JPH02144761A - Sentence producing device - Google Patents

Abstract

PURPOSE:To improve the operability of a correction back-up function by minimizing the unnecessary point-out actions based on the operating information on an operator and carrying out the learning so as to minimize unnecessary operations. CONSTITUTION:A morpheme analyzing part 10 serves as a polish/correction back-up function attaining means and checks the propriety of the word composition and the grammer of the sentences. A syntax structure analyzing part 11 checks the propriety of the arrangement of the sentence component elements. A deciding/retrieving part 12 corrects the wrong words and handles the information for decision. Then a control information memory part 13 handles the overall control information on the operation history of an operator and the polish/ correction back-up. The input information received from a keyboard input part 2 and the KANA (Japanese syllabary)/KANJI (Chinese character) conversion are processed by an input control part 6. While the editing/storing printing processes are carried out by a word processor control part 8 and an input/output control part 9. In addition, a learning information memory part 14 performs the learning under the control of the part 7.

Description

[Detailed Description of the Invention] The present invention relates to a text creation device, and more particularly, to a text creation device for creating a Japanese input sentence, proofreading, editing, etc., for example, a word processor. ,computer,
It can also be applied to general office automation equipment such as personal word processors.

Word processors with a proofreading support function or text editing support function have been released since the end of 1987 as writing devices for Jumeishiyo Japanese, and proofreading support software for PCs has also been released. Furthermore, software on general-purpose computers or systems for companies specializing in calibration began to become popular around 1986.

On the other hand, similar software in foreign languages (mainly English) has been on the market for quite some time. The software scans and analyzes the strings of sentences that have been created in some way, and the parts that deviate from the predetermined rules of the text.

It is designed to point out incorrect terms or provide useful information for revising the text. The processing format is interactive, which calls the function while editing the text, and performs revisions and proofreading, or - batch processing, which reads the document file and checks preset items, and then checks the results. There are some that print , and others that do both.

On the other hand, operations related to the elaboration/proofreading support function include ■ operations for selecting the function the operator wishes to execute, ■ operations for changing evaluation parameters for several functions such as checking sentence length, and ■ operations for changing evaluation parameters for several functions such as checking sentence length. Operations that specify the range of work, ■
In each function, ``Help'' (reference) allows you to view related information for the points pointed out, ``Correction'' is done by the operator, ``Replacement'' is selected by the operator by presenting correction candidates and paraphrase candidates from the system side. There are editing operations such as ``Continue'' (skip) that move on to the next step without making any corrections or replacements if the problem is not correct.

Especially in kana-kanji conversion for one Japanese input, if the first conversion result is not the notation that the operator wants to input, a homophone candidate is selected. At this time, the operator's selection operation is learned, and the next conversion There are ways to optimize the results.

In addition, the points pointed out by proofreading support include: ■ Mistakes in punctuation due to input errors, mistakes in okurigana, or
Brackets can be used for things that are quite clear and effective, such as forgetting things, and things that are not very clear and useful, such as points out for easily misleading homonyms or homonyms. The degree of effectiveness varies depending on the writing style and the content of the writing.

In particular, in the case of (■) above, the algorithm for pointing out the problem is generally rudimentary, unlike when it is elaborated by a human, and there are still many unnecessary points. However, in conventional systems, the selection of a function and the range in which the function works can only be specified based on the form, and once you start pointing out a function, if there are parts of the same description in a sentence, the same point will be made uniformly. The process was repetitive, and the operator had to skip that part one by one, which was a major cause of inconvenience.

The present invention was made in order to solve the above-mentioned drawbacks, and it learns to minimize unnecessary indications and avoid unnecessary operations as much as possible based on operator operation information. The purpose of this invention is to provide a text creation device that improves the usability of the proofreading support function.

Tsubaki - Ikkai In order to achieve the above-mentioned object, the present invention provides a separation proofreading means that has a proofreading function for a text being written or after it has been written, and an operator operation information in a writing device for writing Japanese or English writing. In addition, when the above-mentioned points are moved to the next point without being corrected, The system is characterized in that it stops making subsequent points, and furthermore, it displays the points pointed out, but makes the next point without requesting the operator to make corrections. below,
An explanation will be given based on an example of the present invention.

