QUERYING OF DISTRIBUTED DATABASES USING NEUTRAL ONTOLOGY MODEL FOR QUERY FRONT END

• US 20080077598A1
• Filed: 07/04/2007
• Published: 03/27/2008
• Est. Priority Date: 09/27/2006
• Status: Active Grant

• Current Assignee: The Boeing Co.
• Original Assignee: The Boeing Co.
• Inventors: Yuan, Jun, Wilson, John G., Van Rossum, David A., White, James M., Herrmann, Joseph R., Wilmering, Timothy J.
• Granted Patent: US 7,693,900 B2
• Time in Patent Office: Days
• US Class Current: 707/100
• CPC Class Codes: G06F 16/367 Ontology

Abstract

According to an embodiment, a method includes constructing a neutral ontology model of a query front end characterized by ontology schemata which subsume the plurality of different databases on the network in order to provide a common semantic interface for use in generating queries for data from any of the different databases, importing respective database metadata representing logical and physical structures of each database subscribed for receiving queries for data from the database using the query front end, constructing mappings of the database metadata representing the logical and physical structures of each subscribed database to the ontology schemata of the query front end, and storing the constructed mappings for use by the query front end for queries through the common semantic interface of the neutral ontology model for data from any of the different databases.

Claims

1. A method for enabling queries for data from any of a plurality of different databases on a network comprising:

constructing a neutral ontology model of a query front end characterized by ontology schemata which subsume the plurality of different databases on the network in order to provide a common semantic interface for use in generating queries for data from any of the different databases;

importing respective database metadata representing logical and physical structures of each database subscribed for receiving queries for data from the database using the query front end;

constructing mappings of the database metadata representing the logical and physical structures of each subscribed database to the ontology schemata of the query front end; and

storing the constructed mappings for use by the query front end for queries through the common semantic interface of the neutral ontology model for data from any of the different databases.

2. A method according to claim 1, wherein the constructing of the neutral ontology model includes:

collecting data that will be represented in the databases with corresponding data structure definitions to be used as a basis for conceptual ontology terminology used in the neutral ontology model;

analyzing data relationships of the databases for the conceptual ontology terminology;

defining an initial ontology in agreement with all included databases; and

developing a lexical ontology of the conceptual ontology terminology to be used for the common semantic interface to the included databases.

3. A method according to claim 1, wherein the importing of the database metadata includes:

physical models identifying the layout of data in each database;

logical models identifying conceptual relationships within the data of each database; and

data dictionaries providing textual descriptions of the defined data elements.

4. A method according to claim 1, wherein the constructing of mappings includes:

mappings between ontology and lower level schemata elements;

mappings between ontological queries and lower level queries; and

mappings between local security policies and global security policies.

5. A method according to claim 1, wherein the constructed mappings of database metadata to ontology schemata are stored in a repository used by the query front end, and at run time mapped ontology relationships are combined with controlling logical assertions in a forward chaining manner to infer connections between conceptual query terminology and the data structures of the subscribed databases on the fly.

6. A method according to claim 1, wherein the constructing of mappings includes storing:

metadata modeling each database to be mapped to the ontological schemata for the neutral ontology model;

mappings correlating database metadata to the neutral ontology model; and

any changes to the data content of the databases that must be performed in order for the mappings to be accurately represented by the neutral ontology model.

7. A method according to claim 1, wherein the constructing of mappings includes steps of importing local metadata for each of the subscribed databases, importing the ontology schemata for the neutral ontology model, and defining mappings between the local database metadata and the ontology schemata.

8. A method according to claim 1, wherein the stored mappings of the query front end are used by a semantic query engine to translate semantic queries in high-level semantic terminology into low level database queries, and transform retrieved data from the low level queries into instantiations of the neutral ontology model.

