JPH10275154A - System and method for supporting language dictionary production - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the language dictionary production support system which efficiently shows objective information needed to produce a language dictionary. SOLUTION: This system is equipped with a document registering means 1 which classifies and registers existing documents according to various parameters including their sources, a morpheme analyzing means 2 which converts the registered documents into sequences of morphemes consisting of word notation and parts of speech, a frequency totalizing means 3, which totalizes the appearance frequencies of the respective morphemes by the parameters at the time of the classification by the document registering means 1, and an example display means 4 which displays examples of the registered documents by the morphemes. Consequently, accurate information on the appearance frequencies of words in the documents and examples of specified words is provided as an assistant for the producing operation of the language dictionary.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and a method for supporting the production of a language dictionary such as a Japanese language dictionary, a character dictionary, a thesaurus, etc., and more particularly, to the selection of a headword in a language dictionary, the definition of semantic content, and the selection of examples. The present invention relates to a language dictionary production support system and method for supporting the processing of (1).

[0002]

2. Description of the Related Art Conventionally, when producing various language dictionaries, a language expert manually edits the dictionary based on his / her own knowledge and experience. For example, when selecting a headword in a language dictionary, explaining the meaning of the selected headword, selecting an example of a set headword, and the like, a person having specialized knowledge is based on his / her own knowledge and experience. They do their job depending on their personal judgment.

[0003]

However, such a conventional method has the following problems. In the entry word selection work, thoughts by the dictionary creator's own experience are included, and it is difficult to accurately select frequently used words. In the work of explaining the meaning of the selected headword, it is difficult to prepare a large number of documents to be referenced so that there is no bias in the usage of the headword. Is high. In the example entry work of the set headword, the reference document may be biased to a specific field, and there is a possibility that an appropriate example cannot be selected. Further, in order to appropriately perform each of such processes, a lot of human resources and time are required.

The present invention has been made in view of the above circumstances, and presents information on the frequency of use of each word in a document and an example of each word, thereby selecting a headword when creating a language dictionary. It is an object of the present invention to provide a language dictionary production support system and method that assist in various tasks including determination of the meaning and content of a headword and selection of an example of a headword.

[0005]

A language dictionary production support system according to claim 1 is a system for supporting the production of a language dictionary, wherein a document is classified and registered according to various parameters including its source. Registration means, a morphological analysis means for converting the registered document into a sequence of morphemes composed of word notation and part of speech by dividing the registered document into words and determining the part of speech of the divided words, It is characterized by comprising a frequency counting means for counting the appearance frequency of each morpheme for each parameter at the time of classification in the document registration means, and an example display means for displaying the examples in the registered document for each morpheme.

According to a second aspect of the present invention, there is provided a language dictionary production support system according to the first aspect, wherein a character aggregating means for aggregating a plurality of characters having the same content but different notations contained in a document to be registered as the same content; A synonym aggregating means for aggregating a plurality of morphemes in the notation as one synonym.

According to a third aspect of the present invention, in the linguistic dictionary production support system according to the first or second aspect, the morphological analysis means converts the words of the inflected words into morphemes in addition to their inflected forms, and the frequency counting means includes A feature of the present invention is that words having the same stem are summed up as the same morpheme even if the notation of the word is different.

According to a fourth aspect of the present invention, in the language dictionary production support system according to any one of the first to third aspects, the example display means includes means for designating the number of examples to be displayed.

According to a fifth aspect of the present invention, in the language dictionary production support system according to the third or fourth aspect, the frequency totalizing means is
The present invention is characterized in that the appearance frequency is totaled for each part of speech of a morpheme, for each inflected form of a morpheme of a conjugation word, and for each character or morpheme aggregated by the character aggregation means or the synonym aggregation means.

[0010] According to a sixth aspect of the present invention, there is provided a language dictionary production support method.
In a method for supporting the production of a language dictionary, a step of classifying and registering a document according to various parameters including its source, and a step of dividing the registered document into words and determining the parts of speech of the divided words. Converting the registered document into a sequence of morphemes composed of word notation and part-of-speech; and aggregating the appearance frequency of each morpheme for each parameter at the time of classification in the document registration means,
Displaying the example in the registered document for each morpheme.

