WO2005018207A1

WO2005018207A1 - Appratus and method for inputting alphabet characters from keypad

Info

Publication number: WO2005018207A1
Application number: PCT/KR2004/002080
Authority: WO
Inventors: Min-Kyum Kim
Original assignee: Min-Kyum Kim
Priority date: 2003-08-18
Filing date: 2004-08-18
Publication date: 2005-02-24

Abstract

This invention shows that a target word or phrase can be inputted effectively using an index and applying a full input method (FIM). Moreover, this invention shows an example in which a (language restricted) concurrent input method (CIM) is applied effectively. In addition, this invention proposes an input system for inputting various characters by combination of a long stroke of a specific button and a repeat stroke of the specific button, or by combination of a long stroke of a specific button and a repeat stroke of a control button.

Description

APPARATUS AND METHOD FOR INPUTTING ALPHABET CHARACTERS FROM KEYPAD

BACKGROUND OF THE INVENTION

(a) Field of the Invention The present invention relates to an apparatus and method for inputting characters from a keypad. More specifically, the present invention relates to an apparatus and method for inputting characters from a keypad having a small number of buttons such as a telephone keypad. (b) Description of the Related Art With the progress of mobile communications, a function of receiving and sending digital information such as text messages is added to a mobile station chiefly used for voice calls. Hence, the keypad provided on the mobile station for the entry of a telephone number additionally has a function of entering characters, thus reducing the size of the keypad used as an input means in the mobile station and hence limiting the number of buttons included on the keypad. Alphabets of every language are usually much more than 12 keys on the keypad. Therefore a need exists to represent every character with buttons on a telephone keypad alone or in combination of two or more different types.

SUMMARY OF THE INVENTION The present invention suggests the improvement of the prior documents of the applicant (Application No. PCT/KR00/00601 , PCT/KR01/00076, PCT/KR01/02267, PCT/KR02/01355, and PCT/KR03/01433). The contents of the prior documents will be written in this description in part, but refer to the prior documents in connection with necessary parts, which are not written in this description of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate an embodiment of the invention, and, together with the description, serve to explain the principles of the invention:

FIG. 1-1. An example of Base Keypad in English

FIG. 2-1. An example of Succession Keypad in Japanese (1) FIG. 2-2. An example of Succession Keypad in Japanese (2)

FIG. 2-3. Graph of 2-dimensional Multi-use (or Reuse) Cross Control

Processing Method

FIG. 2-4. Graph of 2-dimensional Cross Control Processing Method

FIG. 3-1. (NONE) FIG. 4-1. (NONE)

FIG. 4-2. (NONE)

FIG. 4-3. (NONE)

FIG. 4-4. (NONE)

FIG. 4-5 - 4-8. An example of Keypad using vowel elements in Korean FIG. 5-1. An example of inputting Chinese character using Pinyin index

FIG. 5-2. An example of inputting English word or phrase using Fully- Associated Simple Code

FIG. 5-3. An example of recognizing and processing input values as Simple

Code in Concurrent Input Method (1) FIG. 5-4. An example of recognizing and processing input values as Simple

Code in Concurrent Input Method (2)

FIG. 6-1. (NONE)

FIG. 7-1. (NONE)

FIG. 8-1. (NONE) FIG. 9-1. (NONE)

FIG. 10-1 ~ 10-4. An example of CVSK (Consonant-Vowel Separated Keypad) in a language using Roman alphabet

FIG. 10-5. An example of Short-cut Input Method in CVSK, and an example of recognizing and processing input values as Simple Code in Concurrent Input Method (Chinese)

FIG. 10-6. An example of CVSK in a language using Roman alphabet (5)

FIG. 10-7. (NONE) FIG. 10-8. (NONE)

FIG. 10-9. (NONE)

FIG. 11-1. An example of system configuration when (an) input value(s) is/are interpreted at clinet side FIG. 11-2. An example of system configuration when (an) input value(s) is/are interpreted at server side

FIG. 11-3 ~ FIG. 11-10. Examples of processing Simple Code (in Concurrent

Input Method)

FIG. 1 1-1 1. An example of the type of Simple Code (i.e., short-cut code or short-cut input value)

FIG. 12-1. (NONE)

FIG. 13-1 ~ FIG. 13-4. Examples of Concurrent Input Method for Inputting Word or Phrase by Full input Method in Part and by Short-cut Input Method in Part

FIG. 14-1 ~ FIG. 14-2. Examples of Control Processing Method using Long Pressing (graph)

FIG. 14-3. An example using short pressing after long pressing (graph)

FIG. 14-4. An example using long pressing after short pressing (graph)

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS It is preferable to be refered to prior documents when it is needed. It is apparent that although not specifically described, the content of the prior documents and the contents related to a certain language is also applicable to other languages. 1. Method for using Vowel Elements in Korean FIGS. 4-5 to 4-8 are views showing an example in which vowel elements in Korean is used and the vowel button is utilized as a control button.

Refer to the prior document of the present applicant in connection with the present invention.

2. FIM and CIM in an Input System Storing a Full Code The prior document describes that it is somewhat difficult to discriminate a correct word to the input value on a specific keypad using a specific language restricted FIM (for example, English input and pinyin input on a standard English keypad by the RSM). As described in the description of Chinese, if the system has indexes of all words, the input value is first interpreted as a full code, and second interpreted as a simple code of an appointed type if the input value does not exist in the index. It can be applicable not only to Chinese but also to all languages. Hereinafter, expediently, Chinese will be described. As shown in FIG. 10-14, if the system storing all Chinese Romanization(pinyin) (all words in case of other languages) stores all full code, the system can simply search words or phrases corresponding to the input value, and recognize a target alphabet. If the FIM-based CIM is applied, the system regards and processes the input value as the simple code of the appointed type the moment the system recognizes that there are no more corresponding full code in the index. A process for inputting " ff." of FIG. 10-14 shows a process for interpreting and processing the input value as a full code. Just as the full code is stored in the index, the full code is interpreted through search of the index. If the FIM according to the language restricted input method, which can discriminate words without ambiguity, is applied, it is not useful to store the full code in the index. However, it is useful in case of that the FIM such as the RSM, which may cause ambiguity, is applied, and all target words are stroed in index. As described in the description of "Language Restricted Input Method in

Incomplete CVSK", if all words intended by the user to be inputted are stored in the index when the FIM, which may cause ambiguity, is applied (it can be applied to other languages as well as Chinese), the input value can be processed using the full code index. It is obvious that the above can be applied to other languages besides

