JPH02148265A - Automatic indexing system - Google Patents

Abstract

PURPOSE:To take out only a more important noun as the index word by taking out nouns accompanied with case particles from analysis results of morphemes and eliminating unnecessary words from taken-out nouns. CONSTITUTION:A morpheme analysis dictionary file 2 is used to divide a sentence 1 into words by a morpheme analyzing part 3, and information of parts of speech are given to respective words to generate word data 4. A case particle taking-out rule file 5 is used to extract nouns accompanied with case particles by a noun extracting part 6, and noun data 7 is generated. Since nouns accompanied with case particles are extracted, nouns having important meanings in the sentence are taken out. An unnecessary word dictionary file 8 is used to eliminate unnecessary words such as pronouns from noun data 7 in an unnecessary word eliminating part 9, and an index word 10 is extracted. Thus, a more important noun is taken out as the index word when the number of index words given to the object sentence is limited.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to automatic indexing systems.

[Prior art] Conventionally, in automatic indexing systems, target sentences are
After elemental analysis, index words were extracted by extracting all nouns and removing unnecessary words from the extracted nouns.

[Problems to be solved] In the conventional system described above, all nouns other than unnecessary words are extracted as index words, but many index words are extracted for the target sentence, and the importance of each index word is determined. The problem was that there was a lack of information to measure the degree of

Therefore, in the conventional system, in the target sentence,
There was a problem that it was not possible to extract only the more important nouns.

The present invention has been made in view of the above-mentioned problems.
The purpose of the present invention is to provide an automatic indexing system that can extract only more important nouns as index words.

[Means for solving the problem] In order to achieve the above object, the automatic indexing system of the present invention includes a morphological analysis means that morphologically analyzes a target sentence to create word data, and a morphological analysis means that creates word data from the word data obtained by this morphological analysis means. A configuration comprising a noun extraction means for extracting a noun accompanied by a case particle to create noun data, and an unnecessary word removal means for extracting an index word by removing unnecessary words from the noun data obtained by the noun extraction means. It's Toshide.

[Example] Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of an automatic indexing system according to the present invention. In the automatic indexing system of this embodiment, a morphological analysis unit first analyzes a sentence l using a morphological analysis dictionary file 2. In step 3, word data 4 is created by dividing the word into words and adding part-of-speech information to each word.

Next, a noun extraction unit 6 extracts nouns accompanied by case particles using the case particle extraction rule file 5, and creates noun data 7. By extracting nouns accompanied by case particles, it becomes possible to extract nouns that have important meanings in the text.

Next, an unnecessary word removal unit 9 removes unnecessary words from one noun data 7 using the unnecessary word dictionary file 8. The unnecessary words here refer to nouns such as pronouns.

By the above operations, the index word IO is extracted.

[Effects of the Invention] As explained above, the present invention provides, in an automatic index, a means for morphologically analyzing a target sentence, a step for extracting a noun with a case particle from the result of the morphological analysis, and a step for extracting a noun accompanied by a case particle from the result of the morphological analysis. Having a means for removing unnecessary words has the effect of being able to extract only the more important nouns, rather than all the nouns in one sentence.

This has the effect that if the number of index words to be added to a target sentence is limited, it is possible to extract more important nouns as index words.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of an automatic indexing system according to the present invention. ■: Sentence 2: Morphological analysis dictionary file 3: Morphological analysis unit 4; Word data 5: Case particle extraction rule file 6: Noun extraction unit 7: Noun data 8: Unnecessary word dictionary file 9: Unnecessary word removal unit IO: Index Agent Patent Attorney Kihei Watanabe

Claims (1)

[Claims] a morphological analysis means for morphologically analyzing a target sentence to create word data; a noun extraction means for extracting nouns with case particles from the word data obtained by the morphological analysis means to create noun data; An automatic indexing system comprising: an unnecessary word removing means for extracting index words by removing unnecessary words from noun data obtained by the noun extracting means.