JP3005531B1 - Dictionary data search method and apparatus, search dictionary and index creation method - Google Patents

Abstract

An object of the present invention is to reduce the number of search characters in dictionary data for an input character string, thereby realizing high-speed data search. An index search unit (21) includes a character string input from an input device (1) and a character string of a record (for example, ans, ant @, ant) stored in an index storage unit (31).
By comparing with a), it is checked whether or not there is a divided dictionary corresponding to the input character string. Dictionary search means 22
Stores the divided dictionary corresponding to the divided dictionary found based on the record by the index search means 21 in the dictionary storage unit 32
, For example, "antarctic" as the input string
Is input from the input device 1, it is assumed that the index search means 21 has extracted the leading character string "anta". The dictionary search unit 22 retrieves a divided dictionary whose head character string is “anta” from the dictionary storage unit 32. Since the extracted dictionary is a set of headwords whose first character string starts with "anta",
The first four characters anta of the character string "antarctic" to be searched need not be compared.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dictionary data retrieval method and apparatus for retrieving a character string input from an input device such as a keyboard key from a dictionary entry word, a retrieval dictionary, and a method of creating an index. It is.

[0002]

2. Description of the Related Art Conventionally, when a dictionary entry word is searched using an index, a dictionary entry word has been used as it is as an index entry word.

[0003] When searching for a headword, it is possible to determine whether the headword is in the dictionary or not in the dictionary on the index. The number of words increases by the same number, and a problem remains in terms of memory efficiency.

[0004] Therefore, the dictionary is divided into several small dictionaries, an index is set so as to point to the head of the divided dictionaries, and a search in two stages, a search on the index and a search on the dictionary, is performed. Thus, a method of searching for many headwords with a small memory has been developed.

[0005]

The search method described above is disclosed in Japanese Patent Application Laid-Open No. Sho 62-197822. However, in the technology disclosed in Japanese Patent Application Laid-Open No. Sho 62-197822, a dictionary is used. For example, words that start with "b" may be included in the split dictionary specified by the index "azure" due to mechanical division.
Therefore, to search an input character string from a divided dictionary,
There is a problem that comparison must be performed from the first character.

A search method for solving the problem of the technology disclosed in Japanese Patent Application Laid-Open No. 62-197822 by reducing the number of search characters when searching for an input character string is disclosed in Japanese Patent Application Laid-Open No. 07-129600. It has been disclosed.

FIG. 11 is a diagram showing a data structure of an index and a dictionary used in the search method disclosed in Japanese Patent Application Laid-Open No. 07-129600. JP-A-07-12960
In Japanese Patent Publication No. 0, an example of a Japanese dictionary is described, but for the sake of comparison with the present invention, an English-Japanese dictionary will be used.

The index C1 shown in FIG. 11 has a structure in which head words of the divided dictionary are collected. Dictionary C2
Has a structure in which a plurality of divided dictionaries are aggregated, and each divided dictionary indicates the number of matching characters in which a headword other than the headword captured in the index C1 matches the headword immediately before it. Code (for example, 1, 2, 3)
And a character (for example, back, ndon, te, etc.) subsequent to the character string corresponding to the number of matching characters.

Although it is not described in JP-A-07-129600, it is disclosed in JP-A-07-129600 for easy understanding.
No. 129600 describes a word (for example, aback, abandon, etc.) C3 corresponding to the divided dictionary.

Next, based on FIG.
In the search method disclosed in Japanese Patent Application Publication No. 9600, the operation of searching from a divided dictionary will be described in detail using a specific example.

The index C1 and the dictionary C shown in FIG.
The case where the character string "abduce" is searched using "2" will be described.

In FIG. 12, the first entry word is "abbe
y "(Japanese: monastery)
In step J1, the number of matching characters is reset to "0" as initial processing, and in step J2, the first data "abbey" is obtained as dictionary data. Since the obtained data does not have the code shown in FIG. 11, the process proceeds to step J6, and the keyword “abbey” is extracted by the dictionary keyword extracting process.

Then, the process proceeds to step J7, in which the character string "abduce" to be searched and the dictionary data "abbey" are sequentially compared from the first character "a". As a result, the number of matching characters becomes two characters (a, b), and “search data large” indicating that the search data “abduce” is longer is output.

Next, the process proceeds to step J10, where it is determined whether or not the data is completed.
Proceed to 1 to update the number of matching characters from "0" to "2".
Then, in step J12, the next search data is set to "r".
eviate "and proceed to step J2.

