JPH11250045A - Document preparing device, input processing method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To share two input means by voice and a keyboard and to efficiently prepare a document by means of making the best use of respective merits. SOLUTION: A character string which is keyboard-inputted through a keyboard input part 21 is inputted and a character string which is voice inputted through a voice input part 22 is inputted. The speech-inputted character string is speech-recognized by a speech recognition part 23 and is stored in a data storage part 28 with attribute information showing speech input data. The character string which is keyboard-inputted is stored in the data storage part 28 with attribute informant showing keyboard input data. At the time of KANA (Japanese syllabary)/KANJI(Chinese character) conversion, the type of the character string stored in the data storage part 28 is judged based on attribute information and a conversion part 26 executes a corresponding KANA/KANJI conversion processing. The conversion result is outputted by an output part 29.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document creation device such as a personal computer or a word processor, and more particularly to a document creation device having two input means including a voice and a keyboard, and an input processing method and recording used in the device. Regarding the medium.

[0002]

2. Description of the Related Art Conventionally, a document creation device such as a personal computer or a word processor usually has a keyboard as an input means, and converts a character string inputted through the keyboard into a Kana-Kanji conversion process to convert a sentence mixed with Kanji and Kana characters. Output.

[0003] On the other hand, there is also an apparatus provided with input means by voice. In this method, a user inputs a character string by voice using a microphone. In this case, in order to inform the device of the delimitation of the character string, for example, provide a short silence between single syllables, such as "Tokyo", "Ha", "Good", "Tenki", "is". You need to speak.

[0004] Here, Japanese Patent Application Laid-Open No. Hei 9-198382 discloses that a voice generated by a user is captured, converted into a voice signal and digitized, and the power value of the digitized voice signal is changed at regular intervals. Calculate and detect each silent part by comparing it with a preset threshold value, determine the section sandwiched by each silent part as a single syllable, and perform speech recognition processing on each single syllable in order and store it in a register Is disclosed. The character strings stored in this register are converted into Kanji-Kana-mixed sentences by Kana-Kanji conversion processing and output.

[0005]

In a conventional document creation apparatus, any one of a keyboard and a voice is used.
Japanese input was performed using two input means.
However, the keyboard and the voice each have advantages and disadvantages, and there are the following problems when performing Japanese input.

[0006] That is, the keyboard input is a long-read character string that can be input at a relatively high recognition rate by voice input, for example, “Tokyo” or “Tenki” of “Tokyo”. Even you have to type all of that string on the keyboard. On the other hand, with voice input, for example,
As described above, even if a character string with a short reading that can be easily input by using a keyboard is input, the input must be made by voice. In this case, the shorter the reading is, the higher the probability of erroneous recognition is, so that the input efficiency is reduced.

As described above, the method using only one input means has a problem that it is impossible to provide a user with an input environment with a small load in every scene. SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and provides a document creating apparatus which can use two input means by voice and a keyboard, and can create a document more efficiently by taking advantage of each of them. An object of the present invention is to provide an input processing method and a recording medium.

[0008]

According to a first aspect of the present invention, there is provided a document creating apparatus including two input means including a voice and a keyboard, wherein a character string input by voice is recognized by voice. Voice recognition means for processing, and attribute information indicating that the data is voice input data is added to the character string recognized by the voice recognition means, and attribute information indicating that the character string is input from the keyboard is keyboard input data. Storage means for adding and storing, and whether the character string stored in the storage means is voice input data or keyboard input data is determined based on the attribute information, and a kana-kanji conversion process corresponding to each is performed. It has a conversion means for executing, and an output means for outputting a conversion result obtained by the conversion means.

According to such a configuration, it is possible to input a reading character string for kana-kanji conversion using two input means of a keyboard and a voice. In this case, the character string input by voice is subjected to voice recognition processing, and then stored in the storage unit together with attribute information indicating that the character string is voice input data. The character string input by the keyboard is stored in the storage unit together with attribute information indicating that the character string is keyboard input data. At the time of Kana-Kanji conversion, the type of the character string stored in the storage means is determined based on the attribute information, and the conversion result is output by the Kana-Kanji conversion process according to each.

According to the present invention, it is determined whether the character string stored in the storage means is voice input data or keyboard input data based on the attribute information, and the character string is displayed in a display form corresponding to each. And a character string display unit.

According to such a configuration, the character string input by the keyboard and the character string input by voice are displayed in different display forms. Thereby, the input means used for inputting the character string can be visually determined.

