JPS63768A - Japanese sentence producer - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention is directed to a Japanese text creation device that supports the creation of Japanese texts.

BACKGROUND OF THE INVENTION Computer application systems such as word processors and editing devices for Japanese language are popular as systems to support the creation of Japanese text. In these systems, the user first inputs a character string in Roman characters or kana for each phrase, and then converts it into hiragana or kanji. These inputs are normally performed using a keyboard, and the user selects and hits a predetermined key and inputs while checking the characters displayed on a CRT or liquid crystal display.

Problem to be solved by the invention In this case, Japanese, especially kanji, has many words and characters with the same sound and is ambiguous, so kana-kanji conversion cannot be uniquely determined, and there are many words and characters that have the same sound. The user must select and input an appropriate kanji from among the kanji candidates. Therefore, the input operation is very troublesome and has the disadvantage of being inefficient.

The present invention aims to reduce the troublesomeness of inputting Japanese sentences and to enable efficient input of sentences.

Means for Solving the Problems In order to solve the above problems, the present invention provides an input device, a sentence meaning interpretation device that interprets the meaning of an input phrase, and a phrase group prediction device that predicts the next input phrase. This is a Japanese sentence creation device equipped with the following.

Operation In the above configuration, the phrase inputted from the input device is semantically interpreted by the sentence meaning interpretation device in parallel with the input operation, and the phrase group prediction device predicts and presents the next inputted phrase accordingly, Since the user can input a phrase simply by selecting it, input effort and time can be significantly reduced.

EXAMPLES Hereinafter, examples of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing the basic configuration of an embodiment of a Japanese text creation device according to the present invention.

In the figure, 1 is an input device such as a keyboard, and 4 is a CRT.
A display device Q such as a display or a liquid crystal display; 2 an input/output control device for controlling the operations of the input device 1 and the display device 4; 7 is a sentence meaning interpretation device that interprets the meaning of a sentence, and 7 is a phrase group prediction device that predicts a phrase group based on the knowledge of the knowledge database 8.

When creating Japanese sentences, if you use a keyboard as a character input device, you can input characters from kana or romaji to hiragana,
Considering the ambiguity of Japanese, it is considered impossible to avoid conversion into kanji at all. Therefore, text input is simplified by reducing the number of repetitions of this input cycle, that is, the process of inputting kana or romaji → conversion to hiragana or kanji → selection. 1st
The structure of the diagram is such that a sentence meaning interpretation device analyzes the meaning of a group of phrases input from an input device in parallel with the input, understands the meaning in the context, and predicts the next group of phrases input from the user. , n and allows the user to select an appropriate one, thereby greatly reducing the time and effort required to enter text.

The character string input by the user from the input device 1 is sent to the sentence meaning interpretation device 3 through the input/output control device 2, and
It is displayed on the display device 4.

The sentence meaning interpretation device 3 creates a new context semantic structure to which a group of phrases has been added based on the context structure already interpreted and the knowledge stored in the electronic dictionary 5 and the knowledge database 6, and generates a group of phrases. It is sent to the prediction device 7.

The phrase group prediction device 7 compares the received semantic structure with the knowledge in the knowledge database 8 to predict a phrase group, and sends this to the text display device 4 via the input/output control device 2 for the user's selection. wait.

When a selection is input from the input device 1, the information is sent to the sentence meaning interpretation device 3 through the input/output control device 2. The sentence meaning interpretation device 3 incorporates the structure of the phrase group previously sent from the phrase group prediction device 7 into a new context structure.

In addition, if no selection is made even after prediction,
Receives direct character input from the user.

FIG. 2 shows a concrete example. Reference numeral 11 denotes a keyboard through which the user inputs selection commands for Japanese sentences and predicted phrase groups, and the input Japanese sentences and predicted phrase groups are displayed on the CRT display 14.

