JP4689636B2 - Ontology database update method and ontology database update system - Google Patents

Description

  The present invention relates to a configuration of an ontology DB (Database), and more particularly to an ontology DB update method and an ontology DB update system for updating an ontology DB based on metadata mapping.

  As a conventional ontology DB configuration technique, there is one in which operation metadata groups are categorized for each purpose, and the categories are linked bidirectionally based on the relationships between categories (equivalent relationships, inheritance relationships).

  OWL-S (Web Ontology Language for Services: see Non-Patent Document 1, for example) is applied to the description of the operation metadata here. OWL-S consists of three descriptions: Profile (Profile), Position Model (Process Model), and Grounding (Granding).

  The profile description describes what service is provided. In the Process Model (process model) description, an atomic process that expresses the operation of the service component in an abstract manner, and parameters of the atomic process are described. The Grounding description describes how to access the service, that is, mapping information between the Process Model description and the interface information of the actual operation. Accordingly, the atomic process described in the process model and the parameters of the process correspond to the operation metadata.

  From the above, the Process Model description is categorized for each purpose, and the ontology DB is configured by describing the equivalence relation or the inheritance relation between the categories in which the Process Model description is registered. This makes it possible to detect a Process Model description that matches the search condition by following the equivalence relation or the inheritance relation (see, for example, Non-Patent Document 2).

Further, as a metadata mapping technique, there is a technique that performs mapping between semantic information schemas of a plurality of systems by using names, class attributes, or accumulated past mapping results. In addition to the name, the class attributes and past mapping results can be used to improve accuracy. This makes it possible to share information among a plurality of systems individually constructed with different concepts (see, for example, Non-Patent Document 3).
OWL-S web site, [Search February 8, 2007], Internet <URL: http: //www.daml.org/services/owl-s/1.0/owl-s.html> Y. Yamato, Y. Tanaka, and H. Sunaga, “Context-aware Ubiquitous Service Composition Technology”, The IFIP Internal ional Conference on Research and Practical Issues of Enterprise Information Systems (CONFENIS2006), pp.51-61, Apr.2006 . Makoto Nakajo, Yu Miyoshi, Yasuyuki Kimura, “Proposal and Evaluation of Message Mapping Method Based on Semantic Information for Flexible System Cooperation”, Database Society of Japan Letter Vol. 4, no. 1, pp. 37-40, 2005

  However, in the above-described prior art (Non-Patent Document 2), the ontology DB is updated by the administrator, and there is no mention of automatic updating. Nor is it mentioned how to find equivalence relations. Further, in the above-described prior art (Non-Patent Document 3), there is no mention of updating the ontology DB by mapping.

  As described above, the ontology DB update technology based on the conventional ontology DB configuration and metadata mapping cannot update the ontology DB by mapping, and cannot update the ontology DB automatically. There's a problem.

  The present invention has been made in view of such circumstances, and an object thereof is an ontology database that can solve the above problems and can automatically update an ontology DB based on metadata mapping. An update method and an ontology database update system are provided.

  In order to solve the above-described problem, the present invention categorizes operation metadata representing the operation of a service component for each purpose and describes the relationship between categories between a plurality of category trees formed by describing relationships between categories. An ontology database update method for extracting equivalence relations and registering the equivalence relations, wherein a category tree selection step of selecting a plurality of category trees from which equivalence relations are extracted, and equivalences from each of the selected category trees Based on the calculation result of the category selection step for selecting a target category for extracting the relationship, the degree of approximation calculation step for calculating the degree of approximation between the selected categories, and the degree of approximation calculation step, the equivalence relationship between the categories is obtained. An equivalence relation extraction step to extract, Is ontology database update method characterized by including the equivalence relation registration step of registering an equivalence relationship category between extracted by the relationship extraction step in the category tree.

