JP2001092827A - Data management apparatus and method - Google Patents

Abstract

(57) [Summary] [Problem] It is becoming common to improve the availability of data by adding tags to various types of data using a markup language such as XML, but the semantic information of the tags is specified. We had to rely on what was distributed and published by the creator. A registration metadata receiving unit for receiving data registration from a metadata developer, a registration metadata examination unit for examining a format of the accepted metadata, and a registration meta for loading credit data on the accepted metadata. Data determination means, means for adding the determination result to the metadata and accumulating or distributing it to the database, registration data publishing means for distributing the requested metadata in response to a request from the user, and feedback information on the metadata And registration metadata / feedback means for receiving the information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data management apparatus and method for managing metadata to be disclosed.

[0002]

2. Description of the Related Art In recent years, Web standard language HTML (Hype
r (Text Markup Language), an XML (eXtensible Markup) that allows created content to be exchanged between a variety of information devices such as computers, televisions, and telephones.
Language) is becoming widespread, and in particular, its use is increasing rapidly in the field of e-commerce. HTML and XM
L stands for SGML (Standard Generalized Marku
p Language) is a part of the language specification (ISO 8879), but the greatest feature of XML when compared with HTML is that the tag set to be added to the content is not determined in advance.

[0003] In HTML, a tag set mainly related to a document layout is determined in advance.
In L, the user can define tags for performing not only layout but also semantic structuring of data as necessary. The definition of tag names and interrelationships between tags is DT
D (Document Type Definition), XM
An L-compatible system processes data based on the DTD.

In the DTD, for example, a tag <product> is composed of a tag <product name> and a tag <price>.
Is a character string, and <price> is numeric type data. This is a declarative expression of the attributes of a product, and XML can be regarded as a kind of knowledge expression language.

On the other hand, in the field of knowledge information processing, a technical term of ontology is actively used. An ontology in this field refers to a systematic description of the concepts required to describe a model of a certain object (world). Specifically, it is expressed in various forms, such as a common vocabulary, a hierarchical description of concepts (terms), a conceptual schema in a database, a solarium, and a strict description of concept definitions and inter-concept relationships according to application purposes. Things. As mentioned earlier, X
Since ML can be regarded as a knowledge expression language, there are many attempts to actually describe an ontology in XML.

[0006] Content marked up in XML can be distributed including its semantic content (interpretation method) by sharing the DTD. DTD
In order to share the DTD, there is a method of embedding the DTD directly in the content, or a method in which the entity of the DTD is a separate file and only the reference destination is described in the content. In the latter method, the DTD entity and its name are managed separately, so that the DTD provider and the user have independent characteristics.

For example, if there is a public DTD, D
The user of the TD can make a description using the tag set only by referring to the name. On the other hand, a public DTD provider can make an extension such as appropriately refining the tag definition independently of the user. Such a sharing mechanism has advantages for both DTD providers and users.

[0008] Therefore, the ontology is also the D in XML.
It would be extremely useful to be able to share using the same mechanism as TD share. In fact, ontology development is a costly operation that requires time, effort and wisdom. If these results are provided as a public ontology, the performance of the knowledge processing system possessed by the user will be dramatically improved. At this time, if the independence between the provider and the user is maintained, even if the ontology development is at an intermediate stage, the user can use it, and the provider can continue to expand after that. You can do it.

[0009] The difficulty of ontology development is that it is inevitably incompatible with intensive development. That is, it is impossible for one organization to develop an ontology of the whole world, and it is natural for each organization to develop a minimum ontology for its specific purpose. Therefore, providing ontologies is necessarily inevitably decentralized.

As a method of realizing an ontology share having such a characteristic, client-server type management is considered promising. That is, ontology providers distributed in various places open a server that provides the ontology on the Internet, and publish the name of the ontology. The user (the ontology using system) appropriately obtains the ontology entity from the server according to the name of the ontology to be used, and performs a predetermined process. Such management is also useful for DTD share in XML.

[0011]

However, these metadata are being de facto-ized by individual organizations and leading companies in their respective fields and industries, setting their own standards and promoting de facto conversion. Since metadata is not taken into account, metadata described in one format is not compatible with metadata described in another format and can be used only by respective dedicated applications.

In addition, the ontology management and the D
When performing TD sharing, a problem is to use communication for the exchange of the ontology. Communication is not always stable, and the server is down, cannot reach the server, or data transmission is interrupted on the way. In addition, many ontology data are large-scale, and data compression is also desired.

The present invention has been made to solve such an ontology management problem, and an object of the present invention is to provide an ontology as a substitute when a certain ontology cannot be used based on the nature of the ontology. Another object of the present invention is to provide an ontology management system and a management method for performing data compression by abstracting an ontology.

[0014]

SUMMARY OF THE INVENTION In order to solve these problems, the present invention provides a metadata receiving means for receiving metadata sent from a metadata developer through a network, and a metadata received by the metadata receiving means. A metadata examination means for examining whether or not the metadata satisfies a predetermined condition, and a metadata / data for accumulating metadata determined to satisfy the predetermined condition by the metadata examination means.
A database, metadata publishing means for distributing metadata stored in the metadata database through a network in response to a request from a user, and determining the reliability of the metadata to be distributed by the metadata publishing means By having the metadata determination means, the metadata for the tag is standardized, and can be used for multiple purposes and with peace of mind.

[0015] Further, storage means for storing a standard ontology that has a correspondence relationship between ontologies of different metadata and stores information for converting between the ontologies as an ontology, and a meta-data received via a network. And a distribution unit for distributing necessary conversion information from a standard ontology of the storage unit or a conversion result by the standard ontology through a network for converting data into metadata of a requested format.

Further, a storage means for storing an ontology having an identifier uniquely identified on a computer network, an ontology using system for designating an identifier and requesting the ontology stored in the storage means, An ontology server that extracts an ontology corresponding to a name from the storage unit, and a detection unit that detects an ontology that substitutes for the requested ontology.

