JP2011096027A - Character input device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a character input device that further reduces a burden in character input. <P>SOLUTION: The character input device includes: a character input part; a kana-kanji conversion part creating a character string candidate; and a display part displaying the character string candidate. The character input device further includes: a document holding part holding an input document; and a sentence start decision part deciding whether a character insertion position in the input document held in the document holding part is located at the beginning of a sentence. The kana-kanji conversion part creates the character string candidate according to output from the sentence beginning decision part. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a character input device related to Japanese.

  2. Description of the Related Art Conventionally, there has been known an invention that reduces a user's input operation by predicting a character string to be input when creating a document and presenting the predicted character string as an input candidate. For example, when searching for a candidate for a predicted character string, Patent Literature 1 also searches for an attached word that can be connected to the previously selected candidate and outputs it as a candidate. If the character string “park” has been entered and then “de” is entered, in addition to “train, phone”, etc., the connectable adjunct “de, de, but” is also predicted as a candidate. Then, “Train, phone, and so on” is output.

  Further, Patent Document 2 shortens the time until a desired word is found. When a user inputs a character, first, several characters following the input character are predicted according to a probability table, and candidates are determined. Output. Here, when the user selects a desired character from the candidates, a character string following the next selected character is searched by dictionary and predicted. As described above, the word search is first performed without searching the word dictionary, and then the dictionary search is performed after narrowing down. Therefore, the hit rate can be improved while shortening the processing time.

  In Patent Document 3, as a character string input candidate input at the time of creating an e-mail, when a document (a collection of a plurality of sentences) is started to be written, a portable character input device that presents a predetermined character string is disclosed. An apparatus is described that presents a character string described in another e-mail including at least an e-mail transmission destination as an input candidate when the telephone inputs the title or text of the e-mail.

JP-A-9-274613 JP-A-8-235318 JP 2006-209558 A

  A specific word or phrase such as a greeting is often used in the beginning of writing a document (a collection of a plurality of sentences), particularly in e-mail. In addition, words that express personality such as “I” and “Ms.”, conjunctions such as “But”, nouns that became the topic of the original document, etc. are often used.

  However, since the devices of Patent Documents 1 and 2 output character conversion candidates by simple prediction without considering the character input position in the document, it is difficult to present a desired conversion candidate and the input load of the user May be large. Further, Patent Document 3 presents input candidates when the input cursor is at a predetermined position, and can detect the beginning of a document. However, for example, it is difficult to detect the beginning of a sentence.

  The present invention has been made in view of such circumstances, and by determining the start of writing of a document (a collection of a plurality of sentences) and the beginning of a sentence, and generating prediction candidates specialized for them, character input can be performed. An object of the present invention is to provide a character input device that can further reduce the load.

  A character input device according to the present invention includes a character input unit, a kana-kanji conversion unit that generates a character string candidate, and a sentence start determination unit that determines a sentence start in a character input device that has a display unit that displays character string candidates. The kana-kanji conversion unit creates a character string candidate in response to an output from the sentence determination unit.

  According to the character input device of the present invention, when the character input location is the beginning of a sentence, even if the user does nothing, it is determined whether it is the beginning of the sentence or other than the beginning of the sentence. Since the candidates are displayed, it is possible to facilitate the user's input operation.

  The character input device of the present invention includes a character input unit, a kana-kanji conversion unit that creates a character string candidate, and a document holding unit that holds an input document in a character input device that has a display unit that displays the character string candidate. And a sentence head determining unit that determines whether or not the character insertion position in the input document held in the document holding part is a sentence head, and according to the output from the sentence head determining unit, the kana-kanji conversion unit Character string candidates are created.

  According to the character input device of the present invention, since the document holding unit holds the input document, even if the cursor position is moved and the character input position is changed, the sentence head determining unit can change the changed character input. It becomes possible to determine whether the position is the beginning of a sentence. As a result, when the changed character input position is the beginning of a sentence, input candidates suitable for the beginning of the sentence are displayed, so that it is possible to facilitate the user's input operation corresponding to the movement of the cursor position. .

  The character input device of the present invention is characterized in that the kana-kanji conversion unit has a dictionary unit, and the words stored in the dictionary unit include information on the sentence head.

  According to the character input device of the present invention, when a word stored in the dictionary has information about a sentence head, the kana-kanji conversion unit preferentially uses a character string candidate that is likely to be used at the beginning of a sentence when creating a character string candidate at the beginning of the sentence. Can be created.

  Further, in the character input device of the present invention, when the sentence head determination unit determines that the input character position is the sentence head, the kana-kanji conversion unit gives priority to a word having information on the sentence head among the words stored in the dictionary part. Character string candidates are created.

  According to the character input device of the present invention, when a user performs an operation of inputting a character, a word that is likely to be used at the beginning of a sentence is preferentially displayed as a character string candidate, thereby facilitating the user's input operation. It becomes possible.

  In the character input device of the present invention, the kana-kanji conversion unit is input when the word input from the character input unit is not registered in the dictionary unit when the sentence start determination unit determines that the input character position is the beginning of the sentence. It is characterized in that a word is registered in the dictionary unit together with information on the sentence head.

  According to the character input device of the present invention, the word input from the character input unit is stored in the dictionary when the sentence start determination unit determines the sentence start and displays the input candidates. It is possible to learn words that are not registered and that are easy to use at the beginning of the sentence, and the number of character string candidates increases, so that the user's input operation can be made easier.

  The character input device of the present invention is characterized in that the kana-kanji conversion unit has a dictionary unit for storing words having information on the beginning of a sentence and a dictionary unit not having information on the beginning of a sentence.

  According to the character input device of the present invention, while reducing the memory usage in the character input device of the present invention, the kana-kanji conversion unit preferentially uses the character string candidates that are likely to be used at the beginning of the sentence when creating the character string candidates at the beginning of the sentence. It becomes possible to create.

  Further, in the character input device of the present invention, the sentence start determination unit distinguishes between the sentence start and the document start, and the kana-kanji conversion unit corresponds to the sentence start and the document start according to the output from the sentence determination unit. Character string candidates are created.

  According to the character input device of the present invention, the sentence head determining unit distinguishes between the sentence head and the document head, and the kana-kanji conversion unit creates character string candidates according to the determination result of the sentence head determining unit. By displaying input candidates suitable for each sentence head, the candidates are narrowed down and the user's input operation can be facilitated.

