JPH11353329A - Document retrieval method, execution device therefor and medium recorded with processing program therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the technique capable of improving the retrieval function of a document management system by extending the retrieval specification of the system. SOLUTION: A retrieval equation or a retrieving condition for executing retrieval processing is inputted, whether the retrieval request is an overall document retrieval or attribute retrieval is judged, and when the retrieval request is overall document retrieval, an available retrieving function is checked by referring to overall document retrieving operator definition information defining an operator for instructing overall document retrieval and overall document retrieving class definition information defining a class to be used for the overall document retrieval and a query object for executing the overall document retrieval is generated in accordance with the contents of these defining information. Then the generated query object is transferred to a server to request retrieving processing, a retrieved result is acquired from the server and checked, a document requiring the acquisition of a document substance is specified, and the system is connected to the server for storing and managing the document to acquire the document substance.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document management system for storing and managing digitized documents by using a computer system, and more particularly to an attribute for retrieving a target document from a plurality of documents stored and managed as a document set. The present invention relates to a technology effective when applied to a document management system in which a framework for providing a search function is extended to a framework including a full-text search function.

[0002]

2. Description of the Related Art The Internet and WWW (World Wide Wealth)
With the spread of b), the number of electronic documents created and distributed in a network environment is increasing, and the need for a document management system for storing and managing these documents is increasing.

[0003] AIIM (The Association for Information a
Document Management Alliance (DMA), a task force of nd Image Management International, provides a document management system application programming interface (API).
e) and a document object model, a DMA middleware specification, has been proposed (URL http://www.aiim.org/dm
a).

FIG. 39 is a diagram showing an outline of an architecture shown in a conventional DMA specification. As shown in FIG. 39, in the DMA architecture, the document management system includes the following layers.

[0005] (1) DMA-compliant application program 400 (2) DMA system 410 (3) Document space (service element integration module) 420 (4) repository 430 Interface between program and DMA system ・ Interface between DMA-compliant application program and document space ・ Object class and interface defined in class inside DMA system ・ Object class and interface defined in document space inside document space.

The interface between a DMA-compliant application program, a DMA system, and a document space is an API (Application Programming Interface), respectively.
411 and SII (Service Integration Interface) 421
Called. The API entity is a set of interfaces defined for all object classes inside the DMA system, and the SII entity is a set of interfaces defined for all object classes inside the document space. From now on, in order to clarify the definition of the term "interface", application programs and DMA
AP for system and document space interface respectively
I and SII are described, and the interface defined in the object class is described as “class interface” or “object interface”.

Hereinafter, each of the above layers will be described. (1) DMA-compliant application program The DMA-compliant application program is a program compliant with the interface of the DMA middleware, that is, API and SII. Access the DMA system or document space via API or SII. The repository is hidden by the document space and is not directly accessed by DMA-compliant application programs.

(2) DMA System A DMA-compliant application program executes a DMA operation via an API.
Access the system. The substance of the API is a class interface of a class defined in the DMA system and a method attached to the class interface. Application programs can access objects (class instances) of the DMA system through the API. In addition, the method of the object interface attached to the object can be called.

A DMA-compliant application program uses a method dmaConnectSystemMa prepared as a library.
Call nager to get an entry point to DMA middleware. Thus, the application program first obtains a pointer to the SystemManager object. As a result, the application program is Sy
It means "connected" or "connected" to the stemManager object. The SystemManager object provides a method for the application program to return a pointer to the DMA system. This is the entry point to connect to the DMA middleware. In this way, the application program can connect to the DMA system after connecting to the SystemManager object.

[0010] The DMA system can manage a plurality of document spaces. To access documents inside these document spaces, a DMA-compliant application program must use a DMA
You need to connect to the system first. The application program can make a request to the DMA system as necessary to list a plurality of document spaces managed by the DMA system. Select the document space to use from among them and connect.

The DMA system also provides a search function over a plurality of document spaces. This will be referred to as a global search function in this specification. With this global search function, search requests for a plurality of document spaces can be issued collectively and search results can be integrated.

(3) Document Space (Service Element Integration Module) The document space is a layer in which a document object model is implemented. The DMA-compliant application program requests the DMA system to display a list of document spaces managed by the DMA system, selects a desired document space, and connects to the document space. Alternatively, if the document space to be connected is determined in advance, it is possible to immediately connect to the document space without requesting a list display.

A DMA-compliant application program accesses a document space via a service integration interface (SII). The SII entity is an interface of a class defined in the document space and a method attached to the interface.

Once connected to the document space, the application program can directly access various objects existing inside the document space via SII.
In addition, the method of the object interface attached to an object can be called to access its properties and other objects. Applications can access SII just like APIs, and you do not need to be aware of the differences.

In the document space, documents can be managed using hierarchical folders. A hierarchical folder is realized by combining a container object corresponding to a folder and a relationship object for performing a hierarchical association between the container objects. Each folder stores a lower layer folder or document object. The document object is an object group used to store one document in a plurality of expression formats (creation application, file format, etc.) and to manage a series of versions of the document. Even if a plurality of expression formats and versions exist, if the document is logically one registered document, the document object is configured to be seen as one document from the DMA-compliant application program. A document object that is prepared for each storage unit of a document entity and provides an access control method for the document entity is uniquely associated with the document entity. The DMA-compliant application program refers to the document object,
It is possible to acquire and refer to the target document entity in the target expression format.

A DMA-compliant application program can access a document entity in the document space by the following two methods.

Navigational access to reach a target document while traversing a hierarchical folder. Access by search for issuing a query to collectively obtain documents that meet desired conditions. (4) Repository Repository is a document in the document space. Stores and manages entities, document objects and metadata. The metadata includes document space metadata relating to the entire document space and document metadata relating to individual documents. In many cases,
A file system or a database system is used as a repository. The access to the repository is hidden from the application program, and when accessing the document object in the document space, the access to the repository is performed as an extension of the processing.

Next, a procedure in which a DMA-compliant application program accesses a document in the document space via the DMA middleware will be described below.

FIG. 40 is a diagram showing a procedure by which an application program of a conventional document management system accesses a document in a document space. First, a procedure in which an application program connects to a document space and accesses a document entity in a navigational manner will be described with reference to FIG.

The DmaConnectSystemManager method is called to connect to the SystemManager object and obtain the object interface IdmaSystemManager. Call the IdmaSystemManager :: ConnectSystem method to connect to the DMA system and obtain the object interface IdmaSystem. Call IdmaSystem :: EnumerateDocSpase method to get a list of document space. Select the document space to be connected from the listed document spaces. Call the IdmaSystem :: ConnectDocSpase method to get the object interface of the document space object
Get IdmaDocSpace. The highest level node of the container object forming the hierarchical folder of the document space is acquired, and the hierarchical structure is sequentially followed to reach the container object storing the document object of the target document entity. Call IdmaDocSpace :: ConnectObject method to get the document object in the container object. A target document entity is acquired by selecting and calling a method for acquiring the document entity from the object interface provided by the document object.

Next, a procedure in which the application program searches for and obtains a document in the document space will be described.
As for the document search, there are a local search for limiting the search target to the inside of each textbook space, and a global search for performing a batch search over a plurality of document spaces. First, the procedure of the local search will be described with reference to FIG.

FIG. 41 is a diagram showing a conventional local search procedure. Are the same as the access of the document space object. Call IdmaDocSpace :: GetScope method to get the object interface IdmaScope of the Scope object required for the search. Construct a Query object that describes the Query. Using a Query object that describes the Query as an argument, Idma
Call Scope :: ExecuteSearch method to instruct search execution. Obtain object interface IdmaEnumerateOfObject of ResultSet object corresponding to search result set as search execution result. (a) Call the IdmaEnumerateOfObject :: GetNext method to obtain a ResultRow object corresponding to each search result, and obtain the identifier (id) of the document object described therein. (b) Call the IdmaDocSpace :: ConnectObject method to obtain a document object in the document space. A target document entity is acquired by selecting and calling a method for acquiring the document entity from the object interface provided by the document object.

Next, the procedure of the global search is shown in FIG.
2 will be described. FIG. 42 is a diagram showing a conventional global search procedure. Are the same as the access of the document space object. Call IdmaSystem :: CreateScope method to get a Scope object for global search. Construct a Query object that describes the Query. Using a Query object that describes the Query as an argument, Idma
Call Scope :: ExecuteSearch method to instruct search execution. Obtain object interface IdmaEnumerateOfObject of ResultSet object corresponding to search result set as search execution result. Call the IdmaEnumerateOfObject :: GetNext method to obtain a ResultRow object corresponding to each search result, and obtain the identifier (id) of the document object described therein. Call the IdmaDocSpace :: ConnectObject method to get the document object in the document space. From the object interface provided by the document object,
The target document entity is acquired by selecting and calling the method for acquiring the document entity.

With the above-described procedure, the DMA-compliant application program can access the document space and search and acquire a document managed in the document space.

The Query object specified by the DMA specification adopts an object-oriented description format based on the SQL specification. That is, the Query object has the following properties.

(1) FromExpr Designates a range to be searched. Specify a searchable class. Equivalent to the SQL From statement. (2) Specify the property to be returned as a search result from the properties of the class set in SelectList FromExpr. Equivalent to the SQL Select statement. (3) QueryExpr Set search conditions. Equivalent to an SQL Where statement. (4) OrderByList Specifies criteria for sorting results in ascending or descending order. This is equivalent to the SQL Order By clause.

These properties are represented by an object or a list or tree of objects. For example, QueryExpr uses logical and arithmetic operators as nodes,
It takes a structure that connects a property or a constant to a node as an operand. Furthermore, it is also possible to adopt a structure in which this structure is connected in multiple stages in a tree shape.

Such a document search framework provided by the DMA specification provides a so-called attribute search function such as, for example, "search for a record in which a specific property has a certain value".

[0029]

As a document search method, there is a full text search in addition to an attribute search. Unlike an attribute search in which a search is performed for a specific attribute by designating a condition regarding its value, a full-text search searches for a document that includes the specified search term in the text. However, there is a problem that full-text search cannot be handled in the DMA specification.

In order to make the search function specification in the DMA specification a framework that can handle full-text search, specification expansion is required. That is, an operator is required to return TRUE when a document containing the specified term is found in the text. In addition, it is necessary to set, as operands of the operator, a search term for checking whether or not it is included in the full-text data to be searched and the full-text data.

It is possible to define and register full-text data as one property of a document object. However, when full-text data is specified as an operand of a full-text search operator and a search term is also specified as an operand, it is necessary to refer to index information for full-text search instead of referring to full-text data or a pointer to full-text data. There is.

