elements.The symbol of any form inequality can be used to represent 12 kinds of code elements, including but Be not limited to code element stroke shown in table 4 (one, Shu, Pie ..., ㄋ) or symbol code (1,2,3 ..., b)；

3) most simple expression formula：Refer to described above.

4) word frequency：The frequency that drawn Chinese character occurs is counted according to general language material；

5) separator：It is used for separating the symbol of different field, such as space, comma etc.；

6) other information：Reflect the useful information of other features of Chinese character, such as simplified and traditional body, pronunciation, radical, implication Deng still with separators between different information fields.

For ease of the inquiry of Chinese character, the data row in 12 code element word tables sorts generally according to stroke order, or according to the Chinese Word ISN order sorts.

【5】Keyboard layout

Keyboard used in the present embodiment, generally comprises following necessary button：

1) stroke key：5 to 12, for starting the input of the keystroke sequence, each stroke key represents 12 kinds of codes The code element of one or more in member；

2) Lian Bijian：1, in the input process of the keystroke sequence, for replacing appointing in addition to the 1st stroke key What one group of continuous stroke key sequence, or after the keystroke sequence terminates input, for inputting punctuate to the viewing area Symbol；

3) delete key：1, in the input process of the keystroke sequence, for cancelling the stroke key of last time input Or Lian Bijian, or after the keystroke sequence terminates input, delete and complete the character of input in the viewing area；

4) acknowledgement key：At least one, for terminating the input of the keystroke sequence, or in the keystroke sequence terminates input Afterwards, for inputting space character or line feed character to the viewing area.

Keyboard layout refers to：

1) mapping relations of stroke key and 12 kinds of code elements；

2) position of stroke key, Lian Bijian, delete key, acknowledgement key.

As shown in figure 9, alphabetic keypad is laid out：For alphabetic keypad (ISO/IEC 9995-1), stroke key is any 12 Letter key, corresponds 12 kinds of code elements, and representative configuration refers to table 3.

Such as Figure 10 and Figure 11, numeric keypad layout and telephone keypad layout：For numeric keypad (ISO/IEC 9995-4) With telephone keypad (ISO/IEC 9995-8), stroke key be 3 rows 3 arrange totally 9, be corresponding in turn to from left to right, from top to bottom one, Shu, Pie, Dian, Ya, Yi,  and ㄅ, 亅 and ㄋ, Yin and ㄥ.Refer to table 3.

Table 3

Dummy keyboard：On-screen keyboard can directly simulate alphabetic keypad, numeric keypad, telephone keypad, and directly use it Keyboard layout, is operated by pointing devices such as mouses.Portable equipment typically have it is hand-held can Touch Screen, the present embodiment pin Such keyboard is proposed it is hand-held can touch-control dummy keyboard (hereinafter referred to as dummy keyboard).

The man-machine interface of one typical mobile device is as shown in figure 12, including text filed 1, dummy keyboard 2, candidate word Area 3.Dummy keyboard includes 6 stroke keys, 1 company key, 1 delete key, 1 acknowledgement key, 1 space bar.

Dummy keyboard has following features：

Distinguish different button operations：Dummy keyboard divides into two kinds of short-press and slip, so can be to same button Different operations is carried out to input different strokes.Stroke key is that two rows three arrange totally six, and each stroke key operates mould in short-press Under formula, one, Shu, Pie, second, Dian, Ya are corresponding in turn to from left to right, from top to bottom, each stroke key in slide, from left to right, ㄋ, 亅, Yin, ㄅ, ㄥ,  are corresponding in turn to from top to bottom.

The keypress function definition such as table 4 of above-mentioned dummy keyboard stroke key.

Table 4

Sequence number	Key symbols	Button title	Short-press	Slide
					1.	One	It is horizontal	Entering stroke：One	Entering stroke：ㄋ
2.	Shu	It is perpendicular	Entering stroke：Shu	Entering stroke：亅
					3.	Pie	Skim	Entering stroke：Pie	Entering stroke：Yin
4.	Dian	Point	Entering stroke：Dian	Entering stroke：ㄥ
					5.	Ya	Folding	Entering stroke：Ya	Entering stroke：
6.	Second	Second	Entering stroke：Second	Entering stroke：ㄅ

(2) the searching algorithm based on Controlling UEP

One of the present embodiment it is important the characteristics of, be the searching algorithm based on Controlling UEP, and using even a key generation For any one section of continuous stroke in the code element stroke coding of Chinese character 12 in addition to the first stroke of a Chinese character, to reach the effect for improving input efficiency.