  According to the present invention, in the above invention, the approximation calculating step includes an approximation between process model names including a detailed description related to the operation of the service component, an approximation between atomic process names representing the operation metadata, and the atomic A part of the process is to calculate the degree of approximation between input parameter messages, the degree of approximation between output parameter messages, the degree of approximation between precondition messages, and the degree of approximation between effect messages. And

  According to the present invention, in the above invention, the equivalence relation extraction step has an equivalence relation when there are a plurality of types of parameters of the atomic process, and the degree of approximation between all corresponding parameter messages is higher than a predetermined value. It is characterized by being extracted as being.

  The present invention is characterized in that, in the above invention, the equivalence relation extracting step extracts the equivalence relation between the selected categories based on a past mapping result representing the equivalence relation between the categories.

  According to the present invention, in the above invention, the category tree selection step may be performed based on a category tree in which mapping is performed or a past mapping execution area indicating to which category of the category tree mapping is performed. It is characterized by selecting a tree.

  The present invention is characterized in that, in the above invention, the category tree selecting step selects the category tree in a background unrelated to a search in service provision.

  In order to solve the above-described problem, the present invention categorizes operation metadata representing the operation of a service component for each purpose and describes categories between a plurality of category trees formed by describing relationships between categories. An ontology database update system including a mapping engine that extracts equivalence relations between each other and registers the equivalence relations in the ontology database, wherein the ontology database selects a plurality of category trees from which equivalence relations are to be extracted. Selection means, and equivalence relation registration means for registering an equivalence relation between categories in the selected category tree extracted by the mapping engine in the category tree, and the mapping engine includes the ontology data A category selection unit that selects a target category from which an equivalence relationship is extracted from each of a plurality of category trees selected by a source, an approximation calculation unit that calculates an approximation between the selected categories, and an approximation calculation unit An ontology database update system comprising: equivalence relation extracting means for extracting equivalence relations between categories based on a calculation result.

  According to the present invention, in the above invention, the degree of approximation between process model names including a detailed description regarding the operation of the service component, the degree of approximation between atomic process names representing the operation metadata, and one item included in the atomic process. A certain degree of approximation between input parameter messages, an approximation degree between output parameter messages, an approximation degree between precondition messages, and an approximation degree between effect messages are calculated.

  According to the present invention, in the above invention, the equivalence relation extracting unit has an equivalence relation when there are a plurality of types of parameters of the atomic process and the degree of approximation between all corresponding parameter messages is higher than a predetermined value. It is characterized by being extracted as being.

  The present invention further comprises mapping result storage means for storing a mapping result representing an equivalence relation between category trees in the above invention, wherein the equivalence relation extraction means is stored in the mapping result storage means. Based on the mapping result, the equivalence relationship between the category trees is extracted.

  According to the present invention, in the above invention, the ontology database stores mapping execution area storage means for storing a past mapping execution area indicating a category tree to which mapping has been performed or to which category of the category tree the mapping has been executed. The category tree selecting means selects the category tree based on past mapping execution areas stored in the mapping execution area storage means.

  The present invention is characterized in that, in the above-mentioned invention, the category tree selecting means selects the category tree in a background unrelated to a search in service provision.

  According to the present invention, a plurality of category trees from which equivalence relationships are extracted are selected, a category from which the equivalence relationships are extracted are selected from the selected plurality of category trees, and the degree of approximation between the selected categories is determined. Based on the calculated degree of approximation, the equivalence relationship between categories is extracted, and the equivalence relationship between the extracted categories is registered in the category tree. Therefore, for the end user, the ontology DB is constantly updated, so that the range of service components that can be used is widened, and it is possible to obtain an advantage that a rich service can be used. Further, for the service provider, there is an advantage that the ontology DB can be easily managed even if the number of registered operation metadata is increased.

  In addition, according to the present invention, the degree of approximation between process model names including detailed descriptions of service component operations, the degree of approximation between atomic process names representing operation metadata, and an input parameter that is an item included in the atomic process The degree of approximation between messages, the degree of approximation between output parameter messages, the degree of approximation between precondition messages, and the degree of approximation between effect messages are calculated. Therefore, there is an advantage that the degree of approximation between the selected categories can be calculated more easily.

