JP3281214B2 - Dictionary search device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dictionary search apparatus,
More specifically, the present invention relates to a dictionary search device having a function of efficiently searching a dictionary including a idiom in a headword.

[0002]

2. Description of the Related Art Conventionally, in analyzing a first language sentence, a first language sentence is used.
When words and idioms included in a language sentence are found, a method such as longest match search is used by referring to words and idioms registered in the dictionary. At this time, if the idiom search is simply realized by the longest match search, it is necessary to search the dictionary irrespective of whether or not the arrangement of words in the first language sentence is registered in the dictionary as idioms. There was a problem that the increase.

As a solution, idiom data is described in word data so that idiom processing is applied only when a certain word has idiom data, and unnecessary processing is not performed on words having no idiom. (Japanese Patent Application Laid-Open Nos. 58-92063 and 58-92063)
No. 92064). However, since all idiom data related to the word are described in the word data, one word data may have hundreds of idiom data. Therefore, in order to finally extract the possibility of at most one idiom from one word data, it is necessary to check hundreds of idiom data. In order to avoid this, components are roughly divided into those containing two or more word concatenation and those not containing them. By treating the concatenated two words as one word as an entry word, describe the idiom information data in the word data of the word (treated as one word), while including two or more word concatenations No idioms are registered as data in one word, as in the prior art described above, whereby the number of idioms per headword can be reduced,
The process of checking idiom data is reduced.
2-251873).

[0004]

According to the above-mentioned prior art, useless idiom processing can be avoided. However, the search processing is performed in two stages, ie, a search for a headword and a search for a idiom, and control becomes complicated. . Further, there is a drawback that if a word exists only as a idiom and does not exist only with the constituent word alone, the idiom cannot be registered. SUMMARY OF THE INVENTION The present invention has been made in view of the above situation, and has as its object to provide a dictionary search device having a idiom search function capable of avoiding unnecessary idiom search processing without complicating control. It is a thing.

[0005]

According to the present invention, there is provided an input sentence storing an input first language sentence.
Storage means, and an input sentence stored in the input sentence storage means.
Search target storage means for extracting and storing a part of a character string;
Register the words and idioms of the first language as headwords
Language information registering means, and
The entered character string is a headword in the language information registration means.
A search means for searching whether or not it exists.
The linguistic information registration means further includes, for each word,
Register the maximum number of constituent words of idioms starting with
The means detects a word of the headword that matches the character string.
After searching, the last registered word for the searched word
Until the number of large constituent words is reached, a partial character string of the input sentence
While adding to the search target storage means,
It is characterized by searching for the idiom at the beginning .

[0006]

According to the present invention, a headword (idiom) composed of a plurality of words connected to each other, and information on a head word of the idiom is provided in the future.
Maximum composition of idioms starting with the same word as the first word
After registering the number of words in the dictionary and searching for a headword consisting of a single word, the headwords of the idioms starting with the word are searched until the maximum number of constituent words is reached.
By performing the search while increasing the number of composed words in order, it is possible to avoid useless idiom search processing without complicating the control.

[0007]

FIG . 1 is a dictionary retrieval apparatus having a idiom retrieval function.
Is a block diagram for explaining an example of the input sentence. In the drawing, reference numeral 1 denotes an input sentence storage buffer storing a sentence input by the input device, and 2 denotes a part of the input sentence stored in the input sentence storage buffer 1. A search target storage buffer 3 for extracting and storing character strings, a dictionary file 3 in which words and idioms are registered as headwords, and each headword is provided with linguistic information such as a translation, and 4 a search target storage buffer 2 , A processor that executes a search process to determine whether or not the character string stored in the dictionary file 3 exists as a headword in the dictionary file 3, a display device or a printing device such as a CRT, and 6 a sentence in the first language by a keyboard or a scanner. , An input device 7 for calculating the number of words forming the idiom, and a word counter 8 for counting the position of the word currently being processed in the input sentence. The dictionary of the dictionary file 3 is a word idiom dictionary in which both words and idioms are registered, and there is no special distinction between words and idioms in the headword part.

In this example , basically, a dictionary search is performed by obtaining as many combinations as possible with respect to an input character string. In the dictionary search, the maximum number of words connected to form a idiom (= the number of constituent words) Restriction) is set in advance, and the search is performed while sequentially increasing the number of headwords composed starting from the word to the maximum number of words. The maximum number of words may be set to the number of constituent words obtained from the idioms included in the dictionary that have the largest number of constituent words. Hereinafter, a dictionary search for English and Japanese will be described as an example unless otherwise specified, but the same holds true for other two languages.

[0009] Figure 2 is a flow chart showing a dictionary search process according to the present embodiment, it will be described below steps S1 to S10. S1. The first language sentence is input by the input means 6 and stored in the input sentence storage buffer 1. S2. The word counter 8 is cleared. S3. The word in the input sentence specified by the word counter 8 is stored in the search target storage buffer 2. S4. The constituent word number counter 7 is cleared. S5. The character string (word or idiom) to be searched stored in the search target storage buffer 2 is searched in a dictionary. S6. The constituent word number counter 7 is incremented. S7. If the content of the constituent word number counter 7 is equal to or less than the preset limit of the number of constituent words, the process proceeds to S8, and if it exceeds the limit, the process proceeds to S9. S8. In the input sentence, a word adjacent to the last word in the search target storage buffer 2 is linked to the contents of the search target storage buffer 2, and the process proceeds to S5. S9. The word counter 8 is incremented. S10. If the word counter 8 is equal to or less than the number of words constituting the input sentence, the process proceeds to S3.