When performing a full-text search, a ranking search that presents a numerical result as a score indicating how well the retrieved document matches the full-text search condition set by the user and presents the sorted result in the order of the score. is there. The score in the ranking search is not data that is stored in advance as a property, but is data that is dynamically generated inside the system according to search conditions when the search is executed.

According to the DMA specification, properties acquired as search results are described in SelectList. Furthermore, properties that can be described in SelectList must be properties of the class specified in FromExpr. Except in the case of joining with Join, there is only one class set in FromExpr.

In order to obtain a score as a search result based on the DMA specification, it is necessary to define the score as a property of a document class to be searched in advance. It is necessary to specify that class in FromExpr of Query object and score property in SelectList. However,
Since the score is data generated dynamically at the time of retrieval, it is inappropriate to define the score as a property in an existing document class. A problem arises in that a difference occurs between the entity of the object that is made persistent and actually stored and the class definition.

Therefore, when trying to obtain a score by performing a ranking search within the framework of the DMA specification, the score is
There is a problem that cannot be described in SelectList. This problem applies not only to scores but also to general data that is dynamically generated during retrieval.

The document management system using the DMA middleware described in the above-mentioned prior art, or when searching for document information in the document space, a search is performed using an SQL-based Query object, and a full-text search function is provided. Extends the existing attribute search function to the full-text search function for searching documents or objects in the document space in a document management system using middleware that has the same search function specifications as DMA middleware, which has only the attribute search function without having In the case where the present invention is implemented, there is no operator for full-text search, full-text data cannot be specified, and data generated dynamically during a search such as a score cannot be handled. This is a problem to be solved.

In recent years, Z39.50 has attracted attention as a standard protocol for document search mainly in the United States. Z39.50 specifies an access protocol to a document database, a query language for search, and an attribute set. The query language is a specification that can support both attribute search and full-text search. In the document management system, this Z3
It may be important to support the search function based on the 9.50 specification.

In the DMA specification, "A" is used as a logical operation operator.
nd ”and“ Or ”operators are defined, while Z39.
"And", "Or", "Prox" and "And-Not" in 50 specifications
An operator has been defined. Therefore, in order to support the search function based on the Z39.50 specification, it is necessary to further support an And-Not operator and the like.

For the operation of the And-Not operator, an equivalent function can be realized by combining the And operator and the Not operator. However, when the Not operator is used, a large amount of temporary search result sets may be generated in the course of the calculation, and it is necessary to hold this. Therefore, there is a problem that resources required by the document management system increase. Therefore, without increasing the resources required by the document management system, Z39.
In order to be able to support 50 search specifications, it is necessary to support as an And-Not operator, which is an atomic operator.

An object of the present invention is to solve the above-mentioned problem, and to provide a technique capable of improving the search function of the document management system by extending the search specification of the document management system.

Another object of the present invention is to provide a technique capable of realizing a search function capable of complying with the Z39.50 search specification.

[0042]

According to the present invention, in order to solve the above-mentioned problems, an operator for expanding a DMA-specified attribute search function and designating a full-text search function without impairing a framework of the attribute search function, and a search target. Introduce a class to specify in the SelectList the data that is dynamically generated at the time of retrieval, such as full-text data and score.

When a full-text search is instructed, a Query object is generated using these full-text search operators and classes. Also, when interpreting the Query object, determine whether or not the full-text search operator and class are used, and if used, execute the full-text search by referring to the full-text search index, If not used, execute attribute search by referring to the persisted object.

These introduced operators and classes are registered in the client and server as definition information. Further, using these operators and classes, the client side refers to the definition information based on the search request from the user to generate a Query object, and the server side refers to the received Query object to the definition information. There is provided a processing step for interpreting.

The operators to be introduced are a full-text search operator capable of designating full-text data and search terms as operands, and a full-text search operator to which a function for calculating a score is added. These can be defined as a single operator having a flag for designating a score.

The class to be introduced is an abstract class having data and dynamically generated data at the time of retrieval for each registered document and full-text data as properties, such as a score. No need to persist. By defining dynamically generated data and full-text data in one class, the search target class can be set in FromExpr in one class.

In the processing step of generating a Query object, an appropriate operator and class are selected depending on whether the search request from the user is an attribute search or a full-text search. That is, when full-text search is specified, search conditions are configured using the introduced operator and set in QueryExpr. At this time, the full-text data and search terms defined in the properties of the introduced class are set as operands. The introduced class is From
Specify as a search target in Expr. Further, properties to be obtained as search results, including data dynamically generated at the time of a search such as a score, are listed in the SelectList.

In the processing step of interpreting the Query object, when traversing the tree of the Query object, an instruction to execute a full-text search by recognizing that full-text search is specified when an introduced operator is encountered I do.

As described above, an operator for designating the full-text search function and a class for designating data to be dynamically generated at the time of retrieval, such as full-text data to be searched and a score, in the SelectList are introduced. By using these operators and classes to provide a processing step for generating a Query object on the client side and a processing step for interpreting the Query object on the server side, this can be done without impairing the framework of the attribute search function of the DMA specification. Can be extended to realize a full-text search function.

As described above, according to the present invention, a document management system using DMA middleware, or a search using an SQL-based Query object when searching document information in a document space, and a full-text search function In a document management system using middleware that has a search function specification similar to that of DMA middleware that has only an attribute search function without having a document search function, by introducing a new full-text search operator and class, The full-text search function for searching for an object can be easily realized by extending the existing attribute search function.

As a result, the conventional attribute search framework can be easily extended to a framework that includes full-text search without impairing the framework, and the effect of improving the search function can be obtained.

In the present invention, in order to solve the above-mentioned problems,
Introduce And-Not operator. As a result, a search function capable of complying with the Z39.50 search specification can be realized by extending these attributes without losing the framework of the attribute search function of the DMA specification and the framework of the full-text search function that extends the DMA specification.

The introduced And-Not operator is registered in the client and the server as definition information. Furthermore, using this operator, the client side refers to the definition information based on the search request from the user, and
A processing step for generating a y object and a processing step for interpreting the received Query object with reference to the definition information are provided on the server side.

When the And-Not operator is used, it is possible to suppress the generation of a temporary search result set in the process of operation. That is, an increase in resources required by the document management system can be suppressed. This makes it possible to realize a search function that can support the Z39.50 search specification without significantly increasing the processing cost required by the search function of the document management system.

As described above, according to the document management system of the present invention, since the operator and the class for designating the full-text search function are introduced, the search specifications of the document management system are extended to search the document management system. It is possible to improve the function.

[0056]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (Embodiment 1) An operator and a class for designating a full-text search function are introduced below, and a Query object is generated and interpreted using these operators and classes. Embodiment 1 of extending this to realize a full-text search function without impairing the framework
Will be described.

FIG. 1 is a diagram showing the principle of the document management system according to this embodiment. In the document management system according to the present embodiment, the operator for specifying the full-text search function by extending the DMA-specific attribute search function without impairing the framework of the DMA-specific attribute search function, the full-text data to be searched, and the search of the search such as the score Introduce a class to specify dynamically generated data in SelectList. When instructing a full-text search,
Using these full-text search operators and classes, Qu
Create an ery object. Also, when interpreting the Query object, determine whether or not the full-text search operator and class are used, and if used, execute the full-text search by referring to the full-text search index, If not used, execute attribute search by referring to the persisted object.

These introduced operators and classes are registered in the client and server as definition information. Further, using these operators and classes, the client side refers to the definition information based on the search request from the user to generate a Query object, and the server side refers to the received Query object to the definition information. There is provided a processing step for interpreting.

The operators to be introduced are a full-text search operator capable of specifying full-text data and a search term as operands, and a full-text search operator to which a function for calculating a score is added. These can be defined as a single operator having a flag for designating a score.

The class to be introduced is an abstract class whose properties are data that is dynamically generated at the time of retrieval for each registered document, such as a score, and full-text data. No need to persist. By defining dynamically generated data and full-text data in one class, the search target class can be set in FromExpr in one class.

In the processing step of generating a Query object, an appropriate operator and class are selected depending on whether the search request from the user is an attribute search or a full-text search. That is, when full-text search is specified, search conditions are configured using the introduced operator and set in QueryExpr. At this time, the full-text data and search terms defined in the properties of the introduced class are set as operands. The introduced class is From
Specify as a search target in Expr. Further, properties to be obtained as search results, including data dynamically generated at the time of a search such as a score, are listed in the SelectList.

In the processing step for interpreting the Query object, when traversing the tree of the Query object, an instruction to execute a full-text search by recognizing that full-text search is specified when an introduced operator is encountered I do.

As described above, an operator for designating a full-text search function, a class for designating data dynamically generated during a search, such as full-text data to be searched and a score, in a SelectList are introduced. By using these operators and classes to provide a processing step for generating a Query object on the client side and a processing step for interpreting the Query object on the server side, this can be done without impairing the framework of the attribute search function of the DMA specification. Can be extended to realize a full-text search function.

FIG. 2 is a diagram showing the overall configuration of the document management system of the present embodiment. The document management system according to the present embodiment includes a registration client 101, a reference client 10
2. It comprises a search client 103, a network 16, a global document space management server 201, and document space management servers 301 to 303.

The registration client 101 holds the document to be registered, checks the available document space management server by inquiring the global document space management server 201, determines the registration destination of the document, and registers the document via the network 16. The previous document space management server 301, 302 or 30
Transfer to 3 and request registration.

The reference client 102 requests the document space management server in which the registered document to be registered is registered, obtains it via the network 16, and refers to it.

When performing a global search, the search client 103 uses the global document space management server 201.
When performing a local search, the document space management servers 301 to 303 are requested to perform a search process. When requesting a search process to the global document space management server 201, the global document space management server 201 sends the search client 10 to a plurality of document space management servers defined as global search targets.
3 is issued in a lump, the search results returned from each document space management server are merged, and the search client 103 acquires the search results via the network 16. According to this result, determine the document to be referenced,
The document is requested to be referred to a document space management server in which the document is registered, and the document is acquired via the network 16 for reference. When requesting the search processing from the document space management servers 301 to 303, the document space management server performs the search processing using the search index, and the search client 1
03 acquires the search result via the network 16. A document to be referred to is determined according to the result, and the document is acquired and referenced via the network 16.

