JPS6315633B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention relates to a Japanese word processor, which is a Japanese text creation device, and is a Japanese word processor that uses various input methods such as line input, full key input, and kana input. This paper is particularly concerned with Japanese word processors that use kana input methods.

In a word processor using a kana input method, a kana sentence inputted from a kana keyboard is exchanged with a mixed kana/kanji sentence by the control unit of the word processor to create a complete Japanese sentence. Various methods have been developed to convert kana to kanji, including the following.

1. At the stage of inputting kana characters from the keyboard, enter the designated kanji key in the part that needs to be converted to kanji.

2. The control unit of the word processor analyzes the kana text, automatically determines the parts that need to be converted to kanji, and converts them to kanji.

To explain method 1 in detail, for example, when inputting the sentence ``Go to school'' using the kana keyboard, the display screen will display ``Gatsukou ni Iku'' immediately after inputting from the kana key, and the kanji character will be displayed on the display screen. During the conversion process, "Gatsukou" is changed to indicate that "Gatsukou" is a kanji.
Use the keys for specifying kanji characters before and after the character ``gatsukou'' to enclose it in [ ]. Such an operation is required every time a Kanji part appears, making key input operations inefficient.

In order to improve the complexity of key operations, a method has been considered in which the system machine finds the kanji part, as shown in 2.

To explain method 2 in detail, for example, in response to the sentence ``I read a book'', ``I read the book'' is input from the kana keyboard according to the pronunciation, but this input kana The text is first cut into clauses or phrases by the word processor itself.

In order to perform this cutting operation, the word processor has Japanese grammatical information preset in its internal memory, and parses and cuts the word by comparing it with this information.

In the above example, when cutting the phrase ``watashi wa'', the word ``watashi'' is cut out and it is determined based on the information in the internal memory that this word is a ``noun.'' Next, we look at ``wa'' and note that ``wa'' is a particle, and see if it is possible to disconnect the connection relationship of noun (nominal) + wa (particle) by referring to the grammar table set in internal memory. Decide whether it is possible or not, and say, ``I am the hon.
It is divided into clauses and phrases, such as "Yonda". Thereafter, conversion processing from "kana" to "kanji" is performed for each clause or phrase. In this way, the commonly used method for converting "kana" to "kanji" is to first cut it into clauses and phrases and then convert them to kanji, but this method is better than directly converting kana sentences to kanji. , this is because the conversion accuracy is improved considering the structure of the Japanese language. Therefore, in the input method like ``2'', it is essential to process the separation of clauses and phrases, but in order to perform this process, as mentioned above, a large amount of information such as Japanese grammar information is required. It is necessary to set it in memory in advance.

The present invention has been proposed under the above circumstances, and it does not require special operations to specify kanji when inputting kana, and it does not require a large memory such as Japanese grammatical information to input sentences according to the pronunciation. The present invention provides an input method that makes it possible to automatically cut off clauses and phrases.

Normally, when considering the key input operations performed by an operator, key input is performed in a fixed rhythm, including punctuation marks, particles (wa, wo, ni, and...), clauses such as phrases, and words that separate phrases, and the key input operations are performed while looking at the manuscript. If you are doing this, look at the text you will type next at this break. In addition, when writing a sentence by typing instead of creating it by hand, the next sentence to be input by key is devised at the position where a clause or phrase breaks.

In either case, the key input operation is interrupted at the position that marks the break between clauses and phrases. That is, the key input of words within a clause or phrase is input at a certain fixed time interval, and it is conceivable that this time interval becomes significantly longer between clauses or between clauses.

The present invention focuses on the characteristics of such key input operations, examines the time interval between key inputs, and makes it possible to automatically identify the breakpoints of clauses and phrases based on this time interval. .

FIG. 1 shows an embodiment of the key input section of a word processor according to the present invention, in which 1 indicates a kana keyboard;
2 is a code converter that generates a corresponding character code based on the data from the kana keyboard 1;
is a mono multivibrator whose pulse width can be set by the time constants of R and C, 4 is a code generator that generates a special code according to the signal of mono multivibrator 3, and 5 is a code generator whose output from the code converter 2 and code generator 4 is The given OR gate 6 is an output signal from the keyboard and is given to the word processor as a key input. The operation of the present invention will be described using figures.

When the kana keyboard 1 is operated and any key is pressed, a signal corresponding to the operated key is inputted to the code converter 2 using a pulse S, which is generated once for each keystroke, as a strobe signal. is converted to the code of Further, the output pulse S formed when the key is pressed is also provided as an input to the mono-multivibrator 3.

If a sentence such as ``I read the book'' is entered from the keyboard 1, the timing of the input is assumed to be as shown in the timing chart in Figure 2. This is because, as already mentioned, the input rhythm is extended at punctuation marks, clauses such as particles, and phrase breaks. Therefore, due to the keyboard pulse signal represented by the input signal B in FIG. 2, the output signal from the monomulti 3 becomes the output signal Q in FIG.

