JP2000250938A - Retrieval device for xml document - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a user to effectively retrieve an XML(extensible markup language) data base having the DTD(document-types tag definitions) which are structurally different from each other and semantically similar to each other with no consciousness of the difference of DTD. SOLUTION: An input analysis part 11 extracts an element name of an input retrieval expression which is produced by a data base client 3. A central control part 17 acquires a near-synonym of the extracted element name via a near- synonym extraction part 15 and compares this near-synonym with an element name that is stored in a category analogizing part 12 to select the element names which are coincident with each other. Then the part 17 produces an output retrieval expression by means of the selected element name and retrieves an XML data base 2 by using the output retrieval expression. The retrieval result of the data base 2 is notified to the client 3 via the parts 11 and 17 and an output synthesizing part 16.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an XML document retrieval apparatus, and more particularly, to an XML (extensible Markup Language) database even if a user does not know a tag type definition (DTD: Document Type Definition) of a document to be retrieved. XML that can retrieve desired data from XML
The present invention relates to a document search device.

[0002]

2. Description of the Related Art In recent years, XML has attracted attention as a language for describing and exchanging documents on the Internet or an intranet. Unlike HTML, XML uses a tag for describing a document having a structure, so that the document can be described and managed in units of small elements instead of as a unit. To date, several databases for storing and retrieving documents described in XML have been published. For example, Object Design eX
There is a product named celon.

In an XML document, since a tag can be freely defined and used by a user, a common DTD is not used among all users, but a DTD defined / extended independently by an information sender. It is conceivable that a document may be described by using. As a result, XML documents having different DTDs but having semantically similar DTDs are scattered on the Internet or an intranet.

FIG. 6 shows an example of two types of DTDs which are different in structure but similar in meaning, and an XM based on the two types.
5 shows an example of an L document and a search expression. Fig. 6 (a) shows paper, ti
This is a DTD that defines each tag of tle, author, and date (the name of the tag is called an element name), and indicates that paper includes the remaining three. On the other hand, FIG.
DT that defines each tag of Title, page, and writer
D indicates that the article contains the remaining three.

FIG. 1B shows an expression of an XML document.
r is the starting point (root element name), and the value for each element name is SAMPLE TITLE, john, 1103
Is shown. FIG. 11B shows that the DTD according to the article is used as a starting point, and that the value for each element name is SAMPLE TITLE, 123, john.

FIG. 2C is a search expression described in XML-QL, where paper is the root element name and the value of author is
This indicates that the value of title is acquired from the XML document john. Also, FIG. 13C shows that Title is a Title element from an XML document whose writer value is john, with article as the root element name.
Is obtained.

FIG. 9 is an explanatory diagram of an example of document retrieval using the XML database. The process configuration includes a database client 31 that receives a user's search request and sends the request to the database in a database operation language.
And an XML database 32 for storing an XML document and providing an external operation in a database operation language. In this conventional configuration, when a user or an application program tries to acquire the entire document or a part of the document from the database, the user or the like will be able to obtain all types of documents (eg, paper type,
It is necessary to understand the DTD of each article type and issue a search operation for each of them as shown in 33 and 34 shown in the figure.

[0008]

As described above, on the Internet and intranets, XML databases having different DTDs in structure but having semantically similar DTDs are scattered. If an attempt is made to retrieve an XML document from the XML database, it is necessary to separately describe a retrieval formula for all DTD documents that are considered to have a necessary value, which is not efficient.

For example, taking FIG. 9 as an example, if you want to know the title of a book written by john, in the XML database, the documents defined by paper and article are different. For each, FIG.
As in 33 and 34, it is necessary to separately describe a query and make an inquiry. For this reason, there is also a problem that the cost increases as the number of documents described based on different similar DTDs increases.

[0010] An object of the present invention is to eliminate the above-mentioned problems of the prior art, and to allow a user to be aware of the difference between DTDs in an XML database having a structurally different but semantically similar DTD. An object of the present invention is to provide an XML document search device that can efficiently search without using an XML document.

[0011]

In order to achieve the above-mentioned object, the present invention provides an XML document search apparatus for searching a desired document from an XML document by retrieving a tag element name from an input search formula. Means for extracting, means for extracting a synonym of the extracted element name, and
Means for acquiring a tag element name corresponding to the synonym from the category index in comparison with a category index corresponding to a tag definition (DTD) of a database, and outputting the tag element name using the tag element name acquired from the category index Means for creating a search formula for X, and using the search formula for output,
The feature is that the ML database is searched.