The document space management servers 301 to 303 store and manage documents at the request of the registration client 101. When registering a document, a document entity that is an electronic document to be registered in the document management system, a hierarchical directory in which the document is registered when the document entity is managed by the document management system, and a logical A document object that expresses and manages a simple structure, document space metadata that is attribute information on a document space composed of a set of the above document entities, and document metadata that is attribute information of each of the above document entities are registered. Document space management server is registration client 1
01, reference client 102 and search client 1
In response to a request from the server 03, registration, reference, update, deletion or search of these data is performed. When updating the document space metadata and the document metadata accompanying the registration, update and deletion of the document, the document space metadata is updated by the global document space management server 2.
Call 01.

The global document space management server 201
A plurality of document space management servers or document spaces are grouped and managed. A list of available document space management servers (document spaces) is transmitted via the network 16 in response to a request from each client. In addition, document space metadata and document metadata of each document space management server are collected, and collectively stored and managed as global metadata. When updating the document space metadata and the document metadata accompanying the registration, update and deletion in each document space management server, the updated document space metadata and the document after receiving the notification from the corresponding document space management server Get metadata and update global metadata. A global metadata reference request from the client is received and transmitted via the network 16. Further, it performs full-text search processing in response to a global search request from the search client 103, and transmits search results.

The network 16 is a local area network or a wide area network. The network 16 is an arbitrary combination of the registration client 101, reference client 102, search client 103, network 16, global document space management server 201, and document space management servers 301 to 303. It is connected so that communication of is possible.

FIG. 2 shows the registered client 101,
Although the configuration has one reference client 102 and one search client 103, each client may have two or more configurations. It is also possible to adopt a configuration in which one client has a plurality of functions, for example, registration and reference.

The document space management servers 301 to 30
3. It is possible to adopt a configuration in which the global document space management server 201 is mounted on the same machine, or a configuration in which these servers and clients are mounted on the same machine.

Further, in the document management system of the present embodiment, various commands and data are exchanged via the network 16, but it is also possible to use a portable medium such as a floppy disk, a magneto-optical disk, or an optical disk. It is.

The registration client of the document management system according to the present embodiment, that is, 101 in FIG. 2 will be described below. FIG. 3 is a diagram showing a schematic configuration of the registration client 101 of the present embodiment. The registration client 101 is a keyboard 10, a display 11, a central processing unit (CPU) 1
2, a floppy disk drive 13, a floppy disk 14, a communication control device 15, a main memory 17, a magnetic disk device 18, and a system bus 19.

The keyboard 10 is used to input a command for instructing execution of a document registration process and the like. The display 11 is used to display the execution state of the processing in the client and the like. The central processing unit 12 executes various programs constituting the client. The floppy disk drive 13 has a floppy disk 14
Used to read and write data to The floppy disk 14 stores a registered document in advance and is used to input the registered document to the client. Communication control device 15
Is used to communicate with the document space management server and the global document space management server via the network 16 to exchange commands and data. Main memory 17
Is used to hold various programs and temporary data for executing processing in the client.
The magnetic disk device 18 is used to store a document to be registered in advance. The system bus 19 is used to connect these various devices.

In the main memory 17, a document registration control program 111, a global document space management server connection program 120, a document space list acquisition program 121, a document space management server connection program 122, a document registration request program 123, and a system program 20 are stored. The work area 21 is stored.

In the magnetic disk device 18, a registered document storage area 124 is secured.

The document registration control program 111 controls the execution of the global document space management server connection program 120, the document space list acquisition program 121, the document space management server connection program 122, and the document registration request program 123, and sends them to the document space management server. Register the document.

The global document space management server connection program 120 establishes a connection between the registration client 101 and the global document space management server 201.

The document space list acquisition program 121 requests the global document space management server 201 to transmit the list data of the managed document space management servers 301 to 303 to the registration client 101, and sends this request. get.

Document space management server connection program 122
Are registered client 101 and document space management server 30
Establish a connection with the server 1 to 303.

The document registration request program 123 requests the document space management servers 301 to 303 with which the registration client 101 has established a connection to register a document. In response to this, the document space management server registers the document as a document entity, generates a new document object and document metadata, and updates the document space metadata.

The system program 20 provides basic functions necessary for executing various programs constituting the client on a computer, such as input / output of data with peripheral devices. The work area 21 is used to store data that is temporarily required when executing the program.

The registered document storage area 124 is used for preparing and storing documents to be registered in the document space management server in advance.

Although the document management system according to the present embodiment is configured to input and read a document stored in the floppy disk 14, a document such as a magneto-optical disk or an optical disk may be used.
It is also possible to adopt a configuration in which the document is read from another portable medium, or a configuration in which a document transferred via the network 16 is input. Further, in the document management system of the present embodiment, the document to be registered is transferred to the document space management server via the network 16, but instead, a portable medium such as a floppy disk, a magneto-optical disk, or an optical disk is used. It is also possible to take

Hereinafter, a reference client of the document management system according to the present embodiment, that is, 102 in FIG. 2 will be described. FIG. 4 is a diagram showing a schematic configuration of the reference client 102 of the present embodiment. The reference client 102 includes a keyboard 10, a display 11, a central processing unit 12, a floppy disk drive 13, and a floppy disk 1.
4, a communication control unit 15, a main memory 17, a magnetic disk unit 18, and a system bus 19.

The keyboard 10 is used for inputting a command for instructing execution of document reference processing and the like. The display 11 is used to display the execution state of the processing in the client and the like. The central processing unit 12 executes various programs constituting the client. The floppy disk drive 13 has a floppy disk 14
Used to read and write data to The floppy disk 14 is used to store a reference document referred by the client. The communication control device 15 communicates with the document space management server and the global document space management server via the network 16 and is used to exchange commands and data. The main memory 17 is used to hold various programs and temporary data for executing processing in the client. The magnetic disk device 18 is used to store a referenced document. The system bus 19 is used to connect these various devices.

The main memory 17 includes a document reference control program 112, a global document space management server connection program 130, a document space list acquisition program 131, a document space management server connection program 132, a document reference request program 133, and a system program 20. The work area 21 is stored.

In the magnetic disk device 18, a reference document storage area 134 is secured.

The document reference control program 112 controls the execution of the global document space management server connection program 130, the document space list acquisition program 131, the document space management server connection program 132, and the document reference request program 133. Reference of document.

The global document space management server connection program 130 establishes a connection between the reference client 102 and the global document space management server 201.

The document space list acquisition program 131 requests the global document space management server 201 to transmit the list data of the document space management servers 301 to 303 under management to the reference client 102, and sends this request. get.

Document space management server connection program 132
Refers to the reference client 102 and the document space management server 30
Establish a connection with the server 1 to 303.

The document reference request program 133 requests the document space management servers 301 to 303 with which the reference client 102 has established a connection to refer to a document. In response, the document space management server extracts the stored document entity and transfers it to the reference client 102 via the network 16.

The system program 20 provides basic functions necessary for executing various programs constituting the client on a computer, such as inputting / outputting data to / from a peripheral device. The work area 21 is used to store data that is temporarily required when executing the program.

The reference document storage area 134 is used to store a reference document transferred from the document space management server.

Hereinafter, the search client of the document management system of this embodiment, that is, 103 in FIG. 2 will be described. FIG. 5 is a diagram illustrating a schematic configuration of the search client 103 according to the present embodiment. As shown in FIG. 5, the search client 103 of the present embodiment has a Query object generation program 146.

Query object generation program 146
From the search request input from the keyboard 10 or the search expression previously stored in the search expression storage area 145, the definition information stored in the operator definition information storage area 147 for full text search and the class definition information storage area 148 for full text search is This processing unit generates a Query object for performing a full-text search according to the contents of operator definition information for full-text search and class definition information for full-text search.

A query object generation program 146 for causing the search client 103 to function as a query object generation processing unit is recorded on a recording medium such as a CD-ROM, stored on a magnetic disk or the like, and then loaded into a memory and executed. Shall be. The medium on which the program is recorded may be a medium other than the CD-ROM.

The search client 103 has an operator definition information storage area 147 for full-text search and a class definition information storage area 148 for full-text search.

The full-text search operator definition information storage area 147 is an area for previously storing full-text search operator definition information defining an operator who instructs full-text search. The full-text search class definition information storage area 148 is an area in which class definitions for full-text search that define classes used in full-text search are stored in advance.

The search client 103 uses the keyboard 1
0, a display 11, a central processing unit 12, a floppy disk drive 13, a floppy disk 14, a communication control device 15, a main memory 17, a magnetic disk device 18, and a system bus 19.

The keyboard 10 is used to input a command for instructing execution of a document search process and the like. The display 11 is used to display the execution state of the processing in the client and the like. The central processing unit 12 executes various programs constituting the client. The floppy disk drive 13 has a floppy disk 14
Used to read and write data to The floppy disk 14 stores in advance a search formula for specifying a search condition when requesting a search process, and is used to input the search formula to the client. The communication control device 15 communicates with the document space management server and the global document space management server via the network 16 and is used to exchange commands and data. The main memory 17 is used to hold various programs and temporary data for executing processing in the client. The magnetic disk device 18 is used to store a search formula that specifies search conditions when requesting a search process in advance. The system bus 19 is used to connect these various devices.

In the main memory 17, a document retrieval control program 113, a global document space management server connection program 140, a global metadata reference program 14
1. A document space management server connection program 142, a document search request program 143, a query object generation program 146, and a system program 20 are stored.
Further, a work area 21 is held.

In the magnetic disk device 18, a search result storage area 144, a search expression storage area 145, a full-text search operator definition information storage area 147, and a full-text search class definition information storage area 148 are secured.

The document search control program 113 includes a global document space management server connection program 140, a global metadata reference program 141, a document space management server connection program 142, and a document search request program 1.
43. By controlling the execution of the Query object generation program 146 and requesting the global document space management server 201 for a global search process, a full-text search process for documents of a plurality of document space management servers is performed collectively. It is also possible to directly perform a full-text search process on documents in individual document space management servers.

The global document space management server connection program 140 establishes a connection between the search client 103 and the global document space management server 201.

Global metadata reference program 14
1 is for the global document space management server 201,
A request is made to transmit the global metadata to the search client 103, and this is obtained.

Document space management server connection program 142
Are the search client 103 and the document space management server 30
Establish a connection with the server 1 to 303.

The document search request program 143 requests the global document space management server 201 or the document space management server with which the search client 103 has established a connection to search for a document. Global document space management server 2
01, the global document space management server 201 issues a search request collectively to a plurality of document space management servers defined as targets of global search, and the full text search is performed in each document space management server. I do. After merging the search results returned from each document space management server, the search result set is transferred to the search client 103 via the network 16.