[0011] According to a seventh aspect of the present invention, there is provided a language dictionary production support method.
7. The method according to claim 6, further comprising the steps of: aggregating a plurality of characters having the same content but different notations contained in the document to be registered as the same content; and aggregating a plurality of morphemes having the same content but different notation as one synonym. It is characterized by having.

[0012] The language dictionary production support method according to claim 8 comprises:
8. The method according to claim 6, wherein, in the step of converting to the morpheme string, the inflected words are used as morphemes in addition to their inflected forms, and in the step of counting the appearance frequency, the inflected words are differently expressed. Even though the words have the same stem, they are counted as the same morpheme.

According to a ninth aspect of the present invention, there is provided a language dictionary production support method.
In any one of claims 6 to 8, in the step of displaying the example, the number of examples to be displayed is specified.

A language dictionary production support method according to claim 10 is
10. The method according to claim 8, wherein, in the step of summing up the appearance frequency, the appearance frequency is summed up by each part of speech of a morpheme, by each inflected form of a morpheme of a conjugation word, and by each of the aggregated characters or morphemes. .

FIG. 1 is a diagram showing a configuration of a language dictionary production support system according to the present invention. As shown in FIG. 1, the system according to the present invention comprises a document registration unit 1, a morphological analysis unit 2,
It has frequency counting means 3 and example display means 4.

The document registering means 1 stores a document which has already been published in a newspaper, a novel, an essay, a commentary, a paper, etc., for example, in a newspaper, its publication date, publisher, publication column, etc., in a novel, its genre, author, etc. , By age, etc., and classified hierarchically by appropriate parameters according to the source of the source. By doing so, it is possible to obtain the morpheme tallying result and display the example, which will be described later, for each classified parameter.

The morphological analysis means 2 decomposes the registered document into words, and converts the document into a sequence of a plurality of morphemes composed of the word notation and the part of speech. At this time, if the inflected words in the document are also determined for the words that have inflections, that is, inflection words (specifically, verbs and auxiliary verbs), only the part of speech is displayed in the morpheme tallying results and example display described later. It can also be used without consideration.

The frequency counting means 3 counts the frequency of appearance of morphemes in the registered document based on the analysis result of the morphological analysis means 2 by using a parameter which is a classification criterion in the document registration means 1 as a unit. In this way, data on the appearance frequency of morphemes for each parameter can be easily obtained. At this time, regarding the morphemes of the inflected words, even if the notations are different, that is, even if the inflected forms are different, if the morphemes including words having the same stem are counted as the same morpheme, it is more realistic. The result of summation of various morphemes is obtained. In addition, it is also possible to total the frequency of appearance for each part of speech of a morpheme and for each inflected form of a morpheme of a conjugate word.

The example display means 4 displays, for each specified morpheme, a sentence before and after the morpheme so that the user can understand how the morpheme is used in an actual document. At this time, it is possible to display all the examples for each morpheme in units of a parameter which is a classification criterion in the document registration means 1 and to specify the number of examples to be displayed. In this manner, the dictionary creator can be presented with appropriate examples as desired by the person.

In addition to the above-described structure, a character aggregating means for registering new and old characters, guessing characters, and the like, so that characters having different notations but the same meaning, such as new and old characters, can be treated as identical characters. Is provided before the morphological analysis means 2, a more realistic conversion result into morphemes and a total result of morphemes can be obtained.

If words having different notations but having the same meaning are provided with synonym aggregating means for aggregating them as the same words, the words having such different notations are regarded as the same words. It is possible to obtain the morpheme tally results under various conditions, such as the morpheme tally result in the case and the morpheme tally result in the case where the words of the synonyms are not the same word.

In the language dictionary production support system of the present invention,
As described above, the frequency of use of each word in existing documents,
Information necessary for creating a language dictionary, such as examples of use of each word, can be provided to the creator appropriately without excess or shortage. Therefore,
Since it is possible to provide accurate and appropriate information to assist in tasks such as selection of headwords, determination of semantic contents, and selection of examples when producing a language dictionary, it will be greatly useful in supporting the production of a language dictionary.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be specifically described with reference to the drawings showing the embodiments. In the following description, a system that supports the production of a Japanese dictionary will be described in detail as an example.