According to the present invention, the input search formula is
Based on a synonym of the element name of the tag described in the search expression, it is automatically converted into an output search expression having an element name corresponding to the tag definition of a document existing in the XML database. The client does not need to know the DTD of the type of the document to be searched, so that the search procedure is simplified and the search range can be expanded.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an embodiment of the XML document search system of the present invention. As shown in FIG. 1, the XML document search system includes:
It comprises an XML document search device 1, an XML database 2, and a database client 3.

The XML document retrieval apparatus 1 includes an input analysis unit 11 for receiving an external input and analyzing the input, a category analogization unit 12 for receiving a set of elements and outputting a category name characterizing the set of elements, XML database 2
Category index management unit 14 that manages a category index 13 corresponding to the information of the DTD, a synonym extraction unit 15 that outputs a plurality of synonyms of a given keyword, and an output synthesis unit that sends processing results of the search device 1 to the outside. 16 and a central control unit 17 for controlling the whole of the respective units.

The configuration of the XML document search apparatus 1 will be described in more detail. The input analysis unit 11 receives a database operation request from the database client 3 and extracts parameters of the operation request. Further, a response from the XML database 2 is received. The category analogizing unit 12 receives a set of elements belonging to the same element name from the central control unit 17, estimates a category name characterizing the element set, and sends the category name having the highest reliability to the central control unit 17. . The category index management unit 14 uses XM
The category index 13 corresponding to the DTD information in the L database 2 is managed.

The category index 13 uses a "category name" characterizing a set of elements corresponding to a certain tag of the DTD as an index key and an actual DTD corresponding thereto as a value. The “category name” is derived from an actual value of the XML database 2 by analogy using a thesaurus.

The synonym extracting section 15 outputs a plurality of synonyms of the given keyword. An existing thesaurus DB or the like can be used. For example, QZS Dictionary Ser
A thesaurus DB such as ver can be used. The output synthesizing unit 16 receives each parameter of a search expression that is a processing result performed by the XML document search device 1 in response to a database operation request made by the database client 3, and synthesizes a plurality of search expressions to form the XML database 2 To send to. Also, the X transferred from the input analysis unit 11
A response from the ML database 2 is sent to the database client 3. The central control unit 17 receives parameters from the input analysis unit 11 and performs a database operation process and a category index construction / change process using the category analogization unit 12, the category index management unit 14, and the synonym extraction unit 15, The result is sent to the output synthesizing unit 16.

Next, the operation of the XML document retrieval apparatus 1 having the above configuration will be described below. First, an operation performed by the central control unit 17 when the XML document search device 1 is connected to the XML database 2 for the first time will be described with reference to a flowchart of FIG. 2 and a specific example of FIG. This operation is an operation of constructing the category index 13 from the values of the actual XML database 2.

In step S1, the XML database 2
To all root element names and their corresponding types (DT
D), and issues a DTD registration request to the category index management unit 14. The category index management unit 14 stores the category index 13
Register the DTD. In the example of FIG. 3, the root element name “paper” stored in the XML database 2 and its corresponding DTD “paper, title, author, date”, the next root element name “article” and its corresponding DTD “ art
icle, Title, page, writer ”, the next root element name“ trip ”and the corresponding DTD“ destination, depart
ure, arrival ",... are acquired from the XML database 2 and temporarily registered in the category index 13.

In step S2, for a certain root element name in the root element names, the XML database 2
Get any number of documents (data) from In the example of FIG. 3, the document “SAMPLE, j” corresponding to the root element name “paper”
ohn, 9701, "SAMPLE2, john, 9811" and the like are acquired from the XML database 2.

In step S3, the acquired plurality of documents are sent to the category estimating unit 12, and a category name representing the sent plurality of documents is acquired. The category analogizing section 12 analogizes a category name representing the plurality of documents based on the plurality of documents, and sends the category name having the highest reliability (cname) to the central control section 17. In the example of FIG. 3, the category analogizing unit 12 determines that the document “SA”
It is assumed that the category name "book" is inferred from "MPLE, john, 9701" and "SAMPLE2, john, 9811".

In step S4, the category index management unit 1
4, a registration request for the cname is issued. The category index management unit 14 registers and manages the cname in the category index 13 in association with the root element name. In the example of FIG.
The ame “book” is registered in the category index 13 in association with the root element name “paper”.