Query object generation program 146
The definition information stored in the operator definition information storage area 147 for full-text search and the class definition information storage area 148 for full-text search is obtained from a search request input from the keyboard 10 or a search expression previously stored in the search expression storage area 145. Refer to to generate Query object for full-text search.

The system program 20 provides basic functions necessary for executing various programs constituting the client on a computer, such as input / output of data with peripheral devices. The work area 21 is used to store data that is temporarily required when executing the program.

The search result storage area 144 is used to store search results transferred from the global document space management server 201. The search formula storage area 145 is used to store in advance a search formula to be passed along with a command to specify a search condition when requesting a search process.

The full-text search operator definition information storage area 147 is used to previously store the full-text search operator definition information. The full-text search class definition information storage area 148 is used to previously store the full-text search class definition information.

In the document management system of the present embodiment, the retrieval formula stored in the floppy disk 14 is input and read. However, the configuration may be such that the retrieval formula is read from another portable medium such as a magneto-optical disk or an optical disk. It is also possible to adopt a configuration in which a search formula transferred via the network 16 is input. Further, in the document management system according to the present embodiment, the search result is transmitted to the global document space management server 2 via the network 16.
However, a configuration using a portable medium, such as a floppy disk, a magneto-optical disk, or an optical disk, may be used instead.

Hereinafter, the global document space management server of the document management system according to the present embodiment, that is, 201 in FIG. 2 will be described. FIG. 6 is a diagram showing a schematic configuration of the global document space management server 201 of the present embodiment.

As shown in FIG. 6, the global document space management server 201 of this embodiment has a Query object interpretation program 155. The query object interpretation program 155 includes an operator definition information storage area 156 for full-text search and a class definition information storage area 1 for full-text search.
Checking the available search function by referring to the definition information stored in 57, and interpreting a Query object for performing a full-text search in accordance with the contents of the operator definition information for full-text search and the class definition information for full-text search Department.

When the global document space management server 201 is
Query to function as ery object interpretation processing part
After the object interpretation program 155 is recorded on a recording medium such as a CD-ROM and stored on a magnetic disk or the like,
It shall be loaded into memory and executed. The medium on which the program is recorded may be a medium other than the CD-ROM.

The global document space management server 201
Is the operator definition information storage area 156 for full-text search.
And a class definition information storage area 157 for full-text search.

The full-text search operator definition information storage area 156 is an area for previously storing full-text search operator definition information defining an operator who instructs full-text search. The full-text search class definition information storage area 157 is an area in which class definition information for full-text search that defines a class used in full-text search is stored in advance.

The global document space management server 201 includes a keyboard 10, a display 11, a central processing unit 12,
It comprises a floppy disk drive 13, a floppy disk 14, a communication control device 15, a main memory 17, a magnetic disk device 18, and a system bus 19.

The keyboard 10 is used to input a command for instructing a system management function or the like. The display 11 is used to display the execution status of the processing in the server and the like. The central processing unit 12 executes various programs constituting the server. The communication control device 15 communicates with the registered client 101 via the network 16,
It communicates with the reference client 102, the search client 103, and the document space management servers 301 to 303, and is used to exchange commands and data. Main memory 17
Is used to hold various programs and temporary data for executing processing in the server. The magnetic disk device 18 is used to store a managed document space management server and global metadata.
The system bus 19 is used to connect these various devices.

In the main memory 17, a global document space management program 211, a storage control program 150, a global metadata management program 151, a metadata acquisition program 152, a query object interpretation program 155, and a system program 20 are stored. Area 21 is held.

In the magnetic disk device 18, the document space management data storage area 153, the global metadata storage area 154, the operator definition information storage area 1 for full text search
56 and a class definition information storage area 157 for full-text search are secured.

Global document space management program 211
Are the storage control program 150, the global metadata management program 151, the metadata acquisition program 15
2. It controls the execution of the Query object interpretation program 155 and manages global metadata and document space management data in the global document space management server 201.

The storage control program 150 controls the input and output of data between the magnetic disk device 18 and the work area 21.

Global Metadata Management Program 15
1 collectively manages document space metadata and document metadata collected from each document space management server as global metadata, and controls registration, reference, update, deletion, and search thereof. In addition, in response to a request for full-text search processing from the search client 103, a global search is performed for performing a collective search on a plurality of document space management servers, and the search result is transferred to the search client 103.

The metadata acquisition program 152 requests and acquires the changed metadata from the document space metadata and the document space management server that has notified the update of the document metadata.

Query object interpretation program 155
Refers to the definition information stored in the full-text search operator definition information storage area 156 and the full-text search class definition information storage area 157 and interprets the full-text search Query object.

The system program 20 provides basic functions necessary for executing various programs constituting the server on a computer, such as input / output of data with peripheral devices. The work area 21 is used to store data that is temporarily required when executing the program.

The document space management data storage area 153 is under the management of the global document space management server 201, and stores data relating to the document space management server that can be used by clients connected to the global document space management server 201. Used for

The global metadata storage area 154 is
It is used to collectively manage and store document space metadata and document metadata collected from each document space management server as global metadata.

The full-text search operator definition information storage area 156 is used to store in advance the definition information of the full-text search operator. The full-text search class definition information storage area 157 is used for storing definition information of a full-text search class in advance.

The document space management server of the document management system according to the present embodiment, that is, 301 to 303 in FIG. 2 will be described below. Since all of these document space management servers have the same configuration, description will be made with reference to 301.

FIG. 7 shows the document space management server 3 of this embodiment.
FIG. 1 is a diagram showing a schematic configuration of No. 01. As shown in FIG. 7, the document space management server 301 of the present embodiment has a Query object interpretation program 167. The query object interpretation program 167 refers to the definition information stored in the operator definition information storage area 168 for full-text search and the class definition information storage area 169 for full-text search, checks available search functions, and performs a full-text search. A processing unit that interprets a Query object according to the contents of operator definition information for full-text search and class definition information for full-text search.

A Query object interpretation program 167 for causing the document space management server 301 to function as a Query object interpretation processing unit is recorded on a recording medium such as a CD-ROM, stored on a magnetic disk or the like, and then loaded into a memory. Shall be executed. The medium on which the program is recorded may be a medium other than the CD-ROM.

The document space management server 301 has an operator definition information storage area 168 for full text search and a class definition information storage area 169 for full text search.

The full-text search operator definition information storage area 168 is an area for previously storing full-text search operator definition information defining an operator who instructs full-text search. The full-text search class definition information storage area 169 is an area in which class definitions for full-text search that define classes used in full-text search are stored in advance.

The document space management server 301 uses the keyboard 1
0, a display 11, a central processing unit 12, a floppy disk drive 13, a floppy disk 14, a communication control device 15, a main memory 17, a magnetic disk device 18, and a system bus 19.

The keyboard 10 is used for inputting a command for instructing a system management function or the like. The display 11 is used to display the execution status of the processing in the server and the like. The central processing unit 12 executes various programs constituting the server. The communication control device 15
The registration client 10 via the network 16
1. Reference client 102, search client 103
It communicates with the global document space management server 201 and is used to exchange commands and data. The main memory 17 is used to hold various programs for executing processing in the server and temporary data. The magnetic disk device 18 includes a document entity, a document object,
It is used to store document space metadata, document metadata, and definition information for full-text search. System bus 1
Reference numeral 9 is used to connect these various devices.

In the main memory 17, the document space management program 311, the storage control program 160, the metadata management program 161, the metadata change notification program 1
62, the Query object interpretation program 167 and the system program 20 are stored.
1 is held.

In the magnetic disk device 18, the document entity storage area 163, the document object storage area 164, the document space metadata storage area 165, the document metadata storage area 166, the operator definition information storage area 1 for full text search
68 and a class definition information storage area 169 for full-text search are secured.

The document space management program 311 includes the storage control program 160, the metadata management program 16
1. It controls the execution of the metadata change notification program 162 and the query object interpretation program 167, and manages the documents in the document space management server 301.

The storage control program 160 controls data input / output between the magnetic disk device 18 and the work area 21. The metadata management program 161 manages document space metadata and document metadata, and controls registration, reference, update, deletion, and search thereof. The metadata change notification program 162 notifies the global document space management server when document space metadata and document metadata are updated.

Query object interpretation program 167
Refers to the definition information stored in the full-text search operator definition information storage area 168 and the full-text search class definition information storage area 169, and interprets the full-text search Query object.

The system program 20 provides basic functions necessary for executing various programs constituting the server on a computer, such as input / output of data with peripheral devices. The work area 21 is used to store data that is temporarily required when executing the program.

The document entity storage area 163 is used to store the entity of the registered document. The document object storage area 164 is used to store document objects for managing registered documents.

The document space metadata storage area 165 is used to store document space metadata which is attribute information on the document space. Document metadata storage area 166
Is used to store document metadata which is attribute information of the document itself.

The full-text search operator definition information storage area 168 is used to store the definition information of the full-text search operator in advance. The full-text search class definition information storage area 169 is used to store the definition information of the full-text search class in advance.

Note that the document management system of this embodiment is 3
Although the configuration is made up of three clients and three document space management servers, the number of both can be essentially any number, and they can be configured independently of each other.

Next, a method of implementing a full-text search in the document management system of the present embodiment will be described. After describing the full-text search operator and class introduced in the document management system of the present embodiment, the full-text search execution procedure will be described.

First, an operator for full-text search in the document management system of the present embodiment will be described. In the document management system of the present embodiment, the following operators are newly provided.

Contains: Instructs full-text search.・ Contains_with_score: Full-text search is instructed together with score calculation. Used for ranking search. Equal_Content: Relates search terms in full-text search with full-text data.・ Prox: Specify the neighborhood condition at the time of neighborhood condition search.・ Soundex: Expands the notation of the search term.・ Synonym: Expand synonyms of search terms. QueryOpper of QueryOperator class
Function as an operator by setting to the atorID property. Hereinafter, each operator will be described.

FIG. 8 is a diagram showing operator specifications and operand specifications of the Contains operator of this embodiment. FIG. 9 is a diagram illustrating an example of the Contains operator according to the present embodiment. The Contains operator specifies execution of a full-text search. The Contains operator has the following functions.・ Specify that a full-text search is performed on the search target.・ Specify full-text search conditions in the operand.・ Specify the search target in FromExpr.・ Return TRUE if the full-text search condition is met.