Chinese. Additionally, the above can be applied not only the keypad of FIG. 1- 1 but also the keypad of FIG. 10-* or other keypads. 3. CIM for Simultaneously Processing Input Values as Various Type Codes (i.e., Various Types of Input Values) CIM for simultaneously recognizing an input value as various type codes The CIM for regarding and processing input values as predetermined type codes (input values) is possible. For instance, In FIG. 10-14, if an input value is regarded and processed as a full code(A) and a fully associated simple code(B) at the same time, the system can recognize "bai" interpreted as the full code and "bei" interpreted as the fully associated simple code in connection with the input value "1*#". In this case, the system can provide the user with "bai" and "bei" at the same time. There is a difference in that the system outputs only "bei" when interpreting the input value "1*#" as the full code adopting the FIM-based CIM but outputs only "bai" when interpreting the input value "1*#" as the appointed short-cut code adopting the SIM-based CIM. Also, in such CIM, the system can regard and process the input value as only a different type code the moment the system recognizes that there are no more input values existing in a special type code. As the result, it is included in the category of the above CIM. It is not necessary to store the full code and the fully associated simple code in the indexes of the system as shown in FIG. 10-14, and the input value can be interpreted as various type codes at the same time. 4. Priority Order of Words, When Ambiguity Occurs In a specific input method (for example, a full input method, or a specific type short-cut input method), If the same input values cause ambiguity, as shown in FIG. 10-14 and other drawings, the system can output a target word according to the input order on the basis of the priority order in case of that ambiguity occurs when a specific input method is applied. For instance, in FIG. 101-4, in case of that ambiguity occurs when the full input method is applied, the system outputs a target word according to a "priority order (A)" between full codes(A) and recommend the target word to the user. Next, interpretation of an input value as various type codes will be described. As described above, in FIG. 10-14, if the input value of "1^*#" is interpreted as a full code, it is interpreted as "bai", and if interpreted as a fully associated simple code, it is interpreted as "bei". Like the above, when ambiguity occurs by interpreting the input value as different type codes, the system can provide a target word to the user according to the priority order

"(A)+(B) priority" (i.e., the priority that input value is regarded as (A) and (B) simultaneously) in FIG. 10-14. For instance, in the CIM that the input value is first regarded and processed as a full code, but second regarded and processed as a syllable- based initial code and a syllable-based initial value simple code the moment the full code construction rules are violated, the syllable construction rules of a special language are violated, or the system recognizes that there are no more corresponding full codes in the index, when the input value is regarded and processed as a short-cut code, the input value can be processed based on the priority order made when ambiguity occurs between the "syllable-based initial code" and the "syllable-based initial value simple code" (namely, if the syllable- based initial code is (C) and the syllable-based initial value simple code is (D), the target word is outputted according to the priority order of "(C)+(D) priority"). Chinese is described in the above, but it is obvious that other languages can be applied in the same way if the target word is not Chinese characters but Chinese romanization(pinyin). 5. CIM adopting the CPM(Control Processing Method) as the

FIM(Full Input Method) As shown in FIGS. 2-1 and 2-2, in case of that control buttons are a button [*] and a button [#] and the CPMERC (Control Processing Method Except Representative Character) is applied, the CIM is applied, and if all words including only representative characters in the index and the SIM-based

CIM is applied, the system can regard and process the input value as a full code the moment "*" or "#" is pressed as the input value. In case of Japanese of FIGS. 2-1 and 2-2, it is assumed that all shortcut codes have only numeral values. In this case, the system stores all words, which have only the representative characters in the index (it can be included in the short-cut code index or formed in separation from the short-cut code index), and the CIM is applied. At this time, the system can provide the user with the result that the input value is interpreted as a full code and the result that the input value is interpreted as a short-cut code with reference to the index of words, which have only representative characters, (or the simple code index including the words, which have only representative characters) as target words according to a predetermined priority order. For instance, in FIGS. 2-1 and 2-2, if an input value of "111" is processed as a full code (namely, if the FIM-based CIM is applied), it is interpreted as "fo fo fo". If words starting from ^u ~ does not exist with reference to the index of all words including only the representative characters, but only ^ufo i. " or ^ufo ι^~" exists, the system can provide the user with "&> ^" or

"* ι^~" in priority to "h h~". Refer to FIGS. 10-15 and 10-16. In FIG. 10-15, the input value of "11" is interpreted as the full code, but the system searches the "index of all words including only representative characters" and checks whether or not the word is effective as the word includes only the representative characters. Here, if "h " obtained by interpreting "11" as the full code exists in the index, the system recognizes the input value as an effective word. The system recognizes ^a i. ^•>" corresponding to the input value by interpreting the input value as a simple code again the moment the system checks that there are no more words corresponding to the input value in the index of all words, which include only representative characters. In FIG. 10-15, the system recognizes "-β ^" as the very best word in connection with the input value of "11". As "& ι^ ..." ("..." means characters added to " ^") may represented according to succession input values of "11", the system can output it in the next order of "- Likewise, as the target word may be "fo fo" even though it does not exist in the index, it can be also outputted in the next order of " ^". In FIG. 10-15, "£> ^ ..." and " > >" are indicated with a blur color and in the second and third recommendation orders. It means that the system can output the input word as "* I- .." or "* *", or output only "* W. In FIG. 10-15, (1), (2) and (3) may occur in order, or simultaneously occur as described in the description of the "CIM for simultaneously processing an input value as various type codes". Likewise, the full code and the simple code indicated in the index are to help your understanding. It is obvious that the same process can be carried out even though the full code and the simple code are not stored in the index (if the simple code can be interpreted regularly like the syllable-based initial code or the fully associated simple code). As shown in FIG. 10-16, a case in which the index of all words including only representative characters and the simple code index exist in one index is also identical with the above. The data structure in embodiment may have various types. An operation result of the system is similar to the SIM-based CIM described above. At this time, not the SIM-based CIM but the FIM-based CIM or the general CIM is applied. It can be applied in the same way not only to the CIM adopting the CPMERC as the FIM but also to all CIMs adopting the CPM as the FIM. 6. Concurrent Input Method for Inputting Word or Phrase by Full input Method in Part and by Short-cut Input Method in Part A method for outputting words corresponding to the input of word(pull cord input) from the index and allowing the user to select the target word has been used widely. For instance, there is an "auto-completion" function of a Microsoft windows. Of course, even in case of the full input(entered in full code), the corresponding syllables (words or phrases) are output from syllables (words or phrases) stored in the index (for example, index for concurrent input) and the user can select the target word or phrase. The above will be described on the basis of the input system shown in FIGs. 4-5 to 4-8 as follows. For instance, assuming that words such as "---i^l≡r " and " n - 1 --1" are stored in the index, if "10* = ΞL" is entered, the system can output words corresponding to "Ji... "(namely, words starting with " L" such as