As shown in FIG. 11, search data "reviat"
Since “e” has the code “3”, the process proceeds from step J3 to step J4.In step J4, the number “3” of the code of the search data is compared with the number of matching characters set at that time. , The sign is larger, the process proceeds to step J10, and it is determined whether or not the data is completed. The next search data is set to "dicate", and the process proceeds to step J2.

As shown in FIG. 11, search data "dicate"
Has the code "2", the process proceeds from step J3 to step J4. In step J4, the number "2" of the sign of the search data is compared with the number of matching characters set at that time. In this case, since the sign and the number of matching characters are equal,
Proceeding to step J5, the comparison start position is set to the third character.
In step J6, the comparison keyword "dicate" is extracted.

Next, the process proceeds to step J7, in which the character string "abduce" to be searched is compared with the dictionary data "dicate" sequentially from the third character onward. As a result, the number of matching characters becomes "1", and "Large search data" indicating that the search data is longer is output. In this way, the process proceeds to step J10, where it is determined whether or not the data is completed.
Is updated to "uce" in step J12, and the flow advances to step J2.

As shown in FIG. 11, since the search data "uce" has the code "3", steps J3 to J
Proceed to 4. At step J4, the number "3" of the sign of the search data is compared with the number of matching characters set at that time. In this case, since the code matches the number of matching characters,
Proceeding to step J5, the comparison start position is set to the fourth character.
In step J6, the comparison keyword "uce" is extracted.

Next, the process proceeds to step J7, in which four or more characters "uce" of the character string "abduce" to be searched are compared with the dictionary data "uce" sequentially. As a result, the number of matching characters is “3”
Since the number of characters compared is also three, it is determined that the keywords match (step J8), and
Proceeding to 14, the process ends as the target data is "present".

Thus, the search on the dictionary only needs to be performed for the character string after the number of matching characters, and the number of search characters for one headword can be reduced.

However, Japanese Patent Application Laid-Open No. 07-129600
Since the search method disclosed in the above publication is based on the matching relationship with the immediately preceding headword, the number of search characters is updated each time a process of comparing a character string with search data is performed (step J11). There is a problem that must be.

In the search of dictionary data, not only an input character string is searched, but also, for example, when an arrow key or the like is input from an input device such as a keyboard, the dictionary data immediately before the searched dictionary data is searched. In such a case, since it is necessary to construct a headword retroactively to the beginning of the divided dictionary, there is a problem that it is not suitable to search the data immediately before the data after searching the dictionary data of the input character string. .

Furthermore, even if a word that does not exist in the dictionary is input, it cannot be determined by searching on the index and must be searched from the dictionary. Confirm "None")
In such a case, there is a problem that extra search time is spent.

Therefore, in order to solve the problem of the search method disclosed in Japanese Patent Application Laid-Open No. 07-129600, the number of search characters is reduced to reduce the search time, and the search of data before and after the searched word is also performed. In order to easily perform the search, it is necessary to review not only the search flow but also the dictionary structure itself.

An object of the present invention is to provide a dictionary data search method and apparatus, a search dictionary and an index creation method which solve the conventional problems.

[0026]

To achieve the above object, according to an aspect of dictionary data search method according to the present invention is a dictionary data retrieval method for retrieving input character string from the entry word dictionary is the input Index from the character string data
First character is extracted as a character string, and the first character string is made common.
Consists of multiple headwords, the number of characters in the first character string
To limit the amount of data to within the specified tolerance
Split dictionary and the first character string using the first character string as an index
And the data of the divided dictionary corresponding to the first character string.
Using a plurality of records associated with data, the extracted leading character string data is used as an index, and based on the record, the divided dictionary corresponding to the extracted leading character string is searched, and the input dictionary is searched. Out of the string
The input character string is searched for from the headword of the divided dictionary by comparing the remaining character data excluding the extracted leading character string with the data of the divided dictionary as a comparison target.

