JPH0544700B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method for detecting portions of Japanese text where the katakana notation is distorted.

(Prior Art) Japanese is an extremely flexible language, and it is possible to incorporate words from various other languages. Words introduced into Japanese in this way are called foreign words and are usually written using katakana.

However, the way foreign words are written in katakana is not fixed, and there are often multiple ways to write katakana. For example:

piano → piano, piano violin → violin, violin interface → interface, interface, interface interface
Although guidelines for how to write foreign words have been provided in the March 2019 Japanese Language Council report, ``How to write foreign words,'' the current situation is that the way foreign words are written in katakana differs from person to person. Furthermore, even in sentences written by the same person, the same word may be written in multiple katakana ways.

In recent years, Japanese sentences have been created using word processors, digitized, stored, and reused, and sentences created by multiple people can be combined into a single sentence, and sentences and example sentences created by others can be combined. It has become easy to partially modify and use it. In such cases, the consistency of the text as a whole becomes a problem, and one of the major causes of this is fluctuations in katakana notation.

Previously, the only way to detect such fluctuations in katakana display was to visually check the katakana portion of the written text.

(Problems to be Solved by the Invention) An object of the present invention is to provide a method for automatically detecting fluctuations in katakana notation, for which there has conventionally been no method of detecting other than visually checking.

(Means for Solving the Problem) The present invention provides a text storage means for storing sentences inputted in advance, a katakana string extraction means for extracting a katakana string from the sentences stored in the text storage means, and a katakana string extraction means. katakana string storage means for storing katakana strings extracted by katakana strings; a sequence transformation means, a transformation result storage means for storing the transformation result by the katakana sequence transformation means in association with the katakana sequence stored in the katakana sequence storage means, a transformation result stored in the transformation result means; a transformation result comparing means for comparing and detecting a group of katakana strings with matching transformation results; comparing the group of katakana sequences detected by the transformation result comparing means with the katakana strings stored in the katakana string storage means; a katakana string comparing means for detecting a group of katakana strings whose transformation results match but whose katakana strings are different before the transformation; This is a notation fluctuation detection method characterized by comprising a display means.

(Example) The structure and operation of the present invention will be explained using the drawings.

FIG. 1 is a block diagram showing the structure of an embodiment of the notation deviation detection method of the present invention. In FIG. 1, the text storage means 1 is for storing Japanese texts input in advance as character codes, and the IC
Memory, magnetic disk devices, magnetic tape devices, etc. are used.

The katakana string extraction means 2 sequentially searches for character codes corresponding to katakana from the first character of the sentence stored in the sentence storage means 1, extracts consecutive katakana as a group of katakana strings, and stores the katakana string in the katakana string storage means. Write in 3.

The katakana string storage means 3 stores the katakana string extracted by the katakana string extraction means 2 as well as position information in the sentence stored in the sentence storage means 1. Similar to the text storage means 1, an IC memory, a magnetic disk device, a magnetic tape device, etc. are used.

The katakana string transformation means 4 transforms the katakana string stored in the katakana string storage means 3 by deleting or replacing a specific katakana or partial katakana sequence, and stores the transformation result in the transformation result storage means 5. write to. A specific example of the transformation performed here (deletion or replacement of a specific katakana or partial katakana sequence) is as shown in (column 1). The katakana string transforming means 4 searches for the katakana or katakana string on the left side of the arrow below from the katakana strings stored in the katakana string storage means 3, and replaces it with the katakana or katakana string on the right side of the arrow below, or Perform deletion.

(Example 1) Vua → Bavui → Vivue → Bevuo → Bohua → Ha Hui → Hi Hue → He Huo → Hui → Yi Kii → Ki Nii → Nii Hii →
Heeei → A K → K Nei → Nei Hei →
Heyya → Ia Kiya → Kya Hiya → Hear Piya →
Pia Eya → Air Keya → Care Hair → Hair Peya →
pair → (delete) -→ (delete)... The transformation result storage means 5 is the katakana string transformation means 4
The result transformed by is stored in association with the katakana string stored in the katakana string storage means 3. An IC memory, a magnetic disk device, a magnetic tape device, etc. may be used, and it may be realized by being integrated with the katakana string storage means 3.

The transformation result comparison means 6 compares the transformation results stored in the transformation result storage means 5 and detects a group of katakana sequences with matching transformation results. The transformation results stored in the transformation result storage means 5 are given correspondences with the katakana strings in the katakana string storage means 3, and for a group of matched transformation results, the correspondence information is stored in the katakana string comparison means 7. send to

The katakana string comparison means 7 is the transformation result comparison means 6.
Based on the correspondence information sent from the above, a group of untransformed katakana sequences with matching transformation results is extracted from the katakana sequence storage means 3. Next, the katakana string comparing means 7 compares the untransformed katakana strings with which the deformation results match, and detects a group of katakana strings with which the deformation results match but with which the pre-transformation katakana sequences do not match. Since the katakana string stored in the katakana string storage means 3 is given information on the position in the sentence stored in the sentence storage means 1, the katakana string comparison means 7 detects that the detected deformation results match but are deformed. The previous katakana string sends position information in the sentence stored in the sentence storage means 1 to the fluctuation part display means 8 for a group of different katakana strings.