9. A computerized software-programmed system for accessing data from a plurality of different databases on a network comprising:

a query front end based on a neutral ontology model characterized by ontology schemata which subsume the plurality of different databases on the network in order to provide a common semantic interface for use in generating queries for data from any of the different databases;

a repository for storing mappings of respective database metadata representing logical and physical structures of each database to the ontology schemata of the query front end; and

a semantic query interface which uses the mappings stored in the repository to translate queries in high-level semantic terminology into low level database queries, and transform data retrieved from respective databases in response to the low level queries into instantiations of data retrieval within the neutral ontology model.

10. A system according to claim 9, wherein the system is implemented as a Web Service that allows integration into net-centric database architectures.

11. A system according to claim 9, wherein the query front end includes a graphical user interface operable with a system architecture adaptation of a model-view-controller (MVC) pattern.

12. A system according to claim 9, wherein the query front end includes a mediator component that coordinates interactions between system components to process queries.

13. A system according to claim 9, wherein COTS adapters are used for interfacing between system components and commercial-off-the-shelf (COTS) components in order to enable any COTS component to be adapted and interfaced to the system.

14. A system according to claim 9, having a model mapping and transformation tool providing an offline metadata repository for storing:

metadata modeling each database to be mapped to the ontological schemata for the neutral ontology model;

mappings correlating database metadata to the neutral ontology model; and

any changes to the data content of the databases that must be performed in order for the mappings to be accurately represented by the neutral ontology model.

15. A system according to claim 9, having a mapping management tool providing a graphical interface and associated functions for automating the following tasks:

importing local metadata of the subscribed databases;

importing the ontology schemata for the neutral ontology model; and

enabling graphically defined mappings between local database metadata and the ontology schemata for the neutral ontology model.

16. The system of claim 15, wherein the graphically defined mappings are provided by a user.

17. A system according to claim 13, wherein each COTS adapter is configured to be generic to COTS components as clients, and is provided with an interface to the associated internal system component and a method-invoking-mechanism for passing generic arguments as method calls to an associated COTS component.

18. A system according to claim 17, wherein said method-invoking-mechanism employs a command pattern to represent arguments as objects that execute marshalling and un-marshalling methods on themselves, such that when a COTS client acts to call the adapter interface, said method-invoking-mechanism first builds an AdapterParameter that is sent to the interface via a method call, the interface calls the marshalling method on the parameter, and the AdapterParameter implementation marshals its data into a form the COTS client can receive, and after processing by the COTS client results in a data return, an AdapterReturn object is passed back to the COTS client, and the un-marshalling method is called to un-marshall the command pattern.

19. A system according to claim 18, wherein the AdapterParameter is also used as a formal command object which contains not only parameters and means for marshalling them, but also means to execute an adapter usage pattern specifically for a particular COTS client, such that the adapter usage pattern is encapsulated into the command object and has no impact on the adapter interface or the COTS client.

20. A method of configuring a software adapter for use in interfacing between an internal computerized system component and an external commercial-off-the-shelf (COTS) software product comprising:

providing the adapter with an interface to the internal system component and a method-invoking-mechanism for passing generic arguments as method calls to the associated COTS component as a client, and

employing a command pattern of the method-invoking-mechanism to represent arguments as objects that execute marshalling and un-marshalling methods on themselves, such that when the COTS client acts to call the adapter interface, the method-invoking-mechanism first builds an AdapterParameter that is sent to the interface via a method call, the interface calls the marshalling method on the parameter, and the AdapterParameter implementation marshals its data into a form the COTS client can receive, and after processing by the COTS client results in a data return, an AdapterReturn object is passed back to the COTS client, and the un-marshalling method is called to un-marshall the command pattern.

21. A method according to claim 20, wherein the AdapterParameter is also used as a formal command object which contains not only parameters and means for marshalling them, but also information and/or means to execute an adapter usage pattern specifically for the particular COTS client, such that the adapter usage pattern is encapsulated into the command object and has no impact on the adapter interface or the COTS client.