FIG. 10 shows the Contents_with_scor of this embodiment.
It is a figure which shows the operator specification and operand specification of e operator. FIG. 11 shows Contents_with_scor of the present embodiment.
FIG. 9 is a diagram illustrating an example of an e operator. Contains_with_score
The operator specifies execution of a full-text search involving score calculation and a ranking search that performs sorting by score. When performing a search request to the global document space management server, that is, a global search, the second operand Normaliz
By setting the ation flag, it is possible to specify to normalize the ranking search results for a plurality of document space management servers (repositories) with different ranking algorithms. Contains_with_score operator has the following functions.

[0156] It is specified that a full-text search involving score calculation or a ranking search that performs sorting by score is performed on the search target.・ Specify full-text search conditions in the operand.・ Specify the search target in FromExpr. -Specify the score acquisition in the select list. The score normalization designation (Normalization flag) parameter specifies that the score of the ranking search result is normalized and returned.・ Return TRUE if the full-text search condition is met.

Local / Global Search and Normalization
The following table shows how to specify the normalization of search results by combining flag settings.

[0158]

[Table 1]

FIG. 12 is a diagram showing operator specifications and operand specifications of the Equal_Content operator of the present embodiment. The Equal_Content operator associates a search term in full-text search with full-text data of the text. Equa
The l_Content operator has the following functions.

A search term and full-text data in full-text search are specified as operands.・ Return TRUE for full-text data including search terms.・ Virtual full-text data to be searched in the QueryProperty object of the first operand, that is, ContentSearch
Set the Content property of the class. By setting an instance of the QueryPropertyWithWeightedStruct class, which is a subclass of the QueryProperty class, as the first operand, it is possible to specify a weighted structure specification search. -QueryCon, a subclass of the QueryConstString class
Weighted search terms can be set using the stWeightedString class.・ QueryConstStrings or QueryConstWeightedStrings
If multiple search terms are set by Equal_Cont
The ent operator evaluates according to the OR condition of each search term. That is, if any of the search terms are included
Returns TRUE.

FIG. 13 is a diagram showing a first example of the Equal_Content operator of the present embodiment. FIG. 14 is a diagram illustrating a second example of the Equal_Content operator according to the present embodiment. FIG. 14 shows an example of the Equal_Content operator when the weighted structure designation search is set together with the weighted search term.

A third example of the Equal_Content operator is shown in FIGS. FIG. 15 shows Eq of the present embodiment.
It is a figure which shows the operator specification and operand specification of a ual_Content operator. FIG. 16 shows Equal_Co of the present embodiment.
It is a figure showing the 3rd example of an ntent operator. This third
In the example, it is possible to define the same function without introducing a new class.

FIG. 17 is a diagram showing operator specifications of the Prox operator of the present embodiment. FIG. 18 shows Pr of the present embodiment.
FIG. 4 is a diagram illustrating an operand specification of an ox operator. FIG.
FIG. 3 is a diagram illustrating an example of a Prox operator of the present embodiment. Pr
The ox operator specifies a full-text search based on a condition relating to a distance between search terms appearing in a document, that is, a neighborhood condition in a neighborhood condition search. The Prox operator has the following functions.

Specify the neighborhood condition in full-text search as an operand.・ Specify the search target in FromExpr. A Prox operator is defined as an operator that returns a group of search terms for performing a full-text search for a document that meets the set neighborhood conditions. -When describing neighborhood conditions in QueryExpr, use Equal
QueryProperty as the first operand of _Content operator
Is used to specify full-text data, that is, a search index, and a neighborhood condition using the Prox operator is set as the second operand. -QueryConstString for the first and second operands
A phrase list can be set by using the s object or the QueryConstWeightedStrings object. The search term can be weighted for both the first operand and the second operand, but only the weight of the first operand is evaluated at the time of score evaluation, and the weight of the second operand is ignored.・ This operator is the Contains operator or Contains
Set _with_score operator in a node lower than QueryOperator specified as QueryOperatorID.

FIG. 20 is a diagram showing the operator specifications of the Soundex operator of the present embodiment. FIG. 21 is a diagram illustrating an example of the Soundex operator of the present embodiment. The Soundex operator specifies the expansion of search terms in full-text search. The Soundex operator has the following functions.

The search terms in the full-text search are expanded in different notations, and the expanded search terms are returned as a list. • The operand that specifies the target search term and expansion mode is used as an operand. -Using the QueryConstWeightedString class, it is possible to set weights for search terms. In this case, the set weights are inherited by the expanded search terms.・ This operator is the Contains operator or Contains
Set _with_score operator in a node lower than QueryOperator specified as QueryOperatorID.

FIG. 22 is a diagram showing the operator specifications of the Synonym operator of the present embodiment. FIG. 23 is a diagram illustrating an example of a Synonym operator according to the present embodiment. FIG. 24 shows the Synonym operator and Sou of the document management system according to the present embodiment.
It is a figure showing the example which combined the ndex operator. Synonym
The operator specifies the synonym expansion of the search term in the full-text search. The Synonym operator has the following functions.

The search terms in the full-text search are expanded with synonyms, and the expanded search terms are returned as a list. -The target search term and the synonym dictionary are used as operands. -Using the QueryConstWeightedString class, it is possible to set weights for search terms. In this case, the set weights are inherited by the expanded search terms. -It can be combined with the Soundex operator and set as the operand of the Soundex operator. In this case, after the search terms set for this operator are expanded into synonyms, the Soundex operator expands different notations for each term.・ This operator is the Contains operator or Contains
Set _with_score operator in a node lower than QueryOperator specified as QueryOperatorID.

Next, a class for full-text search in the document management system of this embodiment will be described. In the document management system of the present embodiment, the following classes are newly provided.

ContentSearch: Refers to registered information for full-text search, such as a full-text search index table.
Searchable virtual class to define to specify in Query. QueryConstWeightedString: Used to specify a weighted search term. -QueryConstWeightedStrings: Used to specify multiple weighted search terms.・ QueryPropertyWithWeightedStruct: Used to weight the structure in the structure specification search.

Hereinafter, each class will be described. FIG.
FIG. 4 is a diagram showing property specifications of a ContentSearch class according to the embodiment. FIG. 26 shows the ContentSearch of this embodiment.
It is a figure showing an example of a class. In implementing the full-text search function, the ContentSearch class must refer to registered information for full-text search, such as an index table for full-text search.
It is a class to define to specify in Query. ContentSearch class has the following properties.

[0172] 1. Object_ID The instance ID (OIID) of the object. 2. Class Description Object A class description object that describes the details of this class. 3. The purpose is to be able to describe JOIN between objects in This DMA Query. 4. Content Indicates the full-text data of the registered document when performing a full-text search. 5. DocVersion_ID Document ID of the registered document. 6. Score The score of the registered document when the score calculation is specified. Instead of the information held as static information, it is generated dynamically according to search conditions at the time of search. It is a formal property to get the score. The scores are normalized and returned in the search result set. 7. Raw_Score The score of the registered document when the score calculation is specified. Instead of the information held as static information, it is generated dynamically according to search conditions at the time of search. It is a formal property to get the score. Unlike Score, Raw
_Score is returned without being normalized in the search result set. 8. Doc_Length Returns the length of the searched document. (Unit: bytes) 9. Term_Freq Returns the number of occurrences of the search term evaluated in the score calculation in the document. 10. Term_Weight Returns the weight of the search term evaluated for score calculation. (Range: 0 to 100) The ContentSearch class is a virtual class for specifying registration information for full-text search in Query. Properties defined in the ContentSearch class correspond to various registration information registered for full-text search. However, ContentSearc
There is no persistent object directly corresponding to the h class.

When constructing a Query object, the FromE
Searchable class must be specified as the search target specified in xpr. Search for full-text search
FromExp using ContentSearch class which is able class
Describe the contents of r.

The ContentSearch class has a “Content property”. This Content property corresponds to the full text data of the registered document to be subjected to the full text search, that is, the text.
Using the Content property and the Equal_Content operator, the full-text search condition that the text includes the search term T1 is described as Equal_Content (ContentSearch.content, “T1”).

The Content property does not necessarily need to be actually associated with the text data of the text. Qu
This explicitly indicates to the search engine that ery refers to registration information for full-text search, for example, index data.

In the case of full-text search, Contains or Contains
Set _with_score as QueryOperator. Appropriate search processing is automatically selected according to this setting, and Cont
The property specified in entSearch class is referenced.

FIG. 27 shows QueryConstWeighted of this embodiment.
It is a figure which shows the property specification of String class. FIG.
FIG. 4 is a diagram illustrating an example of a QueryConstWeightedString class according to the embodiment. The QueryConstWeightedString class is used to specify a weighted search term in full-text search. A search term and the weight of the search term are specified in combination.
QueryConstWeightedString class has the following properties.

1. Object_ID The instance ID (OIID) of the object. 2. Class Description Object A class description object that describes the details of this class. 3. The purpose is to be able to describe JOIN between objects in This DMA Query. 4. Create Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that a scratchpad object is to be newly registered as a persistent object. 5. Update Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and will update its properties. 6. Delete Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and that the object should be deleted. 7. Qval String Property that has a String type constant as its value. Write a search term. 8. Property with Qval Weight Integer32 type constant as value. The weight of the search term is described in the range of 0 to 100.

When a weighted search term is specified in a full-text search, the search term and its weight are specified as a pair using the QueryConstWeightedString class.

The QueryConstWeightedString class is a String
Set a search term in the type property Qval String and use Int
Weight 0 to 10 for the Qval Weight property of eger32 type
Set within the range of 0. If no weight is set, Qval W
Set NULL to eight.

FIG. 29 shows QueryConstWeighted of this embodiment.
It is a figure showing the property specification of Strings class. FIG.
0 is a diagram illustrating an example of a QueryConstWeightedStrings class according to the present embodiment. QueryConstWeightedStrings class is used to specify multiple weighted search terms in full text search. A search term and the weight of the search term are specified in combination. QueryConstWeightedStrings class has the following properties.

1. Object_ID The instance ID (OIID) of the object. 2. Class Description Object A class description object that describes the details of this class. 3. The purpose is to be able to describe JOIN between objects in This DMA Query. 4. Create Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that a scratchpad object is to be newly registered as a persistent object. 5. Update Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and will update its properties. 6. Delete Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and that the object should be deleted. 7. Qval Strings List property that has a String type constant as its value. Describe multiple search terms. 8. List property with Qval Weights Integer32 type constant as value.
The weight of a plurality of search terms is described in the range of 0 to 100. When specifying multiple weighted search terms in full-text search, use the QueryConstWeightedStrings class to specify the search terms and their weights in pairs.

The QueryConstWeightedStrings class is Strin
Set multiple search terms in the g-type list property Qval Strings, and set the Integer32-type list property Qval Weig
The weight is set in the range of 0 to 100 at the corresponding position of hts. If you do not want to set a weight, the corresponding Qval Weight
Is set to NULL. The correspondence between the search term and the weight is determined in the order from the head of the list.