) in a proper form. Under circumstances similar with a cellular phone, contents generated by input values are displayed on an input line(the top of an LCD in FIG. 13-1), and candidate words or phrases outputted from the index can be display on the bottom of the LCD in a table form or the like. The user can select, confirm and input the target word using the move button(i.e., navigation button). For example, FIG. 13-1 shows the LCD of the cellular phone. FIG. 13-1 shows an example of outputting the corresponding words of the syllable formed according to the full input onto the bottom of the LCD. Here, to input " L^ ", " i" is entered, and then, the remaining syllables can be inputted by the short-cut input method(short-cut input using the syllable-based initial value simple code). That is, "--1+Π Λ L Π = lθ*+5723" is entered. Here, the target word(namely, "JI^TΓ¹-. ¹ ") which has the first syllable of "J!" and the remaining syllables having the initial value of "u A ^t- τ=" is provided to the user. Meanwhile, half of syllables in Korean have the final consonant, but the residuals do not have the final consonant. In case of

if "1*05723" is entered in succession, it corresponds to "^■§^■ A I_ t=". In this case, the system can recognize "-§- A = 1*057" in a normal full input, and the residual part (for example, "A L C = 723") in the short-cut input by recognizing that a part of the input value does not form normal Korean syllables the moment "2" is entered. The above also belongs to the category of the "concurrent input method(FIM based concurrent input method)", which is presented in the prior document. Syllables recognized in full code (for example, "-§^■") are naturally recognized in full code, and the remaining input values(for example, "A U Π =

723") is recognized in simple code. Here, the corresponding words including the syllable formed by the full input are searched from the index (for example, if

¹ " is stored in the index) and provided to the user, the words corresponding to the simple code input values (for example, if " --. " is stored in the index) are provided to the user, or all the above may be outputted. If the words searched by all the two cases are outputted, a specific word can be outputted on the top of the LCD according to determined rules (for example, according to the priority order of words including parts recognized in full code, the priority order of words, or the priority order of words belonging to a specific group). It is preferable that the output order is set by the user as words can be displayed on the LCD in various ways. In case of that the words searched by all the two cases are outputted, when "^-^ " and " ^ " corresponding to "§λuc^« are outputted on the bottom of the LCD, if the user selects "^ ¹-! 4", all of "^■§^■ A I_ C will be changed into "^^M ", or if the use selects "^^ ^{» »}A I_ T= " corresponding to " ^ ^" is changed into "^^ ". If only

i " is stored in the index,

t " corresponding to the input value of "1*05723" cannot be outputted. Therefore, the system can analyze "1*05723" into "^A L C" and "HD A Π" at the same time, and output the target word (provide to the user) referring to the index. To interpret

"^A U C" into "alt-- A C", jf the syllable in front of " A" which is recognized as the start of the short-cut input is ended to the final consonant, the syllable is excepted from the front consonant(namely, "al"), and the final consonant as unit consonant is put between "-H" and "A", it would be shown complex, but any person who has the basic knowledge of Korean structure can easily know it.

The above can be applied similarly even though the final consonant is double consonant. FIG. 13-2 shows an example of outputting candidate words by applying the priority output order of "words of the specific group designated as priority output, words including parts analyzed in full code, words which do not include the parts analyzed in full code, and so on". Here, "l≡r¹--) ¹ " belongs to the group designated as priority output, and so, is outputted onto the uppermost top. If the user wants to interpret only "HΠ Λ I- C", after "10* = al" is entered, only "aiu T- IT" is interpreted by pressing "5723" after the appointed manipulation for deciding the syllable (for example, non-input for a predetermined period of time, or one-time pressing of a [>] button), and only the corresponding words are searched. Also, in this case, words(for example, "al ^^-r^-W) corresponding to "--iπ Λ u" including " L", which is a syllable previously recognized can be outputted, words(for example, "^HrM ¹ " if it is stored in the index) corresponding to "U I_ T= " excepting "al" can be outputted, or all of the above can be outputted. Refer to FIG. 13-3. Here, the syllable can be decided by once pressing the [>] button for intentional syllable decision. Furthermore, the [>] button can be utilized as a space button, and so, after the decision of syllable, the [>] button can be served to input space. That is, if the user presses the [>] button once after inputting

"JL", the syllable "JL" is decided, and if the [>] button is pressed again, "-H" (' L'+ space) is formed. That is, when "10^*>5723" is entered, the system recognizes it as "31 u A L C", and provides the user with

" which is the corresponding word. Of course, if the system refers to the index of example in FIG. 13-3, the system can output "-H^ ^ " even if only "10*>57 = αLu A" is entered. In such concurrent input method, the output method of syllables including the syllable formed by full input is useful for decision of the target word from the list of candidate words when the corresponding candidate words are provided during the full input. Moreover, the output method of candidate syllables corresponding to only the input values treated by the short-cut input

(namely, input values processed as simple code) excepting the previously formed syllable) is useful for rapidly inputting frequently used suffixes or particles in short-cut input. For example, after "ai¹^-" is entered, (after the input contents is decided by a predetermined tool) "A L C" is entered to output " =f M ή- "(under the assumption that it is stored in the index). As described above, it would be appreciated that suffixes frequently used like "... M ¹^" is stored in a specific group (for example, a suffix group) in the index, and then, words of the corresponding group are first outputted when candidate words are outputted with concurrent input. If "-n^^τ " and "^^ " are all stored in the index, when "ύi τ_ c" is entered, "ϋ^q £}" and "^4 " can be all listed as candidate words. For more example, when