According to a third aspect of the present invention, there is provided instruction means for designating a character to be deleted, and whether the character designated by the instruction means is voice input data or keyboard input data based on the attribute information. In the case of keyboard input data, there is provided a deletion processing means for deleting only the above-mentioned designated character in the case of keyboard input data, and deleting the whole character string including the above-mentioned designated character collectively in the case of voice input data.

According to such a configuration, the characters in the character string input by the keyboard can be deleted one by one, and the entire character string can be deleted by the character input by voice. You can delete them all at once.

According to a fourth aspect of the present invention, there is provided instruction means for instructing the movement of the cursor, and the attribute information indicating whether the character to which the cursor is moved specified by the instruction means is voice input data or keyboard input data. Cursor movement processing means for moving the cursor in units of characters in the case of keyboard input data and moving the cursor to the next or previous character of the character string containing the character in the case of voice input data. It is characterized by having.

According to such a configuration, the cursor can be moved in units of characters for a character string input by a keyboard, and the cursor can be moved to a character string containing the character for a character string input by voice. You can move to the next or previous character in the column.

According to a fifth aspect of the present invention, when the character string to be converted is speech input data, the converting means determines a delimiter position of each phrase of the character string and determines a delimiter position of each phrase. It is characterized in that a conversion candidate having the same is generated.

According to such a configuration, it is possible to obtain a Kana-Kanji conversion result of a character string input by voice as a delimiter of each clause of the character string. Claim 6
In the above, when the character string to be converted is speech input data, the conversion means generates a conversion candidate for each recognition candidate of the character string obtained by the speech recognition means. Displays the conversion candidates corresponding to the recognition candidates selected from the above recognition candidates,
The remaining recognition candidates are displayed together as selection targets.

According to such a configuration, the conversion candidate of the kana-kanji conversion result and the recognition candidate of the voice recognition result are displayed on the same screen, so that the conversion candidate corresponding to each recognition candidate can be selected on the same screen. it can.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a system configuration of a document creation device according to an embodiment of the present invention.
The document creation apparatus according to the present embodiment is realized by a computer that reads a program recorded on a recording medium such as a magnetic disk and the operation of which is controlled by the program.

As shown in FIG. 1, a document creation device according to the present embodiment includes a CPU 11, a memory 12, a keyboard 1
3, display device 14, external storage device 15, printing device 16,
It comprises a communication device 17, an audio processing unit 18 and a microphone 19.

The CPU 11 controls the entire operation of the present apparatus, and operates by activating a program stored in the memory 12. The functions realized based on the program stored in the memory 12 include a word processor by an application program.

The memory 12 is composed of, for example, a ROM or a RAM, and stores various programs and data such as fonts used in the apparatus. The memory 12 has a program area 12a for storing program data such as an application program for realizing a word processor.
In addition, here, an input buffer 12b for storing a character string input by a keyboard or a character string input by voice together with accessory information, and a recognition candidate obtained when performing a voice recognition process on the character string input by voice are stored. And a conversion candidate buffer 12d for storing conversion candidates obtained by performing a Kana-Kanji conversion process on a character string input by keyboard or a character string input by voice.

The keyboard 13 is used to input a reading character string to be converted and to give a conversion instruction (an instruction by operating a "conversion" key) to the input reading character string by key operation. The keyboard 13 includes a "character" key and a "numeric" key for inputting characters, a "cursor" key for instructing cursor movement, a "conversion" key for instructing kana-kanji conversion, and a conversion result. "Next candidate" key for designating the next candidate, "confirm" key for confirming the word obtained as a conversion result, "DEL" key (delete key) for deleting the character at the cursor position, cursor "B" to delete the character before the position
Various keys such as an "S" key (backspace key) are provided.

The display device 14 is, for example, a CRT (Cathode
It consists of a Ray Tube) and an LCD (Liquid Crystal Display), and displays a character string input by keyboard or a character input by voice, and displays the conversion results of those characters.

The external storage device 15 is, for example, an HDD (Hard
Disk Drive) and CD-ROM (Compact Disc Read On)
ly Memory) and stores data such as various programs and documents. The external storage device 15 includes, in addition to an application program for realizing a word processor, a dictionary used in voice recognition processing and a dictionary used in kana-kanji conversion processing (speech recognition dictionary 24 and kana-kanji conversion dictionary 27 shown in FIG. 2). Data is also included.

The printing device 16 is used for printing a document or the like. The printing device 16 may be built in the printing apparatus or may be connected externally. The communication device 17 is, for example, a private LAN (Local
It controls the transmission and reception of data with external devices via an Area Network) or a communication network.

The audio processing section 18 performs audio signal processing such as digitizing an analog audio signal input through the microphone 19. The microphone 19 is for inputting a character string, a function, and the like by voice.