Table 1 shows "This is a read o" as a specific example.
"nly memory." is a table showing the change of each counter and buffer during the dictionary search process in the case of giving "nly memory.". Is 3), and 3.
"This", "is", "a", "r"
ead ”,” only ”,” memory ”,”. ”,” Read only m
emory ”.

[0011]

[Table 1]

As is apparent from the results in Table 1, "thi
s ”,” is ”,” a ”,” read ”,” read only memor
y ”,” only ”,” memory ”,”. "It is obtained as the search succeeds. Thus perform a dictionary search, without complicating the control, it is possible to try all possible words and phrases in the input sentence.

[0013] Description of the invention] However, among the phrases included in the dictionary, are those which consist of mostly consolidated several words, There is also a word without idiom, idioms configuration target word string Fixing the number to the number of words in the idiom consisting of the most words can take all possibilities into account,
It performs a lot of useless searches. Therefore, if the number of constituent words is dynamically changed, useless search can be avoided, and all the possibilities can be searched efficiently.

In the above-described example , the search for the idiom is always performed after the head word is searched, and the search is continued while increasing the number of constituent words to the limit of the number of constituent words. Therefore, when the first word is searched, the maximum value of the number of words constituting the idiom having the word at the beginning is known, and if the number of words constituting the idiom is set to that value, the idiom which is too long (not present in the dictionary) Avoid searching for. To find out the maximum number of words that make up the idiom having the word at the beginning of the search for the head word, check the headword in the dictionary at that time, and search for the headword composed of a single word. The maximum number of idioms starting with "" may be obtained, and that information may be recorded in the linguistic information section of the headword composed of the word.

Table 2 shows the above example "This is a read only me"
"mory." is a table of headwords composed of a single word and information on the maximum number of idioms beginning with the word. The information is input to the dictionary 3.

[0016]

[Table 2]

FIG. 3 is a flowchart of a dictionary search process when the number of constituent words information is added.
Steps S12 to S12 will be described. S1. The first language sentence is input by the input means 6 and stored in the input sentence storage buffer 1. S2. The word counter 8 is cleared. S3. The word in the input sentence specified by the word counter 8 is stored in the search target storage buffer 2. S4. The constituent word number counter 7 is cleared. Set the constituent word limit to the maximum value. S5. The character string (word or idiom) to be searched stored in the search target storage buffer 2 is searched in a dictionary. S6. If the number of constituent words is 1, the process proceeds to S7, and if it is larger than 1, the process proceeds to S8.

S7. The maximum value of the number of constituent words of the idiom starting from the current headword acquired by the dictionary search in S5 is set to the word number limit. S8. The constituent word number counter 7 is incremented. S9. If the content of the constituent word number counter 7 is equal to or less than the preset limit of the number of constituent words, the process proceeds to S10, and if it exceeds the limit, the process proceeds to S11. S10. In the input sentence, a word adjacent to the last word in the search target storage buffer 2 is connected to the contents of the search target storage buffer 2, and the process proceeds to S5. S11. The word counter 8 is incremented. S12. If the word counter 8 is equal to or less than the number of words constituting the input sentence, the process proceeds to S3, and if it exceeds, ends.

Table 3 shows, as a specific example, the same as the previous example.
"a a read only memory." is a table showing the change of each counter and the buffer in the process of the dictionary search processing when "x is a read only memory."

[0020]

[Table 3]

As is apparent from the results in Table 3, if the dictionary search of the present invention is executed, it is possible to test all the possibilities of words and idioms contained in the input sentence without complicating the control. In addition, unnecessary idiom search can be significantly reduced, so that high-speed dictionary search processing can be realized. Even when the first word is not registered in the dictionary, the idiom search is performed up to a preset word number limit, so there is no limit on the registration of idioms.

[0022]

As described above, according to the present invention , a idiom search can be realized without complicating the control of the dictionary search device, so that the dictionary search device and the dictionary structure can be simplified. In addition, unnecessary idiom search processing can be avoided, high-speed dictionary search processing can be performed, and the user's usability can be significantly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of a dictionary search device according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an example of a phrase search .

FIG. 3 is a flowchart for explaining the present invention.

[Explanation of symbols]

1 ... input sentence storage buffer, 2 ... search target storage buffer,
3 dictionary file, 4 processor, 5 display device, 6
... input device, 7 ... constituent word number counter, 8 ... word counter.

Claims (1)

(57) [Claims] An input sentence for storing an input first language sentence
Storage means, and an input sentence stored in the input sentence storage means.
Search target storage means for extracting and storing a part of a character string;
Register the words and idioms of the first language as headwords
Language information registering means, and
The entered character string is a headword in the language information registration means.
Search means for searching whether or not the word exists. The language information registration means further includes, for each word,
Register the maximum number of constituent words of the idiom starting with the word,
The search means includes a word of the headword that matches the character string.
After searching for, before it is registered for the searched word
Until the maximum number of constituent words is reached, some characters of the input sentence
While adding a column to the search target storage means,
A dictionary search device for searching for a idiom starting with a word .