Also, a storage means for storing an ontology having an identifier uniquely identified on a computer network, an ontology using system for designating an identifier and requesting the ontology stored in the storage means, An ontology server that extracts an ontology corresponding to a name from the storage unit, and a conversion unit that converts an ontology abstracted by reducing intermediate nodes of the ontology extracted by the ontology server.

[0018]

BRIEF DESCRIPTION OF THE DRAWINGS The invention will first be described in general with reference to the drawings, after which embodiments of the individual aspects will be given. FIG. 1 is a diagram showing the overall configuration of an embodiment of the present invention. The ontology server 11 is a server that manages various types of metadata as an ontology. The standard ontology 12 stores a standard ontology that enables bridging between various types of metadata, and compatibility between various types of metadata is ensured by a conversion unit 13 that uses the standard ontology.

The ontology server 11 accumulates individual ontologies 14 described in a specific format for a specific industry or application under the management of a management unit 15,
Distribution is performed according to a reference request from the application 16. The application 16 is not only for using multimedia data to which metadata is added,
Authoring tools for creating new multimedia contents 18 by adding new metadata to the multimedia data 17 and authoring tools for creating and editing the ontology itself are also included.

When a specific type of ontology is referred from an authoring tool such as the former, related parts of the standard ontology are also presented at the same time, and authoring according to the description of the standard ontology is recommended and supported. Authoring tools like the latter require the user to input the definition of the correspondence between the newly edited part and the standard ontology,
This is to ensure the compatibility of the created ontology, and to support the creation of metadata that is highly compatible and can be distributed and used in a wide range of fields / industries.

Hereinafter, the first to third aspects of the individual aspects of the present invention will be described.
This will be described with reference to the embodiment. [First Embodiment] FIG. 2 is a diagram for explaining a first embodiment of the present invention. Hereinafter, an embodiment of the management unit 15 of the ontology server 11 will be described with reference to FIG.

The registered metadata receiving unit 21 receives metadata transmitted from a metadata developer via a network. The accepting means has a function as a gateway, such as accepting only data from a user who has been registered in advance and not accepting data from other destinations.

The received data undergoes a basic system examination in a registered metadata examination unit 22. For example, it is essential to identify the developer (company name if it is a company, name if it is an individual), contact information, and in the case of a membership system that only specific developers can register, the member identifier etc. Management information, auxiliary information such as the identifier of the registered metadata that is the basis for updating the registered metadata by adding, modifying, changing, etc., or the product field, function, purpose, etc. that the registered metadata targets Formal examination such as whether index information for metadata search etc. is properly input, and within the range of data elements or attributes that can be taken as the value of data elements or attributes, as to the contents of metadata The basics of the content examination, such as whether the value of the data element or the constraint with the value of other data elements or attributes are satisfied Do such things. As described above, the registered metadata examination section examines whether the registered metadata satisfies at least the minimum conditions in terms of format and content.

On the other hand, the registered metadata determining section 23 determines the property of the metadata to be registered.
For example, past registration results of metadata developers, credit information such as complaints and defect information for registered metadata, checking for consistency with other already registered metadata and existence of conflicts, and other metadata Extraction and determination of requirement information and the like for use, such as determination of the influence on the application, compatibility with the registered metadata of the old version, and the influence on the used application.

It is assumed that the result of extraction is held as tabular data as shown in FIG. 3, for example. Here, the identifier is sample-on # 1, and the version number is ver.
1.1 ontology node no. # 1-1 and below, and identifier # relat-ont # 1 and version number ver1.1 ontology node # 1-1 and below have an application version number viewer # 1. It is described that it cannot be used in combination with those with ver.

If the identifier is sample-ont # 2
, The node number of the ontology whose version number is ver2.2 is node # 2-2 or lower, and the identifier is sample-o.
It is described that nt # 2 and the version number below node # 2-2 of the ontology of ver2.1, that is, all applications are incompatible and cannot be used as they are. If possible, it is possible to provide information that only the part needs to be converted.

Further, not only detailed information on such individual faults, but also a fault ontology that can be used only in a limited range where a large number of such fault information is extracted, and a fault ontology such as this It is also conceivable to use the method such as determining the existence of a bad ontology provider registered at a high rate. If such a determination result is provided to the user as a kind of credit information, it can be a useful information service for the user.

The credit information is also an index indicating the degree of recommendation by a management organization that actually operates the metadata management and publication device and method according to the present invention, and these data are useful to the user as such. It also enables the implementation of a kind of information service related to registration metadata.

For example, when there are a plurality of metadata candidates that can be used for a certain purpose, it is more appropriate to determine which of them is to be selected and present it to the user in the form of a ranking of candidate metadata. It is also intended to enable a service of performing such a service. Such a service may be operated by the management institution itself, or a case where a third party provides it as a value-added service in consideration of other information may be envisaged.

[0030] Regarding the actual operation mode of the metadata management and publication device and method according to the present invention, the subject and the operating organization shall at least manage the metadata database, and otherwise cooperate with the third party organization. Embodiments are also possible. For example, when the number of registered metadata increases, it may be assumed that a situation may occur in which a large operating capacity is required for the main operating organization in the examination processing of the registered metadata. In such a case, as a registration agency service of a third-party organization, an operation mode in which the examination process of the registration metadata is operated and the load on the main body of the institution is reduced is also assumed as an embodiment of the present invention. .

The judgment result adding section 24 is means for adding the judgment result of the registered metadata judging section 23 to the registered metadata. The function to be executed depends on when the judgment of the registered metadata is executed. Are different. That is, the determination of registered metadata is dynamically determined at the time of use, such as interaction with other registered metadata, such as update information for registered metadata, and interaction with other registered metadata. There are things that are better to do.

For this reason, the metadata added at the time of registration
The form to be stored in the database and the form to be added at the time of a distribution request from the user are described. The former is an embodiment in which the response is emphasized, and the latter is an embodiment that mainly reflects the determination based on the latest situation.

However, even in the former case, the latest situation is similarly reflected even if the judgment result of the registered metadata having an effect is updated simultaneously from the judgment result of the newly registered metadata. In some cases, it is possible to accumulate judgment results. However, if the influence is not a whole but a partial one, it is assumed that both the determination and the method of storing the determination result will be complicated. It is assumed that the judgment is more reasonable. Therefore, what embodiment is to be adopted should be appropriately selected according to the property of the target metadata, and is not limited to the form shown in the present embodiment.