Next, a description will be given of a functional configuration of the document creating apparatus realized by the system configuration shown in FIG. FIG. 2 is a block diagram showing a functional configuration of the document creating apparatus realized by the system configuration of FIG. As shown in FIG. 2, the document creation device according to the present embodiment includes a keyboard input unit 21, a voice input unit 22, a voice recognition unit 23, a voice recognition dictionary 24, a control unit 25, a conversion unit 26, a kana-kanji conversion dictionary 27, a data It has a storage unit 28 and an output unit 29.

The keyboard input unit 21 is provided for the keyboard 13
Data (code data) such as a character string or a function input by the above key operation is input. The voice input unit 22 inputs data (voice data) such as a character string or a function that is voice-input from the microphone 19 via the voice processing unit 18.

The voice recognition unit 23 converts a character string (voice data) input by the voice input unit 22 into a voice recognition dictionary 24.
, And passes the recognition result (code data of characters obtained as recognition candidates) to the control unit 25. The speech recognition dictionary 24 stores information necessary for speech recognition processing. Note that there are various methods for voice recognition, but the present invention is not limited to these methods.

The control unit 25 controls the entire apparatus, receives a character string input by a keyboard and a character string input by voice, controls each processing unit, and performs a process according to each. The conversion unit 26 performs a Kana-Kanji conversion process on the character string input by the keyboard and the character string input by voice with reference to the Kana-Kanji conversion dictionary 27, and outputs the conversion result (code data of a character obtained as a conversion candidate) to the control unit. Pass to 25. The kana-kanji conversion dictionary 27 stores information necessary for kana-kanji conversion processing, such as reading and headings corresponding to each word.

The data storage unit 28 stores a character string input by a keyboard or a character string input by voice in a distinguished manner, a recognition candidate obtained when a character string input by voice is recognized by speech, or a recognition input character string input by a keyboard. A conversion candidate obtained when a character string or a character string input by voice is converted to a kana-kanji character, and the like, corresponds to the input buffer 12b, the recognition candidate buffer 12c, and the conversion candidate buffer 12d in FIG.

The output unit 29 outputs a character string input from the keyboard or a character string input by voice to the display device 14, and displays on the display device 14 conversion candidates obtained when the input character strings are converted into kana-kanji characters. Perform processing such as output.

Next, the operation of the present apparatus will be described. Here, in a document creation apparatus having two input means of a voice and a keyboard, (a) display processing of an input character string,
The (b) deletion processing, (c) cursor movement processing, (d) kana-kanji conversion processing, and (e) candidate display processing will be described.

Conventionally, when inputting a reading character string to be subjected to kana-kanji conversion, a method of hitting a key on a keyboard and inputting a voice, using a microphone, and recognizing the input voice by voice. Various methods have been adopted, such as a method of inputting and a method of recognizing and inputting characters written on the screen using a pen, but it is common to adopt only one input method and use each method. Was. However, as described in the above-mentioned [Problems to be Solved by the Invention], each input method has a problem.

In order to solve such a problem, in the present apparatus, a reading character string used for kana-kanji conversion can be input using two input means of a keyboard and a voice.

By making it possible to use a keyboard and two input means by voice, for example, input of a short character string or a character string with poor voice recognition accuracy, editing operation, etc. are performed by keyboard input, and relatively long reading characters are used. It becomes possible to input Japanese using the advantages of each, such as inputting columns by voice input.

(A) Input character string display processing Here, in order to clearly indicate to the user that the character string has been input using the keyboard 13 and the microphone 19 shown in FIG. The displayed character string is distinguished from the character string input by voice. The example of FIG. 3 shows a case where a character string such as “Today”, “Good”, or “Tenki” is input by voice input, and a character string such as “H” or “I” is input by keyboard input. Underlines 31a, 31b, 31c are added to the character string input by voice,
It is displayed separately from the character string entered on the keyboard.

It is to be noted that such a distinction display is not limited to the underline and can be visually distinguished by, for example, changing a font or a display color between a character string input by a keyboard and a character string input by voice. Such a display form is sufficient.

FIG. 4 shows an example of data managed in the Japanese input system. That is, for example, it is assumed that a character string such as “Today”, “Good”, and “Tenki” is input by voice input, and a character string such as “H” and “I” is input by keyboard input. In such a case, the keyboard input character strings “ha” and “is” are added with attribute information in which both the voice flag and the voice character string end flag are set to “0”, and the data storage unit 28 (input buffer 12b ). The voice flag “0” indicates that the data is keyboard input data. For the voice input character strings "today", "good" and "tenki", the value of the voice flag is set to "1", and the last character of the character string indicates the end of the character string. The attribute information with the end flag set to “1” is added and managed in the data storage unit 28 (input buffer 12b). The voice flag “1” indicates that the data is voice input data.