FIG. 2 is a diagram showing the configuration of hardware for realizing the Japanese dictionary production support system of the present invention. A computer 11 and a display 12 for displaying the results of counting the frequency of appearance of morphemes, examples of morpheme use, and the like. And the computer
A keyboard 13 for key input and a mouse 14 for instruction input, a recording medium 15 such as a magnetic disk or an optical disk, and an external storage as a database for storing registered Japanese documents, analysis results of morphemes, etc. Means 16. The present invention relates to a computer-readable storage medium 15 storing a computer program including each process described below.
By loading the computer program recorded on the recording medium 15 into the disk drive 11a, each processing described later is executed.

FIG. 3 is a diagram showing the configuration of a Japanese dictionary production support system of the present invention. The Japanese dictionary production support system of the present invention converts various existing Japanese documents similar to FIG. Document registration means 1 for classifying and registering according to the parameters described above, morphological analysis means 2 for converting a registered document into a plurality of morpheme columns, frequency counting means 3 for counting the appearance frequency of morphemes in the registered document, In addition to the example display means 4 for displaying an example of the morpheme for each morpheme,
It comprises a character consolidation means 5 for consolidating characters having different notations but the same meaning as the same character, and a synonym consolidation means 6 for consolidating words having different notations but the same meaning as the same word.

Next, the operation will be described. FIG. 4 is a flowchart showing the procedure. First, an existing Japanese document to be processed, such as a newspaper, a novel, an essay, a commentary, or an article, is classified and registered by the document registration means 1 (S1). At this time, for example, in a newspaper,
Paper documents are classified according to parameters such as publishers and publication columns, and in novels, documents are classified according to parameters such as genre, writer, and age. FIG. 5 shows an example of a registration screen of the document registration means 1 for hierarchically arranging, classifying and registering documents.

Next, the character aggregating means 5 allows characters having different notations but the same meaning, such as new and old characters and guess letters, to be treated as identical characters in the subsequent processing.
These characters are collected and registered (S2). For example,
"Sawa" and "Sawa" are collected as the same character. The obtained character aggregation result is stored in the morphological analysis unit 2 and the frequency aggregation unit 3
Sent to By performing such a character aggregation function before the morphological analysis unit 2, it is possible to treat a character not registered in the computer code such as a JIS code as the same character as a character registered in the computer code. It is possible to respond more realistically.

Then, the registered document is decomposed into words by the morphological analysis means 2 to form a sequence of morphemes (word notation + part of speech + inflected form (only inflected words)) (S3).
For example, in the case of a document "Red flowers bloom",
The words are decomposed into the words “red”, “flower”, “ga”, and “blooming” to form a row of morphemes as shown in FIG. Such a process is executed for all the registered documents.

Next, based on the analysis result by the morphological analysis means 2, the frequency totalization means 3 totalizes the appearance frequency of morphemes in units of a parameter which is a classification criterion in the document registration means 1 (S4). Here, morphemes that are inflected words, more specifically, verbs (adjectives, adjectives, adjective verbs) or auxiliary verbs, are treated as one morpheme regardless of any inflected form in the document. The appearance frequency is totaled. For example, if we consider the verb that uses “Saki” and uses 5 lines, the conjugation form is “Sakaka” or “Sakiko” (inadvertent form), “Saki” or “Saki” (consecutive form),
"Blooming" (end form), "blooming" (continuous form), "blooming"
(Assumed form) and "bloom" (instructive form), all of which have the same stem as "bloom" and differ only in inflected form. Tally. In addition, as a breakdown of the tally result, the tally result of the appearance frequency in each inflected form is also obtained.

The frequency counting means 3 obtains a counting result for each part of speech. For example, if you tally with the notation "love",
The word "love" in "love is eternal" is a common noun, and the word "love" in "love" is the stem of the varieties. In this way, "love"
In the case of "Ai" which has two kinds of parts of speech, when calculating the appearance frequency, it is separately calculated such as the frequency of the common noun "Ai" and the frequency of the "Ai" I do.

FIG. 7 is a diagram showing a display example of the counting result by the frequency counting means 3. In this example, the morpheme obtained by the morphological analysis means 2 in units of the parameters registered by the document registration means 1. Is counted. The words are displayed in descending order of appearance frequency, and for the inflected words, the breakdown of the appearance frequency in each inflected form is also shown. By looking at the results of counting the appearance frequencies of such morphemes for various parameters of the registration criteria in the document registration unit 1, the dictionary creator can see at a glance which words are frequently used in the document. I can judge.