o = 70##90^*838" is entered, the system recognizes "*§ " as full input values, and "38", which has an incorrect syllable, as short-cut input values. Likewise, if only "^ll^cfl^-" is stored in the index, "^ ^ fl^-" is outputted. On the contrary, if only " fl%^» is stored in the index, only "tfl " is outputted. If both "*\} ι %" and " fl^-^» are stored in the index, they can be all outputted. However, it is preferable that only one is outputted according to regulations(including/not including parts recognized with full input). Furthermore, in case of "€r^ fl ". if "n o" is entered in succession after " E-HΓ" is entered in full input, the system recognizes it as "^■§- TJ. o ". Therefore, if the user wants to enter only "^~ \ " in short-cut input after "Έ-T-" is entered in full input, after the user enters "-g-r-" and confirms the input syllable "^^"(for example, press the button[>] once for termination of the syllable), the user can enter "38 = τ= o ". Here, even though " fl " has a bad priority order in comparison with other words(for example, "4^"), which are stored in the index and correspond to "τ= o = 38", if 'Sπ-s tflΦ" j_s stored in the index, the system can output "tfl " prior to other candidate words(for example, "iW) relative to the input value "5^*025^*0>38". Moreover, also in a case of that the user enters some syllables in short- cut input like "π u ^ι_ u: = 157*623" and some syllables in full input, it is possible that the system recognize it as

j-". When "157 = c is entered, the system interprets the simple code "157"(for example, syllable- based initial value simple code or word-based initial code), and then, searches words corresponding to "^~ι o A" by each syllable. When "*6" is entered, the system can recognize "157*6 = π u ^" and "7*6 = r". That is, the system recognizes that the previously entered "15" is a short-cut input value, and if "23" is entered again, the system recognizes it as short-cut input value as it does not form a proper syllable. Also, in this case, the system can recognize "π ϋ ^ι_ τz:" as the full code in part, and as the short-cut code in part, and so, provide

" to the user. As described above, a syllable in Chinese is composed of "shengmu + yunmu". Furthermore, as set forth, in the notation of Chinese pronunciation, which is based on the Chinese romanization system (i.e., Chinese pinyin), the "shengmu" is consonants of the Roman alphabet, and the "yunmu" is vowels of the Roman alphabet or "vowel+n" or "vowel+ng". that is, as you can see above, a syllable in Chinese characters written by the Romanization system(pinyin) has the form of "consonant+vowel", "consonant+vowel+n", or "consonant+vowel+ng". Rarely, the romanization system has syllables starting from a vowel "a", "e" or "o". Here, as described above, Roman alphabet consonants, which can be located at the end of one syllable in the romanization system(pinyin) are only "n" and "ng". It can be considered as the final consonant in Korean. Namely, the final consonant of one syllable can have only "n" and "ng". For example, if "^--Kzhonghua)", "Kϋ(minguo)", "φ-iK£! (zhonghuaminguo)", and so on are stored in a pinyin index, "^--^(zhonghua = 49***7739##*, see FIG. 10-6) is entered by the Romanization system in full input, "SH-l(minguo)" is inputted by entering "mg = 73 (see FIG.10-6)" in shortcut input (using the syllable-based initial value simple code as the short-cut input value). To input "if--!", the user enters the corresponding romanization system ("zhonghua = 49***7739##*"), selects and confirms one of listed candidate Chinese characters, and enters "mg=73". Alternatively, the user enters "zhonghua = 49***7739##*", and enters "mg=73" in succession. In case of the former, it is natural that the system outputs only "SS(minguo)", "S Ift-(minge)", . . . corresponding to "73=mg" as the candidate words. In case of the latter, it is also natural that the system recognizes the input value of "49***7739##*" as "Φ--g", and outputs "KS" and "φ- KS" corresponding to

"73" recognized as the short-cut input value. When "49***7739##*7" relative to "49***7739##*73 = zhonghua + mg" is entered, the system recognizes the final numeral "7" as the pinyin shengmu inputted in a normal full input, but when the final numeral "3" is entered, the system recognizes the final input value "73" as the short-cut value. It is the same case as Korean. Therefore, the system can output "Φ-^SS" including "Φ__ " corresponding to the full input value "49***7739##*" and "E-l" corresponding to the short-cut input value "73". Refer to FIG. 13-4. On the contrary, even when the user enters "zh = 49" in short-cut input relative to "Φ- " and enters "7#773##^***" in full input relative to "KS" in succession without selection and input of "Φ^" from the candidate words, the system can output "Φ-- S^" corresponding to "z h minguo". Likewise, when the user enters "73##***" to input "R.W (minguo)", the system can output "-^HH" by recognizing "73##^*** = mguo = m guo". The moment "73" is entered, the system recognizes it as the short-cut input, and searches words corresponding to "m g ", and recognizes "##^*** = uo" as vowel forming the second syllable. The input of "uo" next to "mg" serves to restrict the candidate words corresponding to "m g " to "KIH". When the user enters "73" in short-cut input relative to "KS" after entering "φl " in another way(one of all Chinese input ways such as short-cut input, full input, input using stroke of Chinse character, and so on), if the system outputs words corresponding to "m g " by simply recognizing the input value "73" as a simple code, the system outputs a list of lots of candidate words(for example, "KS(minguo)", "K^(minge)", . . .). However, if "Φ--f K HI" is stored in the index, the system can recognize "Φ i+mg = φ I+73" as "Φ-IKi" regardless of storing "Klϋ" in the index. In this case, as "Φ--g" is previously inputted in Chinese, the system can output "SS" corresponding to the input value "mg = 73" prior to other candidate words(for example, "Klfc (minge)", . . .). Here, when "zhong..." is entered, the system recognizes it as "zho" + "n g " or "zhon"+"g ". However, as the syllables starting with "zho" or

"zhon" do not exist in the Chinse dictionary in reality (so, they are not naturally contained in the pinyin index), the system recognizes it as "Φ" relative to "zhong..." by referring to the index. If they are contained in the index, when "zhong" is entered, the system can output candidate words according to a predetermined output order. In FIG. 10-6, the system may recognize "zhongg...