Further, the dictionary data search device according to the present invention
An input device, a dictionary storage unit, an index storage unit, an index search unit, and a dictionary search unit, a dictionary data search device for searching an input character string from a dictionary entry word, wherein the input data device is used to input a character string to be searched, the dictionary storage section is composed of a plurality of headword the first character string and the common first character
Change the number of characters in a column to limit the amount of data to
The index storage unit stores a record in which the data of the first character string is associated with the data of the divided dictionary corresponding to the first character string using the data of the first character string as an index. The index search means extracts a leading character string as the index from the input character string data, and uses the extracted leading character string data as an index, based on the record, Searching the data of the divided dictionary corresponding to the extracted first character string, wherein the dictionary search means refers to the divided dictionary searched by the index search means, and among the input character strings, By comparing the remaining character data excluding the extracted leading character string with the data in the divided dictionary as a comparison target, The input character string is to find the entry word of the split dictionary.

[0028] Further, a method of creating a search dictionary according to the present invention is a method of creating a search dictionary composed of a plurality of headwords for comparing input character strings. The search dictionary consists of a plurality of split dictionaries,
Each of the divided dictionaries is composed of a plurality of headwords having a common leading character string, and a plurality of basic dictionaries are defined in units of a common leading n (where n is an integer) characters. Then, whether or not the data amount included in each of the divided dictionaries is within the allowable range of the dictionary data is checked, and the data amount of each of the divided dictionaries exceeds the allowable range of the dictionary data. In this case, the divided dictionary having the largest data amount is detected from among the divided dictionaries, and the divided dictionary having the largest detected data amount is further divided into a plurality of units by using a common first (n + 1) character as a unit. If the amount of data included in the divided dictionary exceeds the permissible range of the dictionary data, the divided dictionary having the largest data amount is re-divided. , The number of first characters Further subdivided to, by adjusting the amount of data of all the divided dictionaries within the allowable range of the dictionary data,
This is for creating a divided dictionary.

An index creation method according to the present invention is a method for creating an index for associating a divided dictionary composed of a plurality of headwords having a common leading character string with the leading character string. Is divided into a plurality of divided dictionaries in units of a common first n (where n is an integer) characters, and the first n characters are associated with the divided dictionaries corresponding to the first n characters. Check whether the data amount included in each of the divided dictionaries is within the permissible range of the dictionary data, and when the data amount of each of the divided dictionaries exceeds the permissible range of the dictionary data, Detecting the divided dictionary having the largest data amount, and further dividing the divided dictionary having the largest detected data amount into a plurality of divided dictionaries in units of a common leading (n + 1) character. n + 1) characters and, the leading (n +
1) An index for associating the divided dictionary with the leading character string is created by associating the divided dictionary with a corresponding character.

Further, the data amount of the index, which increases with the number of divided dictionaries to be re-divided, is divided and managed.

[0031]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to the drawings.

(Embodiment 1) FIG. 1 is a configuration diagram showing a dictionary data search apparatus according to Embodiment 1 of the present invention.

In FIG. 1, a dictionary data search device according to the first embodiment of the present invention includes an input device 1 and an information storage device 3.
And a data processing device 2 for retrieving an input character string from a dictionary entry word.

The input device 1 is for inputting a character string to be searched, and a keyboard or the like is used.

The information storage device 3 includes a dictionary storage unit 32 and an index storage unit 31. Dictionary storage unit 32
Are, as shown in FIG. 2, a plurality of divided dictionaries 200 and 20 each composed of a plurality of headwords having a common leading character string.
1, 202, 203, 204,... Are stored.

In the dictionary storage unit 32 shown in FIG. 2, if the headword is a common headword, for example, if the head character string is ans, the divided dictionary 200 is composed of the headword answer, answerable. If the leading character string is ant @, the divided dictionary 201 is composed of the headword ant. The @ in the leading character string ant @ means that there is no character following ant, and a symbol character other than @ may be used.

Further, as shown in FIG.
, The headword antacid, antagonism, antarct
ic, the divided dictionary 202 is formed. Similarly, when the leading character string is ante, the divided dictionary 203 is replaced with the leading character string an.
If it is th, each of the divided dictionaries 204 is configured.

As shown in FIG. 2, the index storage unit 31 stores data of the first character string (for example, ans, ant @, anta,...).
) As an index, the data of the first character string (for example, ans, ant @, anta ...) and the first character string (for example, a
ns, ant @, anta ...)
Records 100, 101, 102, 103, 104 in which data 1, 202, 203, 204,.
.. Are stored.

The data processing device 2 includes an index search unit 21 and a dictionary search unit 22.