  The character input device of the present invention has a dictionary creation unit that creates a temporary dictionary, and the dictionary creation unit extracts words used at the beginning of a sentence from application-related information and registers them in the dictionary unit. Yes.

  According to the character input device of the present invention, the number of character string candidates is increased by using words that are easy to use at the beginning of a sentence from other than the built-in dictionary, so that the user can easily perform input operations.

  In the character input device of the present invention, the application-related information is a received mail, and the dictionary creation unit registers the head of the received mail in the dictionary unit.

  According to the character input device of the present invention, when the dictionary creation unit learns the word written at the top of the received mail in the dictionary and creates a reply mail, the character string candidate is displayed at the top of the reply mail. . Since the word used at the head of the received mail is often used again for the reply mail, it is possible to facilitate the user's input operation.

  In the character input device of the present invention, the application-related information is position information, the character input device has a position information acquisition unit that acquires position information, and the dictionary creation unit registers words related to the position information in the dictionary unit. It is characterized by that.

  According to the character input device of the present invention, the dictionary creation unit acquires position information from the position information acquisition unit, learns from the dictionary, and displays the position information learned at the beginning of the document at the time of document creation as a character string candidate. At the beginning of a sentence such as a mail, a word related to the position is often input, so that the user can easily perform an input operation.

  In the character input device of the present invention, the application-related information is time information, the character input device has a time information acquisition unit that acquires time information, and the dictionary creation unit registers words related to the time information in the dictionary unit. It is characterized by that.

  According to the character input device of the present invention, the dictionary creation unit acquires time information from the time information acquisition unit, learns it from the learning dictionary, and displays the time information learned at the beginning of the document at the time of document creation as a character string candidate. Since words related to time are often input at the beginning of a document, it is possible to facilitate the user's input operation.

  In the character input device of the present invention, the application-related information is phone book information, the character input device has a phone book, and the dictionary creation unit registers words related to the information registered in the phone book in the dictionary unit. It is characterized by that.

  According to the character input device of the present invention, the dictionary creation unit acquires the phone book information from the phone book database mounted on the mobile phone, the information terminal, etc., learns it from the dictionary, and learns at the beginning of the document at the time of document creation Display phone book information as a character string candidate. Information that is easy to use at the beginning of emails, such as nicknames and birthdays, is registered in the phone book as phone book information, which can be searched and displayed as character string candidates for easy user input operations. It becomes possible to.

  In the character input device of the present invention, the application related information is phone book information and personal pronoun information, the phone book information has a correspondence table of the relationship between the mail address and the user, and the personal pronoun information is the personal pronoun and user. The character input device is characterized in that the dictionary creation unit associates an e-mail address and personal pronoun with the same relevance with the user and registers them in the dictionary.

  According to the character input device of the present invention, the association with the user is associated with the mail address registered in the telephone directory, and the association with the user is associated with the personal pronoun registered in the personal pronoun correspondence table. By creating a mail to an address associated with the user, the personal pronouns associated with the same user are displayed as input candidates, so not only reducing the user's input operation, It becomes possible to display candidate pronouns corresponding to the relevance with the user.

  According to the character input device of the present invention, when characters are input at the beginning of a sentence or at the beginning of a document, it is possible to preferentially display more suitable candidates by predictive conversion, thereby further reducing user input processing. be able to.

  Further, according to a further invention of the present invention, after moving to the beginning of a sentence or the beginning of a document after the end of a sentence is confirmed or by moving a cursor, even if a character string for conversion is not input, a candidate is displayed. Input processing can be further reduced.

  Further, according to a further invention of the present invention, a word to be displayed at the beginning of a sentence or at the beginning of a document is created from a reference document such as a received mail or data such as position information, time information, and telephone directory information. Thus, words that are more suitable for use scenes can be put out as candidates. As a result, user input processing can be further reduced.

The block diagram which shows the principal part structure in Embodiment 1 of the character input device of this invention. Flow chart of sentence head determination unit, kana-kanji conversion unit and display unit in embodiment 1 Examples of conversion dictionary and learning dictionary in embodiment 1 The block diagram which shows the principal part structure in Embodiment 2 of the character input device of this invention. Flow chart of sentence head determination unit, kana-kanji conversion unit and display unit in embodiment 2 Explanatory drawing about the sentence processed in Embodiment 2 Examples of conversion dictionary and learning dictionary in embodiment 2 The block diagram which shows the principal part structure in Embodiment 3 of the character input device of this invention. FIG. 10 is a diagram for explaining cursor movement according to the third embodiment. Flow chart of sentence head determination unit in embodiment 3 The block diagram which shows the principal part structure in Embodiment 4 of the character input device of this invention. Flow chart of kana-kanji conversion unit in embodiment 4 The block diagram which shows the principal part structure in Embodiment 5 of the character input device of this invention. Explanatory drawing of the telephone directory and dictionary in the fifth embodiment Explanatory drawing of the telephone directory and personal pronoun correspondence table in Embodiment 5

  Below, the form for implementing this invention is demonstrated. The present invention is not limited to these embodiments, and can be implemented in various modes without departing from the scope of the invention.

[Embodiment 1]
The first embodiment will be described. FIG. 1 is a block diagram showing a main configuration of the character input device according to the first embodiment of the present invention. As shown in FIG. 1, the character input device according to the first embodiment includes a character input unit 101, a sentence head determination unit 102, a kana-kanji conversion unit 103, and a display unit 104. The character input unit 101 converts an input from an input device (input means) such as a keyboard into a hiragana string. The conversion method includes a 2-touch method, a 5-touch method for mobile phones, and a Roman character conversion used in personal computers. Input from the character input unit 101 is sent to the sentence head determination unit 102.

  FIG. 2 shows an operation flow from sentence start determination to candidate display by the sentence start determination unit, kana-kanji conversion unit, and display unit shown in FIG. In the figure, those with S at the head indicate steps. For example, S201 is step 201. When the character input unit 101 confirms the input of “carriage return” or a punctuation mark “.” In the character input unit 101, the sentence head determination unit 102 determines that the state at that time is a sentence head (Yes in S201 in FIG. 2). Even when a mark such as “!” Or “?” Or a word or pictograph such as “Bye” that means the end of a sentence is confirmed, the next input may be determined as the beginning of the sentence. A state where nothing is input, that is, the head of the page is also determined as the beginning of a sentence. Thus, when it is determined Yes in S201, the determination result of the sentence head and the input information from the character input unit are sent to the Kana-Kanji conversion unit (S202 in FIG. 2).