= 49^***7733" as "zhonk.." = "zho" + "n_k__" or "zhon"+ "k__". It is similar with the concurrent input method for simultaneously processing the input values in full code and in short-cut input value(simple code), as described above. As set forth, as the short-cut input value (i.e., simple code), for convenience, an example using "syllable-based initial value simple code" is presented, but if "syllable-based initial code" is applied as the short-cut input value (i.e., simple code), the example shown in FIGs. 11-3 to 11-10 can be applied for simple code recognition. 7. Character input method & control processing method using long pressing In general, one object(for example, alphabet, numeral, English alphabet of languages, which do not use English as the native language, character(special character), function, and control of specific purpose are called "alphabet etc.") can be expressed by pressing a button once. Here, the user can express another object, which is different from an object expressed by one-pressing, by pressing a specific button long once. Hereinafter, to press the specific button long is called "long pressing" or "long stroke". Hereinafter, when the "long pressing" is not mentioned, it means normal button pressing, and called "short stoke" to describe the normal button pressing in comparison with the "long pressing". Hereinafter, English will be explained for describing general matters, and each language will be explained for describing applicable cases by each language. In this case, it would be evident that the matters explained in one language can be applied to other languages in the same way. 7. 1. English (and other languages) 7.1.1. Alphabet input using long pressing Presently, a method for inputting the corresponding numeral by pressing a specific numeral button long has been used. However, it is not restricted to only numerals. For example, assuming that one-pressing of each numeral button as shown in FIG. 1-1 means that the first character of characters assigned to the corresponding button is entered(for example, "A" is entered by pressing a button[2] once), the second character(for example, "B") assigned to the same button can be entered by the long pressing. However, it is not preferable that the long pressing is applied for input of frequently used characters as it may break a natural flow of input. Therefore, it is preferable that the input by the long pressing is applied to "alphabet etc." which is not used frequently. Moreover, it is preferable that time for the long pressing can be set by the user according to the user's pressing speed. In the present invention, the long pressing is used appropriately, and thereby, the present invention can simplify input regulations of a specific input method, widen an expression range, and remove ambiguity. For convenience, the long pressing of the button [2] is expressed by "2~" by attaching "~" next to

"2". As described above, for applying the existing repeat selection method, when the button [2] is pressed three times in FIG. 1-1 , there is ambiguity as it cannot be recognized whether "222" means "C", "AAA", "AB", or "BA". In FIG. 1-1 , for example, it is defined that "A" is entered by long pressing or one-pressing of the button [2], "B" is entered by one-pressing of the same button after "A" is entered, and "C" is entered by one-pressing of the same button after "B" is entered. If characters assigned to the same button are entered in succession, when the character next to the second character is entered, it can be entered without ambiguity by starting the entry with the long pressing. For example, when "DACB" is entered, it can be entered by "3 2 2-22 2-2" without ambiguity. When A,C and B assigned to the same button are entered in succession, they can be entered without ambiguity as "C" and

"B" are entered by the long pressing. This input method is similar with the existing repeat selection method, and can remove ambiguity by using the long pressing only when characters assigned to the same button are entered in succession. Of course, it is possible that "A" is entered by the long pressing of the button [2] (namely, "2~") in case of "DACB = 322-222-2". In this case, one(for example, "A") of characters assigned to the specific button(including both clearly expressed assignment and suggestively expressed assignment) is called a representative character, the remaining characters(for example, "B", "C") are called succession characters. The succession characters are entered by pressing the button, to which the representative character and the succession characters are assigned, several times. That is, if characters are assigned to a button in order of "A(representative character) - B(2^nd succession character) - C(3^rd succession character), the system recognizes "A = 2-", "B = A + 2 = "2-2", and "C = B + 2 = A+2 + 2 = 2-22". In addition, that "A" can be entered not by the long pressing but by the normal pressing(namely, short stroke) means that the representative character can be entered by the one-pressing(namely, short stroke) as in the existing repeat selection method if the characters assigned to the same button are not entered in succession. Hereinafter, for convenience, it is called "repeat selection method by long pressing of representative character". If only two characters are entered using only one button, one of the characters is entered by the normal one-pressing(namely, short stroke), and the other is entered by the long pressing(namely, long suurve ^ιuι ex-tκιμιc, m r n . 10-1 , "B = 1", "P = 1-"). However, as shown in FIG. 1-1 , in case of that three characters are assigned to one button(for example, a button [2]) (including both clearly expressed assignment and suggestively expressed assignment), if the first character is entered by one-pressing(for example, "A=2"), the second character is entered by long pressing(for example, "B=2-"), and the third character is entered by two-pressing(for example, "C=22") or by long pressing and one-pressing(for example, "C=2-2"), there is also ambiguity when the third character is entered. The reason is that If "C=22" is defined, it may be recognized as "AA", if "C=2-2" is defined, it may be recognized as "BA". In this case, the first, second and third characters can be selected optionally. Also, the above method can reduce ambiguity more considerably than the typically simple Repeat Selection Method. The Repeat Selection Method using long pressing has several merits in that characters can be entered without ambiguity, and in that alphabets, numerals, characters assigned to the same button can be entered effectively. For example, assuming that specific characters "A1 , A2, A3,... "(including both clearly expressed assignment and suggestively expressed assignment) are assigned to a button [x], and a specific numeral "N1" is also assigned to the button [x], and special characters "S1 , S2, S3,..." are also assigned to the button [x], the characters assigned to the button [x] can be entered by long pressing. For example, the character "A1" is designated as the representative characters and entered by long pressing of the button [x]. The input of "A1 , A2, A3, N1 , S1 , S2, and S3" assigned to the button [x] is shown in FIG. 14-3 in the form of a graph. In FIG. 14-3, "A1 S1" is entered by "xx-xxxx", or by

"x-x-xxxx". Here, in languages that English is not native language, English characters "E1 , E2, E3, ..." can be added and assigned to the same button. Furthermore, to input the objects(for example, "A1 , A2, A3, N1 , S1 , S2, and S3") assigned to the button [x], it is possible that the assigned numerals and characters(special characters) are entered not by the repeat selection method using long pressing of representative character but by the existing repeat selection method. But, it is necessary to avoid ambiguity by "a predetermined time delay" or input of separator(for example, one-pressing of the button [>] described in "complementary matters in concurrent input method") after entry of one object, as there is ambiguity when objects assigned to the same button are entered in succession. For example, A1 , A2, A3, N1 , S1 , S2, and S3 are assigned to the button [x], and when "S1A1" is entered by the repeat selection method, "S1A1" can be entered by "xxxxx>x". It is to avoid ambiguity when the repeat selection method is applied, but it is different from the existing repeat selection method in that special characters, numerals, characters are all assigned to the same button and entered by pressing the button. Till now, when characters of mother tongue, "A1 , A2, A3" are assigned to the button [x], the system recognizes "x=A", "xx=B", and "xxx=C", or if the button is pressed in excess of the defined maximum repeat pressing numeral(for example, the button [x] is pressed four times), the system recognizes "xxxx=A" and "xxxxx=B" according to a circulation toggle method.