FIG. 31 shows the QueryPropertyWithW of this embodiment.
It is a figure showing the property specification of eightedStruct class. FIG. 32 shows QueryPropertyWithWeightedS of the present embodiment.
It is a figure showing an example of a truct class. QueryPropertyWithWe
The ightedStruct class is used to specify the structure and to specify the weight for the structure in the structure specification search of the full text search. The content to be searched is specified in combination with the description that limits the structure and the weight. QueryPropertyWithWeighte
The dStruct class has the following properties.

1. Object_ID The instance ID (OIID) of the object. 2. Class Description Object A class description object that describes the details of this class. 3. The purpose is to be able to describe JOIN between objects in This DMA Query. 4. Create Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that a scratchpad object is to be newly registered as a persistent object. 5. Update Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and will update its properties. 6. Delete Flag If this flag is set, it indicates to ExecuteChange or ExecuteChanges that this object already exists as a persistent object and that the object should be deleted. 7. QS Class Occur A property that has a searchable class occurrence number as its value. 8. Qprop ID Property that has the ID of the specified Query Property as its value. 9. Struct Description Property whose value is a string that describes the structure specification. 10.Weight Description Property that has a constant of Integer32 type as its value. The weight for the structure is described in the range of 0 to 100.

In the structure specification search of full-text search, when specifying a structure and specifying a weight for a structure, Quer
Use yPropertyWithWeightedStruct class.

Using the QueryPropertyWithWeightedStruct class, the content to be searched specified by the Qprop ID property is specified in combination with the structural description specified by the Struct Description property and the weight related to the structure specified by the Weight Description property. Weight D
If NULL value is set in escription property,
It is interpreted that no weight is given to the structure.

Finally, the procedure of a full-text search in the document management system of this embodiment using the above-described full-text search operator and class will be described. For simplicity, a case will be described as an example where a full-text search is performed from the document search control program 113 of the search client 103.

When performing a local search, the document search control program 113 existing in the search client 103
First, the global document space management server connection program 14
0 is used to connect to the global document space management server 201.

In the global document space management server 201, a list of connectable document space management servers is acquired using the global metadata reference program 141, and the global metadata management program 151
To refer to the global metadata.

The global metadata includes a field related to the document managed by each of the document space management servers 301 to 303, a source of the document, and the like. Document search control program 11
3 can first narrow down the document space management server to be subjected to full-text search by referring to the global metadata.

Further, a Query object is generated by the Query object generation program 146 in order to connect to the corresponding document space management server and perform a target search, and the Query object is passed to the document space management server to request full text search processing. In the document space management server, the Query object is interpreted by the Query object interpretation program 167 and passed to a search engine as a search expression to execute a full-text search.

FIG. 33 is a flowchart showing a processing procedure for generating a Query object according to the present embodiment. The procedure for generating a Query object will be described with reference to FIG. A search formula or search condition is input to a work area from a disc or a keyboard. It is determined whether the search request is a full-text search or an attribute search. In the case of full-text search, the available functions are checked by referring to the definition information of the operator and class for full-text search. Generate FromExpr. Generate SelectList. Generate QueryExpr. Generate OrderByList. Generate QueryRoot object. Generate Query object. If there are a plurality of QueryRoot objects in the Query object, the steps are repeated.

FIG. 34 is a flowchart showing a processing procedure for generating a QueryExpr according to the present embodiment. The procedure for generating QueryExpr will be described with reference to FIG. Initialize the current node to be processed as the start node. Check whether the operator set to the node is full text search. If the operator to be set for the node is full-text search, either Containers or Contai
The ns_with_score operator is set in the node, and if the operator to be set in the node is not full-text search, the operator is set in the node. Describe search conditions in the operand of the operator. Checks if there is an operator nested in the operand. If there is a nested operator in the operand,
Moves the current node down (to the level of the nested operator) and returns to. If there is no nest in the operand, it is checked whether the current node is the start node. If the current node is the start node, the QueryExpr generation processing ends. If the current node is not the start node, move the current node to a higher level (upper operator level). It is checked whether there is another operator. If there is another operator, the process returns to the step. If there is no other operator, the process returns to the step.

FIG. 35 is a flowchart showing a processing procedure for interpreting a Query object according to the present embodiment. Figure 3 shows the procedure for interpreting a Query object in the document space management server
5 will be described.

The operator for full-text search refers to the definition information of the class, and inputs information of available functions to the work area. Enter a Query object into the work area. Follow the link of the QueryRoot object FromExpr, S
Extract electList, QueryExpr, OrderByList. Follow the link of the Query object and extract the QueryRoot object. Interpret FromExpr. Interpret SelectList. Interpret QueryExpr. Interpret OrderByList. Generate a search expression for the search engine of the document space management server. If there are a plurality of QueryRoot objects in the Query object, the steps are repeated. This makes it possible to acquire the search result of the full-text search. The search client 103 checks the search result, specifies a document for which the document entity needs to be acquired, and connects to the document space management server to acquire the document entity.

FIG. 36 is a flowchart showing the processing procedure for interpreting QueryExpr according to this embodiment. The procedure for interpreting QueryExpr will be described with reference to FIG. Initialize the current node to be processed as the start node. Interpret the operator operands according to the operator definition information. Store the interpretation result in the buffer. Checks for nested operands. If there are nested operators in the operand,
Moves the current node down (to the level of the nested operator) and returns to. If there is no nesting in the operand, it is checked whether the current node is the start node. If the current node is not the start node, move the current node to a higher level (upper operator level). It is checked whether there is another operator. If there is another operator, the process returns to the step. If there is no other operator, the process returns to the step. If the current node is the start node, the interpretation result stored in the buffer is output.

Similarly, when performing a global search, a Query object for performing a target search is generated after connection to the global document space management server 201, and this query object is passed to the global document space management server 201 to perform a full-text search process. Ask for. At this time, a global search can be executed by issuing a search request collectively to a plurality of target document space management servers.

The generation and interpretation of the Query object at this time are performed in the same procedure as in the case of the local search. This makes it possible to acquire the search result of the full-text search. The application program confirms the search result and specifies a document that needs to acquire the document entity, and then connects to the document space management server to acquire the document entity.

FIG. 37 is a diagram showing an example of a Query object for instructing a ranking search according to the present embodiment. This is an example in which a search condition “a document includes both character strings“ document ”and“ search ”” is set. A ranking score, registered document ID, and document title are returned in the search result set. The second operand of Contains_with_score, N
The normalization of the score is specified by setting the ormalization flag.

FIG. 38 is a diagram showing an example of a Query object for instructing a neighborhood condition search according to the present embodiment. "The document contains both the character strings" document "and" search, "and the distance between them is 10 characters or less. The order of the first word list and the second word list is non-commutative. Is set in the search condition. A ranking score, registered document ID, and document title are returned in the search result set. Co
Normalization f, the second operand of ntains_with_score
Set lag to specify normalization of score.

[0202] By introducing new operators and classes suitable for full-text search without impairing the conventional attribute search framework, a framework including full-text search can be obtained. That is, the conventional search function can be easily extended, and a more sophisticated search function can be realized.

As described above, according to the document management system of the present embodiment, since the operator and the class for designating the full-text search function are introduced, the search specification of the document management system is extended and Can be improved.

(Embodiment 2) An operator capable of designating And-Not conditions is defined below as operator definition information.
A description will be given of a document management system according to a second embodiment which realizes a search function capable of supporting the Z39.50 search specification by extending these attributes without impairing the framework of an attribute search function of the MA specification and a full-text search function that extends the MA specification. .

FIG. 43 is a diagram showing the principle of the document management system of this embodiment. As shown in FIG. 43, in the present embodiment, by introducing the And-Not operator, Z39.50 can be extended and extended without impairing the framework of the attribute retrieval function of the DMA specification and the full-text retrieval function that extends the attribute retrieval function. Implement a search function that can support search specifications.

The introduced And-Not operator is registered in the client and the server as definition information. Furthermore, using this operator, the client side refers to the definition information based on the search request from the user, and
A processing step for generating a y object and a processing step for interpreting the received Query object with reference to the definition information are provided on the server side.

The use of the And-Not operator can suppress the generation of a temporary search result set in the process of operation. That is, an increase in resources required by the document management system can be suppressed. This makes it possible to realize a search function that can support the Z39.50 search specification without significantly increasing the processing cost required by the search function of the document management system.

The configuration of the client and server of this embodiment is the same as that shown in the first embodiment. The operator definition information storage area 147 for full-text search, the operator definition information storage area 156 for full-text search, and the operator definition The information storage area 168 also stores the definition information of the And-Not operator.

[0209] Query object generation program 1
Reference numeral 46 denotes a definition stored in the operator definition information storage area 147 for full-text search and the class definition information storage area 148 for full-text search based on a search request input from the keyboard 10 or a search expression previously stored in the search expression storage area 145. Refer to the information, Query object of full text search and And-Not
Generate a Query object including the operator.

Query object interpretation program 155
Refers to the definition information stored in the full-text search operator definition information storage area 156 and the full-text search class definition information storage area 157, and interprets the full-text search Query object and the Query object including the And-Not operator, The query object interpreting program 167 refers to the definition information stored in the operator definition information storage area 168 for full text search and the class definition information storage area 169 for full text search, and queries the query object and And-N for full text search.
Interpret the Query object containing the ot operator.

A query object generation program 146 for causing the search client 103 to function as a query object generation processing unit including an And-Not operator,
A Query object interpretation program 155 and a Query object interpretation program 167 for causing the global document space management server 201 and the document space management server 301 to function as interpretation processing units of a Query object including a -Not operator are stored in a recording medium such as a CD-ROM. After being recorded and stored on a magnetic disk or the like, it is assumed that the program is loaded into a memory and executed. The medium on which the program is recorded may be a medium other than the CD-ROM.

FIG. 44 is a diagram showing operator specifications and operand specifications of the And-Not operator of this embodiment.
The And-Not operator has the following functions. -Set a Boolean operand for the first and second operands. • Calculates and returns the truth value of the Boolean operand according to the following truth table.

[0213]

[Table 2]

Table 2 shows that the first operand is A [Boolean],
"Not B" when the second operand is B [Boolean]
And "A And-Not B", and T is true (TR
UE) and F indicate false (FALSE) logical values, and U indicates that the logical value is indefinite (UNKNOWN).