  The kana-kanji conversion unit 103 selects candidates from the dictionary unit 105 having the conversion dictionary 110 and the learning dictionary 111 using the determination result and the input information sent from the sentence head determination unit 102, and displays a candidate list on the display unit 104. (S204 in FIG. 2), the display unit 104 displays the candidates (S206 in FIG. 2). For example, if it is determined that the sentence head is the dictionary part of the dictionary shown in FIG. 3, “Kyo (Today)” and “Boku (I)” whose sentence dictionary flag is ○, “And (and)”, “Landing (Netherlands)”,... Are displayed as candidates in order from “Ore (I)” whose sentence flag is “O” and those with the highest sentence frequency in the conversion dictionary.

  On the other hand, if the input that can be determined to be a sentence head is not confirmed from the character input unit 101, the sentence head determination unit 102 determines that the determination result is a sentence (No in S201 of FIG. 2). Even when punctuation marks, symbols, and pictograms that are used when words are continued, such as “,” and “˜”, the next input may be determined to be in the sentence. In this way, if it is determined No in S201, the determination result that is in the sentence and the input information from the character input unit are sent to the kana-kanji conversion unit 103 (S203 in FIG. 2).

  The kana-kanji conversion unit 103 selects candidates from the dictionary unit 105 using the determination result and input information sent from the sentence head determination unit 102, and sends a candidate list to the display unit 104 (S205 in FIG. 2). The display unit 104 displays the candidates (S207 in FIG. 2). For example, if it is determined that the sentence is in the sentence, and the dictionary part is composed of the dictionary shown in FIG. 3, the sentence flag of the learning dictionary is “play” (for play), “do not go (does not go)”, “ . "" Hidoma (in his spare time) ", and the beginning flag of the conversion dictionary is empty" swimming (swimming) "" and (and) "" ande (rice) "" oranda (the Netherlands) " In order from the one with the lowest sentence head frequency, “span (swim)”, “span (rice)”... Are created as candidates.

  A dictionary installed in the present application will be described. In general, the dictionary data built into the electronic device includes a standard built-in dictionary common to everyone, a learning dictionary that is automatically created and stored to reflect the user's usage record, A user dictionary that can register corresponding word / phrase pairs can be considered, but the conversion dictionary in the present invention means a standard built-in dictionary such as a dictionary necessary for Kana-Kanji conversion, a national language dictionary, a Han-Japanese dictionary, an English-Japanese dictionary, The learning dictionary means a dictionary that is automatically created and stored in accordance with the usage record of the user. In FIG. 1, the conversion unit and the learning dictionary are built in the dictionary unit 105, but the present invention is not limited to this, and other dictionary data can be provided.

  FIG. 3 shows an example of the conversion dictionary and the learning dictionary used in FIGS. In the conversion dictionary, frequency information, sentence head frequency, and sentence head flag are added in addition to general information such as reading / notation / part of speech information. The word stored in the dictionary has information about the beginning of the sentence, such as frequency information, beginning frequency, and beginning flag, so that the Kana-Kanji conversion unit gives priority to character string candidates that are likely to be used at the beginning of the sentence when creating character string candidates at the beginning of the sentence It becomes possible to create.

  The frequency information indicates the ease of use of the word, and the sentence head frequency indicates the ease of use at the sentence head. In FIG. 3, the larger number is shown as the higher frequency. In this case, the sentence frequency is set to six stages from 1 to 6, but the present application is not limited to this. The head flag of the conversion dictionary is a flag that is set in advance for a word that is easily used at the head of the sentence.

  On the other hand, the learning dictionary has a sentence flag as information when storing words used by the user. As shown in FIG. 3, the kana-kanji conversion unit in the character input device of the present invention stores the word used at the beginning of the sentence by the user as a circle in the learning dictionary as shown in FIG. Remember. The words stored in the learning dictionary are temporarily stored, and are sequentially stored as the user uses them. The operation of the kana-kanji conversion unit to be stored in the dictionary is executed by a control circuit not shown in FIG. Here, the control circuit may be arranged inside the kana-kanji conversion unit or may be outside.

  By constructing the dictionary as described above, it is possible to distinguish between a conversion dictionary that holds words and sentence head information for a long time, and a learning dictionary that maintains a short term. However, the present application is not limited to this, and the head flag and the head frequency of the conversion dictionary may be stored according to the use of the user, or the head frequency information may be included in the learning dictionary.

  In order to make it possible to update information such as the sentence head flag and the sentence head frequency in the conversion dictionary, the read / notation / part of speech information is stored in a ROM (Read only Memory), and the head flag and head frequency are stored in a RAM (Random Access Memory). Or by storing them in a nonvolatile memory and associating them with each other.

  Moreover, the creation method of the display candidate list by the kana-kanji conversion unit in the present application can be arbitrarily set. For example, giving priority to all learning dictionary data over the conversion dictionary, the word with the head flag added as the top priority candidate, the word stored in the learning dictionary without the head flag next set as the candidate, and then converted It is possible to search the dictionary data and select candidates in the order of the word with the head flag in the conversion dictionary, the word with the highest head frequency, and the word with the higher frequency information. The added word is a candidate, but the conversion dictionary does not use the sentence head flag, uses the sentence head frequency and frequency information, uses only the frequency information, or uses the sentence head flag, the sentence head frequency, or the frequency information. It is also possible.

  In addition, the display number may be limited so that the number of candidate characters does not become enormous, the candidate may be turned, and the display number can be set automatically or manually according to the size of the display unit. You may do it. In addition to displaying only the notation, it is possible to display not only the reading but also both. In addition, in general electronic dictionary data search and the like, candidates are output with priority given to the stored contents of the learning dictionary over the conversion dictionary, so in the example of FIG. ) ”Is displayed preferentially, and then“ I (I) ”and“ Swim (swimming) ”can be displayed in this order.

