JPH0258159A - Proofreading system for japanese sentence - Google Patents

Abstract

PURPOSE:To unify expressions to standard expressions without trouble that an operator selects an expression at each time of detecting fluctuation of an expression by using a standard expression dictionary, where standard expressions are preliminarily determined, to detect the presence or the absence of fluctuation of expressions and automatically correcting the expression to a standard expression by a correcting means at the time of detecting the presence of fluctuation. CONSTITUTION:A document storage means 1 where a Japanese-language document is stored, a word dictionary 3 where expressions of words and information of parts of speech, word discrimination information, and expression discrimination information corresponding to expressions are stored, a connection table 2 for parts of speech where information indicating whether connection between parts of speech is permitted or not is stored, and a standard expression dictionary 4 where sets of word discrimination information, expression discrimination information, and expressions are stored. A word dividing means 5 which divides an input sentence with a word as the unit, an expression fluctuation detecting means 6 which detects expressions other than standard expressions, and an expression fluctuation correcting means 7 which corrects expressions other than standard expressions to standard expressions are provided. The presence or the absence of fluctuation of expressions is detected by the detecting means 6; and the presence is detected, expressions are automatically corrected to standard expressions by the correcting means 7. Thus, expressions are unified to standard expressions without trouble that an expression is selected at each time of detecting fluctuation of expressions.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention is a proofreading method for Japanese text that detects and corrects orthographic deviations contained in machine-processable Japanese text input using a Japanese word processor or the like. Regarding the system.

BACKGROUND OF THE INVENTION Generally, in Japanese texts, there may be multiple ways to represent one word, but it is desirable to use one of the ways to represent one word in one document. However, when a document is created by multiple people, the notation of words tends to be inconsistent. Therefore, variations in the notation occur in which different notations for the same word coexist in the same document, resulting in an unsightly document.

Regarding this problem of variations in Japanese writing, for example, as shown in Japanese Patent Application Laid-Open No. 62-209668, when a word that can be written in multiple ways is detected, all possible spellings are presented to the operator. There is a method that allows you to choose.

Furthermore, as disclosed in Japanese Patent Application Laid-Open No. 62-22176, there is a method of converting incorrect terms into correct terms using a dictionary that stores pairs of incorrect terms and correct terms.

Problems to be Solved by the Invention However, according to the former, the notation used to express a certain word is usually fixed, so it is difficult to select one by one when there is a possibility of variation in the notation. This is troublesome, and as the number of selections increases, the possibility that variations in the notation will remain due to selection errors also increases.

Also, according to the latter, for example, "password""password"
Even if a word can be expressed in any way, as in the case of ``words'', one expression must always be set as the correct expression, and it lacks flexibility.

Means for Solving the Problem The invention according to claim 1 includes a document storage means for recording and writing Japanese documents, and storing word notation and part-of-speech information corresponding to the notation, word identification information, and notation identification information. word dictionary and
A part-of-speech connection table that stores connections between parts of speech and a standard notation dictionary that stores combinations of word identification information, notation identification information, and notations, and word division means that divides an input sentence into word units; A notation deviation detection means for detecting non-standard notation and a notation deviation correction means for correcting non-standard notation to standard notation are provided.

In addition to the invention set forth in claim 1, the invention set forth in claim 2 further includes standard changing means for changing the contents stored in the standard notation dictionary.

According to the invention as claimed in claim 1, the presence or absence of a deviation in the notation is detected by the notation deviation detection means using a standard notation dictionary in which the standard notation is determined in advance, and when there is any deviation, the correction means changes the notation to the standard notation. Corrected automatically. In other words, there is no need for the operator to select a notation every time a deviation in the notation is detected, and the notation is unified to the standard notation.

Further, according to the invention as claimed in claim 2, since the contents stored in the standard notation dictionary, that is, the standard notation can be changed by the standard changing means, the operator can arbitrarily set what he or she desires as the standard notation, and has flexibility. will have.

Example An embodiment of the present invention will be described based on the drawings.

FIG. 1 shows the system configuration according to this embodiment. First, a document storage means 1 is provided, which is a file for storing Japanese documents created by a Japanese word processor or the like. Further, as dictionary documents, a part-of-speech connection table 2, a word dictionary 3, and a standard notation dictionary 4 are provided.

The part-of-speech connection table 2 stores binary information indicating whether or not parts of speech are grammatically connected to each other, and an example thereof is shown in FIG. In the figure, "○" indicates that connection is possible, and "×
” indicates that connection is not possible.

As illustrated in FIG. 3, the word dictionary 3 stores the notation of each word, corresponding word identification information, notation identification information, and part-of-speech information. Here, the word identification information is information for determining whether words with different notations are the same word, and the same number is assigned to the same word. According to FIG. 3, "handling" and "handling" have different notations but are the same word, so the same number r102J is assigned as word identification information. The spelling identification information is number information for distinguishing between different spellings of the same word. According to the example in Figure 3, "0" is assigned to words that have only one notation, such as "drugs," and "handles" and "handles".
A number of ``1'' or higher is assigned to a word that has multiple notations, such as ``1''.

The standard notation dictionary 4 is configured to search for a pair of a standard notation and its notation identification information for a word that has multiple notations, using the word identification information as a key. An example of this is shown in FIG.