The processing for actually displaying the input character string on the screen is shown in the flowchart of FIG. The processing flow will be briefly described below. FIG. 11 is a flowchart showing the operation of the input character string display process in the embodiment.
When a character string is input through the keyboard input unit 21 or the voice input unit 22 (step A11), the control unit 25
First checks whether the input character string is input by keyboard or voice (step A).
12). This can be determined based on whether the character string has come from the keyboard input unit 21 or from the voice input unit 22 via the voice recognition unit 23. Further, when outputting the character string after voice recognition from the voice recognition unit 23 to the control unit 25, the determination can be made by adding an attribute indicating that the character string is voice-inputted.

As a result, in the case of the character string input by the keyboard (No in step A12), as shown in FIG. 4, the control unit 25 sets the voice flag indicating the input format of the character string and the voice character string end. By setting both of the flag values to "0", the character string is stored in the data storage unit 28 as keyboard input data (step A13), and then output to the output unit 2 so as to be displayed in a display form for the keyboard input character string.
9 is controlled (step A14). In this case, as shown in FIG. 3, the keyboard input character string is displayed without underlining.

On the other hand, in the case of a character string input by voice (Yes in step A12), the control unit 25 sets the value of the voice flag indicating the input format of the character string to "1", By setting the voice character string end flag of the last character in "1" to "1", the character string is stored in the data storage unit 28 as a voice input character string (step A1).
5) The output unit 29 is controlled so as to display in a display form for a voice input character string (step A16). In this case, as shown in FIG. 3, the voice input character string is underlined.

In this manner, the user can visually determine the voice input character string and the keyboard input character string, and can efficiently perform subsequent editing operations separately for voice and keyboard. Become like

That is, as shown in FIG. 4, the character string input by the keyboard and the character string input by voice are managed separately in the Japanese input system, so that the character input by voice can be performed even when the reading character string is edited. An operation corresponding to a column or a character string input by a keyboard can be performed.

(B) Deletion Processing Hereinafter, processing for deleting a character string as an editing operation will be described. When deleting a character string, the movement differs depending on the position of the cursor and the type of input delete key. Here, the case where the cursor is after the character string and the input delete key is the "BS" key The cursor is at the beginning of the character string and the input delete key is "DEL"
The difference between the processing of the keyboard input character string and the processing of the voice input character string in the case of a key will be described.

As shown in FIG. 5, when the character string "Mojitsu" is input on the keyboard, the cursor is set after "Mojitsu", and when the "BS" key is pressed, the character "Katsu" is displayed. Is deleted. Also, when the cursor is set to “Moji” of “Mojizetsu” and the “DEL” key is pressed, the character of “Momo” is deleted. As described above, the deletion process is performed on the keyboard input character string in character units.

On the other hand, when the character string "Mojitsu" is input by voice, if the cursor is set after "Mojitsu" and the "BS" key is pressed in the same manner as described above, all characters are deleted. . Further, even when the cursor is set to “Moji” of “Mojizetsu” and the “DEL” key is pressed, all characters are deleted. As described above, the deletion process is performed on the voice input character string in units of input.

The operation of the deletion process will be described below. FIG. 12 is a flowchart showing the operation of the deletion process in the embodiment. When the “BS” key is input through the keyboard input unit 21 (step B1)
If the current cursor position is other than the first character of the character string (No in step B12), the control unit 25 determines that the character immediately before the cursor position is a voice input character or a keyboard input character. It is determined by checking the voice flag stored in the data storage unit 28 (step B13). As a result, when the voice flag is “0”, the control unit 25 determines that the character is a keyboard input character (No in step B14), deletes the character immediately before the cursor position from the data storage unit 28, and The output unit 29 is controlled to delete it from the screen (step B15).

On the other hand, if the voice flag is "1", the control unit 25 determines that the character is a voice input character (Y in step B14).
es) By detecting the range of the voice input character string including the character based on the voice input character string end flag (step B16), the entire voice input character string is stored in the data storage unit 2.
8 and the output unit 29 is controlled to delete from the screen (step B17).

Further, “D” is input through the keyboard input unit 21.
When the "EL" key is input (step B18)
Yes), if the current cursor position is other than the last character of the character string (No in step B19), the control unit 25 determines whether the character at the cursor position is a voice input character or a keyboard input character. 28 (step B2).
0). As a result, when the voice flag is “0”, the control unit 25 determines that the character is a keyboard input character (step B21).
No), the character at the cursor position is stored in the data storage unit 28.
And the output unit 29 is controlled to delete from the screen (step B22).