The registered metadata publishing unit 25 is means for extracting appropriate registered metadata from the metadata database in accordance with a request from the user side and distributing it to the request source. Here, the user side refers to the case where metadata is used, such as when a request is made from a terminal device operated directly by the user, or when an application used by the user requests metadata specified by the data being used. This is a general term for requests from devices that have been issued, and is not limited to a form in which a user issues a request directly.

Further, the present invention is not limited to a request issued directly from a terminal device under the control of the user. That is, unlike a device such as a search engine that intervenes in the form of mediation with the end user, it does not use the metadata itself, but does not request the end user based on the request of the end user. And indirect requests via an intermediary.

This also assumes an information service relating to the aforementioned metadata. Not only services such as the ranking of registered metadata above, but also metadata that can be used on the network, not only those maintained by the management unit given a certain authority described above, Even those that are supposed to be able to perform services such as those that have knowledge about metadata etc. originally developed by venture companies and individual business operators, etc. is there.

The request from the user, which is accepted by the registered metadata publishing unit 25, is basically made by using the identifier of the registered metadata. However, a group of metadata is versatile, the coverage of the target field, Due to the required level of detail, the size becomes large as a whole, and the size is too large to distribute via the network. There may be a problem that the data must be received and stored in a large-scale storage medium or the like.

In addition, when the metadata actually used by the user is only a small part of the group of metadata, it is easily assumed that such inconvenience greatly impairs the practicality. You.

Therefore, in this embodiment, it is assumed that a group of metadata is held as an ontology. The ontology is based on a data structure such as a frame-type knowledge representation or a tree-structured semantic network that represents the concept hierarchy, which has been generally used in the field of knowledge engineering, or more generally, an object-oriented database. This is a possible data holding form.

By adopting such a holding form, it is easy to divide the data into subtrees, and the user requests the entire ontology of the tree structure using the ontology identifier or the ontology name representing the ontology. After specifying the sub-tree, specify the sub-tree root node identifier or node name to specify the sub-tree below, so that the user can specify a subset of the size as close as possible to the data actually used on the user side It becomes easier to do. It also realizes a rational metadata distribution method that minimizes the load on the network during distribution and minimizes the amount of storage media required to be maintained on the user side. is there.

It is also assumed that the group of metadata is not only fixed but also frequently updated. For such metadata, the difference between the metadata held by the user and the metadata requested by the user can be specified by specifying the version identifier of the metadata held by the user. Realizing reasonable distribution by receiving only the description of the conversion method such as data or conversion rules, and updating the metadata held by the user using the description of the conversion method Is also possible.

Also, in terms of ontology management, it is conceivable that storing all versions of the ontology in the ontology database would be irrationally both physically and economically. In such a case,
Set the time limit for keeping the old version, and save the data required for the maintenance of the old version by retaining only the difference data from the oldest version or the description of the conversion method of the previous version for the previous version. A management method that significantly reduces the required amount of media is also envisioned.

Next, the registration metadata feedback section 26 will be described. Registration metadata is registered after being reviewed by the management department at the time of registration, and its quality and safety are guaranteed to some extent,
As described in the explanation of registered metadata examination means and registered metadata judgment means, basically, mechanical and formal examination is assumed, so complete security, especially in combination with other metadata If used, other partial data,
It is assumed that it is extremely difficult to realistically or practically assume all cases when used in combination with partial data of other metadata.

It is not practically or practically possible to anticipate and anticipate all the effects and anticipated defects, etc., on applications developed assuming the use of the old version when the new version is introduced. It is assumed that it is possible.

In such a case, it is an object of the present invention to provide a means for minimizing the influence of a defect by actively accepting defect information from a user and promptly reflecting the information in registration examination and judgment. Is what you do.

Finally, the metadata database 27
This is a database that stores registered metadata to be searched and metadata.

The operation of the metadata management and publication device and method will be described below with reference to the flowchart of FIG.

First, the registration metadata sent from the metadata developer via the network is accepted (S
41). At this point, a method may be considered in which the metadata management / publication device is configured as a gateway that accepts only registration data from developers who are authorized to access the device via the network. If configured as a shared system, at this point registration metadata from all technically accessible developers, eg, IP reachable developers, will be accepted and all registration metadata passing the next review stage will be It is possible to adopt a method of accepting regardless of the above.

The accepted metadata is examined to determine whether the data description format, the minimum required data description items are properly filled in, and whether a permissible value is given for a specific description item. (S42), and if it is determined that the required format is not prepared (S43), the developer is notified to that effect (S4).
4) Then, the metadata registration process is stopped (S49).
In this case, it is not only a notification that the metadata sent to the developer has been rejected, but also the reason for the rejection,
For example, it may be desirable to add information about the deficiencies.

Next, when it is determined that the received metadata satisfies the minimum formal requirement (S4
3) Further, a determination process regarding the property of the metadata is performed (S46). However, the determination to be made at this time is only a static determination, and if there is a dynamic characteristic in the registered metadata, it is necessary to make a separate determination, which will be described later. Here, the determination regarding the property is a determination regarding the usefulness of the metadata to the user, and conversely, a determination regarding restrictions on use of the metadata and defects that may be given to the user. In other words,
This is a process of generating credit information for the metadata.

The determination result is stored in a metadata database in association with the metadata or as additional information of the metadata so that the result of the determination can be distributed together with the distribution to the request source (S47).

The above is the process of registering the metadata sent from the developer. Next, the registered metadata is released according to a request from the user. Hereinafter, the distribution process will be described with reference to FIG.

The user who wants to use the shared metadata requests the metadata management and publication device to distribute the metadata to be used (S48). However, the user's request is not limited to the user's own request by directly operating a terminal device capable of accessing the metadata management / publication device, but is also used by the application used by the user. This includes the case where metadata required for data processing is requested through a network, and also includes requests from a mediation mediation service system that mediates between the user and the metadata management and publication device. It is a thing.