The index retrieving means 21 includes the input device 1
Extracts a leading character string (for example, ans, ant @, anta ...) as an index from the data of the character string input from, and extracts the leading character string (for example, ans, ant @, anta ...)
Record 100, using the data of
Based on 101, 102, 103, 104,..., The divided dictionaries 200, 201, 202, 202, 2 corresponding to the extracted leading character strings (for example, ans, ant @, anta,.
.. Are searched.

The dictionary search means 22 includes the divided dictionaries 200, 201, 202,
With reference to 203, 204,..., The extracted leading character string (for example, an
s, ant @, anta...), and the divided dictionaries 200, 201, 202, 203,
By comparing the character string input from the input device 1 with the divided dictionaries 200, 201,
.. Are searched from the headwords 02, 203, 204.

Therefore, according to the dictionary data search device according to the first embodiment of the present invention shown in FIG. 1, the index search means 21 stores the character string input from the input device 1 and the character string stored in the index storage unit 31. Record 100, 1
By comparing a character string of 01, 102, 103, 104... (For example, ans, ant @, anta.
The divided dictionaries 200, 201,
It is checked whether there are 202, 203, 204,...

The dictionary search means 22 determines that the index search means 21 has records 100, 101, 102, 103, 1
.. Corresponding to the divided dictionaries found on the basis of No. 04... Are retrieved from the dictionary storage unit 32 and the input character strings are searched.

For example, when "antarctic" is input from the input device 1 as an input character string, it is assumed that the index search means 21 extracts the head character string "anta".

The dictionary search means 22 determines that the first character string is "anta"
Is extracted from the dictionary storage unit 32.
Since the extracted divided dictionary 202 is a set of headwords whose head character string starts with "anta", the character string "ant"
The first four characters anta of "arctic" need not be compared.

Therefore, according to the embodiment of the present invention,
Since the number of comparison characters between the input character string and the character string in the dictionary can be reduced, the data search can be performed at high speed.

Next, a method of creating a search dictionary and an index in the embodiment of the present invention will be described.

In the embodiment of the present invention, a search dictionary (a plurality of divided dictionaries 200, 201, 202, 203, 204)
..) Is created using a single dictionary 5 as a base, as shown in FIG.

The basic single dictionary 5 is classified from the a dictionary to the z dictionary in units of single characters a to z (step S1 in FIG. 5).

In the embodiment of the present invention, a search dictionary (a plurality of divided dictionaries 200, 201, 202, 203, 204)
..), First, as shown in FIG. 3, a basic single dictionary 5 is used.
01, a plurality of divided dictionaries, for example, a ， dictionary, aa dictionary,..., In units of leading n (where n is an integer) characters, for example, leading two characters a 文字, aa,. -Divide into an dictionary, ... The dictionary including the a ＠ dictionary, the aa dictionary,... an dictionary,.

Next, it is checked whether or not the amount of data included in each of the divided dictionaries (eg, a ＠ dictionary, aa dictionary,... An dictionary,...) 6 is within the allowable range of the dictionary data. (Step S3 in FIG. 5), when the data amount of each divided dictionary exceeds the allowable range of the dictionary data, the divided dictionary 6
Of the divided dictionaries having the largest data amount, for example, the an dictionary 601 which is the divided dictionary is detected, and the divided dictionaries 601 having the largest detected data amount are assigned to the common head (n +
1) Characters, for example, the first three characters ana,... Ant,.
.. a plurality of divided dictionaries 7 in units of anz, eg, an
a dictionary,... ant dictionary,... anz dictionary (step S4 in FIG. 5), and again the divided dictionary (eg, anana dictionary,... ant dictionary,. 7 is larger than the allowable range of the dictionary data, the divided dictionary having the largest data amount among the divided dictionaries, for example, the ant dictionary 701 which is a divided dictionary is
The number of characters of the first character is changed to, for example, the first four characters and re-divided.
Dictionary (for example, ant @ dictionary, anta dictionary)
Dictionary,...).

By repeating the above processing, the data amount of all the divided dictionaries is adjusted to be within the allowable range of the dictionary data, and the divided dictionaries are created.

Next, a method of creating an index according to the embodiment of the present invention will be described. An index creation method according to an embodiment of the present invention has a basic configuration in which a divided dictionary composed of a plurality of headwords having a common leading character string is associated with the leading character string.

A specific description will be given. That is, the basic single dictionary 5 shown in FIG. 3 is classified from the a dictionary to the z dictionary in units of single characters a to z, and as shown in FIG. , B, ...
z) corresponds to each of the a dictionary to the z dictionary of the dictionary 5. For example, the index 901 of the first character a of the index 9 is associated with the a dictionary 501 of the dictionary 5 (step in FIG. 5). S2).

