JPH05128146A - English morphological analyzer - Google Patents

Abstract

PURPOSE:To eliminate morpheme analysis failure due to morpheme division error by outputting plural analyzing results without selecting single result relating to an English text which contains a complex word, duplicated complex words, etc., and which is expected to obtain plural analyzing results. CONSTITUTION:A morpheme analysis control part 4 executes morpheme analysis with the English text inputted by an input means 1. A word dictionary 3 is referred to for each word (morpheme) constituting the inputted text during the morpheme analyzing process. At that time, if the word-string such as a complex word, etc., which permits plural morpheme divisions exists, a re-analysis position storing part 5 stores that position. Then, analysis is continued with one of plural division patterns. the analyzing result is stored in an analyzing pattern storage part 6. Then, analysis is performed again with another pattern starting with the position stored in the re-analysis position storage part 5, and that result is also stored in the storage part 6 and then collectively displayed on a display part 7.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an English morphological analysis process performed when translating English into Japanese.

[0002]

2. Description of the Related Art In a morphological analysis process, when an English sentence including a compound word or an English sentence that apparently forms a compound word depending on the method of morpheme division is analyzed, this English sentence becomes a compound word. There is a case in which a pattern that cannot be obtained is selected and the morphological analysis is mistaken.

[0003]

As described above, when an English sentence in which the English sentences apparently overlap to form a compound word is analyzed, only one of the patterns of the overlapping compound words is analyzed. However, if the pattern could not be a compound word in this English sentence, this is an error in morpheme division, and the result of the morpheme analysis is already at the time when the pattern of the compound word is selected. It ’s an error,
If a translation process is performed using this morphological analysis result, only an incorrect translated sentence can be issued.

The same applies to a sentence including a phrasal verb. In morphological analysis that does not perform semantic analysis of English sentences, it is not possible to determine which is the correct interpretation, that is, the morpheme splitting that is combined as a phrasal verb, and the one that is not done. It may be lost.

The present invention solves the above problems, and an object of the present invention is to provide an English morpheme analysis apparatus which eliminates the failure of morpheme analysis due to an error in morpheme division.

[0006]

In order to achieve the above object, the present invention outputs a plurality of analysis results without uniquely determining an analysis result when a plurality of analysis results such as overlapping compound words are considered. Have a configuration.

[0007]

According to the present invention, by outputting a plurality of morpheme analysis results with the above configuration, it is possible to eliminate an analysis error for morpheme division.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a functional block diagram of a morphological analysis apparatus according to an embodiment of the present invention. In the figure,
1 is an input means for inputting an English character string, 2 is an input means 1
An input English sentence storage unit for storing English sentences input from 3 is a word dictionary unit for storing headwords and morpheme information, 4 is 3
The morpheme analysis control unit 5 that performs morpheme analysis processing using the word dictionary unit of 5 is a reanalysis position storage unit 6 that stores the position of the morpheme when another interpretation of morpheme division can be considered during analysis. An analysis result storage unit for storing the morphological analysis result, and a display unit 7 for displaying the analysis result.

The relation and operation of each constituent element of the apparatus composed of the above constituent elements will be described. First, the morphological analysis control unit 4 is operated for the English sentence input by the input means 1 to execute the morphological analysis. In the morphological analysis process, the word dictionary unit 3 is searched for each word (morpheme) forming the input English sentence. At this time, if there are a plurality of morpheme-dividable word strings such as compound words, their positions are stored in the reanalysis position storage unit 5,
The analysis is continued for one of the plurality of divided patterns, and when the analysis result is obtained, the analysis result storage unit 6
The analysis result is stored in the memory, the analysis is performed again from the position stored in the reanalysis position storage unit 5 again by interpreting another pattern, and the result is also stored in the analysis result storage unit 6. In this way, one or more analysis results are stored in the analysis result storage unit 6 and displayed on the display unit 7.

The processing when a plurality of analysis results are obtained will be described below with reference to the flowchart of FIG.

First, in step A, the morphological analysis process is started. If a plurality of divisions are possible because the morpheme division process is performed in step B and the division target word string is a compound word, the current morpheme division position is stored as a reanalysis position in step C. .. The analysis process is continued using one of the plurality of division patterns (step D), and when one analysis result is obtained, the morpheme analysis result is stored in step E. After the analysis is completed, the reanalysis position is confirmed, and if reanalysis is necessary, the reanalysis position is obtained in step F, and the reanalysis position is morpheme-divided in a pattern different from the above (step B). When the processing is performed (steps C and D) and the analysis is completed, the analysis result is stored in step E. Thus, one or more morphological analysis results are stored and displayed in step G.

Further, a specific example will be described. Suppose there is a request for morphological analysis for the following English sentences.

When the morpheme division is performed on this English sentence, it is found that there are two division patterns of set and set in in the word set.

First, when analysis is performed with a pattern that is divided into sets, The / sun / set / in / the / west /
The morpheme analysis result with morpheme information divided into.
(In fact, various kinds of morpheme information are added, but only the part of speech is shown here as a representative.) The / sun / set / in / the / west /. Article noun noun preposition article noun symbol verb adverb adjective can get.

Next, when the morpheme is divided by the phrase verb set in and analyzed, it is divided into The / sun / set in / the / west /., And the morpheme analysis result with morpheme information added is The / sun / set in / the / west /. Article Noun Verb Article Noun It becomes a symbol, and the morphological analysis results of two patterns are obtained.

As described above, according to the English morphological analysis apparatus of the embodiment of the present invention, a plurality of morpheme-separable word strings such as compound words are stored in the reanalysis position storage unit, and a plurality of division patterns are stored. Since it is analyzed and displayed, morpheme analysis failure due to morpheme division processing is eliminated.

[0017]

As is apparent from the above embodiments, according to the present invention, analysis of English sentences containing compound words, compound words apparently overlapping, etc., in which multiple analysis results can be considered. Since a plurality of analysis results are output without uniquely determining the analysis result, it is possible to provide an English morpheme analysis device having no analysis error for morpheme division.

[Brief description of drawings]

FIG. 1 is a functional block diagram of a morphological analysis apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a process for obtaining a plurality of analysis results in the same embodiment.

[Explanation of symbols]

 1 input means 2 input English storage unit 3 word dictionary unit 4 morphological analysis control unit 5 reanalysis position storage unit 6 analysis result storage unit 7 display unit

Claims (1)

[Claims] 1. An input unit for inputting a character string of an original sentence, an input English storage unit for storing the character string input by the input unit, and a word dictionary unit for storing a headword of a word, morpheme information and the like. A morphological analysis control unit that performs a morphological analysis process using the word dictionary unit, an angular analysis position storage unit that stores the position of the morphological element when interpretation of another morphological division is considered during analysis, An English morphological analyzer that includes an analysis result storage unit that stores a morpheme analysis result and a display unit that displays the analysis result, and outputs the analysis results of a plurality of patterns without determining one morpheme analysis result.