  According to the present invention, when an input operation that means the beginning of a sentence is performed from the character input unit, words that are likely to be used at the beginning of the sentence before the user performs an operation of inputting characters are preferentially input from the learning dictionary or the conversion dictionary. By displaying as, it becomes possible to facilitate the user's input operation. In addition, the display unit outputs words that are likely to be used at the beginning of the sentence to the display unit. If characters are input in this state, the candidate display may be canceled and normal character candidates may be displayed. You may narrow down and display candidates that match the input characters.

  In any case, the newly determined word is stored in the learning dictionary or the conversion dictionary together with the sentence head flag and the sentence head frequency information. This makes it possible to learn words that are not registered in the built-in dictionary and are easy to use at the beginning of the sentence, and increase the number of character string candidates, thereby facilitating the user's input operation.

[Embodiment 2]
The second embodiment is a modification of the first embodiment. In the character input device according to the first embodiment, the sentence start determination unit determines whether or not the sentence start is based on an input from the character input unit, and the kana-kanji conversion unit determines word candidates to be displayed according to the determination result of the sentence start determination unit. The priority was changed. On the other hand, in the character input device according to the second embodiment, the sentence head determination unit 402 determines whether to distinguish the sentence head, the sentence head, and the sentence.

  An example is shown in FIG. 6, and the sentence head, sentence head, and definition in the sentence will be described. “Line feed” in FIG. 6 indicates the line feed position in the sentence for the sake of explanation, and does not indicate the character “line feed”. A document is distinguished by paragraphs (with a space at the beginning of a line) and is a collection of simple sentences. A new document is added by inputting a character after a line feed is input. A sentence is a simple sentence and is mainly distinguished by “.”. The second document in FIG. 6 has three sentences, sentence B, sentence C, and sentence D. Therefore, in FIG. 6, three documents and five sentences are described. The sentence heads are sentence A, sentence B, and sentence E. The sentence heads are sentence A, sentence B, sentence C, This is the beginning of sentence D and sentence E. In-sentence does not apply to the beginning of a document or sentence. It can be said to be in the middle of a single sentence. However, it does not mean the middle of a document.

  FIG. 4 is a block diagram showing the main configuration of Embodiment 2 of the character input device of the present invention. As shown in FIG. 4, the character input device according to the second embodiment includes a character input unit, a sentence head determination unit, a kana-kanji conversion unit, and a display unit, similarly to the character input device described in the first embodiment. Here, the character input unit and the display unit perform substantially the same operations as in FIG.

  FIG. 5 shows an operation flow from sentence start determination to candidate display by the sentence start determination unit, kana-kanji conversion unit and display unit shown in FIG. When the input of the character input unit is confirmed as “line feed”, the sentence head determination unit 402 determines that the state at that time is the document head (Yes in S501 in FIG. 5). In FIG. 6, this is the state when, for example, the input of sentence A or sentence D is completed and “new line” is entered. After a line feed, the cursor moves to the beginning of sentence B or sentence E by space input. In this state, the state of the document is maintained, and when a character other than a space or a line feed is input, the state of the sentence head of the document is changed to the state in the sentence.

  The head of the page is also determined as the beginning of the document. The top of the page is when a new document is created, and is the state immediately before the sentence A is input in FIG. The state of the beginning of a document is not changed by a space or line break.

  If it is not the beginning of the document (No in S501 of FIG. 5), it is next determined whether it is the beginning of the sentence. The determination is mainly made when the input of the character input unit 101 is confirmed as “.” And no line feed is input (Yes in S502 of FIG. 5). If neither a line feed nor “.” Is input, it is determined that the sentence is in text (No in S502 of FIG. 5). Here, “.” Is used as an example to determine the beginning of a sentence, but a mark such as “!” Or “?” Or a word meaning a sentence delimiter such as “Bye” may be used. However, pictograms and figures may be used.

  In addition, the sentence is determined to be in the sentence when it does not apply to the beginning of the document nor the beginning of the sentence, but it is in the sentence using punctuation marks, symbols, and pictograms that are used when words such as "," and "~" continue. Sometimes it is determined.

  The sentence head determination unit 402 sends the determination result and input information to the kana-kanji conversion unit 403 (S503, S504, and S505 in FIG. 5). The kana-kanji conversion unit 403 selects a candidate from the dictionary unit 405 having the conversion dictionary 410 and the learning dictionary 411 using the determination result and the input information sent from the sentence head determination unit 402, and displays a candidate list on the display unit 404. (S506, S507, S508 in FIG. 2). Finally, the display unit displays the candidates (S509, S510, S511 in FIG. 5).

  Since the dictionary installed in the second embodiment has the same function as the dictionary described in the first embodiment, the description of overlapping parts is omitted. The difference from the first embodiment is that there are three types of determination results of the sentence head determination unit 402 such as the sentence head, the sentence head, and the sentence, and these three kinds can be handled separately.

  4 and 5 will be further described. When it is determined Yes in S501, the determination result of the beginning of the document and the input information from the character input unit 101 are sent to the Kana-Kanji conversion unit (S503). The input information from the character input unit 101 is a line feed indicating that the determination result is the head of the document.

  The kana-kanji conversion unit finalizes the line feed (actually performs a line feed), and selects and displays candidates corresponding to the document head state from the dictionary unit 405. For example, if the dictionary part is composed of the dictionary shown in FIG. 7, “Kyo (Today)” with the sentence flag of the learning dictionary being ◎ and “I (I)” with the sentence flag of the conversion dictionary being ◎. Are sent to the display unit as candidates (S506), and the display unit displays the sent candidates (S509).

  Here, an example is shown in which, when it is determined that the sentence begins with a sentence, only words whose sentence start flag in the learning dictionary or conversion dictionary is ◎ are displayed as candidates. However, a method for creating a display candidate list by the kana-kanji conversion unit is described in the embodiment. It can be arbitrarily set similarly to 1. For example, a word with a head flag of ◎ is a top priority candidate, but a word with a head flag of a learning dictionary or conversion dictionary with a head flag or a word without a head flag is also output as a candidate. Good.

  If NO is determined in S501 and YES is determined in S502, the determination result of the sentence head and the input information from the character input unit are sent to the kana-kanji conversion unit (S504). The input information from the character input unit is “.”, “?”, Etc., indicating that the determination result is the beginning of the sentence. The kana-kanji conversion unit confirms “.” Or “?” (Actually displays “.” Or “?” As a confirmed character), and selects a candidate corresponding to the sentence head from the dictionary unit 405. And display. For example, if the dictionary part is composed of the dictionary of FIG. 7, “I (I)” whose sentence flag in the learning dictionary is ○ and “and (and)” whose sentence flag in the conversion dictionary is ○. The candidate is sent to the display unit (S507), and the display unit displays the sent candidate (S510).