With~Lian Bijian is represented, for example even a key reduces keystroke sequence length to table 5：

Table 5

Sequence number	Stroke after omission	Match word	Reduce touch potential
				1.	Day~people	It is	2
2.	Ren~ Dian	's	3
				3.	Chi~very little		4
4.	Mono- Dian of Pie~Dian	Deng	7
				5.	Yi Dian Ya~mesh	Frost	8
6.	Vertical~tri-	Long	8

【1】Algorithm is sketched

Searching algorithm based on Controlling UEP is referring to Fig. 3.Because keyboard has a variety of number of keys, therefore to 12 code elements 12 code element stroke codings in word table are when carrying out searching, it is necessary to which 12 code element stroke codings are mapped as into stroke key according to keyboard layout Sequence.

, it is necessary to which keystroke sequence is disconnected at Lian Bijian when with a company of including key in keystroke sequence, two sections of initial and end is drawn Subsequence, then keyboard layout mapping is carried out respectively.

The value of the degree of correlation (being abbreviated as R) is as follows：

1) R=0：It is uncorrelated；

2) R=1：I.e. general candidate word；

3) R=2：I.e. stroke number it is consistent with keystroke sequence length or with keystroke sequence head and the tail all consistent candidate words.

【2】Algorithm flow explanation

The input of algorithm：Keyboard layout, 12 code element word tables, keystroke sequence；

The output of algorithm：Candidate list of words；

The interrupt mode of algorithm：Miscue.

Algorithm steps：

1) keystroke sequence：The keystroke sequence of user's input is designated as A；

2) initialize：12 code element word table index i=i0；

3) 12 code element word tables of traversal are judged：

A) the Chinese character Ci and its 12 code element stroke coding bi in 12 code element word tables are taken out；

B) according to keyboard layout, bi is mapped as stroke key sequence Bi；

C) compare A and Bi, determine degree of correlation Ri；

If d) the degree of correlation>0, then Chinese character Ci is added to candidate list of words；

E) judge whether candidate's number of words reaches the 12 code element word tables of the upper limit or traversal of setting；

If so, terminating；Otherwise, i=i+1, is performed a).

4) candidate's number of words is judged, if 0, then output error is pointed out, for example：

A) this button operation is ignored；

B) miscue sound or vibration are sent；

C) in text filed output prompt.

Wherein, referring to Fig. 5, Controlling UEP algorithm steps are as follows：

1) judge whether include Lian Bijian in keystroke sequence A：

A) if it is not, then judging whether stroke key sequence Bi starts that (i.e. the top n element and A of Bi is completely the same, A length with A Spend for N)；If starting with A, Bi=A is continued to determine whether, if then Ri=2, otherwise Ri=1；If not started with A, then Ri =0；

B) if so, then using the part before keystroke sequence A Lian Bijian (being free of) as keystroke sequence A1, Lian Bijian Part after (being free of) judges whether Bi is started with A1 and terminate (the i.e. rear N2 member of Bi with A2 as keystroke sequence A2 Element is completely the same with A2, and A2 length is N2), if Bi is to be started with A1 and terminated with A2, Ri=2, if Bi be not with A1 is started and terminated with A2, then Ri=0；

2) terminate.

【3】Algorithm example

For example, using digital keyboard input character " east ", its process is as follows：

Keystroke sequence A=" 18735 "

Inquire about 12 code element word tables, 12 code element stroke coding bi=" 1a634 " of " east " word correspondence

12 code element stroke codings are mapped as stroke key sequence Bi=" 18735 " through keyboard layout

Compare Bi and A, there is Bi=A, you can know Ri=2.

(3) word fitting algorithm is matched

The characteristics of Chinese character is had gradually in writing process into word, that is, write a Chinese character, typically can undergo some successively The individual intermediateness into word, for example, writing " in " when, " Shu ", " ㄇ ", " mouth " these three intermediatenesses can be undergone successively.And logical Cross during input keystroke sequence carrys out input stroke coding, several typically can be also undergone successively into the subsequence of word, sub- sequence The formed word of row is referred to as matching word.