In step S5, all root element names are
It is determined whether or not cname has been associated. If the determination is negative, the process returns to step S2 and the above operation is repeated. In the example of FIG. 3, next, the root element name “ar
documents "Flower, 101, thomas" and "An
imals, 100, tom "and" Database, 56, john "
From these, for example, the category name “book” is inferred, and cnam
The e-book "book" is registered in the category index 13 in association with the root element name "article".

The above processing is repeated, and if the determination in step S5 becomes affirmative, the processing of constructing the category index ends. By the above operation, for example, the category index 13 as shown in FIG. 5 is created.

It should be noted that the constructed category index needs to be maintained by changing it according to the insertion or updating of the data type, or by improving the accuracy of the category name as the number of stored documents increases or changes. It is. The update of the category name is performed by the central control unit 17, the category index management unit 14, and the category analogization unit 12 in response to a data operation or a data type operation.

Next, the operation of the data search processing of the XML document search apparatus 1 will be described with reference to the flowchart of FIG. 4 and the explanatory diagram of FIG. In step S11, it is determined whether or not a search expression has been input by the search operation of the database client 3. If this determination is affirmative, the process proceeds to step S12, where a certain number i is set to one. In step S13, a root element name, a parameter element name, and its value are extracted from the search expression 21. Number of extracted parameters (root element name + parameter element name)
Is x.

For example, as shown in FIG. 5, when a search formula 21 is input from the database client 3, the search formula is sent to the central control unit 17 through the input analysis unit 11. The central control unit 17 calculates the root element name “document”, the “author” corresponding to the parameter element name and “john” as its value, and the other element name “title” from the search formula 21. Extract. In this case, the number of parameters x
= 3.

In step S14, the synonym extraction unit 15
Then, the extracted root element name and parameter element names are passed to and the corresponding synonyms are obtained. In the example of FIG. 5, the root element name “document” and the parameter element names “author” and “title” are passed to the synonym extraction unit 15.
Then, the synonym extraction unit 15 returns a synonym corresponding to the root element name and the parameter element name to the central control unit 17. Note that the synonym extraction unit 15 includes:
A commercially available thesaurus DB23 can be used.

In step S15, a synonym of the root element name, for example, a book, paper, Pa
Per, Document, article, etc. are stored in the category index management unit 14
To the category index 13 through the category index 13
From the root element name and the DT that have the synonym in the category name
Get D. In the example of FIG.
Root element name "paper" corresponding to category index "book"
And "article". Further, a DTD corresponding to each root element name is obtained.

In step S16, it is determined whether or not a synonym group of the root element name exists in the category index. If this determination is negative, the process ends. On the other hand, when the result is affirmative, the process proceeds to step S17, where the number of root element names obtained from the category index is set to k, and the DTD of the i-th root element name is obtained.
An element name that matches the synonym in D is selected. At this time, the number of the selected element names is set to y.

In the example of FIG. 5, the DTD “paper, title, author, date” of the root element name “paper” is acquired, and the synonyms “author, writer,
Author, ...., Title, title, Theme, ... "is selected from the DTD. In this example, "paper,
title, author '' matches, so the `` paper, title, auth
or "is selected.

In step S18, it is determined whether or not the number of matching element names y = the number of parameters x extracted from the search formula is satisfied. If this determination is affirmative,
Proceeding to step S19, one output search formula is created.
In the example of FIG. 5, one output search formula is created using “paper, title, author”.

In step S20, it is determined whether or not i ≧ k is satisfied. When this determination is negative and when the determination in step S18 is negative, the process proceeds to step S21 and 1 is added to i. Then, returning to step S17, the next root element name (in the example of FIG. 5, “article
)), And selects an element name that matches the synonym in the DTD. In this example, "article, wr
iter, Title "is selected. The above operation is repeatedly performed, and if the determination in step S20 becomes affirmative, the process proceeds to step S20.
Proceeding to 22, the output synthesizing unit 16 synthesizes an output search formula. In the example of FIG. 5, the output search expressions 22a and 22b are obtained by this combination.

In step S23, the search expressions 22a and 22
2b is sent to the XML database 2. In step S24, responses from the XML database 2 are collected and sent to the output synthesizing unit 16 via the input analyzing unit 11,
In step S25, the collection result is sent from the output synthesizing unit 16 to the database client 3.