Next, in the embodiment of the present invention, search dictionaries (a plurality of divided dictionaries 200, 201, 202, 203,
As shown in FIG. 3, a common first n (where n is an integer) characters, for example, the first two characters a ＠, aa,... An,. The dictionary is divided into a plurality of divided dictionaries (for example, a dictionary, aa dictionary,... An dictionary,...), And at this time, as shown in FIG.
0, for example, record 1 corresponding to the first two characters an
001 and an an dictionary 601 of the plurality of divided dictionaries 6 are associated with each other. In this case, a ＠ dictionary of the divided dictionary 6,
.. B dictionary,... Each of the z dictionary and index 10,.
.., Z are also associated with records.

As described above, when the data amount of each divided dictionary exceeds the allowable range of the dictionary data, the divided dictionary having the largest data amount among the divided dictionaries 6 is used.
For example, the an dictionary 601 which is a divided dictionary is detected, and the divided dictionary 601 having the largest detected data amount is identified as a common first (n + 1) character, for example, the first three characters ana,
... Ant,.
For example, the amount of data included in the divided dictionary (eg, ana dictionary,... Ant dictionary,. Is larger than the allowable range of the dictionary data, the divided dictionary with the largest data amount among the re-divided divided dictionaries,
For example, the ant dictionary 701, which is a divided dictionary, is re-divided by changing the number of leading characters to, for example, the first four characters, and divided into units of leading four characters, for example, ant @, anta,. ＠Dictionary, anta dictionary, ...
The work of dividing into 8) is performed. In this process, the division dictionaries 7 and 8 are associated with the records of the indexes 11 and 12 as shown in FIGS. 3 and 4 (step S5 in FIG. 5). .

In the example shown in FIG. 4, the divided dictionary (ant dictionary) in which the first three characters ant of the divided dictionary 7 are used as a unit is associated with the record 1101 corresponding to the first three characters ant in the index 11. In this case, the remaining divided dictionaries of the divided dictionaries 5, 6, 7, and 8 whose description is omitted are associated with the respective records of the indexes 9, 10, 11, and 12.

Therefore, for example, as shown in FIG.
The record 100 of the index storage unit 31 is “answe
A divided dictionary 200 having a common leading character string "ans" of "r" and "answerable" is shown.

In FIG. 2, the first character string in the record 101 of the index storage unit 31 is "
"＠" is a symbol meaning space or blank space, "、"
Is not necessary, but is represented by a symbol other than a character. this is,
While the records 102 to 104 specify the split dictionaries 202 to 204 with the first character string of 4 characters (“ant” +1 character), the record 101 has the first character string 3
If the division dictionary 201 is specified by characters ("ant"),
For example, when searching for an input character string of “antacid”, it is to avoid that a leading character string common to the input character string exists in both the record 101 and the record 102.

Next, a search operation in the embodiment of the present invention will be described with reference to FIGS.

The input character string given from the input device 1 is
It is supplied to the index search means 21.

The index search means 21 retrieves one leading character string from one record of the index from the index storage unit 31 (step S31).

In step S32, it is determined whether or not there is a leading character string.
It is assumed that the input character string is not in the dictionary (step S38), and the processing of the index search means ends.

In this case, the dictionary search means 22 is not executed,
A message that the input character string is not in the dictionary is displayed on the output device 4.

If there is a head character string, the number of characters in the head character string is set as the number of comparison characters (step S33), a comparison character string is extracted from the input character string (step S34), and the head character string and the comparison character string are compared. (Step S35).

As a result of the comparison, if the character strings are not the same, the record of the next index is extracted (step S3).
6), proceed to step S31.

If the character strings are the same, the dictionary indicated by the index is extracted (step 37), and the processing of the index search means is terminated.

Next, the dictionary search means 22 receives the input character string and the first character string of the index, and calculates a comparison start position (step S39).

In step S37, an entry word is extracted from the extracted dictionary (step S40), and the input character string and the entry word in the dictionary are compared one by one from the comparison start position (step S41).

As a result of the comparison between the input character string and the dictionary entry (step S42), if the character string is the same, the dictionary data is extracted (step S44), and the process is terminated.

If the entry word in the dictionary is larger, it is determined that the input character string is not in the dictionary (step S46), and the process is terminated.