On the other hand, if the voice flag is "1", the control unit 25 determines that the character is a voice input character (Y in step B21).
es) By detecting the range of the voice input character string including the character based on the voice input character string end flag (step B23), the entire voice input character string is stored in the data storage unit 2.
8 and the output unit 29 is controlled to delete from the screen (step B24).

(C) Cursor Movement Processing Next, cursor movement processing will be described. In the case of cursor movement as well, the movement differs depending on the cursor position and the type of input movement key. In this case, the cursor position is behind the character string, and the input "cursor" key is the "←" key (left arrow). Key) and when the cursor position is at the beginning of the character string and the input "cursor" key is the "→" key (right arrow key), the difference between the keyboard input character string and the voice input character string processing explain.

As shown in FIG. 6, when the character string "Mojitsu" is input on the keyboard, the cursor is set after "Mojitsu" and the "←" key is pressed. The cursor moves to. Also, when the cursor is set to “Moji” of “Mojiretsu” and the “→” key is pressed, the cursor moves to the character of “Jiji”. As described above, the cursor movement processing is performed on the keyboard input character string in units of characters.

On the other hand, when the character string "Mojitsu" is input by voice, if the cursor is set after "Mojitsu" and the "←" key is pressed in the same manner as above,
The cursor moves to the character of “MO”. Also, when the cursor is set to “mo” of “mojizetsu” and the “→” key is pressed, the cursor moves after “tsu”. As described above, the cursor movement process is performed on the voice input character string in input units. That is, the cursor does not move in the middle of the character string input by voice.

Hereinafter, the operation of the cursor moving process will be described. FIG. 13 is a flowchart showing the operation of the cursor moving process in the embodiment. When the “→” key is input through the keyboard input unit 21 (Yes in step C11), if the current cursor position is other than the end of the character string (No in step C12), the control unit 25 sets the character at the cursor position. Is determined by checking the voice flag stored in the data storage unit 28 (step C13). As a result, when the voice flag is “0”, the control unit 25 determines that the character is a keyboard input character (No in step C14), and controls the output unit 29 to move the cursor on the screen one position to the right. (Step C15).

On the other hand, if the voice flag is "1", the control unit 25 determines that the character is a voice input character (Y in step C14).
es), the output unit 29 is controlled to move the cursor to the right to the next character of the voice input character string including the character (steps C16 and C17). Specifically, while checking the voice input character string end flag, the cursor is moved from the current cursor position to the position after the character whose voice input character string end flag is “1”.

When the "←" key is input through the keyboard input section 21 (step C18).
If the current cursor position is other than the beginning of the character string (No in step C19), the control unit 25 controls the output unit 29 to move the cursor one position to the left (step C20). At this time, the control unit 25 determines whether the character at the cursor position is a voice input character or a keyboard input character by checking a voice flag stored in the data storage unit 28 (step C21). As a result, when the voice flag is “1”, the control unit 25 determines that the character is a voice input character (Yes in step C22), and moves the cursor to the left to the first character of the voice input character string including the character. The output unit 29 is controlled so as to perform the operation (steps C23 to C26). More specifically, the cursor is moved from the current cursor position to the first character of the voice input character string including the character while checking the cursor position, the voice flag, and the voice input character string end flag.

As described above, with respect to a character string input by voice, deletion processing and cursor movement processing are performed in input units. This is because a character string input by voice is a character string input by one utterance,
This is because it is natural to treat the same character as one character input at the time of keyboard input, and it is easy for the user to accept it.

(D) Kana-Kanji conversion process Next, the Kana-Kanji conversion process taking into account the division of the character string to be read by voice input will be described.

In normal kana-kanji conversion, an input character string is converted into a kanji-kana mixed sentence using a kana-kanji conversion dictionary in a Japanese input system. At this time, a plurality of change candidates are evaluated using various types of information such as word frequency information, grammar information, and user learning information. Then, among these conversion candidates, the candidate with the highest evaluation point is preferentially output as the conversion result. It should be noted that the kana-kanji conversion method and the method of assigning evaluation points are not the subject of the present invention, and thus will not be described in detail here.

Here, as shown in FIG. 7, for example, when a reading character string such as “wipe a kimono here” is input using the keyboard 13, the conversion result is as follows.
The one with the highest evaluation score will be output. In this example, three candidates such as “here / take off / wear”, “here / take off / on”, and “swim here” are evaluated as “10”, “5”, and “0”, respectively. , The conversion candidate having the evaluation point “10”, such as “Take off the footwear here”, is output with priority.