Rather, the process of adding credit information, etc.
This processing assumes the existence of such an intermediary service system, and assumes a function of recommending the most suitable one of the candidates to the user when there are a plurality of metadata available to the user.

The delivery of the requested metadata is assumed to be performed through a network. In particular, considering the delivery through a network having a small transmission capacity such as a mobile environment, the capacity of the data to be delivered is required. It is considered that practical measures should be taken to minimize the amount of data. For the delivery of requested metadata, it is always necessary to send the entire series of registered systems at once, including data that the user does not necessarily need. There is a possibility that a useless communication load such as distribution is generated, and when the system of a series of registration metadata has a certain scale, such a useless communication load cannot be ignored in practical use. Therefore, the distribution request from the user can specify not only a series of systems but also a specific part actually used by the user.

As one means for realizing such partial designation, in the present invention, it is assumed that metadata is described and held as an ontology. That is, the user can designate the partial data by designating the data in the metadata system described as the ontology and the data from the specific node to the end. Therefore, the distributing means cuts out the inherited part when the inheritance from the higher node exists in reverse to the nodes below the designated node of the designated ontology (S
410) It shall be distributed.

Further, it is assumed that not only the requested metadata is delivered but also the above-mentioned credit information relating to the metadata is added and delivered. In some cases, there is a dynamic characteristic, and in the case of partial data distribution, some data can be determined only after the part is specified. Therefore, after the distribution data is determined as necessary, a determination regarding the distribution data is performed.

It is assumed that feedback information is used as information that can be used for the determination. In other words, it receives and accumulates from the user, etc., information on failure information related to the specified distribution metadata or information that can be suitably used in fields and applications different from those assumed by the maker, and can be used. If there is (S411), it is used for determination (S412).
Only a dynamic determination is made for other elements (S41)
3).

As a means for reducing the amount of data to be distributed, it is considered that the above method of partial data distribution alone is not practically sufficient. Therefore, it is assumed that the metadata is also cached on the user side and, for those which are used fixedly or frequently, the metadata stored locally is used as much as possible within an allowable cache size. . In addition, the metadata stored in the cache that is stored locally shall be fully utilized, and if an updated version of the metadata stored in the cache is required, the version data required for the upgraded version shall be requested, In the case of the partial data in the previous example, even when the metadata in the cache is further partial data of the necessary partial data, only the difference is distributed, or the data in the cache and the required meta data are transmitted. If the data is in a neighborhood relationship in the field and can be used with only a small correction, only the correction data is distributed.

Therefore, based on the requested metadata and the information of the metadata held by the user in the cache,
It is determined whether or not it can be dealt with only by the distribution of some kind of difference data (S414), and if it can be dealt with only by the description of the difference data and the conversion or correction method, it is searched or generated (S415) and distributed. If such a response is not possible, all the requested data is distributed (S416). Second Embodiment Next, a second embodiment of the present invention will be described with reference to FIGS. First, an embodiment of the distribution of bridging data between ontologies by the ontology server 401 will be described with reference to FIG. Here, format A
Metadata A52 of the corresponding ontology A5
A case will be described in which the application 56 interpreting by referring to FIG. 3 and using the multimedia data can use the multimedia data to which the metadata B 54 of the format B is added.

The reason why the ontology A53 and the ontology B55 are not enclosed in the frame of the ontology server 51 is that the ontology is not limited to being delivered from the ontology server via the network every time the ontology is referred to. Without the application,
Alternatively, this is to indicate that the cached data may be referred to in a local environment on the user side.

The processing flow will be described below with reference to the flowchart shown in FIG. When the application 56 receives the multimedia data to which the metadata B54 has been added, the application 56 transfers the metadata B54 to the ontology server 51.
Send to The conversion unit 57 of the ontology server 51 receives it (S61), and identifies the ontology A53 that the application 56 refers to as standard (S62), and identifies the ontology B55 that the metadata B54 refers to (S63). .

The conversion unit 57 refers to the standard ontology 58 and checks whether the definition of the correspondence between the ontology A53 and the ontology B55 exists and whether or not bridging is possible (S64). The description content of the metadata B54 is analyzed (S65), and the conversion from the ontology B55 to the ontology A53, in other words, the metadata B5, while referring to the standard ontology 58.
4 is searched for in the standard ontology (S
66) The conversion information necessary for conversion into the description of the metadata A52 according to the definition is sequentially extracted (S6).
7).

In order to show the details of this processing concretely and simply, a description will be given using an extremely simplified example. For example, it is assumed that a descriptor (a tag in a markup language) of “price” is included in the metadata B54 and processing corresponding to the descriptor is expected. The conversion unit 57 searches the standard ontology 58 and cuts out a partial ontology as shown in FIG. 7, for example.

It can be seen from the partial ontology of FIG. 7 that the “price” in the ontology B 55 is associated with the “price” in the ontology A 53 using the “price” of the standard ontology 58 as a standard definition. Therefore, the conversion unit 5
7 returns a response in which the descriptor “price” in the metadata B 54 is replaced with “price” or a partial description defining the relationship between “price” and “price” and “price” in the standard ontology 58 The extracted partial description is sequentially extracted as conversion information, or the processing content of “price” is converted into the description format of the ontology A53 based on the partial description.

The extracted conversion information returns an answer in the format required by the application (S68). For example, if the conversion processing can be performed by the application itself, the conversion information sequentially extracted is merged and returned (S610). If the conversion processing cannot be performed by the application itself or is difficult, the ontology server 53 and the ontology server 51 side perform the conversion processing. The conversion process is performed, and the metadata converted into the ontology A53 format or the necessary parts of the ontology B55 converted into the three formats are collectively returned (S69) and returned (S611).

By the above processing, the correspondence between the ontology of different metadata and the core metadata description between them is retained, and information for enabling conversion between them is retained as the ontology. It converts the metadata received via the network into metadata in a format that can be used by the user, and distributes the conversion information or the conversion result necessary for the conversion through the network. [Third Embodiment] Next, a third embodiment of the present invention will be described with reference to FIGS. First, with reference to FIG. 8, an embodiment of a support method for recommending a standard ontology by the ontology server 81 and ensuring compatibility of metadata based on the recommendation will be described.

