Apparatus and method for federated querying of unstructured data

• US 20070203893A1
• Filed: 02/27/2006
• Published: 08/30/2007
• Est. Priority Date: 02/27/2006
• Status: Abandoned Application

• Current Assignee: Business Objects Incorporated (SAP SE)
• Original Assignee: Business Objects SA (SAP SE)
• Inventors: Cras, Jean-Yves, Hassenforder, Marcel, Krinsky, Anthony, Chevrier, Marc
• Application Number:
  US11/364,564
• Publication Number:
  US 20070203893A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/33 Querying
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A computer readable medium is configured to receive a query, to map the query to an unstructured data source, to dispatch a request based on the query to the unstructured data source, to aggregate data returned by the unstructured data source in a structured data store, and to issue the query against the structured data store.

Claims

1. A computer readable medium comprising executable instructions to:

receive a query;

map said query to an unstructured data source;

dispatch a request based on said query to said unstructured data source;

aggregate data returned by said unstructured data source in response to said request in a structured data store; and

issue said query against said structured data store.

2. The computer readable medium of claim 1 wherein said query is in a Structured Query Language (SQL) format.

3. The computer readable medium of claim 1 wherein said unstructured data source includes textual objects included in electronic mail documents, word processing documents, and web pages.

4. The computer readable medium of claim 1 wherein the executable instructions to map include executable instructions to:

create a simplified query based on said query;

parse said simplified query; and

select said unstructured data source based on said simplified query.

5. The computer readable medium of claim 4 further comprising executable instructions to generate an execution plan for said simplified query that resolves dependencies of said simplified query on said unstructured data source, wherein said request is based on said execution plan.

6. The computer readable medium of claim 4 wherein the executable instructions to select said unstructured data source include executable instructions to:

read metadata describing capabilities of said unstructured data source; and

determine compatibility between said unstructured data source and said simplified query.

7. The computer readable medium of claim 4 wherein an input parameter of said unstructured data source is a value parsed from said simplified query.

8. The computer readable medium of claim 6 wherein said data returned by said unstructured data source is a row set including a parameter column and a data column.

9. The computer readable medium of claim 8 wherein said parameter column is based on said input parameter.

10. The computer readable medium of claim 8 wherein said data returned by said unstructured data source further includes a system column with a value independent of said database query.

11. The computer readable medium of claim 4, further comprising executable instructions to:

find dependencies of said simplified query on said unstructured data source;

generate candidate execution plans that resolve said dependencies;

select a lowest cost execution plan from said candidate execution plans; and

use said lowest cost execution plan to obtain said data returned by said unstructured data source.

12. The computer readable medium of claim 11 wherein the executable instructions to generate include executable instructions to:

create a first set of candidate execution plans that resolves a first set of said dependencies; and

create a second set of candidate execution plans that resolves a second set of said dependencies by building on said first set of candidate execution plans.

13. The computer readable medium of claim 11 wherein the executable instructions to generate include executable instructions to determine candidate data sources available to said candidate execution plans, wherein each of said candidate data sources resolves at least one of said dependencies.

14. The computer readable medium of claim 11 wherein the executable instructions to select include executable instructions to determine the aggregate cost of each of said candidate execution plans.

15. The computer readable medium of claim 14 wherein said aggregate cost is based on cost of a method call, and on cardinality of said method call.

16. A computer readable medium comprising executable instructions to:

receive a query;

map said query to a structured data source and an unstructured data source;

dispatch requests based on said query, including a first request to said structured data source and a second request to said unstructured data source;

aggregate data returned by said structured data source and said unstructured data source in response to said requests in a structured data store;

issue said query against said structured data store.

17. The computer readable medium of claim 16 wherein said requests are processed in parallel by said structured data source and said unstructured data source.

18. The computer readable medium of claim 16 wherein the executable instructions to map include executable instructions to factor said query into components including a first query to be applied to said structured data source and a second query to be applied to said unstructured data source.

19. The computer readable medium of claim 18 further comprising executable instructions to generate an execution plan for said second query that resolves dependencies of said second query on said unstructured data source, wherein said second request to said unstructured data source is based on said execution plan.

20. The computer readable medium of claim 19 wherein the executable instructions to generate include executable instructions to:

generate candidate execution plans that resolve said dependencies;

select a lowest cost execution plan from said candidate execution plans; and

use said lowest cost execution plan to obtain said data returned by said unstructured data source.