  Further, according to the present invention, when there are a plurality of types of parameters of the atomic process and the degree of approximation between all corresponding parameter messages is higher than a predetermined value, the parameters are extracted as having an equivalence relationship. Therefore, there is an advantage that the equivalence relationship between categories can be extracted more easily.

  Further, according to the present invention, based on the past mapping result representing the equivalence relation between the categories, the equivalence relation between the categories for extracting the equivalence relation between the selected categories is extracted. Therefore, there is an advantage that the equivalence relationship between categories can be extracted more easily.

  In addition, according to the present invention, the category tree is selected based on a past mapping execution area representing a mapping-executed area. Therefore, there is an advantage that the category tree can be selected more easily.

  In addition, according to the present invention, the category tree is selected in a background unrelated to the search. Therefore, there is an advantage that the category tree can be selected without imposing a burden on the system.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

(System configuration)
FIG. 1 is a block diagram showing a configuration example of an ontology DB (database) update system according to an embodiment of the present invention. In FIG. 1, the ontology DB update system includes a network 1, an ontology DB 2, a mapping engine 3, and a service component providing device 4. The ontology DB 2, the mapping engine 3, and the service component providing device 4 are connected to the network 1.

  The network 1 is a communication network such as telephone exchange, an information network such as the Internet, or a network such as a mobile network. The ontology DB 2 stores a category tree formed by describing the relationship between categories in which the Process Model description is registered. The ontology DB 2 activates the mapping engine 3 in order to extract the equivalence relation between the category trees, and registers the equivalence relation with the categories in the different category trees sent from the mapping engine 3 in the category tree. The ontology DB 2 matches this search condition when the service cooperation engine (not shown) searches for a process model having an atomic process equivalent to the operation metadata when providing service by linking service components. Has a function of detecting the Process Model description.

  The mapping engine 3 has a function of calculating the degree of approximation between categories included in each of the two category trees and extracting an equivalence relationship between the categories. The service component providing device 4 registers the Process Model description in the ontology DB 2, registers the Process Model description and the like in the service component information DB (not shown), and functions of the service component from the service cooperation engine (not shown). When an execution request for (operation) is received, the function is executed and the result is transmitted to the service cooperation engine.

  In FIG. 1, one service component providing apparatus 4 is shown for convenience of drawing, but a plurality of service component providing apparatuses 4 may exist.

(System function)
The function of the ontology DB update system according to the present embodiment is that the ontology DB 2 selects two category trees from which equivalence relations are extracted from the accumulated category trees, and maps the information of the selected two category trees. This is transmitted to the engine 3, receives the equivalence relation information between the categories in the two category trees sent from the mapping engine 3, and registers them in the category tree.

(Ontology DB update system component)
Next, the ontology DB update system component apparatus according to the present embodiment will be described.
FIG. 2 is a block diagram illustrating a configuration example of a configuration device in the ontology DB update system of FIG. FIG. 2 shows a configuration example of the ontology DB 2, the mapping engine 3, and the service component providing apparatus 4 according to the first embodiment.

  2, the ontology DB 2 includes a metadata storage unit 21, a category tree selection unit 22 that is a feature of the present invention, an equivalence relation registration unit 23 that is a feature of the present invention, and a mapping execution region that is a feature of the present invention. And an accumulating unit 24. The metadata accumulating unit 21 accumulates (stores) the category and the Process Model description registered therein including the relationship between the categories. Here, two categories, a category tree A211 and a category tree B212, describing the relationship between the categories are stored.

  The category tree selection unit 22 selects two category trees from which the equivalence relationship is extracted from the metadata storage unit 21 based on the approximation of the category name or the like (category tree selection means). At this time, the selection is made in the background unrelated to the search in the service provision. The equivalence relation registration unit 23 registers the equivalence relation between the category trees transmitted from the mapping engine 3 in the category tree of the metadata storage unit 21 (equivalence relation registration unit).

  Further, the mapping execution area storage unit 24 stores a mapping-performed area (mapping execution area) (mapping execution area storage means). The mapping-performed area stores the category tree in which mapping is performed or up to which category of the category tree. As a result, it is possible to easily find out which combination of category trees has not yet been implemented and which category tree has many unexecuted categories. In addition, it is possible to avoid duplicating what has already been done.

