JPS62221064A - Word processor - Google Patents

Abstract

PURPOSE:To easily correct the misinput in a short time by displaying the corresponding characters in place of hitherto characters on a sequential character display position in a sentence production area and at the same time displaying the characters of lower ranks of recognition with the increment of each rank of recognition. CONSTITUTION:A register R1 works as a display means for the character of the 1st rank of recognition and registers R2-R4 work as the display means for the characters of the 2nd and following ranks of recognition. The storage contents of a display memory 5 are displayed on the display screen of a CRT display 7 by a display circuit 6. While a sentence production part 4 produces sentences based on the storage contents of registers R1-R4 serving as the speech recognition results and the characters, control commands, etc., inputted with the operation of a keyboard 8. A shift pulse generating circuit 9 produces shift pulses in response to the operating frequency of keys on the keyboard 8. Then the storage contents of registers R1-Rn of a shift register 3, i.e., the codes of characters latched respectively are shifted respectively.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a word processor. More specifically, the present invention relates to a word processor, and more specifically, when inputting characters by voice input or handwritten character input instead of inputting character codes by key operation. To,
The present invention relates to a word processor that can easily and quickly correct erroneously recognized speech or handwritten characters.

[Prior art]

Enter text using voice input or handwritten text input,
Word processors have been developed that recognize this to input characters; however, it is impossible for this type of word processor to achieve a recognition rate of 100%.

Therefore, it is an important issue how quickly and easily corrections can be made when incorrect characters are input into a word processor due to erroneous voice recognition or handwritten character recognition.

By the way, conventionally, the following two types of correction methods have been adopted as a method for correcting characters inputted by voice or handwriting into a word processor due to erroneous recognition.

+11 Recognition 8gl1 Only the first recognized character in the recognition ranking is displayed in the character display position of the next item in the sentence creation 6B area, and this is displayed in place of the character in the second recognition ranking by operating a predetermined key. If you press the key again, the third character in the recognition ranking will be replaced and displayed, and so on.For each keystroke, the recognition candidate characters will be displayed one by one according to the recognition ranking, and the correct input character will be displayed. Repeat the above operations until it is displayed.

(2) Display the recognized character with the first recognition rank in the next character display position (main display position) in the sentence creation area, and display the characters with lower recognition ranks, such as the second to fourth recognition ranks. Three characters are arranged and displayed in a predetermined area (sub-display position) of the display screen, such as at the bottom of the display screen, according to the recognition order, and when the correct input character is selected using the number keys, the character is displayed. is displayed in place of the first character in the recognition ranking, which was displayed at the next character display position (main display position) in the text creation area.

[Problem that the invention seeks to solve]

However, in the method (11) mentioned above, key operations must be repeated many times until the correct input character is displayed, and therefore, in some cases, it may take a long time until the correct input character is displayed. It may happen.

In method (2), if there is a correct input character among the characters displayed at the bottom of the display screen (!11 display position) that are ranked second or lower in the recognition order, then the key operation is used to specify that character. You only need to do this once, but if the correct input characters are not displayed, you will need to re-enter the voice input or handwritten characters.

As described above, there are various difficulties in correcting erroneous input using conventional word processors that input characters using voice recognition or handwritten character recognition.

The present invention was made in view of the above circumstances, and is intended to reduce the number of key operations in a word processor that recognizes characters input by voice input or handwritten character input and uses them as input characters to create sentences. To provide a word processor capable of easily correcting erroneous inputs in a relatively short time and without the need to perform voice input or handwritten character input again.

[Means for solving problems]

In the present invention, the first character in the recognition ranking of characters input by voice input or handwritten character input is displayed at the character display position of the next item in the sentence creation area, and the following recognition LJ+ jIff
Display several characters in the lower part of the display screen according to the recognition order, for example, at the bottom of the four display screens, and when the corresponding key is operated, the corresponding character will be displayed at the next character display position in the text creation area. will be displayed instead of the characters, and
Each character with a lower recognition rank is moved up in recognition rank and newly displayed.