FIG. 45 is a diagram showing an example of the And-Not operator of the present embodiment. As shown in FIG. 45, two operators such as operands A and B are set for the And-Not operator.

FIG. 46 is a diagram showing the attribute search according to the present embodiment.
FIG. 9 is a diagram illustrating an example of a Query description. In FIG. 46, “Author (Auth
or property) is “Suzuki” and the organization (Org property)
Is a document saying "it is not Hitachi". . "Represents an example of a Query description when performing an attribute search based on a search condition, and returns the object ID of the registered document and the title of the document in the search result set.

Next, a search procedure in this embodiment will be described. For the sake of simplicity, a case where a search is performed from the document search control program 113 of the search client 103 will be described as an example.

When performing a local search, the document search control program 113 existing in the search client 103
First, the global document space management server connection program 14
0 is used to connect to the global document space management server 201.

In the global document space management server 201, a list of connectable document space management servers is acquired using the global metadata reference program 141, and the global metadata management program 15
1 to refer to the global metadata.

[0220] The global metadata includes the related field of the document managed by each of the document space management servers 301 to 303, the source of the document, and the like. Document search control program 11
3 can first narrow down the document space management server to be searched by referring to the global metadata.

Further, a Query object is generated by the Query object generation program 146 to connect to the corresponding document space management server and perform a target search, and the query object is passed to the document space management server to request a search process. In the document space management server, the Query object is interpreted by the Query object interpretation program 167 and passed to a search engine as a search expression to execute a search.

Next, a procedure for generating a Query object will be described with reference to FIG. A search formula or search condition is input to a work area from a disc or a keyboard. It is determined whether the search request is a full-text search or an attribute search. In the case of full-text search, the available functions are checked by referring to the definition information of the operator and class for full-text search. Generate FromExpr. Generate SelectList. Generate QueryExpr. Generate OrderByList. Generate QueryRoot object. Generate Query object. If there are a plurality of QueryRoot objects in the Query object, the steps are repeated.

For example, in the case of performing an attribute search based on a search condition of “a document that the author is“ Suzuki ”and the affiliated organization is not“ Hitachi ”” in FIG. 46, the search condition is first input to the work area in FIG. I do. In FIG. 33, it is determined whether the search request is a full-text search or an attribute search. Since the search request is for performing an attribute search, FIG.
By performing the following processing of No. 3, a QueryExpr having an And-Not operator in FIG. 33 is generated.

Next, a procedure for generating a QueryExpr will be described with reference to FIG. Initialize the current node to be processed as the start node. It is checked whether the operator to be set for the node performs a full-text search. If the operator to be set in the node performs full-text search, the Contains or Contains_with_score operator is set in the node according to the operator definition information.If the operator to be set in the node is not to perform full-text search, Set the operator to the node. Describe search conditions in the operand of the operator. Checks if there is an operator nested in the operand. If there is a nested operator in the operand,
Moves the current node down (to the level of the nested operator) and returns to. If there is no nest in the operand, it is checked whether the current node is the start node. If the current node is the start node, the QueryExpr generation processing ends. If the current node is not the start node, move the current node to a higher level (upper operator level). It is checked whether there is another operator. If there is another operator, the process returns to the step. If there is no other operator, the process returns to the step.

For example, when generating a QueryExpr for performing an attribute search based on the search condition of FIG.
Therefore, the current node indicating the processing target is initialized to the first point of QueryExpr in the tree in FIG.

In FIG. 34, it is checked whether or not the operator to be set in the node is to perform a full-text search. In this case, an attribute search is performed. Therefore, this And-Not operator is used in FIG. Set to node.

In FIG. 34, a search condition is described in the operand of the And-Not operator. In FIG. 34, it is checked whether or not there is an operator nested in the operand. So lower the current node (nest operator level)
Move to and set the lower level operator.

Next, the procedure for interpreting a Query object in the document space management server will be described with reference to FIG. Refer to the definition information of the search operator and the class and input the information of the available functions to the work area. Enter a Query object into the work area. Following the link of the QueryRoot object, FromExp
Extract r, SelectList, QueryExpr, OrderByList. Follow the link of the Query object and extract the QueryRoot object. Interpret FromExpr. Interpret SelectList. Interpret QueryExpr. Interpret OrderByList. Generate a search expression for the search engine of the document space management server. If there are a plurality of QueryRoot objects in the Query object, the steps are repeated. Thereby, it is possible to obtain the search result. The search client 103 checks the search result, identifies a document that needs to obtain a document entity,
Connect to the document space management server and acquire the document entity.

For example, when interpreting a Query object for performing an attribute search based on the search condition in FIG.
In step 5, referring to the definition information of the search operator and class, the information of the available functions such as the And-Not operator is input to the work area, and the query object in FIG. 46 is input to the work area in FIG.

In FIG. 35, by following the link of the QueryRoot object in FIG. 46, FromExpr, SelectList, Qu
eryExpr and OrderByList are extracted.
Follow the link of the object and extract the QueryRoot object. Next, the following processing of FIG.
Interpret QueryExpr including Not operator,
Generate search expressions for search engines that perform And-Not processing.

Next, the procedure for interpreting QueryExpr will be described with reference to FIG. Initialize the current node to be processed as the start node. Interpret the operator operands according to the operator definition information. Store the interpretation result in the buffer. Checks for nested operands. If there are nested operators in the operand,
Moves the current node down (to the level of the nested operator) and returns to. If there is no nesting in the operand, it is checked whether the current node is the start node. If the current node is not the start node, move the current node to a higher level (upper operator level). It is checked whether there is another operator. If there is another operator, the process returns to the step. If there is no other operator, the process returns to the step. If the current node is the start node, the interpretation result stored in the buffer is output.

For example, when interpreting a QueryExpr for performing an attribute search based on the search condition in FIG.
Therefore, the current node indicating the processing target is initialized to the first point of QueryExpr in the tree in FIG.

In FIG. 36, A described in each node of QueryExpr in FIG. 46 according to the definition information of the operator
The operand of the nd-Not operator is interpreted, and the interpretation result is stored in a buffer in FIG.

In FIG. 36, it is checked whether or not there is a nest in the operand. If there is a nest operator in the operand, the current node is moved to the lower level (the level of the nest operator) in FIG. Is interpreted.

Similarly, when performing a global search, a Query object for performing a target search is generated after connection to the global document space management server 201, and the query object is passed to the global document space management server 201 to perform search processing. Ask. At this time, a global search can be executed by issuing a search request collectively to a plurality of target document space management servers.

The generation and interpretation of the Query object at this time are performed in the same procedure as in the case of the local search. This makes it possible to obtain a search result. The application program confirms the search result and specifies a document that needs to acquire the document entity, and then connects to the document space management server to acquire the document entity.

In the above example, the And-Not operator is used for the attribute search, but the And-Not operator may be used for the full-text search.

FIG. 47 is a diagram showing an example of full-text search of the And-Not operator according to the present embodiment. In FIG. 47, And-Not
Query to instruct ranking search using operator
The object is set with a search condition of “a document includes a character string“ Suzuki ”and does not include a character string“ Hitachi ””.

When performing a full-text search based on the search condition, first, the search condition is input to the work area in FIG. In FIG. 33, it is determined whether the search request is a full-text search or an attribute search. Since the search request is to perform a full-text search, the following processing in FIG.
In FIG. 33, a QueryExpr having an And-Not operator is generated.

That is, the current node indicating the processing target in FIG. 34 is initially set to the first point of the QueryExpr in the tree in FIG. 47, and in FIG. 34, it is checked whether or not the operator to be set to the node performs full-text search. ,
Here, since a full-text search is performed, the following processing in FIG.
The ntains_with_score operator and the And-Not operator are set in the QueryExpr node as shown in FIG.

When interpreting a Query object for performing a full-text search based on the search condition of FIG. 47, first, FIG.
Therefore, referring to the definition information of the search operator and the class, the information of the available functions such as the And-Not operator is input to the work area, and the query object of FIG. 47 is input to the work area in FIG.

In FIG. 35, by following the links of the QueryRoot object in FIG. 47, FromExpr, SelectList, Qu
eryExpr and OrderByList are extracted.
Follow the link of the object and extract the QueryRoot object. Next, the following processing of FIG.
Interpret QueryExpr including Not operator.

That is, the current node indicating the processing target in FIG. 36 is initially set to the first point of QueryExpr in the tree in FIG. 47, and And described in each node of QueryExpr in FIG. 47 according to the definition information of the operator in FIG. The operand of the -Not operator is interpreted and the result of the interpretation is stored in the buffer in FIG.

In FIG. 36, it is checked whether or not there is a nest in the operand. If there is a nest operator in the operand, the current node is moved to the lower level (the level of the nest operator) in FIG. Is interpreted.

When the interpretation of QueryExpr or the like is completed, a search expression for a search engine for performing And-Not processing is generated as shown in FIG. 35, and a full-text search is performed using the And-Not operator.

As described above, in this embodiment, a search function that can support the Z39.50 search specification can be realized by introducing a new And-Not operator without impairing the conventional search framework. It becomes possible. That is, the conventional search function can be easily extended, and a more sophisticated search function can be realized.

As described above, according to the present embodiment, a document management system using DMA middleware, or an SQL-based Query object when searching for document information in a document space, and an attribute search To search for documents in the document space by introducing a new And-Not operator in a document management system using middleware that has the same search function specifications as DMA middleware, which has a function, a full-text search function, or both. Z39.50
It is possible to easily realize a search function that can support a search specification by extending an existing search function. As a result, it is possible to easily expand to a framework that also includes the Z39.50 search without damaging the framework of the conventional search specification, and obtain an effect that the search function can be improved.

As described above, according to the document management system of the present embodiment, since the search specification is extended and the And-Not operator can be used, a search function compatible with the Z39.50 search specification is realized. It is possible.

[0249]

According to the present invention, since an operator and a class for designating the full-text search function are introduced, the search specification of the document management system can be extended to improve the search function of the document management system. It is. Further, according to the present invention, since the And-Not operator can be used by extending the search specification, it is possible to realize a search function that can support the Z39.50 search specification.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a principle diagram of a document management system according to a first embodiment.

FIG. 2 is a diagram illustrating an overall configuration of a document management system according to a first embodiment.

FIG. 3 is a diagram illustrating a schematic configuration of a registration client 101 according to the first embodiment.

FIG. 4 is a diagram illustrating a schematic configuration of a reference client 102 according to the first embodiment.

FIG. 5 is a diagram illustrating a schematic configuration of a search client 103 according to the first embodiment.

FIG. 6 is a global document space management server 2 according to the first embodiment.
FIG. 1 is a diagram showing a schematic configuration of No. 01.