  Here, an example is shown in which if a sentence is determined to be the beginning of a sentence, only words with a sentence head flag of ○ in the learning dictionary or conversion dictionary are displayed as candidates. However, for how to create a display candidate list by the kana-kanji conversion unit, It can be arbitrarily set in the same manner as described in the case of the sentence head.

  When it is determined No in S502, the determination result that is in the sentence and the input information from the character input unit are sent to the kana-kanji conversion unit (S505). The input information from the character input unit is “,” “˜”, etc., indicating that the determination result is in the sentence.

  The kana-kanji conversion unit confirms “,” and “˜” (actually displays “,” and “˜” as confirmed characters), and selects candidates corresponding to a state in the sentence from the dictionary unit 405. indicate. If the dictionary part is composed of the dictionary of FIG. 7 at this time, the sentence flag of the learning dictionary is “play” (for play), “do not go (does not go)” “.” ”), And in order from the least frequent sentence in the conversion dictionary,“ Gone (rice) ”,“ Land (Netherlands) ”... Are sent as candidates to the display unit (S508), and the display unit is sent. Candidates are displayed (S511). Here, as in the case of the document head described above and the case of the sentence head, the method of creating the display candidate list by the kana-kanji conversion unit can be arbitrarily set.

  FIG. 7 shows an example of the conversion dictionary and the learning dictionary of the second embodiment. The difference from FIG. 3 is that the sentence head flag is subdivided. In FIG. 7, ◎ is a flag for preferential display as a candidate at the beginning of a document. ○ is a flag for preferential display as a candidate at the beginning of a sentence.

  In addition, the display unit outputs a word that is easy to use according to the document head, the sentence head, and the state in the sentence to the display unit, but if the desired word is not displayed in the candidate, the user Can cancel the candidate display by performing a normal character input by an input operation, and can perform a normal input operation. At this time, the kana-kanji conversion unit of the character input device of the present invention adds or changes the sentence head information to the newly input word and stores it in the learning dictionary or conversion dictionary.

  For example, if a new word is input at the beginning of the document, the beginning flag ◎ is added to the newly input word, and if a new word is input at the beginning of the sentence, the beginning of the sentence is added to the newly input word. Each flag ○ is set and stored in the learning dictionary or the conversion dictionary. When a new word is input in the state of the sentence, the learning dictionary and the conversion dictionary are updated by deleting the head flag of the newly input word or lowering the head frequency.

  In this way, it is possible to learn words that are not registered in the built-in dictionary and are easy to use at the beginning of sentences, and to make it difficult to select candidates that are difficult to use at the beginning of sentences, Since the quality of candidate creation is improved, user input operations are facilitated.

  In the case of a sentence, use only the normal dictionary function without using information such as the sentence head flag and sentence head frequency, or use information such as the sentence head flag and sentence head frequency together with the normal dictionary function. Is also possible. If it is determined that it is in a sentence, it is possible to display candidates that have learned word connections, and candidates that have learned word connections and candidates in sentences can be displayed. Without waiting for the user's input, if there is a user's input, it is also possible to perform priority candidate display and refined search using the functions of the normal learning dictionary and conversion dictionary. Further, words that have information in the conversion dictionary and exclude words that appear only at the beginning of the sentence, such as impression verbs and part of conjunctions, may be used as candidates in the sentence.

  As explained above, the sentence head judgment unit distinguishes between the sentence head and the document head, so that input candidates suitable for the sentence head and the document head can be displayed. It is possible to facilitate the input operation of the user.

[Embodiment 3]
A third embodiment will be described. FIG. 8 is a block diagram showing the main configuration of the character input device according to the third embodiment of the present invention. As shown in FIG. 8, the character input device according to the third embodiment includes a character input unit 801, a sentence head determination unit 802, a kana-kanji conversion unit 803, a dictionary unit 805, a conversion dictionary 810, a learning dictionary 811, a display unit 804, and document holding. A unit 806 and a determination information unit 808 are included. Each configuration also has the functions described in the first and second embodiments, but the description thereof is omitted.

  The third embodiment is an example when a cursor movement operation is performed from the character input unit 801. The cursor can be moved by a sensor device such as a cross key or a jog dial, or by moving a plurality of keys in conjunction with each other such as pressing the Shift key and an arrow key at the same time. There is no particular limitation on the method, and any method may be used. Further, here, the character input unit 801 is similarly defined as an operation unit for performing an operation for moving the cursor, a character key for inputting characters, a touch panel, and the like.

  As shown in FIG. 8, this configuration includes a document holding unit 806 that holds a created input document 807, and holds input sentences and documents. When the cursor input is instructed from the character input unit 801, the display unit moves the cursor position in response to the instruction. At this time, the document holding unit 806 calculates the character insertion position in response to the cursor movement instruction from the character input unit, and sends the character string information immediately before the moved cursor to the sentence head determining unit 802.

  The sentence head determination unit 802 determines whether or not it is a sentence head from the information on the character string insertion location according to the cursor movement instruction from the character input unit and the character string information sent from the document holding unit 806, and the determination result is used as the determination information unit. Stored in 808. The determination information unit 808 sends a determination result to the kana-kanji conversion unit 803.

  Processing when the cursor position is moved will be described with reference to FIG. The movement of the cursor position means an operation of moving the cursor to correct the document when the document creation has already been completed up to the sentence J shown in FIG. In FIG. 9, (A), (B), (C), (D), and (E) are not for the input characters but for explaining the position of the movement destination of the cursor. When the cursor is moved to (A), (B), (C), (D), or (E), the document holding unit 806 holds the document in FIG. 9 and sends the immediately preceding character information at each position to the sentence head determining unit 802. .

  The immediately preceding character information is “space” when the cursor is at the position (A), “,” at the position (B), “space” at the position (C), and at the position (D). Is “input character”, and “.” At the position (E). The sentence head determination unit 802 determines whether or not each cursor position is a sentence head from the character information immediately before, and stores it in the determination information unit 808 as determination information.