The present invention provides a word processor for creating a sentence by sequentially displaying characters input by voice or number input and recognized in a sentence creation area of a display screen, and the present invention provides voice recognition of input characters. Or, a character display means having the first recognition rank for displaying the first character in the recognition rank in the text creation area of the display screen at the character input position of the next item in the text creation area of the display screen; an integer of 2 or more) for displaying recognition candidate characters in a predetermined area of the display screen according to recognition #a rankings; When any of the recognition candidate characters is designated as a character to be input, the recognition candidate character is displayed at the next character display position in the sentence creation area by the first recognition ranking first character display means. In addition, the present invention is characterized by comprising means for incrementing the respective ranks of recognition candidate characters of lower rankings in the second or lower recognition ranking character displaying means and displaying them in the predetermined area.

[Effect]

According to the present invention, incorrect input when inputting characters into a word processor by voice input or handwritten character input can be quickly corrected by a relatively small number of key operations without having to perform voice input or handwritten character input again.

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on drawings showing embodiments thereof.

FIG. 1 is a block diagram showing the configuration of a speech input word processor to which the method for correcting erroneous input according to the present invention is applicable;
The figure is a block diagram showing the configuration of the voice recognition unit 2, FIG. 3 is a schematic diagram showing the numeric keypad of the keyboard 8, and FIG.
~(C1 is a schematic diagram of the display screen of the CRT display 7.

The voice uttered by the operator is input through the microphone 1, is recognized by the voice recognition section 2, and the recognition results of kana characters up to the 0th position are output for each syllable of the input voice. Specifically, the speech recognition section 2 is configured as shown in FIG. 2, and the input speech is converted by the feature extraction circuit 21 into a human speech pattern whose features are extracted. The standard pattern memory 23 is for pre-registering the operator's voice as a standard voice pattern. Then, the input speech pattern is compared with the standard speech patterns stored in the standard pattern memory 23 in the identification circuit 22, and a recognition ranking is determined in descending order of similarity, and n standard patterns are selected. , the codes of the characters are sequentially outputted as recognition &lIl@complementary characters and latched in each register R1 to Rn of the shift register 3 according to the recognition 1B order.

The data of these n recognition results are stored in the shift register 3.
The data are latched and shifted in order according to the shift pulses applied from the shift pulse generation circuit 9 to the n registers R1 to Rn constituting the registers R1 to Rn.

Registers R1 to R4 of the shift register 3 are connected to the display memory 5, and are located at the next character display position (hereinafter referred to as main display position) in the text creation area on the display screen of the CRT display 7, which is a display device. cursor (4th
The memory contents of register R1 are written to the position (underlined in the figure), and the contents of registers R2 to R4 are written to the part corresponding to the bottom line of the display screen of the CRT display 7 (hereinafter referred to as the sub-display position). The memory contents are written respectively (see FIG. 4), that is, register R1 operates as a display means for the first character in the recognition order, and register R2
~R4 operates as a display means for characters ranked second or lower in recognition ranking.

The stored contents of the display memory 5 are displayed on the display screen of the CRT display 7 by the display circuit 6 as shown in FIGS.

It is displayed as shown in (C).

On the other hand, the sentence creation unit 4 receives the register R which is the voice recognition result.
Create sentences according to the memory contents of 1 to R4, characters input by operating the keyboard 8, control commands, etc.
The contents are written into the display memory 5. This allows CRT
Characters input on the display screen of the display 7 are created into sentences and displayed on the display screen of the CRT display 7.

The shift pulse generation circuit 9 generates the corresponding number of shift pulses when the keys "1", "2'+"3'" on the numeric keypad (see FIG. 3) on the keyboard 8 are pressed. As a result, the contents stored in each of the registers R1 to Rn of the shift register 3, that is, the codes of the characters latched therein, are shifted.

Next, the method of the present invention implemented by the voice input word processor configured as described above will be explained with reference to the schematic diagram of the display screen shown in FIG. 4 and the flowchart showing the processing procedure shown in FIG. 15.

Now, for example, if the operator inputs "Only" through the microphone 1, this voice input is recognized by the voice recognition unit 2, latched in the shift register 3, and then written into the display memory 5. Rarely,
displayed on the CRT display 7. After converting these characters into kanji, when the operator inputs "Nyoru", the same message will be displayed on the CRT display 7, but as shown in FIG. The recognition m ranking of the last letter "ru" is "bu", "
``gu'', ``mu'', ``nu'', ``u'', ``ru'', ``zu'', ``<
'', ``su'', as shown in Figure 4+al, ``bu'' is displayed at the cursor display position which is the main display position as the recognized character for the input character ``ru'', and the overall result is ``speech recognition''. This means that ``to buy'' has been entered.