The synonym aggregating means 6 aggregates words having different notations but the same meaning, ie, synonyms with different notations, as the same word (S5). For example, the only difference between "do" and "do" is the sending.
The difference between "computer" and "computer" is only the extension or non-extension, and the difference between "integrated circuit" and "IC" is either Japanese or English notation. Are aggregated as one word.
The obtained word aggregation result is sent to the frequency counting means 3.

The frequency aggregation means 3 receives the character aggregation result from the character aggregation means 5 and the word aggregation result from the synonym aggregation means 6. Then, when it is desired to reflect these aggregation results in the appearance frequency of the morpheme (S6: YES), the aggregation results of various types of appearance frequencies can be obtained with reference to the aggregation results (S7). . That is, the character aggregation means 5
Thus, an accumulation result can be obtained in units of characters in which different characters are the same. For example, if "Sawa" and "Sawa" are combined as the same character, and "Sawada" is specified, "Sawada" and "Sawada"
Is obtained by summing the appearance frequencies of In addition,
Individual results for "Sawada" and "Sawada" are also obtained. In addition, an accumulation result in a word unit, which is converted into a synonym of the synonym by the synonym aggregation means 6, that is, an accumulation result after synthesizing the synonym of the alias is obtained. For example, "integrated circuit" and "IC"
When "integrated circuit" is designated and "integrated circuit" is designated, a totaling result obtained by summing up the appearance frequencies of "integrated circuit" and "IC" is obtained. It should be noted that the results of individual summation of “integrated circuit” and “IC” are also obtained.

As described above, the frequency counting means 3 can obtain various counting results, and can provide the counting results to the dictionary creator. Recognition can be minimized.

The specified example morpheme and the sentences before and after the morpheme are displayed by the example display means 4 (S8).
FIG. 8 is a diagram showing a display example by the example display means 4,
In this example, the verb “to” is specified, and examples of using the “to” (including inflected forms) are displayed in a list. By looking at examples of such words, the dictionary creator can determine at a glance how the word is used in the document and provide a wealth of reference information in determining the meaning of the word Will be done.

The example display means 4 can also display examples in all registered documents.
It is also possible to display examples in all documents in units of parameters at the time of classification by the document registration means 1,
Further, it is also possible to display examples of characters or synonyms morphemes collected by the character collecting means 5 or the synonym collecting means 6.

Since it is difficult to see all the examples of morphemes having a high frequency of appearance, the example display means 4 has a function of narrowing down the number of displayed morphemes. In this case, narrowing down is performed in consideration of the frequency of appearance of morphemes for each parameter classified by the document registration unit 1, so that only examples biased by the narrowing down are not displayed. When displaying examples of synonymous characters or characters or synonyms aggregated by the character aggregating unit 5 or the synonymous aggregating unit 6, not only the appearance frequency of the morpheme for each parameter but also these synonymous characters or synonyms are displayed. It is possible to narrow down in a form that also considers the frequency of appearance of word morphemes. For example, when the appearance frequencies of “Sawada” and “Sawada” are as shown in FIG. 9A, it is difficult to see all the examples at a glance because a total of 200 appear. Therefore, the number of examples to be displayed is 40 (1
/ 5), the number of examples to be displayed is also adjusted to 1/5 as shown in FIG. 9B.
Narrow down each. By doing so, it is possible to see an example that takes into account the appearance frequency of each parameter or synonymous character or synonym, without looking at all the examples of morphemes with a high appearance frequency.

If it is difficult to judge the meaning of the word using only the preceding and following sentences, the entire document can be displayed by selecting a document that is the source of an example using the word. FIG. 10 is a diagram showing a display example of such an original sentence. In this example, the original sentence of an example using "incorporate" is shown. In this way, it is possible to easily display necessary and sufficient information for the dictionary creator to make an accurate decision.

In the above-described example, a system for supporting the production of a Japanese dictionary has been described. However, the present invention can be similarly applied to the case of supporting the production of a dictionary in a language other than Japanese. Of course.

[0040]

As described above, according to the present invention, accurate information on the frequency of occurrence of a word in a document and an example of a specified word can be easily provided to a language dictionary creator. Provide excessive support for work.
According to the present invention, it is possible to select the headword of the language dictionary based on the statistical data of the words in the document, so that the work of selecting the headword can be significantly reduced, and the words having a high appearance frequency can be searched. Since it can be selected as a word, it is possible to create a language dictionary that is more suitable for actual social life taking into account the current language situation. In addition, since the meaning of the headword selected based on the example actually used can be determined, it is possible to determine the correct meaning in a short period of time. Further, since many examples are displayed for each word, an appropriate example can be selected. As described above, the present invention can greatly contribute to support for producing a language dictionary.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a language dictionary production support system of the present invention.