The circulation toggle method is to allow entry of the target character by repeatedly pressing the button [x] when the user mispresses the button [x] in applying the repeat selection method. However, as presented by the applicant, when the user presses the delete button not to delete the entire syllable but to delete the "latest (or last) input value", the circulation toggle method is not needed. That is, when the user wants to enter "xx=B" but enters "xxx=C", the user can enters "B" by pressing the delete button once without additionally pressing "xx" to input "B". FIG. 14-4 shows a process for inputting character by long pressing after short pressing of a specific button. In FIG. 14-4, the row of A1 , A2, A3, ... shows an input process by the existing repeat selection method. B2 is entered by long pressing of the specific button(for example, button [x]) after A1 is entered by short pressing of the same button. That is, "A1" is entered by "x", and "B2" is entered by "xx-". Here, if entry of a specific character is not defined by long pressing of the button [x], and assuming that only A1 and B2 are entered using the button [x], it can be interpreted without ambiguity. Even though the long pressing of the button [x] is defined to a specific character(for example, character "E1"), if "A1" and "E1" are not simultaneously used (i.e., do not appear consecutively), A1 , B2 and E1 can be entered without ambiguity by using the language restriction(language restriction in which A1 and E2 are not used simultaneously). In FIG. 14-4, if the long pressing of the button [x] is not defined as input of the specific character, A1 , B2 and C3 can be recognized without ambiguity. Moreover, in case of A1 , A2 and D3, there may be ambiguity in "A1=x" and "A2=xx=A1A1", but if A1 in language is not used in succession, A1 , A2 and D3 can be entered without ambiguity using the language restriction (language restriction in which A1 is not used successively). For convenience, the method as shown in FIG. 14-4 is called "repeat selection method by long pressing after short pressing". For example, in FIG. 10-6, "i" is entered by "#" and "u" is entered by "#-", and a certain vowel(for example, " u" - character "u" having affix "..") can be entered by "##-" without ambiguity ("I" and "u" are not used in succession in Chinese pinyin). Of course, as "i" is not used in succession, it is possible that other object can be entered by "##". It can be applied to other languages in he same way. In FIG. 14-4, it can be applied to enter next characters of rows of A1 , A2 and A3(for example, S1 , S2, S3, ...). 7.1.2. Control processing method using long pressing Next, an example in which the method is applied to selection of control will be described. The example will be described assuming that control is set to be selected after representative character. In the above description, the example, in which "A" is entered by both the short stroke and the long stroke of the button [2], is to show that "A" can be entered similarly to the existing repeat selection method. However, a specific object can be inputted by one-pressing of a specific button, and another object can be inputted by long pressing of the same button. For example, a specific object(for example, character x - "x" means not the real character "x" but any character) can be inputted by one- pressing of a specific button(for example, a button [1]), and another object(for example, any character "y") can be inputted by long pressing of the same button. That is, "character x" is entered by "1", and "character y" is entered by "1~". Here, as "character x" and "character y" are notional, even though "A=2-" or "A=2" is defined by the "repeat selection method by long pressing of representative character", "A" by the normal pressing(namely, short stroke) corresponds to "character x", and "A" by the long pressing corresponds to "character y". The above can be applied to selection of control in the same way. For example, a specific object(for example, "control a1") can be inputted by one- pressing of a specific button(for example, a button [*] as a control button), and another object(for example, "control b1") can be inputted by long pressing of the same button, "control a2", "control a3" can be selected by repeat pressing of the control button. Here, it is important that "control b2", "control b3, "control b4" can be selected not by long pressing but by repeat pressing after "control b1" is selected by long pressing of the control button. At this time, there is no ambiguity. For more detailed explanation, refer to the previously descried example, in which the representative character "A" is entered by long pressing and next characters are entered by short stroke. For convenience, an object inputted by combination of the specific button(for example, the button [1]) and "control b1" is called "B1", and it is as follows: "A1 = x + {control a1} = 1^*", "A2 = x + {control a2} = 1*^*", "A3 = x + {control a3} = 1 *^**", . . . , "B1 = x + {control b1} = 1*-", "B2 = x + {control b2} = 1*-*", "B3 = x +

{control b3} = 1 *-**", . . .

Of course, it is possible that control b1 , control b2, and control b3 can be selected by one long stroke, two long strokes, and three long strokes respectively. However, it is natural that frequent use of the long pressing is not good. As in the example of B1 , B2, B3, ..., for convenience, the input by the normal pressing after the one long pressing of the control button is called "repeat pressing after long pressing", and such control processing method is called "control processing method by repeat pressing after long pressing".

