Method For Optimizing And Executing A Query Using Ontological Metadata

• US 20080256026A1
• Filed: 10/16/2007
• Published: 10/16/2008
• Est. Priority Date: 10/17/2006
• Status: Abandoned Application

• Current Assignee: Modus Operandi, Inc.
• Original Assignee: Modus Operandi, Inc.
• Inventors: Hays, Michael Glen
• Application Number:
  US11/873,137
• Publication Number:
  US 20080256026A1
• Time in Patent Office: Days
• US Class Current: 707/2
• CPC Class Codes: G06F 16/242 Query formulation
• CPC Class Codes: G06F 16/245 Query processing

Abstract

A method is provided for optimizing a query. The method includes providing metadata, and inputting an initial query including at least one initial class. The method further includes processing the initial query with the metadata. Additionally, the method includes obtaining an optimized query based on the processing of the initial query, where the optimized query provides at least one subsequent class based on the at least one initial class.

Claims

1. A method for optimizing a query, comprising:

providing metadata;

inputting an initial query;

processing the initial query with the metadata; and

obtaining an optimized query based on said processing of the initial query, said optimized query providing at least one subsequent class based on said at least one initial class.

2. The method of claim 1, wherein said optimized query further provides a respective physical table location of said at least one subsequent class within a respective data source.

3. The method of claim 2, wherein said metadata comprises an upper level ontology language including a plurality of classes and data to link said at least one subsequent class within said upper level ontology to said respective physical table within said respective data source.

4. The method of claim 2, wherein said metadata comprises an upper level ontology language including zero classes and data, said metadata being provided to develop at least one database.

5. The method of claim 3, wherein said upper level ontology language comprises at least one ontological relationship between said plurality of classes, wherein one of said classes is said initial class within said initial query.

6. The method of claim 3, wherein said processing comprises:

parsing said initial query into said at least one initial class and at least one initial attribute of said initial class;

identifying said subsequent class as an ontological equivalent of each initial class based upon said upper level ontology language of said metadata, said subsequent class having said respective physical table location within said respective data source; and

identifying at least one attribute of said subsequent class, said at least one attribute based upon said at least one initial attribute.

7. The method of claim 5, wherein said processing comprises utilizing said at least one ontological relationship of said upper level ontology language to convert said initial query into said optimized query comprising a plurality of queries, said plurality of queries each including said at least one subsequent class linked to said respective physical table location within said at least one data source.

8. The method of claim 7, wherein said processing converts a language of said initial query into a language of said optimized query, such that each of said queries language is compatible with a language of said respective data source having said respective physical table of the respective class.

9. The method of claim 8, wherein said initial query is provided in a SPARQL language, said optimized query is provided in a SQL language to be compatible with a SQL data source

10. A method for executing an optimized query, said optimized query based on processing an initial query with metadata, said method comprising:

providing said optimized query, said optimized query including at least one subsequent class and a respective physical table location of said at least one subsequent class within a respective data source;

providing an interface layer to access said respective data source;

obtaining data of said at least one subsequent class from said respective physical table location within said respective data source; and

returning a data result based on said optimized query.

11. The method of claim 10, further comprising:

requerying each data from said data result of said optimized query against said at least one physical table location to filter out data which fails to satisfy the optimized query; and

returning a final data result set in response to said optimized query.

12. The method of claim 10, wherein said at least one subsequent class includes at least one respective attribute included within said initial query, said obtaining data includes obtaining data of each respective attribute from said physical table location of said data source for each subsequent class.

13. The method of claim 12, wherein said returning said data result comprises comparing said data of each attribute of each subsequent class with a filter included within said optimized query, said comparing for eliminating data which fails to satisfy said optimized query.

14. The method of claim 11, wherein said requerying comprises querying each attribute data of said subsequent class with said respective physical table location to eliminate attribute data of said subsequent class which fails to satisfy said optimized query.

15. A method for executing a query, comprising:

parsing the query into a syntax tree;

identifying an initial class of said query within said syntax tree;

identifying an ontological equivalent class of said initial class, said ontological equivalent class having a physical table located within a data source;

identifying an attribute of said ontological equivalent class, said attribute having data located within said physical table;

determining if a remaining initial class requires identification of an ontological equivalent class;

obtaining said attribute data for an ontological equivalent class from said physical table within said data source;

appending said attribute data for said ontological equivalent class to a result group;

determining if a remaining ontological equivalent class requires the obtaining of the attribute data; and

returning said result group in response to said query.

16. The method of claim 15, further comprising:

requerying said result group by comparing each attribute data for each ontological equivalent class in said result group with said respective physical table location to eliminate attribute data of said ontological equivalent class which fails to satisfy said optimized query.