The fluctuation part display means 8 is a katakana string comparison means 7.
The katakana sequence of the position sent from is displayed as a candidate that is likely to have a spelling error. As the display means 8, a CRT display, a liquid crystal display, a kanji printer, etc. can be used. Display methods include marking the detected katakana string by highlighting it or displaying it in color, or displaying the detected katakana string as a keyword in KWIC format ( A method of cutting out and displaying keywords together with their surrounding context) is possible.

Next, the operation of this embodiment will be explained using an example.

First, it is assumed that the following sentence is stored in the sentence storage means 1.

It is important to increase the independence of program modules. To this end, the key is to make the interface between modules as simple as possible. A module for converting kana to kanji may have an interface that receives a kana string and returns a string containing kanji and kana, but in the case of this interface, the problem is how to handle selection operations.

Next, the katakana string extraction means 2 extracts a katakana string from the character storage means 1 and writes it into the katakana string storage means 3, resulting in the following. [ ]
The numbers in parentheses are the serial numbers of the katakana strings, and the numbers in parentheses are information about the position of the katakana string in the sentence, and are the number of characters counted from the beginning of the sentence.

[1] (1-5) Program [2] (7-11) Module [3] (36-40) Module [4] (43-50) Interface [5] (58-61) Simple [6] ( 68-71) Point [7] (83-87) Module [8] (117-124) Interface [9] (135-141) Interface For this katakana string, the katakana string transforming means 4 (for example As shown in 1), delete or replace katakana or katakana strings. That is,
"Mojiyuru" becomes "Mojiyuru" by deleting "-"
become. "Interface" becomes "interhesu" by replacing "hue → to", and becomes "interhesu" by deleting "-".
“Interface” becomes “interheis” by replacing “hue→he”, and then “hei→he”.
By replacing "he" it becomes "interhes", and by deleting "-" it becomes "interhes". "Interface" becomes "Interhesu" by replacing "Fue → He", and becomes "Interhesu" by deleting "-". The other katakana strings "program,""simple," and "point" are not transformed by the katakana string transforming means 4 of this embodiment, which performs deletion and replacement as shown in (Example 1). As a result, the following results are obtained and are stored in the transformation result storage means 5. In addition, the numbers in [ ] are
This corresponds to that of the katakana string storage means 3. However, the deletion/replacement method of the Katakana string transforming means 4 is not limited to (Example 1). In addition to (Example 1),
For example, methods such as deleting "n" or changing lowercase letters to uppercase letters may be considered, and if such a method is adopted, a modified result different from this example will be obtained.

[1] Program [2] Module [3] Module [4] Interface [5] Simple [6] Point [7] Module [8] Interface [9] Interface Next, the deformation result comparison means 6 stores the information in the transformation result storage means. A matching result is detected among the stored transformation results. As a result, a group of [2], [3], and [7] and a group of [4], [8], and [9] match, and these numbers are sent to the katakana string comparing means 7.

Therefore, the katakana string comparison means 7 uses [2][3]
The katakana string storage means 3 is searched to determine whether or not the katakana strings before transformation match [7] and [4], [8], and [9]. As a result, [2] [3]
It turns out that [7] matches, and [4], [8], and [9] differ, and the transformation results match, but the sentences related to [4], [8], and [9] that have different katakana sequences before transformation Location information (43-50), (117-124), (135-
141) is sent to the fluctuation portion display means 8.

The fluctuation part display means 8 is a katakana string comparison means 7.
Display the katakana string of positions sent from.
Examples of the display by the shaking portion display means 8 are shown in FIGS. 2 and 3. Figure 2 shows an example where the detected part is displayed with underlining, and Figure 3 shows the
This is an example of a KWIC display centered on the detected part.

(Effect of the invention) As explained above, according to the present invention, conventionally,
With regard to fluctuations in katakana display, which there is no other way to detect than by visual inspection, it is now possible to automatically detect areas where there is a high possibility of fluctuations in katakana notation, improving the efficiency of proofreading work. can be achieved.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, and FIGS. 2 and 3 are examples of a wobbling portion display means. In the figure, 1... text storage means, 2... katakana string extraction means, 3... katakana string storage means, 4
. . . Katakana string transformation means, 5 . . . Transformation result storage means, 6 . . . Transformation result comparison means, 7 .

Claims (1)

[Claims] 1 Text storage means for storing sentences input in advance, Katakana string extraction means for extracting Katakana strings from the sentences stored in the sentence storage means, Katakana strings for storing Katakana strings extracted by the Katakana string extraction means. a column storage means, a katakana sequence transformation means for deforming a katakana sequence stored in the katakana sequence storage means by deleting or replacing a specific katakana or partial katakana sequence, the katakana sequence transformation means; A transformation result storage means for storing the transformed results in association with the katakana strings stored in the katakana string storage means, and a group of katakana strings whose transformation results match by comparing the transformation results stored in the transformation result storage means. a deformation result comparison means for detecting a deformation result, by comparing a group of katakana sequences detected by the deformation result comparison means with the katakana sequence stored in the katakana sequence storage means, the deformation results match, but the deformation The preceding katakana string is characterized by comprising a katakana string comparing means for detecting a group of different katakana strings, and a fluctuation part display means for displaying a group of katakana strings detected by the katakana string comparing means. Shake detection method.