The group of A1 , A2,... means the "the existing control processing method" proposed by the present applicant, and the group of B1 , B2,... means the "control processing method by repeat pressing after long pressing". FIG. 14-1 shows the above in the form of a graph. The above is an example showing that the one long pressing is applied to only the first control selection as in the repeat selection method by long pressing of representative character. In addition, as shown in FIG. 14-2, it is possible to provide various expansions(for example, C group, and D group) through the long pressing. Here, of course, another object(e.g., character "y" in the above example) can be inputted by long pressing of not the control button but other button(e.g., the button [1] in the example). Furthermore, as the characters (In FIG. 1-1 , three characters) assigned to each button of the keypad can be recognized without ambiguity by the repeat selection method by long pressing(for example, "A = 2-", "B = 2-2", "C = 2-22"), other characters of other group can be inputted by repeat pressing after the repeat pressing or long pressing of the control button after one of the three characters are selected by manipulation of a numeral button. That is, in FIG. 1-1 , an character (including character, numeral, mother tongue, English alphabet, etc.) of a specific group(for example, group 1) can be entered by repeat pressing of the control button after "A" is entered, and then, another character of other specific group(for example, group 2) can be entered by the "repeat pressing after long pressing". Likewise, an character of the group 3 can be entered by combination of "B" input and the repeat pressing of the control button, and an character of the group 4 can be entered by the repeat pressing after long pressing. In FIG. 1-1 , it is also possible that characters of any group are entered by the repeat pressing of the control button, or it is also possible that characters of any group are entered by the repeat pressing after long pressing relative to three characters indicated on the button on the keypad and defined regulation(for example, "2-222"). As shown in the examples of Korean case(FIGs. 4-5 - 4-8) and of Japanese case(FIGs. 2-1 , 2-2), the control processing method by repeat pressing after long pressing is very useful or input of numerals and English characters (in languages which do not use English as mother tongue), which are used intermittently (infrequently). Hereinafter, applications by language will be described simply, and it can be applied to other languages. 7.2. Korean In FIGs. 4-5 - 4-8, the method can be applied to a case, in which affixed characters(for example, aspirated consonants, tense consonants, and the consonants assigned with vowel) are entered by repeat pressing of a button to which basic consonants are assigned, in a similar way(for example, "-1 =1", "τι _{= 1 1}„ ^««=, ₌I I I ^» OΓ " Π =1 "₁ "=1 =1 1", "71=111"). That is, if characters assigned to the same button are used in succession, characters after the second character are entered by long pressing. For example, when FIG. 4-5, if the continuous pressing of the button [1] is defined as "=. ", if "^~i " is used in succession, it can be recognized not as "^~ι + -ι " but as "=ι ", and at this time, if the second "π" is entered by long pressing, it may be recognized as "n + -ι ". That is, "^~ι " of the second syllable in "^■^^■7]-" is entered by long pressing. Also, "=ι " of the second syllable in ^u^f}" is entered by also long pressing, i.e., "=ι = 1-1". In this case, the method can be replaced with the control processing method for inputting affixed characters (one of the two methods or both of the methods can be applied). Another character excepting the representative character, for example, aspirated consonant or tense consonant, can be entered by long pressing. If the aspirated consonant is entered by long pressing, the tense consonant can be entered by combination of basic consonants. Of course, as described in the prior document, in case of " L 7} ", there is ambiguity as it may be recognized as "-$. --: 7l"("ι--:" is the final consonant of ".SL"). Likewise, the long pressing can be applied to input consonants(for example, "-ar" assigned to a button [o] in FIGs. 4-5 and 4-6) assigned with vowels. That is, "^■&" is entered by "0~". Based on FIG. 4-5 as presented in the prior document, I think that "input of -σ = 8**" or "-QΓ = 0**" or "-s- = 0(it is possible in specific cases presented in the prior document) is more effective than the input using long pressing. The input method of the consonants(for example, "^■&") assigned with vowel and the input method using long pressing can be applied selectively. Next, an example using the "control processing method by repeat pressing after long pressing" will be described. A representative Korean input system by the same inventor of the present invention is introduced on http://www.simplecode.net together with a simulator (in the homepage, input of old Korean characters, and concurrent input technology are also introduced), and, it is also introduced that input of numerals and English alphabets is possible without mode transition. In the general order of use frequency, input is performed by "mother tongue alphabet-various special characters (i.e., various kinds of symbols) - numerals-English alphabets". As shown in FIG. 4-5, there is no control button for inputting numerals and English. Here, it is possible that numerals are entered by combination of the numeral button and long pressing of the button [*](that is, "*~=1*~"). That is, numeral "1" is entered by "1*~". The button [*] is a vowel button at first, and at the same time, can be used as the control button. If the button [^*] is pressed once, it indicates a vowel " 1 ", but if the button [*] is pressed long once, it indicates a certain object("numeral control" in the example). English alphabets can be inputted by various methods. For example, "A" can be entered by "2#~", "B" can be entered by "2#-#", and "C" can be entered by "2#-##". "2#~#" is recognized not as "A" and vowel " 1 " (i.e., "A ] ") but as "B". The reason is that there is no syllable starting from vowel " 1 " in Korean(i.e., Korean restriction). If the long pressing and repeat pressing of the button [#] is used for another purpose, English alphabet is considered as succession character of numeral, and so, various definition such as "A = 2*-^*" or "A = 2^*~# (applying Cross Control Processing Method)" , "B = 2^*-^**" or "B =

2^*~##" can be made. Another example of the Control Processing Method by long pressing will be described. In FIGs. 4-5 - 4-8, if vowel "T", or "TT" is entered by repeat pressing of a button to which vowel "— " is assigned (i.e., "— = *", "T = **", "TT = ***^»)_ j js required to press the button [^*] maximum three times(the "maximum possible repeat number" is three) in succession to input a character. In this case, an aspirated consonant control can be selected by pressing of "the maximum possible repeat number+1 "(namely, four-pressing of the button [*]). That is, "=. " is entered by "1 ***^*". In this case, to reduce the number of pressing times, the aspirated consonant control is defined as "*~", and so, "=ι = ^~ι +{aspirated consonant control} = 1*-" can be defined. The above can be applied also to tense consonants and the consonants assigned with vowel, which are considered as affixed characters, in the same way. Furthermore, in the control processing method, it can be applied to all cases in which aspirated consonant control is selected by the repeat selection method. In addition, it is also possible that affixed characters can be inputted by long pressing and/or by the repeat pressing selectively. In this case, numerals can be inputted using other buttons(for example, the button [#]). 7.3. Japanese As shown in FIGs. 2-1 and 2-2, to input Japanese, lots of succession characters are inputted, and the system is useful for inputting the succession characters using the repeat selection method. That is, " > = 1" or "£> = 1-",