FIG. 7 is a diagram illustrating a schematic configuration of a document space management server 301 according to the first embodiment.

FIG. 8 is a diagram illustrating an operator specification and an operand specification of a Contains operator according to the first embodiment.

FIG. 9 illustrates an example of a Contains operator according to the first exemplary embodiment.

FIG. 10 is a diagram illustrating an operator specification and an operand specification of a Contains_with_score operator according to the first embodiment.

FIG. 11 is a diagram illustrating an example of a Contains_with_score operator according to the first embodiment.

FIG. 12 is a diagram illustrating an operator specification and an operand specification of an Equal_Content operator according to the first embodiment.

FIG. 13 is a diagram illustrating a first example of an Equal_Content operator according to the first embodiment.

FIG. 14 is a diagram illustrating a second example of the Equal_Content operator according to the first embodiment.

FIG. 15 is a diagram illustrating an operator specification and an operand specification of an Equal_Content operator according to the first embodiment.

FIG. 16 is a diagram illustrating a third example of the Equal_Content operator according to the first embodiment.

FIG. 17 is a diagram illustrating operator specifications of a Prox operator according to the first embodiment.

FIG. 18 is a diagram illustrating operand specifications of a Prox operator according to the first embodiment.

FIG. 19 is a diagram illustrating an example of a Prox operator according to the first embodiment.

FIG. 20 is a diagram illustrating operator specifications of a Soundex operator according to the first embodiment.

FIG. 21 is a diagram illustrating an example of a Soundex operator according to the first embodiment.

FIG. 22 is a diagram illustrating operator specifications of a Synonym operator according to the first embodiment.

FIG. 23 is a diagram illustrating an example of a Synonym operator according to the first embodiment.

FIG. 24 is a diagram illustrating an example in which a Synonym operator and a Soundex operator of the document management system according to the first embodiment are combined.

FIG. 25 is a diagram illustrating a property specification of a ContentSearch class according to the first embodiment.

FIG. 26 is a diagram illustrating an example of a ContentSearch class according to the first embodiment.

FIG. 27 is a diagram illustrating property specifications of a QueryConstWeightedString class according to the first embodiment.

FIG. 28 is a diagram illustrating an example of a QueryConstWeightedString class according to the first embodiment.

FIG. 29 is a diagram illustrating property specifications of a QueryConstWeightedStrings class according to the first embodiment.

FIG. 30 is a diagram illustrating an example of a QueryConstWeightedStrings class according to the first embodiment.

FIG. 31 is a QueryPropertyWithWeightedStr of the first embodiment.
It is a figure showing the property specification of a uct class.

FIG. 32 shows QueryPropertyWithWeightedStr in the first embodiment.
It is a figure showing an example of a uct class.

FIG. 33 is a flowchart illustrating a processing procedure for generating a Query object according to the first embodiment.

FIG. 34 is a flowchart illustrating a processing procedure for generating a QueryExpr according to the first embodiment.

FIG. 35 is a flowchart illustrating a processing procedure for interpreting a Query object according to the first embodiment.

FIG. 36 is a flowchart illustrating a processing procedure of QueryExpr interpretation according to the first embodiment.

FIG. 37 Quer instructs a ranking search in the first embodiment.
FIG. 9 is a diagram illustrating an example of a y object.

FIG. 38 is a query for instructing a neighborhood condition search according to the first embodiment.
FIG. 4 is a diagram illustrating an example of an object.

FIG. 39 is a diagram showing an outline of an architecture indicated by a conventional DMA specification.

FIG. 40 is a diagram showing a procedure in which an application program of a conventional document management system accesses a document in a document space.

FIG. 41 is a diagram showing a procedure of a conventional local search.

FIG. 42 is a diagram showing a conventional global search procedure.

FIG. 43 is a diagram illustrating the principle of the document management system according to the second embodiment.

FIG. 44 is a diagram illustrating an operator specification and an operand specification of an And-Not operator according to the second embodiment.

FIG. 45 is a diagram illustrating an example of an And-Not operator according to the second embodiment.

FIG. 46 is a diagram illustrating an example of a Query description when performing an attribute search according to the second embodiment.

FIG. 47 is a diagram illustrating an example of full-text search of an And-Not operator according to the second embodiment.

[Explanation of symbols]

146: Query object generation program, 147 ...
Storage area for operator definition information for full-text search, 148 ... Class definition information storage area for full-text search, 155 ... Query object interpretation program, 156 ... Storage area for operator definition information for full-text search, 157 ... Class definition information storage area for full-text search, 167 ... Query object interpretation program, 168 ... operator definition information storage area for full text search, 169 ... class definition information storage area for full text search.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tohru Takahashi 2-2-2 Kagahara, Tsuzuki-ku, Yokohama-shi, Kanagawa Prefecture Inside System Development Division, Hitachi, Ltd. (72) Inventor Tetsu Wakayama 5030 Totsukacho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Address: Hitachi, Ltd.Software Development Division (72) Inventor: Akio Higashi

Claims (16)

[Claims] In a document search method for searching for a target document from a plurality of documents stored and managed as a set of documents, a search formula or search condition for performing a search process is input, and the search request is a full-text search. And if the search request is a full-text search, the operator definition information for the full-text search that defines the operator who instructs the full-text search and the class for the full-text search that defines the class used in the full-text search Checking the available search function by referring to the definition information, generating a Query object for performing a full-text search according to the contents of the operator definition information for the full-text search and the class definition information for the full-text search, It is necessary to hand over the object to the server to request search processing, obtain the search result from the server, confirm the search result, and obtain the document entity. Document search method characterized by identifying the document to obtain a document entity to connect to a server that stores manage the document. 2. A document retrieval method for retrieving a target document from a plurality of documents stored and managed as a document set, wherein a Query object for performing a retrieval process is received from a client, and the retrieval request is a full-text retrieval. To check the available search function by referring to the operator definition information for full-text search that defines the operator who instructs full-text search and the class definition information for full-text search that defines the class used in full-text search, and perform the full-text search Interpret the Query object according to the contents of the full-text search operator definition information and the full-text search class definition information, generate a search expression for a search engine according to the interpretation result of the Query object, and execute the search process. A document search method characterized by passing search results to a client. 3. The system according to claim 1, wherein the operator designates a neighborhood condition at the time of a neighborhood condition search.
Alternatively, the document search method according to claim 2. 4. The document search method according to claim 1, wherein the operator develops a different term of the search term. 5. The document search method according to claim 1, wherein the operator develops a synonym of the search term. 6. The document search method according to claim 1, wherein the operator is an And-Not operator. 7. In a document search method for searching for a target document from a plurality of documents stored and managed as a document set, a search formula or search condition for performing a search process is input, and the search request is a full-text search. If the search request is an attribute search that includes the operation of the And-Not operator, the available search function is checked by referring to the operator definition information that defines the And-Not operator. And
A Query object including an And-Not operator is generated according to the contents of the operator definition information, the generated Query object is passed to a server to request a search process, a search result is obtained from the server, and the search result is confirmed. A document for which a document entity needs to be acquired by using the method, and connecting to a server that stores and manages the document to acquire the document entity. 8. In a document search method for searching for a target document from a plurality of documents stored and managed as a document set, a Query object for performing a search process is received from a client, and the search request is calculated by an And-Not operator. In the case of attribute search including, check the available search function by referring to the operator definition information that defines the And-Not operator,
Interpret a Query object including an And-Not operator according to the contents of the operator definition information, generate a search expression for a search engine according to the interpretation result of the Query object, execute a search process, and send the search result to the client. A document search method characterized by passing. 9. A client for searching for a target document from a plurality of documents stored and managed as a document set, wherein operator definition information for full-text search defining an operator instructing full-text search and a class used in full-text search are defined. The full-text search class definition information and the full-text search operator definition information and the full-text search class definition information are referred to confirm available search functions, and a Query object for performing a full-text search is assigned to the full-text search operator. A client, comprising: a query object generation processing unit that generates the definition information and the contents of the class definition information for full-text search. 10. A server for retrieving a target document from a plurality of documents stored and managed as a document set, wherein operator definition information for full-text retrieval defining an operator who instructs full-text retrieval and a class used in full-text retrieval are defined. The full-text search class definition information, the full-text search operator definition information and the full-text search class definition information are referred to confirm available search functions, and a Query object for performing a full-text search is assigned to the full-text search operator. A server comprising: a query object interpretation processing unit that interprets according to the content of definition information and class definition information for full text search. 11. A client for retrieving a target document from a plurality of documents stored and managed as a document set, comprising: operator definition information defining an And-Not operator;
Quer to check the available search function by referring to the operator definition information and generate a Query object including an And-Not operator according to the contents of the operator definition information
a client comprising a y object generation processing unit. 12. A server for retrieving a target document from a plurality of documents stored and managed as a document set, comprising: operator definition information defining an And-Not operator;
Quer interprets a query object including an And-Not operator according to the contents of the operator definition information by checking the available search function with reference to the operator definition information.
A server comprising: a y object interpretation processing unit. 13. A full-text search operator that defines an operator who instructs a full-text search on a medium that records a program for causing a computer to function as a client that searches for a target document from a plurality of documents stored and managed as a document set. The available search function is checked by referring to the definition information and the class definition information for full-text search that defines the class used in the full-text search, and a Query object for performing full-text search is defined as the operator definition information for full-text search and the full-text search. Generated according to the contents of class definition information for
y A medium recording a program for causing a computer to function as an object generation processing unit. 14. A full-text search operator that defines an operator who instructs a full-text search on a medium that records a program for causing a computer to function as a server that searches for a target document from a plurality of documents stored and managed as a document set. The available search function is checked by referring to the definition information and the class definition information for full-text search that defines the class used in the full-text search. Interpreted according to the contents of class definition information for
y A medium in which a program for causing a computer to function as an object interpretation processing unit is recorded. 15. A medium storing a program for causing a computer to function as a client for retrieving a target document from a plurality of documents stored and managed as a document set, referring to operator definition information defining an And-Not operator. And confirming a search function that can be used, and recording a program for causing a computer to function as a Query object generation processing unit that generates a Query object including an And-Not operator according to the contents of the operator definition information. Medium to do. 16. A medium for recording a program for causing a computer to function as a server for retrieving a target document from a plurality of documents stored and managed as a document set, referring to operator definition information defining an And-Not operator. And confirming a usable search function, and recording a program for causing a computer to function as a Query object interpretation processing unit that interprets a Query object including an And-Not operator in accordance with the content of the operator definition information. Medium to do.