Match word fitting algorithm automatic in the input process of keystroke sequence can carry out real-time retrieval and display to matching word, So that keystroke sequence is converted to readily discernible spelling text.For example, as the Shu of input stroke coding " Shu Ya mono- " when, in same text This input position display for a user successively " Shu ", " ㄇ ", " mouth ", " in ", simulate " in " writing process.For another example, " profit " word Input process is as shown in table 6：

Table 6

Touch potential

1

2

3

4

5

6

7

Input

Pie

One

Shu

Pie

Dian

Shu

亅

Output

Pie

Pie mono-

Thousand

Thousand Pie

Standing grain

Standing grain Shu

Profit

【1】Algorithm is sketched

Referring to Fig. 4, the principle of matching word fitting algorithm is to carry out matching degree analysis to candidate list of words and keystroke sequence, so Afterwards according to the different situations of matching degree maximum, to the different result of text filed output.

Matching degree (being abbreviated as M) value is as follows：

1) M=3：

If a) certain candidate word word frequency highest in all R=2 candidate word, makes its M=3；

If b) certain candidate word R=1 and be unique candidate word, makes its M=3；

2) M=4：If the existing R=2 of certain candidate word, it is unique candidate word again, then makes its M=4；

3) M=R：Other all situations, M=R.

【2】Algorithmic descriptions

The input of algorithm：Candidate list of words

The output of algorithm：Match word

The interrupt mode of algorithm：Into component sequence fitting algorithm

Algorithm steps：

1) matching degree analysis is (referring to Fig. 6)：

If a) candidate word is unique, its M=R+2 is made；

If b) certain candidate word word frequency highest in all R=2 candidate word, makes its M=3；

C) other all situations, M=R；

2) to the Chinese character of text filed output M >=3, and completion is automatically confirmed that to M=4 Chinese character；

3) when the Chinese character quantity of M >=2 is more than 1, the Chinese character of all M >=2 is exported to candidate word region；

4) when the matching degree of all Chinese characters in candidate list of words is 1, it is transferred to component sequence fitting algorithm；

【3】Algorithm example

1) " the Shu Pie Dian Shu 亅 " of Pie mono-, draw candidate list of words to input keystroke sequence：Sharp pears plough Li Li, wherein, matching word only has " profit ", is unique match word.

2) for another example, " Pie Dian Shu Ya Shu " one by one one by one, draw candidate list of words to input keystroke sequence：Stupid candidate word is unique, then It is " stupid " to match word.

3) keystroke sequence " Shu Ya Shu " one by one for another example, are inputted, candidate list of words is drawn：Chrysomelid Shen is told, wherein, matching word has three It is individual：" chrysomelid Shen ", is first matching word " leaf " to the matching word of text filed output.

(4) component sequence fitting algorithm

When keystroke sequence is longer and without matching word, in order to provide preferable visual feedback, the present embodiment proposes one kind Fitting algorithm, can be converted to keystroke sequence the orderly sequence arrangement (hereinafter referred to as component sequence) of Chinese character, part or stroke.

For example, during " washing " is inputted, after component sequence fitting algorithm, having drawn correct component sequence " Rui The Pie of Http bis- ", are component sequence fit procedure example as shown in table 7：

Table 7

When keystroke sequence cannot both be fitted to matching word, cannot be fitted to multiple parts again, just with match word and The combination of stroke is as component sequence, such as " thousand Pie " in table 6.

【1】Algorithm is sketched

Referring to Fig. 7, the principle of component sequence fitting algorithm is to estimate that user expects the group of Chinese character using one group of candidate word Into, by the analysis to the progressive expression formula of candidate word, and component sequence is formed dynamically, reach the effect of stroke dynamic group word.

【2】Algorithmic descriptions

The input of algorithm：Candidate list of words, keystroke sequence, current spelling text (hereinafter referred to as spelling text)

The output of algorithm：Component sequence

The interrupt mode of algorithm：Nothing

Algorithm steps：

1) stroke sequence length N is calculated；

2) the progressive expression formula of each candidate word in candidate list of words is calculated；

3) judge whether all candidate words have same progressive expression formula：

A) if so, then component sequence is equal to the progressive expression formula that all candidate words have；

B) if it is not, there is the progressive expression formula that at least one is differed, then current spelling text and most end are once defeated The stroke entered is by order arrangement one after another as component sequence.