  The determination information has at least three states: the beginning of the file, the sentence, and the end of the sentence, and the initial state is the beginning of the file. In FIG. 9, “Space” in (A) is the head of the file, “Space” in (C) and “.” In (E) are the end of the sentence, and “,” and (D in (B). ) Is stored in the determination information unit 808 with the text “input character” in the sentence.

  When the judgment information is sent to the kana-kanji conversion unit, the last part of the sentence is replaced with the beginning of the sentence, and a dictionary search is performed. It is also possible to recognize “line feed” before “space” in (C) and distinguish it from (E), with (C) as the beginning of the document and (E) as the beginning of the sentence. For the selection of conversion candidates in these kana-kanji conversion units, the same method as in the first or second embodiment can be applied.

  FIG. 10 is a processing flow of the sentence head determination unit 802 in the present embodiment. In step S1001, the sentence head determination unit 802 determines whether the input character from the character input unit 801 is a character input to be converted by the kana conversion unit. In the case of a conversion target character (Yes in S1001), the sentence head determination unit 802 sends input character information to the Kana-Kanji conversion unit 803 without changing the state (S1002), and the Kana-Kanji conversion unit converts the conversion dictionary 810 installed in the dictionary unit 805, A candidate is selected from the learning dictionary 811 and sent to the display unit 804. The display unit 804 displays the candidate and waits for a new input from the user. For example, when “Nihon” is input from the character input unit 801, “Japan, two, two” or the like is displayed.

  If it is not a conversion target character (No in S1001), the sentence head determination unit 802 determines whether or not the input is confirmed in S1003. “Confirmed input” corresponds to a case where a line feed or a space is input from the character input unit 801, or a case where a word is determined from a candidate selected from the Kana-Kanji conversion unit and output from the display unit. For example, this is the case where the punctuation mark “.” Is entered and confirmed by return, or “Nihon” is entered and “Japan” is selected from “Japan, 2 or 2” and determined.

  When the input is confirmed in this way, the sentence head determining unit 802 determines whether or not the confirmed punctuation (such as “.”) Or the word is a word representing the end of the sentence in S1004. If it is determined that the sentence ends (Yes in S1004), the determination information unit 808 sets the last state of the sentence (S1006), and the determination information is sent to the kana-kanji conversion unit 803. The kana-kanji conversion unit changes the state from the end of the sentence to the beginning of the sentence and outputs the candidate. On the other hand, when the sentence head determination unit 802 determines that the word does not represent the end of the sentence (No in S1004), the determination information unit 808 sets the state in the sentence (S1007), and the determination information is sent to the kana-kanji conversion unit 803. The kana-kanji conversion unit performs candidate output corresponding to the state in the sentence.

  In S1003, the fact that the input is not confirmed means that the cursor is instructed from the character input unit 801. In this case, the character immediately before the position where the cursor has moved and stopped is acquired from the document holding unit 806 (S1005). The sentence determination unit determines whether the character acquired from the document holding unit 806 is a word representing the end of the sentence, such as a line feed or a punctuation mark “.” (S1008).

  When it is determined that the sentence ends (Yes in S1008), the determination information unit 808 sets the last state of the sentence (S1009), and the determination information is sent to the kana-kanji conversion unit 803. The kana-kanji conversion unit changes the state from the end of the sentence to the beginning of the sentence and outputs the candidate. On the other hand, when the sentence head determination unit 802 determines that the word does not represent the end of the sentence (No in S1008), the determination information unit 808 sets the state in the sentence (S1007), and the determination information is sent to the kana-kanji conversion unit 803. The kana-kanji conversion unit performs candidate output corresponding to the state in the sentence.

  As described above, FIG. 10 shows a case where a character is input from the character input unit and a conversion candidate is displayed, a case where the input is confirmed and it is determined whether the sentence is at the end or in the sentence, and the cursor is moved. The flow in the case where it is determined whether the moved position is at the end of the sentence or in the sentence is shown. In accordance with each case, the kana-kanji conversion unit selects a converted character candidate based on the determination information sent from the determination information unit 808. The operation of the kana-kanji conversion unit at this time is as described in the first and second embodiments. In addition, as described in the explanation of FIG. 9, even when the determination is made from the cursor movement position, it is possible to distinguish the beginning of the document from the beginning of the sentence, and the confirmation process is performed according to the state of the beginning of the document or the beginning of the sentence. When this is done, it is possible to learn a word for which the kana-kanji conversion unit of the character input device of the present invention has been established in the sentence learning dictionary.

  The invention according to the third embodiment displays a word that is likely to be used at the beginning of a sentence in response to a cursor movement instructed from a character input unit, by preferentially displaying it as a character input candidate from a learning dictionary or a conversion dictionary. The input operation is facilitated. In the present invention, the sentence head is determined using the input confirmation information and the character information immediately before the cursor position. Therefore, the determination can be made at any cursor position, and the character input candidates can be determined not only at the head of the document but also at the head of the sentence. It is possible to display.

[Embodiment 4]
A fourth embodiment will be described. FIG. 11 is a block diagram showing the main configuration of the character input device according to the fourth embodiment of the present invention. In FIG. 11, except for the dictionary unit 1105, the description is omitted because it is almost the same as the first, second, and third embodiments. This configuration is characterized in that the dictionary unit 1105 has a sentence head conversion dictionary 1112 and a sentence head learning dictionary 1113 that are used at the head of a sentence separately from the conversion dictionary 1110 and the learning dictionary 1111 used in the sentence.

  FIG. 12 is a flowchart illustrating the kana-kanji conversion unit. In S1201, it is determined from the information from the sentence determination unit whether or not the sentence begins. If it is determined that the sentence is the beginning of the sentence (Yes in S1201), candidates are created from the conversion dictionary for sentence head / learning dictionary for sentence beginning in descending order of the sentence head frequency in S1202, and then the head frequency is determined from the conversion dictionary / learning dictionary in S1203. Create candidates in descending order. The created candidates are sent to the display unit and displayed. On the other hand, if it is not the beginning of a sentence (No in S1201), candidates are created in descending order of sentence head frequency from the conversion dictionary / learning dictionary in S1203. Needless to say, the document head dictionary and the sentence head dictionary may be further divided.