  The mapping engine 3 is an engine that characterizes the present invention, and includes a category selection unit 31, an approximation degree calculation unit 32, an equivalence relation extraction unit 33, and a mapping result accumulation unit 34. The category selection unit 31 accumulates information on the two category trees selected by the ontology DB 2, and sequentially selects combinations of all categories from which equivalence relations are extracted (category selection means). The category is supplied to the approximation calculation unit 32.

  The approximation calculation unit 32 calculates the approximation between categories selected by the category selection unit 31 (approximation calculation means), and supplies the calculated approximation to the equivalence relation extraction unit 33. Details will be described later. The equivalence relation extraction unit 33 extracts which category and which category have the equivalence relation based on the approximation degree calculated by the approximation degree calculation unit 32 (equivalence relation extraction unit). Details will be described later.

  The mapping result accumulation unit 34 accumulates the mapping result (mapping result accumulation unit). The mapping result includes information on a combination of categories automatically determined to be in an equivalence relationship and a combination of categories input by an administrator or the like offline. As a result, the determination can be made quickly and the same error can be prevented from being repeated.

  The service component providing apparatus 4 includes a category reference unit 41, a metadata creation unit 42, an operation 43, interface information 44, a grounding description 45, a process model description 46, and a profile description 47. The category reference unit 41 refers to the category information of the ontology DB 2 published on the network and the Process Model description registered therein.

  The metadata creation unit 42 creates a process model description and the like based on the reference result of the category reference unit 41. Operation 43 is a function provided by the service component. The interface information 44 describes the interface information of the operation 43. The Grounding description 45 describes how to access the service, that is, mapping information between the Process Model description 46 and the interface information 44.

  The Process Model description 46 describes the Atomic Process parameters and the Atomic Process parameters that abstractly represent what the operation 43 of the service component is, and is registered in the ontology DB 2. The profile description 47 describes what service is provided.

(Mapping process)
Next, the mapping process of the ontology DB update system will be described.
FIG. 3 is a conceptual diagram for explaining an example of mapping processing in the mapping engine 3 in FIG. The category tree selection unit 22 of the ontology DB 2 selects the category tree A 211 and the category tree B 212 from the category trees registered in the metadata storage unit 21, and the category selection unit 31 of the mapping engine 3 selects the category tree A 211. The category A2 is selected from the category B2, and the category B2 is selected from the category tree B212.

  FIG. 3A shows the Process Model description registered in the category A2. The Process Model name is AAA, indicating that it has two Atomic Process α and β. Here, Atomic Processβ will be described. Atomic Process β has a name, input parameters, output parameters, preconditions and effects. The name is β, and there are three types of input parameters: “location”, “price”, and “vehicle type”, and the output parameter has “reservation number”. , "Card" and, as an effect, "temporary transfer of usage rights".

  FIG. 3B shows a Process Model description registered in the category B2. The Process Model name is BBB, indicating that it has two atomic processes a and b. Here, Atomic Process b will be described. Atomic Process b has a name, input parameters, output parameters, preconditions, and effects, similar to Atomic Process β. The name is b, the input parameter has “location” and “cost”, the output parameter has “reception number”, and the precondition is “credit card” It shows that it has “delegation of usage right” as an effect.

  The approximation calculation unit 32 (1) the approximation between the process model names AAA and BBB, (2) the approximation between the atomic process names β and b, and (3) between the input parameter messages “location” and “location” Approximation degree, Approximation degree between “price” and “cost”, (4) Approximation degree between “reservation number” and “reception number” of the output parameter message, (5) Precondition message “card” and “credit card” (6) The degree of approximation between the effect messages “temporary transfer of usage rights” and “transfer of usage rights” is calculated. For calculating the degree of approximation, existing techniques such as the message mapping method described in Chapter 4 of Non-Patent Document 3 can be used. Non-Patent Document 3 describes a method for mapping between semantic information schemas of a plurality of systems and automatically mapping between message formats based on the result, mapping only by name, class attribute usage method, class attribute usage In addition to the method, there are a method of reusing the accumulated mapping result, a method of pattern mapping in addition to the method of using various class attributes. In the class attribute usage method, the degree of approximation between classes belonging to semantic information schemas of different systems is obtained from class attributes such as a class name, an instance set, and a class component.