".,-, = 1~1^{» «} -5 = 1—11", " . = 1-11 1", and "fc = 1-1111" are defined. As described above, if characters assigned to the same button are used in succession, even though only the second character of the characters assigned to the button is entered by long pressing(namely, "1~") and the remaining characters are simply entered by one-pressing, there is no ambiguity. However, in FIGs. 2-1 and 2-2, Japanese words can be inputted with small input strokes without ambiguity by applying the succession (character) control processing method, and a succession (character) control processing method can be applied together with the repeat selection method by long pressing of representative character. FIGs. 2-1 and 2-2 shows the example in which affixed characters (i.e., transfomed characters or succession characters) are inputted not by using the repeat selection method by long pressing of representative character but by using succession (character) control processing method by long pressing. When the succession character is inputted by the control processing method in FIG. 2-2, and when affixed characters(for example, ". ", ". ") of the representative character (for example, "l±") is inputted as shown in FIG. 2-3, the number of the input strokes is too large, and the input is unnatural. In this case, the affixed character(for example, "Ji^'") of the representative character can be defined by long pressing of the corresponding button(a button [6] in FIG. 2-2). The remaining affixed characters can be defined by ".£ = (2^* + a predetermined control button(for example, the button [*]) = 6-*". Here, if the affixed character is not inputted using the long pressing of the button to which the characters are assigned, both the repeat selection method by long pressing of representative character and the succession character input method by control processing method can be applied. The affixed character (for example, ".£", ". ") can be inputted by combination of the long pressing of any control button(for example, the button [^*]) and the representative character "li". That is, ", (any affixed character) = .± + ^*~ = 6*~" and, ".£ (another affixed character) = .2 + *~* = 6*-*" can be defined. Even though there are lots of affixed characters of "( ", of course, it is possible that the affixed characters can be inputted without ambiguity by using the repeat pressing after long pressing, like "6*-**", "6*-***", and so on. The reason is that "**" and "*-*" indicate different objects(here, different controls) as "^*" and "^*~" indicate different objects. Here, it is important that any control can be selected by the long pressing, and different objects(for example, different controls) can be indicated without ambiguity by the repeat pressing of the button(for example, the control button) used or the long pressing after the button is pressed long once. Likewise, the what mentioned in Korean can be applied to input numerals and English alphabets. For example, "numeral 2=2#~", , "A = 2#~#", "B = 2#~##", . . .can be defined. 7.4. Chinese FIGs. 10-1 to 10-6 show the input of the romanization system(pinyin) of Chinese by the keypad without ambiguity(or little ambiguity). However, when "bb" is entered by "11" for short-cut input of words composed of "b_b_"(the underlined parts are pinyin syllables containing vowels), if the Chinese restriction in which "b" is not used in succession is applied, "11" will be recognized as "p". In this case, as described above, after "b" is entered by "1", the input value "1" is intentionally decided as "b" by a predetermined time delay or a specific manipulation(for example, press the button [>]), and then, "b" is entered by "1". It is the ambiguity problem generated when the characters assigned to the same button are entered in succession. Also, the problem can be solved by applying the control processing method by long pressing of representative character (the representative character "b" of the button [1] is entered by "1-"). For example, when character "b" is used in succession to input "bb", the second character can be entered by long pressing of the same button, namely, "bb" can be entered by "1-1-" or "11~". Next, input of numerals and characters will be described. Refer to FIG.

10-6. Input of English will not be described as it can be inputted without conversion after the input of English alphabets in Chinese. In FIG. 10-6, "numeral 1= 1*-"" can be defined. In FIG. 10-6, one-pressing of the button [^*] is recognized as "a", and "^*~"(long pressing) indicates different object("numeral control" in the example). In FIG. 10-6, it is preferable that the control processing method using long pressing and repeat pressing of the button [#] is applied to input various characters. For example, dot(.), comma(,), colon(:), and semi-colon(;) are associated to a button [2], and "dot(.) = 2#~", "commaO = 2#~#", "colon(:) = 2#~##", and "semi-colon(;) = 2#-###" can be defined. "2#~#" is recognized not as "dot(.) + i" but as "comma(,)" because of the Chinese restriction in which words of the romanization system(pinyin) is not started from "i" (refer to the Korean case). Likewise, "2#~##" can be also recognized not as "dot(.)+i+i" or "comma(,)+i" but as "colon(:)". If the user wants to input "dot(.)+i", after "2#~" is entered, the user decides "dot(.) after the predetermined time delay or by the specific manipulation(for example, press the button [>]), and then enters "i" by pressing "#". In FIG. 10-6, the button [^*] is not used when characters of a specific group are inputted. The reason is that there may be ambiguity when the "repeat pressing after long pressing" is applied as there are rarely pinyin words starting from "a", "e" or "o" in Chinese. Additionally, the reason is that numerals can be inputted by pressing the button [*] long once as only one numeral is associated (i.e., assigned) to one button. In FIGs. 10-1 to 10-5, it is always possible that numerals and characters can be inputted by the control processing method using repeat pressing after long pressing of the buttons [*] and [#] as there are no pinyin words starting from "i" and "u" assigned to the buttons [*] and [#]. It would be appreciated that the above method can be applied not only to the keypad of FIGs. 10-1 to 10-6 but also to other keypads in a similar way. 19.5. Languages using Roman characters European languages using roman characters include lots of affixed characters having affix to basic characters. In input of the European languages using roman characters, the affixed characters and special characters can be inputted by applying various kind of long pressing described above.

EFFECT OF THE INVENTION This invention shows that a target word or phrase can be inputted effectively using an index and applying a full input method(FIM). Moreover, this inventions shows an example in which a (language restricted) concurrent input method(CIM) is applied effectively. In addition, this invention proposes an input system for inputting various characters by combination of a long stroke of a specific button and a repeat stroke of the specific button, or by combination of a long stroke of a specific button and a repeat stroke of a control button.

Claims

WHAT IS CLAIMED IS:

1. A method for inputting characters from a keypad, the method comprising the steps of: (a) processing an input value by a concurrent input method(CIM), and at this time, processing a part of the input value as a predetermined full code; and (b) processing a part of the input value after the input value, which is recognized as the full code, as a simple code when the input value after characters produced by the full code does not form a proper syllable in a specific word.

2. The method for inputting characters from a keypad according to claim 1 , wherein the (b) step includes a step of separating a syllable excepting the final consonant of the syllable produced by the full code, and processing both of the input value used for inputting the final consonant and the input value, which is entered after the final consonant and considered as a simple code, as a simple code.

3. A method for inputting characters from a keypad, the method comprising a step of recognizing characters corresponding to repeat pressing of a button to which a number of characters are assigned, and at this time, recognizing the characters in a predetermined order in correspondence with a long stroke of the button and the repeat pressing of the button after the long stroke.

4. The method for inputting characters from a keypad according to claim 3, further comprising a step of recognizing a representative character assigned to the button by a short stroke or the long stroke of the button, and at this time, inputting the second succession character and the other succession characters by the long stroke of the button when the characters assigned to the same button are inputted in succession.

5. A method for inputting characters from a keypad, the method comprising a step of recognizing an appointed character by combination of an input value of a button for inputting a specific character and a long stroke of a specific control button.

6. A method for inputting characters from a keypad according to claim 5, further comprising a step of recognizing a predetermined character by combination of the long stroke of the control button and a normal short stroke of the control button.

7. A method for inputting characters from a keypad, the method comprising a step of recognizing a predetermined character by combination of a normal short stroke of a specific button and a long stroke of the specific button.

8. The method for inputting characters from a keypad according to claim 5, further comprising a step of recognizing the character(including number and specific character) by combination of the long stroke of a specific control button and a normal short stroke of the control button if the specific control button is a vowel button and a specific word is not started from a vowel.