The creator of the multimedia content uses the authoring tool 84 to write the multimedia data 86
When producing the metadata 85 to be added to the application, there is an ontology 82 which is referred to by the industry that the producer intends to distribute and a series of applications assumed as targets, and as it is, it is compatible with other industries and applications. If the multimedia content does not have such characteristics, the created multimedia content can be used only in the industry or a series of applications, and the distribution range is limited.

Therefore, when the metadata creator attempts to describe metadata in a certain format, not only the ontology 82 conforming to the format but also the information of the standard ontology 83 is presented at the same time. Deliver information that supports the creation of multimedia content that is highly compatible, can be distributed in a wider industry, and can be used by more applications.

Hereinafter, this series of processing will be described with reference to the flowchart shown in FIG. When the metadata creator begins to create metadata using the authoring tool 84, the authoring tool 84 instructs the ontology server 81 to refer to the metadata used in creating the metadata. Issue a reference start request. Upon receiving the reference start request, the ontology server 81 identifies the ontology for which the authoring tool 84 has requested reference (S91).

The authoring tool 84 issues a reference request for extracting necessary information from the ontology as the creation of the metadata progresses. Upon receiving the reference request, the compatibility management unit 87 of the ontology server 81 (S9)
2) In addition to identifying the part to be referenced in the ontology for which the authoring tool requires reference, referencing the standard ontology 83 and determining whether information corresponding to the reference part exists in the standard ontology. Is searched (S94).

If there is no corresponding information, the next reference is waited for (S97). If there is the corresponding information, the information required by the authoring tool to describe the metadata to be described is requested by the reference. Not only is it extracted from the ontology 82 that is being used, but information corresponding to it is also extracted from the standard ontology (S95).
The reference information thus extracted is returned to the authoring tool and presented to the user (S96).

In order to show the contents of the above processing concretely and simply, a description will be given using an extremely simplified example used in the second embodiment.

Here, assume that the ontology 82 that the creator of the multimedia content intends to refer to and use is an ontology A. Therefore, it is assumed that the creator intends to create metadata relating to the “price” of the article with reference to the ontology A. At this time, the ontology compatibility management unit 87 refers to the description in the standard ontology 82 related to “price” (here, the same as FIG. 7), and is defined as a standard vocabulary for “price” of the ontology A. The "price" is presented as a recommendation descriptor at the same time, and it is recommended that the creator use the standard "price" instead of "price" for the description of metadata.

By such processing, the creator of the metadata uses the ontology 8 which has been conventionally referred to and used conventionally.
It is recommended to shift from description based on No. 2 to description of highly compatible metadata based on the standard ontology 83, and support creation of more compatible multimedia contents.

In addition to not only the creation of metadata but also the authoring of modifying or newly creating the referenced ontology itself, the compatibility management unit
When the modified part of the reference ontology is identified, the corresponding description part of the standard ontology is also identified, and the creator is encouraged to change the definition of the correspondence relationship after the modification. Also, in the case of new creation, referencing the standard ontology, overwriting only the parts that need to be modified, and prompting the definition of the correspondence with the description of the original standard ontology with the modification Thus, a method of modifying an ontology with high compatibility and supporting creation of an ontology with high compatibility can be considered. [Fourth Embodiment] An embodiment of the present invention will be described below with reference to the drawings. FIG. 10 is a block diagram illustrating a configuration of a data management system according to the fourth embodiment.
The data management system includes an ontology server 101, a name server 102, an ontology proxy 103, and an ontology processing system 104.

The ontology server 101 is the ontology proxy 103 or the ontology processing system 1
In response to a request from the server 04, the contents of the ontology database managed by itself are transmitted to the requester. The first configuration of the ontology server 101 includes an ontology server body 111, a meta-ontology database 11
2. The ontology database 113 consists of three persons.

The ontology database 113 is a collection of ontology having an identification name uniquely specified on the computer network. FIG. 11 shows a configuration example of the ontology database 113. The ontology database includes a header section 1131 and a body section 1132. The header part expresses an identifier that uniquely identifies the ontology on the computer network, an ontology description format (predicate logical format, XML format, frame format, etc.) stored in the body part, and an ontology description object. The language (Japanese, English, etc.) is described.

For example, in the example of FIG. 11A, the identifier is “o”.
ntology#1-1@ont.org1.or.jp ", indicating that the description format is a predicate logical format and the object expression language is Japanese. The entity of the ontology according to is stored.

For example, in the example of FIG. 11A, the parent concept of “maker” and “department store” is “company”, and “company A” and “company B” are instances of “maker” and “department store”. "Is an instance of" Company C ". In the example of FIG. 11B, the parent concept of “company” is “organization” or “building”, which is expressed in XML format. Also, in FIG. 11C, the fact that the objects “company” and “company” are equivalent is shown in FIG.
Indicates that the Japanese objects "company" and "organization" correspond to the English objects "enterprise" and "organization", respectively.

The meta-ontology database 112
This is knowledge describing the relationships between these individual ontologies, for example, the inclusion relationships and translation relationships. From these relationships,
It is also possible to provide another alternative ontology when one ontology is unavailable. FIG.
FIG. 2 shows an example of a meta-ontology database.

The knowledge shown in FIG. 12A corresponds to the identifier “ontology #
Ontology of 1-3@ont.org1.or.jp "is the identifier" ontol
ogy#1-1@ont.org1.or.jp ". More precisely, the ontologies of the first and second terms have the same object expression language and 2
All objects included in the ontology of the term are the first
It appears in the ontology of the term, and indicates that the ontology description (axiom) of the second term can all be derived from the ontology description of the first term.

The knowledge in FIG. 12 (2) indicates that the first and second terms are the same, that is, there is an alias definition of the identifier. The knowledge in FIG. 12 (3) indicates that the first and second terms are identical. Two terms are equivalent, that is, the first term includes the second term, and the second term includes the first term. In addition, the knowledge of FIG.
This indicates that the ontology of the first and second terms can be mutually translated (generated) by the ontology of the third term (translation knowledge of the object language). The knowledge of FIG.
Is the same as The knowledge in FIG. 12 (6) indicates that the ontology of the first and second terms can be consistently combined by the ontology of the third term (equivalence knowledge of objects).