FIG. 1 is an overall configuration diagram for explaining an embodiment of a text creation device according to the present invention. It has a similar configuration to general office automation equipment such as personal computers and word processors. At the center is a control section 5 that controls the entire configuration, and is composed of a general-purpose microcomputer or a dedicated LSI. There are peripheral devices that handle input/output and storage, such as a display section 1, a keyboard input section 2, an external storage section 3 such as a floppy disk, and a printing section 4 such as a printer. Also,
Under the control unit 5, document input, editing, storage, printing, revision,
There is a part that controls calibration, and these parts are composed of firmware, dedicated LSI, etc. Furthermore, in a Japanese system, a dictionary is required to perform kana-kanji conversion for Japanese input, or to realize the proofreading support function, which is the basic function of the present invention, and these are required in the ROM.
firmware such as or a hard disk device is used.

In the embodiment shown in FIG. 1, as means for realizing the proofreading support function, there is a morphological analysis part 1o that checks the correctness of word structure and grammar in a sentence under the proofreading support control part 7, and a morphological analysis part 1o that checks the correctness of the word structure and grammar in a sentence. a syntax analysis unit 11 that checks whether the alignment is appropriate; a judgment search unit 12 that corrects incorrect words to correct words; and a judgment search unit 12 that handles judgment information for judgment; and operator operation history and overall control information for proofreading support. It has a control information storage unit 13 that handles. The input information coming from the keyboard input section 2 and the kana-kanji conversion are performed under the input control section 6, and the editing,
Storage, printing, etc. are performed under the word processor control section 8 and output control section 9.

In addition, the present invention includes a learning information storage section 14 under the revision/proofreading support control section 7. This learning information storage section 1
4 may be included in the control information storage unit 13, but it has a different configuration to make the function clear.

FIG. 2 is an example of a conventional proofreading support system pointing out points in a text. In this system, a text to be revised and proofread is given, and when the execution operation is performed, the system registers words from the beginning of the text in the dictionary, referring to the system's proofreading dictionary, and sequentially searches words and phrases in the text that match certain words. Search for.

- If there is a part with M, the word is displayed on the spot by changing it from other parts.There are various ways of displaying it, but common ones include inverting or surrounding it with a frame, and if it is in color, it is displayed in different colors. There are things to do. ), there are also those that provide a display area at the bottom to display replacement information, etc., and those that provide a window on the screen. ), wait for instructions from the operator. Relevant points are pointed out one by one.

Here we point out homophones that are easy to confuse.

As shown in Figure 2 (a), “sakusei” means “sakusei” in the text.
There is a word ``create'' that is easily confused with ``create'', and ``create'' in the sentence can be immediately replaced with ``create'' in the lower row with a keystroke.

At this point, the operator presses the confirmation key (the same applies to the "continue", "search", and "proceed to the next pointed out" keys) since there is no particular problem with the word "create" in the text. Then, Figure 2 (b
), the next ``create'' in the same sentence is pointed out again. Furthermore, as shown in FIG. 2(c), there is another "creation", which is pointed out again. Then, the next search begins as shown in FIG. 2(d). With this system, it takes a lot of time and effort when there are consecutive words and phrases that have no particular problem in a sentence, as described above.Furthermore, it takes a little too much time to search for the relevant words. It takes a lot of time to edit and proofread short texts using this system. In order to improve this point, there is a system that allows you to specify a range of such places one by one and display them after processing them as a whole, or to process the same words at once, but it is still not possible to do so. It's not enough.

FIG. 3 is for explaining one embodiment of the text creation device according to the present invention, and is an example of the pointing out of the revision/proofreading support system l\ realized using the same search method as in FIG. 2. Here, in order to compare with the conventional example shown in FIG. 2, an explanation will be given using the same display method (it does not particularly depend on the screen configuration method). The beginning of the point is the same as in Figure 2. If the operator presses the confirmation key here, no further indication will be made regarding "creation".

FIG. 4 shows another embodiment of the present invention. The difference from Fig. 3 is that once the confirmation key is pressed, the same point will be displayed, but the search will move on to the next one without waiting for the operator's instructions. In this example, as shown in FIG. 4(b), a distinction is made between a display when waiting for an instruction and a display when pointing out and proceeding without waiting for an instruction. This does not necessarily have to be a distinction, but it is easier to understand if you do.

FIG. 5 shows another embodiment of the present invention applied to an elaboration/proofreading support system that collectively processes the corresponding portions explained in FIG. 2 above. Since this system processes all at once, as shown in FIG. 5(,), all applicable locations are displayed at the first indication, and the cursor is initially positioned at "Create". Normally, the first song cursor moves to each corresponding location.