【3】Algorithm example

For example, input falls, keystroke sequence is " a Shu Shu Dian Dian mono- Pie Ya Dian Shu Ya mono- ", emphasis is defeated with wherein the 3rd to the 8th Enter process analysis procedure analysis, it is as follows：

1) when being input to the 3rd, candidate list of words is：Lv Nian generation taro skills, matching number of words is 2, has drawn first With word " Lv ", spelling text for "Lv”；

2) the 4th is inputted, candidate list of words is：Sub- tight core sesame awns, no matching word, therefore calculated into component sequence fitting Method, but these candidate words do not have common progressive expression formula (N=4), therefore use current spelling text " Lv " and the 4th " Dian " as component sequence, make spelling text be changed into "Lv Dian”；

3) the 5th is inputted, candidate list of words is：Fan Mang Limonene ramies are swung, still without matching word, but are occurred in that for the first time common Progressive expression formula (N=5) " Lv Dian Dian ", thus component sequence be changed into "Lv Dian Dian”；

4) the 6th is inputted, candidate list of words is：The boundless and indistinct marshland full of water weeds of model Sha is swung, still without matching word, common progressive table is at this moment occurred in that Up to formula (N=6) " Lv Rui ", thus component sequence be changed into "Lv Rui”；

5) the 7th is inputted, candidate list of words is changed into：Luo Mo Pai Kuo Pao, still without matching word, also without common progressive expression Formula (N=7), thus component sequence be changed into "Lv Rui Pie”；

6) the 8th is inputted, candidate list of words is changed into：Fall, as unique candidate word, component sequence fitting calculation is not entered back into Method, matching word is " falling ".

It can be seen that, during input " falling ", component sequence fitting algorithm is in the interval that matching word " Lv " and " falling " occurs Filled using the component sequence of gradual change, the transition of matching word is more relaxed.User sees successively text filed From " one " to " Lv Rui Pie " writing process, obtains the visual feedback close with writing process.

Although disclosed herein embodiment as above, described content is only to facilitate understanding the present invention and adopting Embodiment, is not limited to the present invention.Any those skilled in the art to which this invention pertains, are not departing from this On the premise of the disclosed spirit and scope of invention, any modification and change can be made in the implementing form and in details, But the scope of patent protection of the present invention, still should be subject to the scope of the claims as defined in the appended claims.

Claims (8)

1. a kind of Chinese character input method of stroke dynamic group word, it is characterised in that methods described is to use input through keyboard button , will by the searching algorithm based on Controlling UEP, matching word fitting algorithm and component sequence fitting algorithm during sequence Expect Chinese character by order of writing strokes, from part to it is overall, be immediately shown in viewing area so that simulate stroke dynamic combined into The process of Chinese character, and form visual feedback；

The searching algorithm based on Controlling UEP is to compile 12 code element strokes in the keystroke sequence and 12 code element word tables Code does Controlling UEP, and the character of the degree of correlation >=1 in 12 code element word tables is added into candidate list of words；When in 12 code element word tables not When there is the character of the degree of correlation >=1, miscue is produced；

The matching word fitting algorithm is that the keystroke sequence and candidate list of words are done into matching degree analysis, to text filed output The Chinese character of M >=3, and M=4 Chinese character automatically confirms that completion；When the matching degree maximum of character in the candidate list of words is 1 When, it is transferred to component sequence fitting algorithm；Matching degree >=2 and total quantity in candidate list of words described in the instant playback of candidate word region Character more than 1.

2. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that the 12 code element stroke is compiled Code is that Chinese character is encoded by order of writing strokes by pen by 12 kinds of code elements, and 12 kinds of code elements include five basic strokes and seven extension pens Draw；Five basic strokes are：First, Shu, Pie, Dian, Ya；Seven extensions stroke includes：, Yi, 亅, Yin, ㄅ, ㄋ, ㄥ.

3. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that the keyboard is arrived including five 12 stroke keys, a company key, a delete key and at least one acknowledgement key,

The stroke key, for inputting the keystroke sequence, each stroke key represents one or more code elements in 12 kinds of code elements；

The company key, in the input process of the keystroke sequence, for replacing any one in addition to first stroke key The continuous stroke key sequence of group, or after the keystroke sequence terminates input, for the viewing area inputting punctuation mark；

The delete key, in the input process of the keystroke sequence, for cancelling the last time button of the keystroke sequence Operation, or after the keystroke sequence terminates input, delete and complete the character of input in the viewing area；

The acknowledgement key, for terminating the input of the keystroke sequence, or after the keystroke sequence terminates input, for institute State viewing area input space character or line feed character.

4. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that in the 12 code element word table With a complete character set, and 12 code element stroke corresponding with each element in the complete character set is compiled Code, most simple expression formula and word frequency evidence；

Element of any one stroke number more than 1 can be all combined into the other elements in the set in the complete character set Come, the other elements at least include：Constitute the part and stroke of element of the stroke number more than 1；

The most simple expression formula is according to stroke order and the word without interlocking for the element that stroke number in the complete character set is more than 1 Length obtained by shape principle of invariance is split is most short, and can most save the orderly character string of font style characteristic from damage, and most abridged table reaches Each character in formula is also the element of the complete character set；It is equal to 1 for stroke number in the complete character set Element, be exactly its own.

5. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that the 12 code element stroke is compiled The Controlling UEP of code is, by 12 code element stroke codings, stroke key sequence to be mapped as by keyboard layout, when the keystroke sequence During not comprising company key：If stroke key sequence is started with the keystroke sequence but length is different, 12 code element stroke codings correspondence The degree of correlation of character is 1；If stroke key sequence is identical with the keystroke sequence, the correlation of 12 code element stroke codings correspondence character Spend for 2；If the stroke key sequence is not started with keystroke sequence, the 12 code element stroke codings correspondence character degree of correlation is 0；

When the keystroke sequence is comprising company key, keystroke sequence is divided into two subsequences of initial and end by boundary of Lian Bijian, if pen Picture key sequence is started with first subsequence and with odd amount in addition to the round number the sequence ends, then the 12 code element stroke codings correspondence character degree of correlation is 2； If stroke key sequence is not started with first subsequence or not with odd amount in addition to the round number the sequence ends, 12 code element stroke codings correspondence character phase Guan Du is 0.

6. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that the keystroke sequence is with waiting Word selection list match degree is analyzed as follows：

1) M=3：

If a) certain candidate word word frequency highest in all R=2 candidate word, makes its M=3；

If b) certain candidate word R=1 and be unique candidate word, makes its M=3；

2) M=4：If the existing R=2 of certain candidate word, it is unique candidate word again, then makes its M=4；

3) M=R：Other all situations, wherein M=R, M are matching degree, and R is the degree of correlation.

7. the Chinese character input method of stroke dynamic group word as claimed in claim 1, it is characterised in that the component sequence fitting Algorithm is the progressive expression formula for calculating each character in candidate list of words, if in candidate list of words all characters have it is same Progressive expression formula, then using the progressive expression formula as component sequence, the progressive expression formula is chosen in most simple expression formula Since the 1st character, and total stroke number is suitably larger than the subsequence equal to keystroke sequence length, and more than situation when, Constantly repeat " last character in the subsequence to be replaced with the most simple expression formula of the character, and chosen again from the 1st Character starts, and total stroke number is suitably larger than the subsequence equal to keystroke sequence length " operation, the total pen of one finally drawn Draw the subsequence that number is equal to keystroke sequence length；Otherwise using currently spell text and most last input stroke order arrangement as Component sequence.

8. the Chinese character input method of stroke dynamic group word as claimed in claim 5, it is characterised in that the keyboard layout, fixed The mapping relations and key position of adopted stroke key and 12 kinds of code elements；

For alphabetic keypad, stroke key is any 12 letter keys, and corresponds 12 kinds of code elements；

To numeric keypad and telephone keypad, stroke key is that three rows three arrange totally nine, be corresponding in turn to one from left to right, from top to bottom, Shu, Pie, second, Dian, Ya,  and ㄅ, 亅 and ㄋ, Yin and ㄥ；

For dummy keyboard, the stroke key be two rows three arrange totally six, each stroke key under short-press operator scheme, from a left side to It is right, be corresponding in turn to one, Shu, Pie, second, Dian, Ya from top to bottom, each stroke key in slide, from left to right, from top to bottom according to Secondary correspondence ㄋ, 亅, Yin, ㄅ, ㄥ, .