If the input character string is larger, it is checked whether or not the extracted divided dictionary still has dictionary data (step S43).
Proceed to 46. If dictionary data still exists, the process moves to the next data (step S45) and proceeds to step S40.

That is, the comparison check is repeated from the comparison start position calculated in step S39 until the input character string is found, the dictionary data ends, or the dictionary entry becomes larger.

Next, the operation of the embodiment of the present invention will be described using a specific example. As shown in FIG. 2, for example, the index storage unit 31 has a record 102 having the first character string “anta”, and the dictionary storage unit 32 has a headword “a”.
Assume that there is a split dictionary 202 beginning with "nta".

Assume that the input character string "antarctic" is given.

When the index search unit 21 retrieves the leading character string "anta" of the index from the index storage unit 31 (step S31), it stores the number of characters 4 as the number of comparative characters (step S33), and stores the input character string in the comparative character string. "a
The first four characters of "ntarctic" are stored.

The comparison character string "anta" is compared with the character string "anta" of the index (step S35), and the dictionary indicated by the index is extracted because the character string is the same (step S37).

The dictionary search means 22 stores the start position 5 as the comparison start position (step S39), and extracts the headword "antacid" from the divided dictionary 202 (step S4).
0). The fifth and subsequent characters "rctic" of the input character string "antarctic" and the 5th of the headword "antacid" extracted from the divided dictionary 202
"Cid" is compared after the first character (step S41).

The input character string "rctic" and the dictionary entry word "cid"
Are compared in order from the first character (step S42),
The input character string is larger. Specifically, the first character "r" of the input character string is located after and after the first character "c" of the dictionary entry word. It is checked whether dictionary data still exists in the existing divided dictionary (step S43).

Since there is still dictionary data, the next data "a
Move to the position of "ntagonism" (step S45), and extract the headword "antagonism" from the dictionary.

A comparison is made between "rctic", which is the fifth and subsequent characters of the input character string, and "gonism", which is the fifth and subsequent characters of the dictionary entry word (step S41). Since the input character string is larger, the process moves to the next data (step S45), and the headword "antar" is read from the dictionary.
Take out the ctic ".

A comparison is made between "rctic" of the fifth and subsequent characters of the input character string and "rctic" of the fifth and subsequent characters of the dictionary entry (step S41). Since they have become equal, the dictionary data is extracted (step S44), the search processing is terminated, and data such as a Japanese translation of the input character string is displayed on the output device 4.

In the embodiment of the present invention shown in FIG.
The headwords of the divided dictionaries are stored in the form that also includes the first character string indicated by the index.However, the headwords of the divided dictionaries are stored with the first character strings that are common to each divided dictionary deleted. It may be. That is, in FIG. 2, the headword of the divided dictionary indicated by the leading character string “ans” of the index is “wer” instead of “answer” and “werable” instead of “answerable”.

When the dictionary is changed in this way, the processing procedure shown in FIG. 7 changes step 39 to "change input character string to character string from comparison start position" and changes step S41 to " This can be dealt with simply by changing the processing content of "data comparison". By changing the processing content in this way, there is an advantage that the size of the divided dictionary can be further reduced.

As described above, according to the embodiment of the present invention, by making the divided dictionary a group of words starting with the index entry word, the search on the divided dictionary is performed from the entry word on the divided dictionary. The comparison test need only be performed on the remaining character strings obtained by subtracting the number of characters of the index word of the index indicating the index, and the number of search characters with the dictionary data can be reduced. Therefore, the speed of data search can be increased.

Further, since the headwords of the divided dictionaries can be stored in a form in which the common leading character string of each of the divided dictionaries is deleted, the size of the divided dictionaries can be reduced.
The memory usage can be reduced, and more dictionary data can be secured using the reduced memory.

(Embodiment 2) FIG. 8, FIG. 9 and FIG.
It is a figure showing Embodiment 2 of the present invention.

Next, a second embodiment of the present invention will be described in detail with reference to the drawings. Referring to FIG. 8, in a second embodiment of the present invention, the data processing device 2 includes a higher-order index search unit 23 in addition to the configuration of the data processing device 2 in the embodiment shown in FIG. 1 in that the index storage unit 31 of the information storage device 3 in the embodiment shown in FIG. 1 is subdivided into an upper-level index storage unit 311 and a divided index storage unit 312.