  Thus, by constructing the conversion dictionary and the learning dictionary for the sentence head separately from the dictionary for the sentence, the sentence head frequency information of the conversion dictionary, the sentence head flag, and the head flag of the learning dictionary are not required as in the first and second embodiments. It becomes. The number of words in the conversion dictionary / learning dictionary for the beginning of the sentence is overwhelmingly less than the number of words registered in the normal conversion dictionary / learning dictionary. For example, the same words in the normal conversion dictionary / learning dictionary and the conversion dictionary / learning dictionary for the beginning of the sentence Is registered, there are more memory deletions due to flag deletion, and the memory efficiency is improved as a whole. For example, in Shift_JIS which is a Japanese code, 2 bytes are used for one full-width character. One Japanese word is hiragana and is said to consist of 3 to 5 characters. Assuming that this is an average of 4 characters, when 4 hiragana characters are converted to kanji, it is almost equivalent to 2 kanji characters. Thus, for example, in the conversion dictionary shown in FIG. 3, 8 bytes (= 2 bytes × 4 characters per character), 4 bytes (= 2 bytes × 2 characters per character), part of speech information, frequency information, If, for example, 1 byte each is used for the head frequency and the head flag, the total is 16 bytes. If the sentence head flag is deleted, the entire dictionary can be reduced by (number of dictionary registration words) × (1/16 byte). On the other hand, the number of words used at the beginning of a sentence is much smaller than 1/16 of the entire dictionary. Thus, memory efficiency can be improved by separating the conversion dictionary for sentence head and the learning dictionary.

[Embodiment 5]
Embodiment 5 will be described. FIG. 13: is a block diagram which shows the principal part structure in Embodiment 5 of the character input device of this invention. In FIG. 13, the character input unit, sentence head determination unit, kana-kanji conversion unit, display unit, document holding unit, dictionary unit, input document, and the like realize substantially the same functions as the configurations described in the first to fourth embodiments. Description is omitted, and the dictionary creation unit 1309, reference document 1314, position information acquisition unit 1315, standard radio wave receiver 1316, telephone directory 1317, personal pronoun correspondence table 1318, and the like will be described.

  In Embodiment 5, the dictionary creation unit 1309 registers words and character strings in the sentence learning dictionary 1313 based on application-related information related to various applications, and starts up a Japanese input system (FEP, IME, etc.). The kana-kanji conversion unit 1303 selects words and character strings registered in the sentence learning dictionary 1313 as candidates and displays them on the display unit 1304.

  An embodiment will be described in which the associated application is an email and the application related information is a received email. In this embodiment, the reference document 1314 is a received mail. The dictionary creation unit 1309 extracts the word at the head of the received mail document (reference document 1314) and registers it in the sentence learning dictionary 1314. When creating a reply mail for the received mail, the kana-kanji conversion unit 1303 sends the words registered in the sentence learning dictionary 1314 to the display part and displays them as candidates on the head of the reply mail.

  For example, if “XXXX” is displayed at the top of the received mail, a candidate “XXX” is displayed at the top of the reply mail, or “Good morning” is displayed at the top of the received mail. Good morning "is displayed. Candidate display is executed even if nothing is input at the start of creating a reply mail, as in the first and second embodiments. When sending and receiving e-mail, especially at the beginning of a sentence, words used by the other party at the beginning of the sentence, such as greetings, personal pronouns, and nicknames, are often used as they are, and the configuration of the present application makes it possible to reduce user input operations.

  An embodiment will be described in which the related application is location information acquisition and the application related information is location information. The position information acquisition unit 1315 acquires position information from a positioning system such as a global positioning system (GPS) or Galileo positioning system, and the dictionary creation unit 1309 acquires the position acquired from the position information acquisition unit 1315. Based on the information, words related to the regional information are registered in the sentence learning dictionary 1313, the kana-kanji conversion unit 1303 sends the registered words to the display unit, and the display unit displays them as input candidates.

  For example, if you are in Shinjuku-ku, Tokyo, words such as “I am in Shinjuku” or “Please come to Shinjuku” are displayed as input candidates at the beginning of the document to be created. When you are waiting for a person, you often use email to communicate your current location, especially when there is a limit to the number of characters that can be entered, such as a short message on a mobile phone, In some cases, only a short sentence such as “~~~” is input and transmitted, and the user's input operation can be reduced by the configuration of the present application.

  An embodiment will be described in which the related application is time information acquisition and the application related information is time information. For obtaining time information, a standard radio wave (JJY) receiver 1316 is used. Based on the time information received by the standard radio wave (JJY) receiver 1316, the dictionary creation unit 1309 registers words related to the time and month and day in the sentence learning dictionary 1313, and the kana-kanji conversion unit 1303 is registered. Words are sent to the display unit and displayed as input candidates.

  For example, when mail is sent in the morning, “Good morning” is displayed as an input candidate at the beginning of the document. In many cases, greetings related to time such as “Good morning” and “Good evening” are written at the beginning of the mail, and the user's input operation can be reduced by configuring as in the present application.

  An embodiment will be described in which the related application is a phone book and the application related information is phone book information. For explanation, FIG. 14 shows an example of a telephone directory and a dictionary. When creating a mail, if a nickname is registered as phonebook information corresponding to the mail address of the transmission partner in the telephone book 1317, the dictionary creation unit obtains the nickname, and a word related to the nickname is used as a learning dictionary for the beginning of a sentence. The kana-kanji conversion unit 1303 sends the registered word to the display unit and displays it as an input candidate.

  For example, as shown in FIG. 14, if the nicknames “Wakadaisho” and “Aoiokame” are associated with the address A and registered, the dictionary creation unit uses the morpheme conversion to indicate “young” or “blue”. Is registered in the dictionary, and if it is not used, it is registered in the dictionary as "Wakadaisho" or "Aoiokame", and the display unit outputs it as a candidate to the beginning of the sentence. Whether or not to use morpheme conversion can be set by the user. Since there are many cases where the nickname of the transmission partner is described at the head of the mail, it is possible to reduce the user's input operation by configuring as in the present application.

  Further, the phone book information is not limited to the nickname of the transmission partner, and a schedule, birthday, etc. can be registered. For example, if you have set a meeting in the schedule, if you create an e-mail for the other party on the meeting date, the message will be displayed as a candidate for `` Today's meeting '' at the beginning of the e-mail document. When a mail corresponding to the birthday is created, a candidate such as “Happy Birthday” can be displayed at the head of the sentence of the mail document.