For example, FIG. 13 shows that the ontology of the first term (FIG. 11 (a)) and the ontology of the second term (FIG. 11 (b)) are changed based on the knowledge of FIG. 12 (6). According to FIG. 11 (c), the joining node is changed to “company” (=
"Company") to show a consistent connection.

Note that this meta-ontology database is not necessarily required for all ontology servers, and may be configured to include only the ontology server main body 121 and the ontology database 123.

When the ontology server 101, the ontology proxy 103, or the ontology processing system 104 gives an ontology identification name and issues a request, the name server 102 notifies the location of the ontology management server. The association between the identification name and the location is performed by drawing a correspondence table in which the ontology identifier 201 and the ontology server name 202 are paired, as shown in FIG. Note that these correspondences may be one-to-many correspondences. That is, the same ontology may be served by a plurality of ontology servers.

The ontology proxy 103 includes the ontology using system 104 and the ontology server 101
Manages the transmission and reception of the ontology. The configuration of the ontology proxy 103 includes the ontology proxy body 311 and the meta-ontology database 31
2. The ontology database 313 is composed of three persons.
The meta-ontology database 312 and the ontology database 313
This is a cache of received data.

The ontology using system 104 is an information retrieval system that performs query expansion (synonym / synonym expansion) using the ontology, a machine translation system that performs semantic ambiguity resolution using the ontology, and a document that presents information based on the ontology. Browsers, electronic catalogs and electronic commerce systems.

Hereinafter, the operation of the ontology server 101 will be described with reference to the flowchart of FIG.

First, the ontology server checks whether there is a connection request from the ontology proxy or the ontology using system, and if there is a request, establishes a connection with the request source (S151). Next, a request from the request source is accepted. In the present embodiment, the only accepted requests are an ontology transmission request (S152) and a session end request (S153). If there is a request other than these, an error is notified to the request source (S154).

When an ontology transmission request is made, it is determined from the given ontology identifier whether the ontology is usable (S155). More specifically, it is determined whether or not the database body of the ontology is on a file system managed by itself and whether it is in a transmittable state. If the requested ontology is not available, a process of detecting an alternative ontology is performed (S156).

The detection processing of the alternative ontology is performed according to the flowchart of FIG. First, the meta-ontology as shown in FIG. 12 is loaded on the main storage (S161). Next, an alternative ontology detection rule as shown in FIG. 17 is applied to the meta-ontology.

The alternative ontology detection rule in FIG. 17 is divided into a left side and a right side by an implication symbol “:-”. The right side is
This is a sequence of terms divided by the logical conjunction ",". Each rule indicates that if the right side is established, the left side is established. For example, rule (1) states that if ontology A subsumes ontology O, then ontology O
Can be substituted by ontology A (is # alternative). The third term on the left side indicates a sequence of translation knowledge used when substituting the ontology, and is described in rules (1) and (1).
In (2) and (3), the columns are empty. Rule (4) indicates that if ontology A and ontology O can be translated by ontology M (translatable), then ontology O can be replaced with a translation of ontology A with translation knowledge sequence M.

The rule (5) is that if the ontology O can be replaced with a translation of the ontology X with the translation knowledge sequence D, and if the ontology X can be replaced with a translation of the ontology A with the translation knowledge sequence M, , Ontology O
Indicates that the ontology A can be replaced with a translation of the ontology A sequentially with the translation knowledge sequences M and D (transition rule).

Now, the requested ontology is "ontology"
#1-1@ont.org1.or.jp ", which is not available. Then, metaontology knowledge (1) (FIG. 12)
By applying the alternative ontology detection rule (1) to ".", "Ontology#1-3@ont.org1.or.jp" can be detected as an alternative ontology. Alternatively, when the alternative ontology detection rules (1), (4), (5) are applied to the meta-ontology knowledge (4), (5), "ontology # 1-
1e@ont.univ1.edu "Translation knowledge" ej-dictionary @ ontolo
gy.or.jp "can be detected as an alternative ontology. In this way, a plurality of alternative ontologies are detected depending on how rules are applied.

Returning to the description of the flowchart of FIG. In the next step, it is determined whether or not an alternative ontology has been detected by applying the rule (S163).
If detected, all detected ontologies are pushed onto the output stack (S164); otherwise, an error is set (S165) and the process ends.

Returning again to the flowchart of FIG.
The continuation of the operation of the ontology server 101 will be described. When the above-mentioned alternative ontology detection process (S156) is completed, the process branches depending on whether or not the alternative ontology has been detected (S157). If not detected, the process proceeds to step S154, and an error is notified to the request source. On the other hand, if it can be detected, it is checked again whether or not they can be used for each alternative ontology (S155).

Then, when a usable alternative ontology is detected, the ontology processing step (S15)
Go to 8). The processing of the ontology is to reduce the amount of communication related to the transfer of the ontology, and is performed according to a processing form additionally specified by the request source together with the ontology identification name. Processing forms include ontology abstraction and synonym / synonym expansion.

The ontology abstraction is required by a machine translation system or the like, and enables an appropriate translation word selection or the like by eliminating semantic ambiguity. In general, it is said that a concept system (thesaurus) of about 3000 nodes is necessary and sufficient for the purpose of semantic ambiguity. Even more detailed knowledge does not affect the meaningful resolution.

For example, when there is a system as shown in FIG. 13, it is not always necessary that the “company” is subdivided into “manufacturers” and “department stores”. It is often sufficient as an instance of a node.
The operation of eliminating intermediate nodes unnecessary for the purpose and reducing the number of nodes to a necessary and sufficient number of nodes is ontology abstraction. Methods of mechanically abstracting the ontology include the equal depth method (to abstract only the nodes within a certain depth from the root node) and the equal size method (to make the number of dominated nodes uniform, A certain number of nodes are combined into one node), the uniform class probability method (abstract so that the appearance probability of each node is uniform)
Etc. are known.