However, when a reading character string is input in words such as "here", "in", "kimono", "wo", and "nugu", and the reading character string is input for each phrase, the reading character input on the keyboard If you convert it like a column,
Despite the fact that the user is conscious of the result of the conversion and input the phrase separated, the conversion result with a different phrase separation is output, which is a result contrary to the user's desire.

In order to prevent such a situation, a kana-kanji conversion is performed on a reading character string input by voice with consideration given to the delimitation of the phrase, so that the conversion result desired by the user can be preferentially output. To In other words, if a reading character string is input by voice, such as "here", "in", "kimono", "wo", "nug", etc. Here, conversion candidates such as "Take off the kimono" are preferentially output.

The operation of the Kana-Kanji conversion process in consideration of the segmentation by voice input will be described below. FIG.
9 is a flowchart showing an operation of a kana-kanji conversion process in the embodiment, which is aware of a phrase break by voice input. First, a read character string input by keyboard or voice is supplied from the control unit 25 to the conversion unit 26. The conversion section 26 refers to the kana-kanji conversion dictionary 27, performs the kana-kanji conversion processing on the read character string as usual, attaches an evaluation point to each candidate obtained as the conversion result, and returns the candidate to the control section 25 (step D11). . The conversion candidates obtained at this time are stored in the data storage unit 28 (conversion candidate buffer 12d shown in FIG. 1).
Temporarily.

Here, the control unit 25 checks whether or not there is a character string input by voice in the read character string used for the conversion process (step D12). This is performed by checking the voice flag of the input character string stored in the data storage unit 28 shown in FIG. As a result, if the character string input by voice is not included (N in step D12)
o) The control unit 25 controls the output unit 29 based on the evaluation points provided by the conversion unit 26 as it is, so as to preferentially output the conversion candidate having the highest evaluation point as a conversion result (step D19).

On the other hand, if the voice-input character string is included (Yes in step D12), the control unit 25 obtains a phrase break position of the voice-input character string (step D13). This is performed by checking the voice character string end flag of the input character string stored in the data storage unit 28 shown in FIG. If there is a conversion candidate having the delimiter position (Yes in step D14), the control unit 25
Raises the evaluation score for the conversion candidate (step D1
5). Further, the control unit 25 checks the headline character string of the read character string input by voice (step D16), and raises the evaluation point of the conversion candidate that matches the headline character string (step D18).

As described above, for the character string input by voice, the evaluation points for the conversion candidates in the sections delimited by the clauses are increased. Then, when outputting the conversion result, the control unit 25 controls the output unit 29 so that the conversion candidate with the highest evaluation point is preferentially output as the conversion result (step D19).

In the normal kana-kanji conversion process performed first as described above, since it is not expected that a conversion candidate having a segment break as input by the user is necessarily output, the kana-kanji conversion process is performed. At this time, it is also conceivable to perform a kana-kanji conversion process in which a delimiter position is specified in advance by voice input and a phrase is cut at the delimiter position.

(E) Candidate Display Processing Next, the processing of displaying homophones after kana-kanji conversion will be described. For example, when a character string such as “I” is input by voice input, the voice recognition unit 23
, And outputs a recognition result as shown in FIG. In this case, the reading character string for each recognition candidate,
The corresponding headline character string and the evaluation score indicating the certainty are output. In the example of FIG. 8, as a result of speech recognition of the character string "I" input by voice, "I"
-"I", "I"-"hand over", "wadachi"-
It is shown that a plurality of recognition candidates such as "rut", "wagashi"-"wagashi", and "wagashi"-"waga" have been obtained, and each of them is given an evaluation score. These recognition candidates are stored in the data storage unit 28 (recognition candidate buffer 12c shown in FIG. 1).

At the time of Kana-Kanji conversion, Kana-Kanji conversion is performed using the read character strings of these recognition candidates. When the conversion result is output, as shown in FIG. Give priority to columns. That is,
In the example of FIG. 8, first, kana-kanji conversion is performed using a reading character string of a recognition candidate such as “I”. In this case, conversion candidates (heading character strings) for homophones such as “I”, “pass”, and “me (no conversion)” are obtained.
Among them, the evaluation score at the time of speech recognition for "I" is "pass"
Since it is higher, the priority of each conversion candidate is “I” → “Pass” → “I”.

Similarly, the reading character strings of the recognition candidates such as "Wadachi" and "Wagashi" are sequentially converted into Kana-Kanji characters, and the priorities of the conversion candidates are determined by the evaluation points at the time of speech recognition.