  The equivalence relation extracting unit 33 weights the approximations calculated in the above (1) to (6) and superimposes (calculates a linear sum), and those having a higher score than a predetermined threshold for determining that they are equivalent. The equivalence relation is extracted by determining that the relation is equivalent. This weighting coefficient can be set arbitrarily. In addition, at this time, in the inheritance relationship in the category tree, between the categories where the upper categories and the lower categories are in the equivalence relationship, the determination of being in the equivalence relationship is enhanced (for example, to increase the score) Multiply a predetermined factor). The range of upper and lower categories to be considered in order to enhance the determination, that is, the number of steps from the category currently selected as the target for extracting the equivalence relationship is arbitrary, and the equivalence of the upper and lower categories of one step Only the relationship may be considered, or the equivalence relationship up to the upper and lower categories of two or more steps may be considered. When considering the equivalence relationship between the upper and lower categories of two or more steps, the contribution ratio may be changed to increase the determination according to the number of steps from the category currently selected as the target for extracting the equivalence relationship. . For example, the smaller the number of steps from the currently selected category to the higher / lower categories in the equivalence relationship, the larger the coefficient for increasing the score, etc. make low. Also, if there are multiple types of parameters, and all the approximations between all corresponding parameter messages are higher than a predetermined threshold value for determining that they are equivalent, it is determined that they are in an equivalent relationship and extracted. To do. In addition, when it is registered that the currently selected categories are in an equivalence relationship with reference to information registered offline or information of past determination examples based on the mapping result accumulation unit 34. These categories are extracted as equivalence relations.

(System operation)
Next, the operation of the ontology DB update system according to this embodiment will be described.
FIG. 4 is a sequence diagram showing an operation example of the ontology DB update system in FIG. The sequence between the mapping engine 3, ontology DB2, and service component providing apparatus 4 is shown.

  The service component providing apparatus 4 registers the Process Model description in the ontology DB 2 (Step S1). As a result, in the ontology DB2, several category trees including the Process Model registered by the service component providing device 4 are formed and registered in the metadata storage unit 21 (step S2). The category tree selection unit 22 of the ontology DB 2 refers to the mapping execution region described in the mapping execution region storage unit 24 (step S3), and the category that has not been mapped yet or whose equivalence relationship has not been determined is selected. Two category trees are selected from the many category trees, and information on the selected category trees is transmitted to the mapping engine 3 (step S4). Further, the category tree selection unit 22 updates the mapping execution region data in the mapping execution region storage unit 24 to indicate that the selected category tree has been mapped (step S5). In step S4, the selection is made in the background unrelated to the search in the service provision.

  When the category selection unit 31 of the mapping engine 3 stores the received category tree information therein (step S6), the category selection unit 31 of the categories included in each of the two category trees has not yet been selected. A combination is selected (step S7). The degree-of-approximation calculator 32 calculates the degree of approximation between the categories selected by the category selector 31 (step S8). In step S8, the degree of approximation between Process Model names, the degree of approximation between Atomic Process names, the degree of approximation between input parameter messages, the degree of approximation between output parameter messages, the degree of approximation between precondition messages, and effects Calculate part or all of the degree of approximation between messages.