On the other hand, synonym / synonym expansion is required by an information retrieval system or the like. By knowing synonyms or synonyms of terms used in a query (a sentence, a keyword list, etc.) given to the system, The purpose is to improve search performance (recall / accuracy). Mechanical synonyms
As a method of synonym expansion, a method of selecting words with similar semantic similarity (distance) using the above concept system (Resnik, P .: "Using Information Content to Evaluate
Semantic Similarity ina Taxonomy, "Proc. IJCAI-9
5. Hearst, M. And Sch "utze, H.:" Customizing a Lexico
n to Better Suit a Computational Task, "Proc. ACL
SIGLEX Workshop, 1993.).

Using these methods, the ontology server mechanically abstracts the ontology and expands synonyms and synonyms to reduce the amount of communication data (S158).
However, if there is no specification from the request source, nothing is done. Then, the processed ontology is transmitted to the request source (S159), and the next request is waited for (S152). And
When a termination request comes from the request source (S153), the session is terminated (S1510). The above is the operation of the ontology server 1011.

Next, the operation of the ontology proxy 103 will be described with reference to the flowchart of FIG.

First, the ontology proxy checks whether there is a connection request from the ontology using system, and if there is a request, establishes a connection with the request source (S181). Next, it searches its own cache (182) and checks for the presence of the requested ontology (S183).

When the requested ontology is found in the cache, the ontology is loaded on the main memory (S184), and the ontology is processed as necessary (S185) as in the case of the ontology server, and the requested ontology is requested. The original is transmitted (S186), and the session ends (S187).

On the other hand, if the cache is not hit in S183, the process proceeds to S188 and thereafter, and a session with the ontology server is started. First, an ontology identifier is given to the name server 102, and an ontology server that manages the ontology is identified (S188). Next, issue a connection request to the server,
The connection with the server is established (S189). Then, the ontology is requested by giving the identifier (S1).
810), and receives this and stores it in the cache (S181).
1), end the session with the server (S181)
2).

At this time, if the reception ends normally (S1
813) The process advances to S185 to proceed as described above, and ends the session. On the other hand, if the connection with the server is not successful and the server notifies that the ontology cannot be provided, an alternative ontology is detected (S
1814). The detection process of the alternative ontology is performed first in FIG.
6 as described above. When the alternative ontology is detected, it is checked again whether or not each alternative ontology hits the cache (S18).
2) Repeat the above processing until ontology data is obtained. On the other hand, if the alternative ontology cannot be detected, an error is notified to the ontology using system (S1816), and the session ends. The above is the operation of the ontology proxy 103.

Next, the operation of the ontology using system 104 will be described with reference to the flowchart of FIG.

First, the ontology using system checks whether an ontology proxy is set (S19).
1). If an ontology proxy is set, a connection request is issued to the proxy to establish a connection (S192), the necessary ontology is received from the proxy (S193), and the session with the proxy is terminated (S194). finish.

On the other hand, if the ontology proxy has not been set, the process proceeds to S195 and thereafter, as in the above-described operation of the proxy, checks its own cache, and starts a session with the server if necessary.

First, the cache owned by itself is searched (S
195), the presence / absence of the requested ontology is checked (S
196). If the ontology is found in the cache, the ontology is loaded on the main storage (S1).
97), and the process ends.

On the other hand, if no hit is found in the cache in S196, the flow advances to S198 and thereafter to start a session with the ontology server. First, an ontology identifier is given to the name server 2 to identify an ontology server that manages the ontology (S1).
98).

Next, a connection request is issued to the server, and a connection with the server is established (S199). Then, the ontology is requested by giving the identifier (S1).
910), and receives this and stores it in the cache (S191).
1), end the session with the server (S191)
2). At this time, if the reception ends normally (S191)
3) The process proceeds to S197 and ends.

On the other hand, if the connection with the server is not successful and the server notifies that the ontology cannot be provided, an alternative ontology is detected (S191).
4). The alternative ontology detection processing is as described above with reference to FIG. When the alternative ontology is detected, whether or not each alternative ontology hits the cache is checked again (S195), and the processing up to this point is repeated until ontology data is obtained. On the other hand, if the alternative ontology cannot be detected, an error occurs (S1916), and the process ends.

The above is the embodiment according to the present invention. In the above embodiment, the ontology is managed in a decentralized manner. However, regarding the meta-ontology, an ontology management that centrally manages the ontology is also conceivable.
That is, a dedicated meta-ontology server that exists only on the network is provided, and each ontology provider always registers a meta-ontology relating to its own provided ontology in the dedicated server. According to this management method,
Since all the relationships between the ontologies available on the network are obtained, the possibility of discovering without yielding a useful ontology is increased. On the other hand, there is also a problem that the search space for the alternative ontology is widened and the calculation cost is high because many ontologies unrelated to the purpose are included.

According to the present invention, since the ontology is managed in a client-server type decentralized management mode, the ontology development with a high development cost can be performed in a distributed manner, and the ontology can be serviced as a public ontology. Thus, the performance of the knowledge processing system possessed by the user can be improved, and the provider can continue to improve and expand independently of the user.

Further, when an ontology is exchanged, the server may be down or not reachable,
Even if a problem such as interruption of data transmission occurs, a framework for providing an alternative ontology is provided, so that the user's knowledge processing system is not adversely affected.

Further, since there is a framework for abstracting the ontology in accordance with the nature of the provided ontology and the form in which the ontology is used, the ontology can be obtained with a minimum necessary communication amount without reducing the performance of the knowledge processing system. Can be exchanged.

The processing in the embodiment of the present invention can be realized by a computer-executable program, and the program can be realized as a computer-readable storage medium.

The storage media in the present invention include a magnetic disk, a floppy disk, a hard disk, and an optical disk (CD-ROM, CD-R, DVD).
Etc.), a magneto-optical disk (MO or the like), a semiconductor memory, or any other storage medium that can store a program and that can be read by a computer.

An OS (operation system) running on the computer, database management software, MW (middleware) such as a network, etc., according to the instructions of the program installed in the computer from the storage medium realizes the present embodiment. May be executed.