As described above, according to the above-described embodiment, the user is not conscious of the difference or arrangement of the DTD element names,
An XML database can be efficiently searched.

Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 6 shows the category index 1
3 is an explanatory diagram of an operation for constructing No. 3; FIG. This embodiment uses XM without using the category inference unit 12 as shown in FIG.
An arbitrary number or all of the root element names and the corresponding DTDs stored therein are acquired from the L database 2 and registered in the category index 13. According to this method, the root element name and the DTD having the contents as shown in FIG. 7 are registered as the category index 13.

Next, the operation of the data search process of the XML document search device 1 will be described with reference to FIG. The operation of the present embodiment is different from the operation of FIG. 5 in that the central control unit 17 searches for a root element name of the category index 13 based on a synonym of the root element name acquired from the synonym extraction unit 15. The other points are the same as those in FIG.

According to this embodiment, the accuracy of XML database search is considered to be slightly lower than that of the first embodiment, but there is an advantage that the category index 13 can be constructed with a simple configuration and at low cost. are doing.

[0039]

As is apparent from the above description, according to the present invention, an element name of a tag is extracted from an input search expression, and the element name is stored in an XML database based on its synonyms. Since the output search formula is created by converting to the element name of the tag that exists, the user does not need to know in advance the DTD of the document type of the XML database to be searched, and the search formula can be easily performed. Can be created. Therefore, users can search efficiently,
Search results can be obtained with high accuracy.

Further, the category index is automatically updated when there is an update such as addition, change or deletion in the document of the XML database, so that no maintenance is required.
Can be maintained in the best condition.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an embodiment of the present invention.

FIG. 2 is a flowchart showing an operation of constructing a category index according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram of an operation of constructing a category index according to the first embodiment.

FIG. 4 is a flowchart illustrating an operation of a data search process of the XML document search device according to the first embodiment of the present invention.

FIG. 5 is an explanatory diagram of an operation of a data search process of the XML document search device of the first embodiment.

FIG. 6 is a diagram illustrating an operation of constructing a category index according to the second embodiment of this invention.

FIG. 7 is a diagram illustrating an operation of a data search process of the XML document search device according to the second embodiment of the present invention.

FIG. 8 is an explanatory diagram of an example of a DTD, an XML document, and a search expression.

FIG. 9 is an explanatory diagram of a conventional XML document search method.

[Explanation of symbols]

1 ... XML document retrieval device, 2 ... XML database, 3
... Database client, 11 ... Input analysis unit, 12
... Category analogization unit, 13 ... Category index, 14 ... Category index management unit, 15 ... Synonymous word extraction unit, 16 ... Output synthesis unit, 21 ... Input search formula, 22a, 22b ... Output search formula.

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Sadao Obana 2-1-15 Ohara, Kamifukuoka-shi, Saitama F-term in Kadidi Laboratory Co., Ltd. (Reference) 5B075 KK02 KK07 KK13 KK37 KK39 ND03 ND35 NK02 NK32 NK35 PP23 PP25 PP26 PR06 QM07 QP03 UU06

Claims (5)

[Claims] 1. An XML document retrieval apparatus for retrieving a desired document from a plurality of XML documents, means for extracting a tag element name from an input retrieval formula, and extracting a synonym of the extracted element name Means to perform the synonym, tag definition (DTD) of XML database
Means for acquiring a tag element name that matches the synonym from the category index in comparison with the category index corresponding to the category index, and a means for creating an output search formula using the tag element name acquired from the category index An XML document search apparatus, comprising: searching the XML database by using the search expression for output. 2. The XML document search device according to claim 1, wherein the input search expression has a root element name, and an element name of a tag matching a synonym of the root element name is obtained from the category index. XML, characterized in that it is acquired
Document search device. 3. The XML document search device according to claim 1, wherein the category index includes a category name and an element name of a tag located below the category name, and a root of the input search expression. The synonym of the element name is compared with the category name of the category index, and if the two match, a synonym of the element name below the root element name and the tag of the category index related to the category name are further compared. An XML document search device, wherein the XML document is compared with an element name. 4. The XML document search device according to claim 3, wherein the category name of the category index is determined by analogy based on tag data stored in the XML database. XML document search device. 5. The XML according to claim 1, wherein
In the document search device, the category index is updated according to a change in the content of the XML database.
ML document search device.