The upper index search means 23 is adapted to search the input character string input from the input device 1 to which of the divided index storage units the character string belongs.

Referring to FIG. 9, the internal structure of the higher-order index storage unit will be described. Upper index storage unit 3
Reference numeral 11 denotes a head character string of the first record of the divided index storage unit 312 and a number indicating the divided index storage unit 312.

The divided index storage unit 312 has the same configuration as the index storage unit 31 in FIG.

The operation of the second embodiment according to the present invention will be described in detail with reference to the drawings. In the embodiment of FIG. 1, no measure is taken when the divided dictionary is subdivided and the index becomes exaggerated accordingly.

In the second embodiment, in such a case, the upper index search unit 23 searches the upper index storage unit 311 to specify the divided index storage unit 312.

The index search means 21 checks whether or not there is a divided dictionary from the specified divided index storage unit 312. This process is the same as the process in which the index search means 21 in FIG. Also, the dictionary search means 22 performs the same processing as the dictionary search means 22 in FIG.

The processing of the upper index search means 23 will be described with reference to the flowchart of FIG.

First, the number of the divided index to be taken out is stored so that the index starting from a can be taken out (step S61).

Next, one entry word of the divided index is extracted from the upper index storage unit 311 (step S62), and is compared with the input character string input from the input device 1 (step S63). If the input character string is smaller, the divided index storage unit 312 immediately before the compared dictionary is to be extracted, so the number of the divided index to be extracted is obtained (step S67), and the divided index is extracted (step S68).

If the input character string is larger or the spelling of the index word and the input character string are the same in step S63, the number representing the divided index is updated (step S64). Then, it moves to the next index record (step S65), checks whether there is still an index record (step S66), and if there is an index record, returns to step S62. Repeat until an index record to retrieve is found or there are no more index records.

As described above, according to the second embodiment of the present invention, the index is divided even if the index headword increases as the dictionary data increases and the index size exceeds the allowable range. In addition, there is an advantage that a large amount of data can be searched by using a small amount of memory.

[0100]

As described above, according to the present invention, since the number of search characters with dictionary data can be reduced, the speed of data search can be increased. The reason is that the split dictionary is made up of words that start with the index words in the index, and searches on the split dictionary are obtained by subtracting the number of characters in the index words that indicate the split dictionary from the index words in the split dictionary. This is because it is only necessary to perform a comparison test on only the character string of the above.

Further, since the size of the divided dictionary can be reduced, the amount of memory used can be reduced, and more dictionary data can be secured using the reduced memory. The reason is that the headwords of the divided dictionaries can be stored in a form in which the common leading character string of each of the divided dictionaries is deleted.

Further, as shown in the reason of the first effect, the comparison test of the present invention uses the remaining character string obtained by subtracting the number of characters of the index word of the index indicating the divided dictionary from the index word on the divided dictionary. Therefore, the processing speed of the comparison operation is not reduced even if the headwords of the divided dictionaries are stored in such a manner that the common leading character strings of the respective divided dictionaries are deleted.

Further, even when the index entry words increase with the increase in the dictionary data and the index size exceeds the allowable range, the index is divided to reduce the amount of memory used and to store a large amount of data. You can search.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a dictionary data search device according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram illustrating a data structure of an index storage unit and a dictionary storage unit according to the first embodiment of the present invention.

FIG. 3 is a transition diagram illustrating a method for creating a dictionary storage unit according to the first embodiment of the present invention.

FIG. 4 is a transition diagram illustrating a method of creating an index storage unit according to the first embodiment of the present invention.

FIG. 5 is a flowchart illustrating a method of creating an index storage unit and a dictionary storage unit according to the first embodiment of the present invention.

FIG. 6 is a flowchart illustrating an operation of an index search unit according to the first embodiment of the present invention.

FIG. 7 is a flowchart illustrating an operation of a dictionary search unit according to the first embodiment of the present invention.

FIG. 8 is a configuration diagram illustrating a dictionary data search device according to a second embodiment of the present invention.

FIG. 9 is a configuration diagram illustrating an internal structure of a higher-order index search unit according to the second embodiment of the present invention.

FIG. 10 is a configuration diagram illustrating an operation of a higher-order index search unit according to the second embodiment of the present invention.

FIG. 11 is a configuration diagram showing a structure of an index and a dictionary in a conventional example.