  Next, the equivalence relation extraction unit 33 of the mapping engine 3 refers to the past mapping result stored in the mapping result accumulation unit 34, and whether or not the equivalence relation information of the currently selected categories is registered. In addition, it is detected whether there is information on equivalence relations between the upper and lower categories of the currently selected category (step S9). And if the information on the equivalence relation between the currently selected categories is registered, it is extracted as the equivalence relation. In addition, if the information on the equivalence relationship between the currently selected categories is not registered, the currently selected category is selected based on the equivalence relationship between the upper and lower categories and the degree of approximation calculated by the approximation degree calculating unit 32. It is determined whether or not certain categories have an equivalence relationship. In this way, a new equivalence relationship is extracted, and if it can be extracted, information on the equivalence relationship between the categories is sent to the ontology DB 2 (step S10). Further, the equivalence relation extracting unit 33 adds the equivalence relation information between the categories, and updates the mapping result data in the mapping result accumulating part 34 (step S11). If mapping for all categories has not been completed, the process returns to step S7 to select a new category (step S12). In step S10, when there are a plurality of types of parameters and the degree of approximation between all corresponding parameter messages is higher than a predetermined value, the parameters are extracted as having an equivalence relationship.

When the equivalence relation registration unit 23 of the ontology DB 2 receives the information of the equivalence relation of the category from the mapping engine 3, the equivalence relation registration information of the selected categories of the selected category tree is stored in the metadata accumulation unit 21. Is registered in the selected category tree (step S13).
For example, if it is determined that the category A2 in the category tree A211 and the category B2 in the category tree B212 shown in FIG. 3 are equivalent, the mapping result storage unit 34 stores the category A2 in the category tree A211 and the category tree Information that the category B2 in B212 is the same value is written, and the data of the category tree A211 in the metadata storage unit 21 in the ontology DB2 has the same value in the category A2 and the category B2 in the category tree B212. The information indicating that the category B2 and the category A2 in the category tree A211 have the same value is written in the data of the category tree B212.

  According to the above-described embodiment, the ontology DB can be automatically updated based on the metadata mapping.

  In addition, this invention is not limited to the form demonstrated using FIGS. 1-4, Various changes are possible in the range which does not deviate from the main point.

  The ontology DB 2, the mapping engine 3, and the service component providing device 4 described above have a computer system inside. The operation processes of the ontology DB 2, the mapping engine 3, and the service component providing device 4 are stored in a computer-readable recording medium in the form of a program, and the computer system reads and executes the program. The above processing is performed. The computer system here includes a CPU, various memories, an OS, and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  As described above, it is obvious that the present invention can be used industrially. That is, since the ontology DB is constantly updated for end users, the range of service components that can be used is expanded, and a rich service can be used. For service providers, the number of registered operation metadata is increased. Consistency management has become difficult as the number increases. However, since the ontology DB can be easily managed by automation, the service will expand and the demand for systems and related devices will increase further. I can expect.

It is a block diagram which shows the structural example of the ontology DB update system by embodiment of this invention. It is a block diagram which shows the structural example of the component apparatus in the ontology DB update system by this embodiment. It is a conceptual diagram for demonstrating the example of a mapping process in the mapping engine 3 by this embodiment. It is a sequence diagram which shows the operation example of the ontology DB update system by this embodiment.

Explanation of symbols

1 Network 2 Ontology DB
21 metadata storage unit 22 category tree selection unit (category tree selection means)
23 Equivalence Relationship Registration Unit (Equivalence Relationship Registration Unit)
24 Mapping execution area storage unit (mapping execution area storage means)
3 Mapping Engine 31 Category Selection Unit (Category Selection Unit)
32 Approximation degree calculation unit (approximation degree calculation means)
33 Equivalence relation extraction unit (equivalence relation extraction means)
34 Mapping result storage unit (mapping result storage means)
4 Service Component Providing Device 41 Category Reference Unit 42 Metadata Creation Unit 43 Operation 44 Interface Information 45 Grounding Description 46 Process Model Description 47 Profile Description

Claims (8)