These input/outputs are controlled by a microcomputer 12. The input Japanese text is processed by microcomputer 13.
The meaning of the sentence is interpreted based on the terms from the electronic dictionary 15 and the knowledge from the knowledge database system 16. The electronic dictionary 15 stores terms necessary for understanding Japanese sentences, and provides predetermined terms to the microcomputer 13 according to instructions from the microcomputer 13. The knowledge database system stores the knowledge and rules necessary for understanding Japanese sentences, and operations such as changing, adding, deleting, searching, etc. of this knowledge and rules are possible at will. Based on the semantic structure generated by the microcomputer 13 by interpreting the sentence meaning, the microcomputer 17 generates the semantic structure of a group of phrases that are predicted to be input next. The knowledge database system 18 stores the knowledge necessary for this prediction and operates on the knowledge. Microcomputer 17
The plurality of phrase groups predicted in the above are displayed on the CRT display 14 under the control of the microcontroller 12. When the user sees this display and issues a p selection command using the keyboard 11, the microcomputer 13 selects a predetermined group of phrases from a plurality of predicted phrase groups generated by the microcomputer 17 based on that command and creates a new context. Incorporate it into the structure. In this way, the user can complete the input of phrases simply by issuing a selection command.

Below, we will explain the system's responses while looking at the screen, taking several user input phrase groups as examples.

As shown in FIG. 3(a), when the user inputs ``Aobe ordered coffee. Me too'' on the keyboard 11, the microcomputer 17 makes a prediction and displays the message at the bottom of the screen of the CRT display 1. [I ordered 1 coffee. ” “2. I had coffee.” “3. I had the same thing.”

'', three predicted phrases are generated and presented, and the system waits for the user's selection. If the user presses a key to select, for example, ``I ordered 2 coffees.'', the message ``Tabe ordered coffee. I also ordered coffee.'' is displayed as shown in Figure 3(b).

” is taken into the microcomputer 13,
As soon as the data is input, it is displayed on the CRT display 1.

FIG. 4(a) shows that when the user inputs "I took a bath with my child for the first time in a while," the microcomputer 17
I made a prediction and said, ``There was one person.'' I decided to enter I-2. ” A state in which three predicted phrases such as “I thought of entering 3.” are generated and presented is shown. If you select ``1. I took a bath.'', the message ``I took a bath with my child after a long time'' is input as shown in the figure (-).

FIG. 5 shows an embodiment in which correction is made after selection is input. In other words, when you input "I was tired last night," the microconopij is displayed as shown in (a) of the same figure.

Based on Ri17's predictions, ``1 I didn't do anything.'' ``2 I went to bed early.'' ``3 I went to bed without doing anything.'' [4 I slept soundly. ] Predict and present the four phrases.

Here [3 I slept without doing anything. ” is inputted as shown in FIG. 4(b). In addition, this input sentence “
slept. '' to ``I had a good night's sleep.'' If you want to change ``I had a good night's sleep,'' you can directly input ``good night'' using an appropriate screen editing function, and the result will be input as shown in Figure (C).

Effects of the Invention As described above, the present invention predicts the next input phrase by interpreting input Japanese sentences in parallel with the input, and displays this to select the next phrase to input. This allows you to reduce the number of times you have to repeat the tedious task of direct input using a series of operations such as inputting Romaji or kana → converting to hiragana or kanji → selecting from a group of candidates on conventional Japanese text creation devices. , the effort and time required to create Japanese sentences can be significantly reduced, and the effect is significant.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the overall configuration of the Japanese text creation device of the present invention, and Fig. 2 is a block diagram showing a specific embodiment of the Japanese text creation device of the invention.・
・Input/output control device, 3... Sentence meaning interpretation device, 4...
- Display device, 5 - Electronic dictionary, 6... Knowledge database, 7... Bunsetsu group prediction device, 8... Knowledge database. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure 27 Figure 3 Ma Figure 4

Claims (1)

[Claims] (1) An input device for inputting sentences to be input and selections from a group of phrases presented by the system, a sentence meaning interpretation device for interpreting the meaning of the input phrases, and information necessary for performing the sentence meaning interpretation. an electronic dictionary and a first knowledge database that provide information, a second knowledge database that provides information necessary for predicting the next input phrase group after performing the semantic interpretation; , a phrase group prediction device that predicts the next phrase group, a display device that displays an input sentence and the predicted phrase group, and an input/output control device that controls input and output of these. Sentence writing device.