FIG. 12 is a flowchart showing the operation of the conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input device 2 Data processing device 21 Index search means 22 Dictionary search means 23 Upper index search means 3 Information storage device 31 Index storage unit 311 Upper index storage unit 312 Divided index storage unit 32 Dictionary storage unit 4 Output device

Continuation of front page (56) References JP-A-5-54077 (JP, A) JP-A-3-118661 (JP, A) JP-A-60-225273 (JP, A) JP-A-64-58018 (JP) JP-A-5-334365 (JP, A) JP-A-59-136858 (JP, A) JP-A-7-129576 (JP, A) JP-A-60-263236 (JP, A) 55-83962 (JP, A) JP-A-5-266067 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06F 17/30 G06F 17/28

Claims (5)

(57) [Claims] 1. A dictionary data retrieval method for retrieving an input character string from a dictionary entry word, comprising:
Multiple headings that extract the first character and share the first character string
And the number of characters in the first character string is changed.
A plurality of divided dictionaries whose data amount is limited within a predetermined allowable range,
Using the first string as an index and the data of the first string
Relates to the data of the divided dictionary corresponding to the first character string
Using a plurality of attached records, using the data of the extracted leading character string as an index, based on the record, searching the divided dictionary corresponding to the extracted leading character string, the input character string And searching the input character string from a headword of the divided dictionary by comparing the extracted character string with the data of the divided dictionary with the remaining character data excluding the extracted leading character string as a comparison target. Dictionary data search method. 2. A dictionary data search device comprising an input device, a dictionary storage unit, an index storage unit, an index search unit, and a dictionary search unit, and searches for an input character string from a dictionary entry word. The input device is for inputting a character string to be searched, and the dictionary storage unit is configured with a plurality of headwords having a common first character string, and changes the number of characters in the first character string. De
The index storage unit stores a plurality of divided dictionaries in which the data amount is limited within a predetermined allowable range , and the index storage unit uses the data of the first character string as an index to correspond to the first character string data and the first character string. The index retrieval means extracts a leading character string as the index from the input character string data, and stores the data of the extracted leading character string. As an index, based on the record, to search for data of the divided dictionary corresponding to the extracted leading character string, wherein the dictionary search means refers to the divided dictionary searched by the index search means, In the input dictionary, the remaining character data excluding the extracted leading character string is compared with the divided dictionary. By comparing the data, dictionary data retrieval apparatus characterized by a character string the input is to search from the entry word of the split dictionary. 3. A search dictionary creating method for creating a search dictionary composed of a plurality of headwords for comparing input character strings, wherein the search dictionary comprises a plurality of divided dictionaries, Each of the divided dictionaries is composed of a plurality of headwords having a common head character string. A single dictionary serving as a base is divided into a plurality of common head n (where n is an integer) units. Next, whether or not the data amount included in each of the divided dictionaries is within the allowable range of the dictionary data is checked, and the data amount of each of the divided dictionaries exceeds the allowable range of the dictionary data. If
Among the divided dictionaries, a divided dictionary having the largest data amount is detected, and the divided dictionary having the largest detected data amount is further divided into a plurality of divided dictionaries in units of a common first (n + 1) character. Again, when the data amount included in the re-divided divided dictionary exceeds the permissible range of the dictionary data, the divided dictionary having the largest data amount among the re-divided divided dictionaries is changed to the number of characters of the first character. And generating a divided dictionary by adjusting the data size and re-dividing the data, adjusting the data amounts of all the divided dictionaries within the allowable range of the dictionary data. 4. A method for creating an index for associating a divided dictionary composed of a plurality of headwords having a common leading character string with the leading character string, wherein a base dictionary is provided with a common leading n ( Where n is an integer)
Divided into a plurality of divided dictionaries in units of characters,
Associating a character with the divided dictionaries corresponding to the first n characters, and then checking whether or not the amount of data included in each of the divided dictionaries is within the permissible range of the dictionary data. If the data volume exceeds the allowable range of dictionary data,
Among the divided dictionaries, a divided dictionary having the largest data amount is detected, and the divided dictionary having the largest detected data amount is further divided into a plurality of divided dictionaries in units of a common first (n + 1) character. and, said first (n + 1) character, by associating the said dividing dictionary corresponding to the top (n + 1) character, in that, characterized in that to create an index that associates the said dividing dictionary top string < How to make dex. 5. The index creation method according to claim 4, wherein the data amount of the index, which increases with the number of divided dictionaries to be re-divided, is divided and managed.