Further, the storage medium in the present invention is not limited to a medium independent of a computer, but also includes a storage medium in which a program transmitted via a LAN or the Internet is downloaded and stored or temporarily stored.

Further, the number of storage media is not limited to one, and the case where the processing in the present embodiment is executed from a plurality of media is also included in the storage media of the present invention, and the configuration of the media may be any configuration. Good.

The computer according to the present invention is:
The computer executes each process in the present embodiment based on a program stored in a storage medium, and may have any configuration such as an apparatus such as a personal computer or a system in which a plurality of apparatuses are connected to a network. Is also good.

The computer in the present invention is not limited to a personal computer, but also includes an arithmetic processing unit, a microcomputer, and the like included in information processing equipment, and generically refers to equipment and devices capable of realizing the functions of the present invention by a program. are doing.

[0126]

As described above, according to the present invention, standard metadata examined and managed and credited by a third party can be suitably shared and used, and server-client distributed management can be performed. .

[Brief description of the drawings]

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

FIG. 2 is a diagram illustrating a configuration of a first exemplary embodiment of the present invention.

FIG. 3 shows a result of extracting requirement information in a registration metadata examination unit according to the first embodiment of the present invention.

FIG. 4 is a flowchart showing the operation of the first embodiment of the present invention.

FIG. 5 is a diagram showing a configuration of a second exemplary embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of the second exemplary embodiment of the present invention.

FIG. 7 is an example of a partial ontology extracted from the standard ontology of the second embodiment of the present invention.

FIG. 8 is a diagram illustrating a configuration of a third exemplary embodiment of the present invention.

FIG. 9 is a flowchart showing the operation of the third embodiment of the present invention.

FIG. 10 shows the configuration of an ontology management system according to a fourth embodiment of the present invention.

FIG. 11 is a configuration example of an ontology database according to a fourth embodiment of the present invention.

FIG. 12 shows an example of a meta-ontology database according to a fourth embodiment of the present invention.

FIG. 13: Ontology connection of the fourth embodiment of the present invention

FIG. 14 is an illustration of the ontology server identification by the name server according to the fourth embodiment of the present invention.

FIG. 15 is a flowchart showing the operation of the ontology server according to the fourth embodiment of the present invention.

FIG. 16 is a flowchart illustrating a detection process of an alternative ontology according to the fourth embodiment of this invention.

FIG. 17 shows an example of an alternative ontology detection rule according to the fourth embodiment of the present invention.

FIG. 18 is a flowchart showing the operation of the ontology proxy according to the fourth embodiment of the present invention.

FIG. 19 is a flowchart showing the operation of the ontology using system according to the fourth embodiment of the present invention;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Ontology server 12 ... Standard ontology 13 ... Conversion part 14 ... Each ontology 15 ... Management part 16 ... Application 17 ... Multimedia data 18 ... Multimedia content 21 ... Registered metadata receiving part 22 ... Registered metadata examination part 23 ... Registration Metadata judging unit 24 ... Judgment result adding unit 25 ... Registered metadata publishing unit 26 ... Registered metadata feedback unit 27 ... Metadata database 101 ... Ontology server 102 ... Name server 103 ... Ontology proxy 104 ... Ontology processing system

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hideki Hirakawa 1st station, Komukai Toshiba-cho, Saiwai-ku, Kawasaki-shi, Kanagawa Prefecture Inside the Toshiba R & D Center (72) Inventor Miwako Doi Toshiba Komukai-shi, Kawasaki-shi, Kanagawa No. 1 town Toshiba R & D Center F-term (reference) 5B075 KK03 KK07 ND04

Claims (10)

[Claims] 1. Metadata receiving means for receiving metadata sent from a metadata developer through a network, metadata checking means for checking whether the metadata received by the metadata receiving means satisfies predetermined conditions, A metadata database that stores metadata determined to satisfy predetermined conditions by the metadata examination means, and distributes metadata stored in the metadata database through a network in response to a request from a user. 1. A data management apparatus comprising: a metadata publishing unit that performs metadata transmission; and a metadata determination unit that determines reliability of metadata to be distributed by the metadata disclosure unit. 2. The data management apparatus according to claim 1, further comprising feedback information receiving means for receiving information such as a defect or usefulness of the metadata. 3. The data management device according to claim 2, wherein said feedback data is used as judgment data of data judgment means. 4. A storage means for storing a standard ontology which has a correspondence relationship between ontologies of different metadata and stores information for converting between the ontologies as an ontology, and a meta-data received via a network. A data management device comprising: a distribution unit that distributes necessary conversion information from a standard ontology of the storage unit or a conversion result based on the standard ontology through a network for converting data into metadata in a requested format. . 5. When a metadata creator attempts to describe metadata in a certain format, it not only provides an ontology conforming to the format, but also refers to the ontology corresponding to the standard ontology when the ontology is referred to. 5. The data management apparatus according to claim 4, further comprising a providing unit that also provides information. 6. A method for receiving metadata sent from a metadata developer via a network, examining whether the received metadata satisfies predetermined conditions, and converting the metadata determined to satisfy predetermined conditions into metadata. Data stored in a database and, when the metadata stored in the metadata database is distributed through a network in response to a request from a user, determining reliability of the metadata to be distributed; Management method. 7. A storage means for storing an ontology having an identifier uniquely identified on a computer network; an ontology using system for designating an identifier and requesting the ontology stored in said storage means; A data management device comprising: an ontology server that extracts an ontology corresponding to a name from the storage unit; and a detection unit that detects an ontology that substitutes for the requested ontology. 8. The data management apparatus according to claim 1, further comprising an ontology proxy that manages transmission and reception of the ontology between the ontology using system and the ontology server. 9. A storage means for storing an ontology having an identifier uniquely identified on a computer network; an ontology using system for designating an identifier and requesting the ontology stored in said storage means; A data management device comprising: an ontology server that extracts an ontology corresponding to a name from the storage unit; and a conversion unit that reduces intermediate nodes of the ontology extracted by the ontology server and converts the intermediate node into an abstracted ontology. . 10. The data management apparatus according to claim 2, further comprising an ontology proxy that manages transmission and reception of the ontology between the ontology using system and the ontology server.