FIG. 2 is a configuration diagram of hardware for realizing the Japanese dictionary production support system of the present invention.

FIG. 3 is a configuration diagram of an embodiment of a Japanese dictionary production support system of the present invention.

FIG. 4 is a flowchart showing an operation procedure of the Japanese dictionary production support method of the present invention.

FIG. 5 is a diagram illustrating an example of a registration screen of a document registration unit according to the present invention.

FIG. 6 is a diagram showing an example of a morpheme column according to the present invention.

FIG. 7 is a diagram showing a display example of a tally result by the frequency tally means in the present invention.

FIG. 8 is a diagram showing a display example by an example display means in the present invention.

FIG. 9 is a diagram showing an example of narrowing down an example display in the present invention.

FIG. 10 is a diagram showing a display example of an original sentence according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Document registration means 2 Morphological analysis means 3 Frequency totaling means 4 Example display means 5 Character collecting means 6 Synonym collecting means

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FIG06F 15/403 320D 340B 15/413 310B (72) Inventor Kazuhiko Tsuda 4-5-33 Kitahama, Chuo-ku, Osaka-shi, Osaka Sumitomo Metal (72) Inventor Masami Nakamura 4-5-33 Kitahama, Chuo-ku, Osaka-shi, Osaka Sumitomo Metal Industries, Ltd.

Claims (10)

[Claims] 1. A system for supporting the production of a language dictionary, comprising: a document registration unit for classifying and registering a document according to various parameters including a source of the document; and a document unit for dividing a registered document into words. Morphological analysis means for converting the registered document into a sequence of morphemes composed of word notation and part of speech by determining the part of speech of the word, and the parameter at the time of classification in the document registration means. A language dictionary production support system, comprising: frequency counting means for counting each time; and example display means for displaying an example of the registered document for each morpheme. 2. A character aggregating means for aggregating a plurality of characters having the same content and different notations contained in a document to be registered as the same content, and a synonym for aggregating a plurality of morphemes having the same content and different notation as one synonym. 2. The language dictionary production support system according to claim 1, further comprising an aggregation unit. 3. The morphological analysis means sets a morpheme in addition to the inflected form of the inflected word, and the frequency counting means determines whether the inflected word has a different stem but has the same stem. 3. The language dictionary production support system according to claim 1 or 2, wherein the language dictionary is counted as the same morpheme. 4. The language dictionary production support system according to claim 1, wherein said example display means has means for designating the number of examples to be displayed. 5. The method according to claim 1, wherein the frequency counting means determines a part of speech of a morpheme.
5. The language dictionary production support system according to claim 3, wherein the frequency of occurrence is totaled for each inflected form of the morpheme of the inflected word and for each character or morpheme aggregated by the character aggregation means or the synonym aggregation means. . 6. A method for supporting the production of a language dictionary, wherein a step of classifying and registering a document according to various parameters including a source of the document, a step of dividing the registered document into word units and the part of speech of the divided words By converting the registered document into a sequence of morphemes composed of word notation and part-of-speech,
A language dictionary production support method, comprising: a step of totalizing each parameter at the time of classification in the document registration means; and a step of displaying an example of the registered document for each morpheme. 7. The method according to claim 1, further comprising the steps of: aggregating a plurality of characters having the same content but different notations contained in the document to be registered as the same content; and aggregating a plurality of morphemes having the same content different notation as one synonym. 7. The language dictionary production support method according to claim 6, comprising: 8. In the step of converting into a morpheme string, the words of the conjugation word are used as morphemes in addition to their conjugation forms, and in the step of totalizing the appearance frequency, the notation of the conjugation word is different. 8. The method according to claim 6, wherein words having the same stem are counted as the same morpheme. 9. The language dictionary production support method according to claim 6, wherein in the step of displaying the examples, the number of examples to be displayed is specified. 10. The method according to claim 8, wherein, in the step of totalizing the appearance frequency, the frequency of appearance is totaled for each part of speech of a morpheme, each usage form of a morpheme of a conjugation word, and each character or morpheme that is collected. Language dictionary production support method.