Categorize operation metadata representing the operation of the service components for each purpose, extracts the equivalence relation between categories among a plurality of category trees that are formed by describing the relationship between categories, in order to register the equivalence relations And an ontology database update method in a system comprising category tree selection means, category selection means, approximation calculation means, equivalence relation extraction means, and equivalence relation registration means ,
A category tree selecting step in which the category tree selecting means selects a plurality of category trees from which equivalence relations are extracted;
A category selecting step in which the category selecting means selects a target category for extracting an equivalence relationship from each of the selected plurality of category trees;
The approximation degree calculation means is an item included in the atomic process, an approximation degree between process model names including a detailed description related to the operation of the service component, an approximation degree between atomic process names representing the operation metadata, The degree of approximation between the selected categories is calculated by calculating part or all of the degree of approximation between input parameter messages, the degree of approximation between output parameter messages, the degree of approximation between precondition messages, and the degree of approximation between effect messages. An approximation calculating step to calculate;
The equivalence relation extraction means weights and superimposes the approximation degree calculated in the approximation degree calculating step , and judges that the score is higher than a predetermined threshold value for judging that they are equivalent. An equivalence relationship extraction step for extracting equivalence relationships between categories,
The equivalence relationship registration means has an equivalence relationship registration step of registering an equivalence relationship between categories extracted in the equivalence relationship extraction step in the category tree ,
The equivalence relationship extraction step refers to a past mapping result representing an equivalence relationship between categories, determines whether or not equivalence relationship information between currently selected categories is registered, and is registered. If the registered equivalence relation information is extracted, if not registered, the equivalence relation information between the currently selected categories is added, and the mapping result information is updated. How to update.   In the equivalence relation extraction step, when there are a plurality of types of parameters of the atomic process and the degree of approximation between all corresponding parameter messages is higher than a predetermined value, the equivalence relation is extracted as being in equivalence relation. The ontology database update method according to claim 1.   The category tree selecting step selects the category tree based on a category tree on which mapping is performed or on a past mapping execution area indicating to which category of the category tree mapping is performed. The ontology database update method according to claim 1.   2. The ontology database update method according to claim 1, wherein the category tree selection step selects the category tree in a background unrelated to a search in service provision. Categorize operation metadata representing service component behavior for each purpose, extract equivalence relations between categories among multiple category trees formed by describing relations between categories, and store the equivalence relations in the ontology database An ontology database update system with a mapping engine to register,
The ontology database is
Category tree selection means for selecting a plurality of category trees from which equivalence relationships are to be extracted;
Equivalence relation registration means for registering the equivalence relation between the categories in the selected category tree extracted by the mapping engine in the category tree;
The mapping engine is
Category selection means for selecting a target category for extracting an equivalence relationship from each of a plurality of category trees selected by the ontology database;
Approximation between process model names including detailed description of the operation of the service component, approximation between atomic process names representing the operation metadata, and approximation between input parameter messages as one item included in the atomic process An approximation degree calculating means for calculating an approximation degree between the selected categories by calculating a degree of approximation between output parameter messages, an approximation degree between precondition messages, and a part or all of an approximation degree between effect messages ; ,
By weighting the degree of approximation calculated by the degree-of-approximation calculating means and superimposing and judging that the score is higher than a predetermined threshold value for determining that they are equivalent, Equivalence relation extracting means for extracting equivalence relations ,
The mapping engine is
Further comprising mapping result storage means for storing a mapping result representing an equivalence relationship between the category trees;
The equivalence relation extracting means includes:
With reference to the past mapping result representing the equivalence relationship between the categories accumulated in the mapping result accumulation means, it is determined whether or not the equivalence relationship information between the currently selected categories is registered, If registered, the information of the registered equivalence relation is extracted, and if not registered, the information of the equivalence relation between the currently selected categories is added, and the information of the mapping result is updated. Ontology database update system. The equivalence relation extracting means extracts a plurality of types of parameters of the atomic process, and if the degree of approximation between all corresponding parameter messages is higher than a predetermined value, the equivalence relation is extracted as having an equivalence relation. The ontology database update system according to claim 5 . The ontology database is
A mapping execution area accumulating means for accumulating a past mapping execution area indicating the mapping of the category tree in which mapping has been performed or to which category of the category tree has been implemented;
The category tree selection means includes:
6. The ontology database update system according to claim 5 , wherein the category tree is selected based on past mapping execution areas stored in the mapping execution area storage means. 6. The ontology database update system according to claim 5 , wherein the category tree selection means selects the category tree in a background unrelated to the search in service provision.

Images