The display in the relevant area will not disappear unless the operator specifically corrects it. When the first ``Create'' is confirmed without modification in Figure 5(a), the following ``Create'' is displayed, but the cursor does not stop and moves to point out another word in the next sentence. The relevant points are indicated separately, including the first point.

FIG. 6 is an example of another conventional proofreading support system. In this example, processing is performed on a sentence-by-sentence basis.

Even in this case, the same points of the present invention can be realized.

FIG. 7 is a flowchart for explaining the internal operation of the system for realizing the above-mentioned points in terms of operator operations and corresponding formats. This is described in accordance with the embodiments shown in FIGS. 3 and 4 (although they are omitted with respect to FIGS. 5 and 6, they are similar). The operation thereof will be described below. First (Step 1), the operator starts executing the elaboration/proofreading function and specifies the execution function and execution target range.

(Step 2) The system starts executing each function.

Here, the processing proceeds in a form in which each function is executed repeatedly. Regarding the function, the corresponding part is searched while sequentially referring to the dictionary from the beginning of the target range.

When the first relevant location is found, it is pointed out and waits for instructions from the operator. Specifically, this is the display shown in FIG.

(Step 3) Next, when the operator selects 5 skips, the information of the corresponding location is stored in the "learning information storage section". Simply, for example, the character string "Create" described above may be stored. Alternatively, instead of the character string, the corresponding location in the revision/proofreading dictionary may be stored, or the corresponding position on the text may be stored.

(Step 4) When the operator's selection is other, the editing and proofreading dictionary performs operations such as selection similar to conventional operations related to the proofreading support function.

(Step 5) Next, search is performed again from the next position in the target range. If there is no corresponding part, the function ends and moves on to the next function, or the function ends altogether.

(Step 6) If there is a corresponding location, it is checked whether there is information about the corresponding location in the "learning information storage".

(Step 7) When there is something that matches the learning information, point out according to that information. In the embodiment shown in FIG. 3, the corresponding part is not stopped and the process proceeds to the next step without displaying any display. Although the example shown in FIG. 4 does not stop, the display of the corresponding part is changed.

(Step 8) If there is no matching learning information, the corresponding location is pointed out in the usual way.

As is clear from the above explanation, the present invention is capable of pointing out functions that are not very clear and whose effectiveness varies depending on the user's writing preferences and content, as well as unnecessary pointing out due to the immaturity of the algorithm. It is possible to flexibly respond to such situations, and improve the usability of the editing and proofreading support functions when creating texts.

[Brief explanation of the drawing]

FIG. 1 is an overall configuration diagram for explaining an embodiment of a text creation device according to the present invention, FIG. 2 is a diagram showing an example of indications made by a conventional proofreading support system, and FIG. FIG. 4 is a diagram showing an example of the points made by another embodiment. FIG. 5 is a diagram showing an example of the points made by another embodiment. FIG. 7 is a diagram showing an example of pointing out according to another conventional embodiment, and FIG. 7 is a flowchart for explaining pointing operation. 1...Display section, 2...Keyboard input section, 3...
External storage unit, 4...Printing unit, 5...Control unit, 6...
- Input control section, 7... Revision proofreading support control section, 8...
Word processor control section, 9... Output control section, 10...
- Morphological analysis unit, 11... Syntactic analysis unit, 12... Judgment search unit, 13... Control information storage unit, 14... Learning information storage unit, 15... Morphological dictionary, 16... Syntax Dictionary, 17...Edited proofreading dictionary. 18... Control information storage, 19... Learning information storage. (0) Pointed out 1 (a) Pointed out 1 (a) Pointed out 1 Figure (b) Pointed out 2 Figure (b) Pointed out 2 Figure (b) 1 pointed out 2 Figure (a) Pointed out 1 (b) Pointed out 2 (C) Point 3 (d) Point 4 Figure (a) Point 1 (b) Point 2 Figure

Claims (1)

[Claims] 1. In a writing device for Japanese or English writing,
Equipped with a proofreading means that has a proofreading function for a document being written or after it has been written, and a learning means that has a learning function that allows unnecessary operations to be avoided as much as possible regarding indications to be subsequently carried out based on operator operation information. A writing device characterized by: 2. The text creation device according to claim 1, wherein when moving on to the next pointing out without correcting the pointing out, the writing device stops making the subsequent pointing out. 3. The text creation device according to claim 2, wherein, regarding the above-mentioned point, the point pointed out is displayed, but the next point is made without requesting the operator to correct it.