Furthermore, in order to carry out actual detection and correction processing, word division means 5, spelling deviation detection means 6, spelling deviation correction means 7 are provided, and furthermore, standard changing means 8 is provided. The word division means 5 extracts one sentence from the document storage means 1, and uses the word dictionary 3 and the part-of-speech connection table 2 to divide the sentence into word strings that can be grammatically connected to each other. At this time, word identification information and spelling identification information are assigned to each m word. For example, if the input sentence is "handling drugs", it is divided into "drugs" + "wo" + "handling j".

Regarding the words thus divided by the word dividing means 5, the spelling deviation detection and correction processing by the spelling deviation detecting means 6 and the correcting means 7 is performed according to the flowchart of FIG. First, it is checked whether the notation identification information of the word is "O". If it is "o", there is only one spelling of that word, so there is no variation in spelling. However, if it is "1" or more, the standard notation dictionary 4 uses the word identification information as a key.
Search inside and obtain a pair of notation identification information and notation.

Then, the notation identification information given to the word is compared with the notation identification information obtained from the standard table layout s4. If the two match, it is determined that the notation in the input sentence is the standard notation and there is no variation in the notation, and the notation is left as is. On the other hand, if the two notation identification information are different, it is determined that the word in the input sentence has a variation in the notation, and the word is replaced with the notation obtained from the standard notation dictionary 4.

For example, consider the case where the word "handling" is targeted. This word "handling" has word division means 5.
When dividing the word by , word identification information r102J and notation identification information "1" are given using the word dictionary 3. Therefore, when the detection means 6 searches the standard notation dictionary 4 using the word identification information r102J as a key, a set of ``2 pieces of notation identification information and the notation ``handle'' is obtained. In this way, since the orthographic identification information obtained by word segmentation and the orthographic identification information obtained by searching the standard table = dictionary 4 are different, it is determined that there is a deviation in the orthography, and the is replaced with the standard notation J, which handles r.

In this way, even if there are spelling variations in the input sentence,
It will be automatically detected and corrected to a standard notation format. In particular, since the standard notation is determined in advance by the standard notation dictionary 4, there is no need for the operator to select which notation form to correct when correcting deviations in the notation, and the notation is automatically unified to the standard notation. Can be done.

By the way, the standard notations in this standard notation dictionary 4 are:
In the initial state, it is predetermined in the system, but in Oshi 1, the user may decide to use a different notation as the standard notation depending on the word.

In such a case, the contents of the standard notation dictionary 4 may be changed at the user level by the standard changing means 8.

FIG. 6 is a flowchart showing the standard changing process performed by this standard changing means 8. First, enter the notation you want to make the new standard notation. Next, the word dictionary 3 is searched for the notation, and corresponding word identification information and notation identification information are obtained. Then, a corresponding record in the standard notation dictionary 4 is obtained using the word identification information as a key. Therefore, the notation identification information and notation of that record are rewritten to the input notation and notation identification information corresponding thereto. FIG. 7 shows the result of manually changing the notation ``handle'' to the standard notation dictionary 4 in its initial state as shown in FIG. 4 to the standard notation.

Effects of the Invention As described above, according to the invention described in claim 1, the present invention provides a standard that stores a combination of word identification information, notation identification information, and notation in addition to a document storage means, a word dictionary, and a part-of-speech connection table. a word dividing means for providing a notation dictionary and dividing an input sentence into word units; a notation deviation detection means for detecting non-standard notation;
Since a notation deviation correction means for correcting a non-standard notation to a standard notation is provided, the presence or absence of a notation deviation is detected by a notation deviation detection means using a standard notation dictionary in which standard notation is determined in advance, When there is a deviation, it is automatically corrected to the standard notation by the correction means, so there is no need for the operator to select a notation every time a deviation is detected, and it is possible to unify the notation to the standard notation. According to the invention set forth in claim 2, in addition to the invention set forth in claim 1, a standard changing means for changing the contents stored in the standard notation dictionary is provided, so that the standard notation in the standard notation dictionary is changed by the standard changing means. It is possible to use the standard notation as desired by the operator, providing flexibility.

[Brief explanation of the drawing]

The drawings show one embodiment of the present invention, and FIG. 1 is a block diagram;
Figure 2 is a configuration diagram of a part-of-speech connection table, Figure 3 is a configuration diagram of a word dictionary, Figure 4 is a configuration diagram of a standard notation dictionary, Figure 5 is a flowchart showing the process of detecting and correcting deviations in orthography, and Figure 6 is a configuration diagram of a word dictionary. The figure is a flowchart showing the standard changing means, and FIG. 7 is a configuration diagram of the changed standard notation dictionary. ■...Document storage means, 2...Part-of-speech connection table, 3...M-word dictionary, 4...Standard notation dictionary, 5...Word division means, 6
・Notation deviation detection means, 7...Notation deviation correction means, 8
・Means for changing standards

Claims (1)

[Scope of Claims] 1. A document storage means for storing Japanese documents, a word dictionary storing word notation and part-of-speech information corresponding to the notation, word identification information, orthography identification information, and a word dictionary that stores the connection between parts of speech. A part-of-speech connection table that stores information about whether or not it is possible, a standard notation dictionary that stores pairs of word identification information, orthographic identification information, and orthography, a word division means that divides input sentences into word units, and a notation that detects non-standard notations. A Japanese text proofreading system comprising a deviation detection means and a notation deviation correction means for correcting non-standard notation to standard notation. 2. The Japanese text proofreading system according to claim 1, further comprising standard changing means for changing the contents stored in the standard notation dictionary.