That is, a negative pulse is generated when the input time interval from the kana keyboard becomes larger than the time constant set by the resistor R and capacitor C connected to the mono-multivibrator 3. Here, by making the resistor R a variable resistor, individual differences in input time intervals can be corrected. In this embodiment, the code generator 4 to which the output of the mono-multivibrator 3 is applied is designed to be turned on at the falling edge of the pulse, and is turned on at the time when the negative pulse of the output signal Q is generated. , generates a code signal, that is, a code indicating a break section where the input rhythm changes. The code signal is combined with the output signal of the code converter 2 at the OR gate 5 and output from the kana keyboard output signal terminal 6. The output signal is
In the example sentence above, between "wo" and "yo" and "."
A delimiter code is inserted after 3. The control unit of the word processor executes the conversion process from "kana" to "kanji" based on the output signal into which the above code signal has been inserted, but the inserted code is used as the cutting point for clauses and phrases. This eliminates the need to previously set a large amount of grammatical information in memory and analyze it into clauses and phrases based on this memory information, as was the case in the past.

Next, FIG. 3 shows another embodiment that is an improvement on FIG. 1. 7 is a kana keyboard, 8 is a kana keyboard 7
9 is a mono multivibrator, 10 is a register in which a specific word is set, and 11 is given the output signal of code converter 8 and data register 10. Compare and match circuit, 12
is mono multivibrator 9 and comparison matching circuit 11
OR gate to which the output signal of is given, 13 is OR
A code generator generates a delimiter code according to the signal from the gate 12; 14 is an OR gate to which the output signals of the code converter 8 and the code generator 13 are applied; and 15 is an output signal from the keyboard 7.

In the figure, a certain specific code is set in advance in the data register 10, and when a key is pressed on the keyboard 7, a comparison is made to see if the pressed key input matches the code set in the register 10. The circuit 11 makes the decision.

If they match, a low level signal is output from the comparison match circuit 11, and if they do not match, a high level signal is output from the comparison match circuit 11. It is designed to operate the code generator 13 with a signal obtained by ORing the output of the matching circuit 11 and the output of the monomulti 9. The operation will be explained using the timing chart shown in FIG. When the sentence ``I read the book.'' is input at the timing of input signal E, the output from monomulti 9 becomes output signal Q'.

Now, the data register 10 contains characters used at the break points of clauses and phrases, such as punctuation marks, particles,
The code is set. Therefore, the output signal from the comparison and coincidence circuit 11 has a timing as shown in the comparison and coincidence circuit output signal P in FIG.

The output signal obtained by ORing these two signals is OR
The timing is as shown in the gate output signal W. The code generator 13 performs a code generation operation at the fall of the output of the OR gate.

As a result, it can be seen that delimiter codes occur only between "wo" and "yo" and after ".". In the embodiment shown in FIG. 1, since the delimiter code generator is operated by the output signal Q, a special code is generated between "n" and "da", and there is a possibility that a mistake may be made in cutting a clause or phrase. On the other hand, according to the embodiment shown in FIG. 3, it can be seen that a delimiter code is generated only when a word serving as a break point of a clause or phrase is input with an extended time interval. This makes it possible to eliminate erroneous disconnection of phrases or phrases when the operator interrupts key input for some reason after inputting a word other than the breakpoint of the phrase or phrase.

As described above, according to the present invention, in a word processor in which sentences are input from a kana keyboard according to the pronunciation, a delimiter code is generated when a predetermined time has elapsed between one key operation and the next key operation. Unlike conventional devices, there is no need to preset complex text creation information for segmentation, the capacity of the control device for text creation can be reduced, and the configuration of the word processor system has been simplified. At the same time, it is possible to create sentences containing kana and kanji without forcing the operator to perform operations such as specifying kanji when entering keys.

According to the present invention, when used by a skilled operator, the key-in intervals are constant and the number of break points in sentences is reduced. Even in that case, the conversion accuracy of kana and kanji conversion can be further improved by combining it with the algorithm of the kana and kanji conversion method for the written input text.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the main parts of one embodiment according to the present invention, FIG. 2 is a timing chart for explaining the same embodiment, and FIG. 3 is a block diagram showing the main parts of another embodiment according to the invention. 4 are timing charts for explaining the operation of the same embodiment. 1, 7... Keyboard, 2, 8... Code converter, 3, 9... Mono multivibrator, 4, 13
...Code generator, 6, 15...Keyboard output terminal, 10...Data register, 11...Comparison matching circuit.

Claims (1)

[Scope of Claims] 1. In a word processor in which kana is input sequentially from a kana keyboard according to the pronunciation of text information, the next kana is input within a preset time after the kana is input by key operation of the kana keyboard. 1. An input method for a word processor, comprising: generating a specific code from a code generating means when the input data is not being input; and using the specific code signal as a delimiter code for kana-kanji conversion. 2. In claim 1, when a predetermined kana key provided on a kana keyboard is operated and the next kana input is not made within a preset time, An input method for a word processor, characterized in that a specific code is generated from a code generating means.