By the way, as shown in Fig. 4(a), the "bu" displayed at the main display position is the "nyoru" input by voice.
This is the first recognition candidate character in the recognition ranking for the last character "ru", and the recognition candidate characters in the lower tgiaa position are displayed three characters at a time on the bottom row of the display screen of the CRT display 7, which is the sub display position. Each item is displayed with a number specifying it.

At this time, if the correct character is displayed in the sub-display position, if the operator presses the corresponding key,
The character was previously displayed in the main display position.
It is displayed in place of the character at i position 1, and the correction of the incorrect input is completed. However, if the correct character is not displayed in the sub-display position, as in Figure 4 (al), press the appropriate key,
In this case, when you press the "3" key, the corresponding character "nu" in the 4th place of recognition #IAiII will be displayed in the main display position, and each character "u" in the recognition rank below it will be displayed.
, "ru", "zu", "<", and "su", the three characters with the highest recognition rankings will be displayed at the bottom of the display screen. ``gu'', ``mu'', ``nu''
will be displayed instead.

Specifically, in this operation, when the key '3# on the numeric keypad is pressed, a shift pulse is applied from the shift pulse generation circuit 9 to the shift register 3 three times, and thereby each of the shift registers 3 This is because the stored contents of registers R1 to Rn are each shifted three times. That is,
First, the character "bu" in position 1 is stored in the first register R1, the character "gu" in the second position is stored in the second register R2, and il! The character "mu" in the third place in the ia order is latched in the third register R3, etc., but by being given three times from the shift pulse generation time VII9, the fourth character is latched in the third register R3. It was latched in register R4! It turns out that the character "nu" in the fourth place in the ga ranking was latched in the first register RL and in the fifth register R5! The character “u” in the 5th place in the li ranking is the second register R2
Then, the character "ru" in the 6th place in the recognition order, which was latched in the 6th register R6, is shifted and latched in the 3rd register R3, etc., so that the main display position on the display screen is shifted and latched. ``nu'' is displayed in the sub-display position, and the characters ``u'', ``ru'', and ``zu'' are displayed in order in the sub-display position.

This state is the 4th YI! As shown in J(b), when you press the "2" key that corresponds to the number that indicates the correct character "ru" among the recognition candidate characters displayed in the sub-display position, the recognition rank is 6th. The correct character ``ru'', which is the digit character, is displayed in the main display position. FIG. 4C shows a state in which this correct character "ru" is displayed at the main display position, that is, at the character display position for the next item in the text creation area of the display screen of the CRT display 7.

At this time, the character ``zu'' whose recognition rank is lower than the correct character ``ru'' which is displayed in the main display position at the same time and is ranked 6th,
The display positions of "ri" and "su" are moved up according to their respective recognition rankings.

(Effects) As described above, in the present invention, compared to conventional word processors that use voice input or handwritten character input, incorrect input can be corrected more quickly and without having to perform voice input or handwritten character input again. It is.

In the above embodiment, an example was shown in which the present invention was applied to a voice input word processor, but it goes without saying that the present invention is also applicable to a word processor that inputs handwritten characters.

[Brief explanation of drawings]

The drawings show embodiments of the present invention, and FIGS. 1 and 2 are block diagrams showing the configuration of a word processor according to the present invention, and FIG. 3 is a schematic diagram of the numeric keypad of the keyboard.
FIG. 4 is a schematic diagram showing the display state of the display screen, and FIG. 5 is a flowchart showing the processing procedure.

Claims (1)

[Claims] 1. In a word processor configured to create a sentence by sequentially displaying characters input by voice or input by hand and recognized in a sentence creation area of a display screen, A recognition rank first character display means for displaying a character ranked first in recognition rank by recognition or handwritten character recognition at the next character input position in a sentence creation area of a display screen; character display means for displaying characters of the second or lower recognition rankings (m is an integer of 2 or more) in a predetermined area of a display screen according to the recognition ranking; When any of the characters is specified as a character to be input, displaying the recognition candidate character at the next character display position in the text creation area by the first recognition ranking first character display means; , means for incrementing the respective ranks of recognition candidate characters of lower rankings in the second or lower recognition ranking character displaying means and displaying them in the predetermined area.