  An embodiment will be described in which a personal pronoun correspondence table is used as a related application in addition to the telephone book, and the application related information is telephone book information and personal pronoun information. In the present embodiment, the relevance with the user associated with the mail address registered in the telephone book 1317 as the telephone book information, and the personal pronoun associated with the personal pronoun correspondence table 1318 as the personal pronoun information Use the relationship with the person. Here, the relevance with the user is an intimacy or a usage scene. For explanation, FIG. 15 shows an example of a telephone directory and a personal pronoun correspondence table.

  In the phone book shown in FIG. 15, in addition to the e-mail address, the relevance with the user is registered in association with the e-mail address, and the relevance 1 with the user indicates that the usage scene is a business scene. Intimacy is assigned as 2 is a friend and 3 is a family. On the other hand, in the personal pronoun correspondence table, relevance 1 with the user corresponds to personal pronouns that are intended to be honored, such as “Wakakushi” and “you”, where the usage scene is a business scene and the closeness is a friend. Only “2” corresponds to “Yatsu”, and 2 and 4 correspond to “You”.

  Here, when the user creates an email, the dictionary creation unit detects the association with the user corresponding to the email address from the phone book, and selects the personal pronoun corresponding to the detected association with the user from the personal pronoun correspondence table. The kana-kanji conversion unit 1303 sends the registered word to the display unit and displays it as an input candidate. For example, when a user creates a mail for C, it detects that the familiarity is 3 from the phone book, selects “I, I” corresponding to the familiarity 3 from the personal pronoun correspondence table, and creates a mail document. As a candidate at the head of the sentence. By configuring as in the present application, not only the user's input operation is reduced, but also the candidate display according to the familiarity with the user and the usage scene becomes possible.

  However, in these fifth embodiments, the example in which the dictionary creation unit is registered in the sentence head learning dictionary has been described, but it may be registered in the sentence head conversion dictionary, or a dictionary configuration as shown in FIG. 1 or FIG. In this case, it may be registered in a conversion dictionary or a learning dictionary.

  As shown in FIG. 13, the fifth embodiment also includes a document holding unit 1306 that holds an input document 1307. When the position of the cursor moves on a document whose input has been confirmed, the moved position is the beginning of the sentence. Since the character string candidates can be displayed, the application related information can be displayed not only at the beginning of the document as described above but also at the beginning of the sentence. As a result, it is possible to reduce user input operations while optimizing the dictionary corresponding to various usage scenes.

  The character input device according to the present invention can be applied to a character input device that presents candidates such as words and sentences to be input when creating a document such as an e-mail on a mobile phone.

101, 401, 801, 1101, 1301 Character input unit 102, 402, 802, 1102, 1302 Beginning determination unit 103, 403, 803, 1103, 1303 Kana-Kanji conversion unit 104, 404, 804, 1104, 1304 Display unit 105, 405, 805, 1105, 1305 Dictionary part 110, 410, 810, 1110, 1310 Conversion dictionary 111, 411, 811, 1110, 1111 Learning dictionary 806, 1306 Document holding part 807, 1307 Input document 808 Determination information part 812, 1112 1312 Text conversion dictionary 813, 1113, 1313 Text learning dictionary 1309 Dictionary creation unit 1314 Reference document 1315 Location information acquisition unit 1316 Standard radio wave (JJY) receiver 1317 Phone book 1318 Personal pronoun correspondence table

Claims (13)

In a character input device having a character input unit, a kana-kanji conversion unit that creates a character string candidate, and a display unit that displays the character string candidate,
A character input device, comprising: a sentence head determining unit for determining a sentence head, wherein the kana-kanji conversion unit creates a character string candidate in response to an output from the sentence head determining unit. In a character input device having a character input unit, a kana-kanji conversion unit that creates a character string candidate, and a display unit that displays the character string candidate,
A document holding unit that holds the input document, and a sentence head determining unit that determines whether or not the character insertion position in the input document held in the document holding part is a sentence head. Accordingly, the character input device wherein the kana-kanji conversion unit creates a character string candidate.   The character input device according to claim 1, wherein the kana-kanji conversion unit includes a dictionary unit, and words stored in the dictionary unit include information about a sentence head.   When the sentence head determination unit determines that the input character position is the sentence head, the kana-kanji conversion unit preferentially creates a character string candidate among words stored in the dictionary unit, the word having information about the sentence head. The character input device according to claim 3.   If the word input from the character input unit is not registered in the dictionary unit when the sentence determination unit determines that the input character position is the beginning of the sentence, the kana-kanji conversion unit combines the input word with information about the sentence head The character input device according to claim 3, wherein the character input device is registered in the dictionary unit.   The character input device according to claim 3, wherein the kana-kanji conversion unit includes a dictionary unit that stores words including information related to a sentence head and a dictionary unit that does not include information related to a sentence head.   The head determining unit distinguishes between the head of a sentence and the head of a document, and the kana-kanji conversion unit creates a character string candidate corresponding to the head of the sentence and the head of the document in accordance with the output from the head determining unit. The character input device according to claim 1, wherein the character input device is a character input device.   6. The character input according to claim 5, further comprising a dictionary creation unit for creating a temporary dictionary, wherein the dictionary creation unit extracts words used at the beginning of a sentence from application-related information and registers them in the dictionary unit. apparatus.   The character input device according to claim 8, wherein the application related information is a received mail, and the dictionary creation unit registers a head part of the received mail in the dictionary unit.   The application-related information is position information, the character input device has a position information acquisition unit that acquires position information, and the dictionary creation unit registers a word related to position information in the dictionary unit. The character input device according to claim 8.   The application-related information is time information, the character input device has a time information acquisition unit that acquires time information, and the dictionary creation unit registers words related to the time information in the dictionary unit. The character input device according to claim 8.   The application related information is phone book information, the character input device has a phone book, and the dictionary creation unit registers words related to information registered in the phone book in the dictionary unit. The character input device according to claim 8.   The application related information is phone book information and personal pronoun information, the phone book information has a correspondence table of relevance between an email address and a user, and the personal pronoun information is correspondence of a personal pronoun and a user. 9. The character input device according to claim 8, wherein the character input device has a table, and the dictionary creation unit associates an e-mail address having the same relevance with the user and a personal pronoun and registers them in the dictionary. .