Here, the homophone display of the voice input character string is as follows.
As shown in FIG. 10, the recognition candidates output from the speech recognition unit 23 are displayed together with the homophone candidates of the normal kana-kanji conversion result. That is, for example, when the voice input “I” is converted to Kana-Kanji and the same phonetic candidate is displayed, the same phonetic candidate “I”, “Pass”, “Password” of the Kana-Kanji conversion result using the recognition candidate “I” is displayed. Along with "I", "Watachi", another recognition candidate for voice input "I",
"My" is also displayed.

In this state, by pressing the “↓” key (down arrow key) of the “cursor” key, the recognition candidate “wadachi” is
When you select, the phonetic candidates "rut" and "wadachi" of the kana-kanji conversion result using the recognition candidate "wadachi" are displayed, and the remaining recognition candidates "me" and "wagashi" are displayed there. Is done. Similarly, if the recognition candidate "Wagashi" is selected, the phonetic candidates "Waganashi", "Wagashi" and "Wagashi" of the kana-kanji conversion result using the recognition candidate "Wagashi" are displayed, and the remaining The recognition candidates “I” and “Rudachi” are displayed.

In this way, even when the kana-kanji conversion is performed using the wrong speech recognition candidate, the correct speech recognition candidate can be selected from the homophone display, and the kana-kanji character using the correct speech recognition candidate can be selected. The converted homophone candidate can also be selected on the same homophone display.

The operation of the candidate display process will be described below. FIG. 15 is a flowchart showing the operation of the candidate display process in the embodiment. The control unit 25 first outputs a candidate display in a form in which the remaining recognition candidates are combined with the kana-kanji conversion homophone candidates using the reading character string of the first recognition candidate in the speech recognition candidate group. The section 29 is controlled (step E11).

Here, the control unit 25 is the keyboard input unit 2
The type of the candidate selection key input through 1 is checked (step E12), and if the end key is input (Yes in step E13), the candidate display processing here ends. When the “↓” key or the “↑” key is input as the recognition candidate movement key, the control unit 25 selects and recognizes the next or previous recognition candidate according to the position of the currently selected recognition candidate. Move the candidate (from step E14)
E21), display the same phonetic word candidate of the kana-kanji conversion using the reading character string of the recognition candidate and the remaining recognition candidates together (step E11).

If the homophone candidate selection key is input (Yes in step E22), the control unit 25 sets the next homophone candidate as a selected homophone candidate (step E23), and displays a candidate with the homophone candidate. Redisplay (step E1)
1).

As described above, keyboard and voice 2
Even when the two input methods are used in combination, input, editing, and conversion processes can be performed by making use of the characteristics of each input method, and a document can be efficiently created.

In the above embodiment, the case where the keyboard and the voice are used is described as an example. However, for example, a pen (handwriting input method) can be used together. However, in the case of the pen (handwriting input method), the processing for the character string after input (character string after character recognition) is the same as that of the keyboard, so that the combination of the keyboard and the pen is
The combination of voice and pen is more effective.

Further, the method described in the above-mentioned embodiment can be executed by a computer as a program such as a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM,
It is also possible to write the data on a recording medium such as a DVD or a semiconductor memory and apply it to various devices, or to transmit it via a communication medium and apply it to various devices. A computer that realizes the present apparatus reads the program recorded on the recording medium, and executes the above-described processing by controlling the operation of the program.

[0082]

As described above, according to the present invention, it becomes possible to input a reading character string for kana-kanji conversion using two input means of a keyboard and a voice, thereby improving the input efficiency.

Further, by displaying the character string input by the keyboard and the character string input by voice in different display forms,
The user can immediately understand the input means used for inputting the character string.

When editing a character string input by keyboard and a character string input by voice, the character string input from the keyboard can be deleted in units of characters. The entire character string can be deleted at once.

When editing a character string input by keyboard and a character string input by voice, the cursor can be moved in character units with respect to the character string input by keyboard. You can move the cursor to the next or previous character in the string containing that character.

Further, it is possible to obtain a kana-kanji conversion result according to the delimiter of each phrase of the character string input by voice, and it is possible to improve the input efficiency. In addition, by displaying the conversion candidate of the kana-kanji conversion result and the recognition candidate of the voice recognition result on the same screen, the conversion candidate corresponding to each recognition candidate can be selected on the same screen.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a system configuration of a document creation device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a functional configuration of a document creation device realized by the system configuration of FIG. 1;

FIG. 3 is an exemplary view showing a display example of an input character string in the embodiment.

FIG. 4 is an exemplary view showing an example of input character string management in the embodiment.

FIG. 5 is an exemplary view showing an example of deleting an input character string in the embodiment.

FIG. 6 is an exemplary view showing an example of movement of a cursor in the embodiment.

FIG. 7 is an exemplary view showing an example of a kana-kanji conversion result of an input character string in the embodiment.

FIG. 8 is an exemplary view showing an example of a speech recognition result for the speech input character string in the embodiment.

FIG. 9 is an exemplary view showing an example of a kana-kanji conversion result for a voice input character string in the embodiment.

FIG. 10 is an exemplary view showing a candidate display example for a voice input character string in the embodiment.

FIG. 11 is an exemplary flowchart illustrating the operation of input character string display processing in the embodiment.

FIG. 12 is an exemplary flowchart illustrating the operation of a deletion process in the embodiment.

FIG. 13 is an exemplary flowchart showing the operation of cursor movement processing in the embodiment.

FIG. 14 is an exemplary flowchart showing the operation of kana-kanji conversion processing in the embodiment with consideration given to segmentation by voice input.

FIG. 15 is an exemplary flowchart showing the operation of candidate display processing in the embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... CPU 12 ... Memory 12a ... Program area 12b ... Input buffer 12c ... Recognition candidate buffer 12d ... Conversion candidate buffer 13 ... Keyboard 14 ... Display device 15 ... External storage device 16 ... Printing device 17 ... Communication device 18 ... Audio processing part 19 ... Microphone 21 ... Keyboard input unit 22 ... Speech input unit 23 ... Speech recognition unit 24 ... Speech recognition dictionary 25 ... Control unit 26 ... Conversion unit 27 ... Kana-Kanji conversion dictionary 28 ... Data storage unit 29 ... Output unit

Claims (8)

[Claims] 1. A document creating apparatus comprising two input means including a voice and a keyboard, comprising: a voice recognition unit configured to perform a voice recognition process on a character string input by voice; and a character string recognized by the voice recognition unit. Storage means for adding attribute information indicating that the input data is voice input data to a character string, and adding attribute information indicating that the input data is keyboard input data to a character string input by a keyboard, and storing the character string. A conversion unit that determines whether the character string is voice input data or keyboard input data based on the attribute information, and executes a kana-kanji conversion process according to each of them; and outputs a conversion result obtained by the conversion unit. A document creating apparatus comprising: 2. A character string display means for judging whether a character string stored in the storage means is voice input data or keyboard input data based on the attribute information, and displaying the character string in a display form corresponding to each. The document creation apparatus according to claim 1, further comprising: 3. An instructing means for instructing a character to be deleted, and determining whether the character specified by the instructing means is voice input data or keyboard input data based on the attribute information. 2. The document according to claim 1, further comprising a deletion processing unit that deletes only the specified character in the case of data, and collectively deletes the entire character string including the specified character in the case of voice input data. Creating device. 4. Instructing means for instructing the movement of a cursor, and determining whether the character to which the cursor is moved specified by the instructing means is voice input data or keyboard input data based on the attribute information. Cursor movement processing means for moving a cursor in units of characters in the case of keyboard input data and moving the cursor to the next or previous character of a character string containing the character in the case of keyboard input data. The document creation device according to claim 1, wherein 5. When the character string to be converted is voice input data, the conversion means determines a delimiter position of each phrase of the character string and generates a conversion candidate having a delimiter position of each phrase. 2. The document creation apparatus according to claim 1, wherein 6. When the character string to be converted is speech input data, the conversion means generates a conversion candidate for each recognition candidate of the character string obtained by the speech recognition means. 2. The document according to claim 1, wherein the output unit displays a conversion candidate corresponding to the recognition candidate selected from the recognition candidates and also displays the remaining recognition candidates as selection targets. Creating device. 7. An input processing method used in a document creation device having two input means by voice and a keyboard, wherein a character string input by voice is subjected to voice recognition processing, and the character string is recognized by the voice recognition processing. Attribute information indicating voice input data is added to the character string, and attribute information indicating keyboard input data is added to the character string input by the keyboard, and the character string is stored in the storage unit. Is determined based on the above attribute information whether the character string stored in the Kana-Kanji is the voice input data or the keyboard input data, and performs the Kana-Kanji conversion process according to each of them. The conversion result obtained by this Kana-Kanji conversion process An input processing method characterized by outputting 8. A recording medium used in a computer provided with two input means using a voice and a keyboard, comprising: a procedure for performing a voice recognition process on a character string input by voice; and a character recognized by the voice recognition process. A procedure of adding attribute information indicating that the data is voice input data to a column, adding attribute information indicating that the data is keyboard input data to a character string input by a keyboard, and storing the data in a storage unit; A procedure for determining whether the stored character string is voice input data or keyboard input data based on the attribute information, and executing a kana-kanji conversion process according to each of the steps; and a conversion obtained by the kana-kanji conversion process. Computer-readable recording recording a program for causing a computer to execute a procedure for outputting a result Body.