Methods and apparatus for focusing search results on the semantic web

• US 8,315,998 B1
• Filed: 03/18/2004
• Issued: 11/20/2012
• Est. Priority Date: 04/28/2003
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Raytheon BBN Technologies Corp. (Rtx Corporation), Verizon Corporate Services Group Incorporated (Verizon Communications Inc.)
• Inventors: Dean, Michael Anthony
• Primary Examiner(s): Pham, Hung Q
• Application Number:
  US10/803,551
• Time in Patent Office: 3,169 Days
• Field of Search: 707/5, 707/3, 707/705, 707/715
• US Class Current: 707/705
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

Methods and apparatus for searching Semantic Web structured resources and presenting the search results to a user can increase the likelihood that the search can produce meaningful results. An index database can be constructed by parsing literal values from statements from the resources to obtain component words of the literal values, with the index relating the component words to the statements from which they were parsed. A search engine can obtain component words from the database that match search terms in a query prepared by the user and can obtain the related statements for the matching words. A servlet or other application using the search engine and database can obtain predicates, instances, types of said instances, and full literal values of said instances for the related statements. The search results can be presented to the user as a summary of the predicates, instances, types, and full literal values.

Claims

1. A method of obtaining search results, comprising:

identifying, in a computing device that includes a processor and a memory, tagged statements in at least one Semantic Web structured resource, wherein the tagged statements each include subject/object/predicate triples;

parsing the tagged statements from the at least one Semantic Web structured resource to identify component words;

constructing an index from said component words, said index relating said component words to said tagged statements;

comparing said component words to a search term to identify matching words;

identifying related ones of said tagged statements for said matching words based on said index;

obtaining predicates, instances, types of said instances, and literal values of said related ones of said tagged statements, wherein each of the predicates, instances, types of said instances, and literal values is found in at least one of the at least one Semantic Web Structured resources; and

summarizing said obtained predicates, instances, types of said instances, and literal values for presentation to a user as summarized search results;

receiving user input selecting a summarized subject of interest; and

displaying a list of predicates associated with the selected summarized subject of interest.

2. The method of claim 1, wherein summarizing comprises:

arranging said predicates, instances, types, and literal values into one or more graphical representations; and

grouping said one or more graphical representations according to at least one of said types and said literal values.

3. The method of claim 1, comprising:

identifying Semantic Web structured resources to obtain identified Semantic Web structured resources;

gathering statements from said identified Semantic Web structured resources to obtain gathered statements;

presenting said gathered statements for parsing of said gathered statements;

wherein constructing an index comprises updating said index based on the parsing of said gathered statements; and

wherein said identifying, gathering and presenting are iteratively performed.

4. The method of claim 3, wherein summarizing comprises:

arranging said predicates, instances, types, and literal values into one or more graphical representations; and

grouping said one or more graphical representations according to at least one of said types and said literal values.

5. A system for obtaining search results for a query prepared by a user, comprising:

at least one parser identifying tagged statements in at least one Semantic Web structured resource, wherein the tagged statements each include subject/object/predicate triples, and receiving statements from Semantic Web structured resources and identifying component words of said statements;

constructing an index from said component words, said index relating said component words to at least one of said tagged statements;

a database for storing said index;

a search engine for matching search terms of said query to said component words to obtain matched words, said search engine identifying said tagged statements related to said matched words;

a servlet for obtaining predicates, instances, types of said instances, and literal values of said tagged statements related to said matched words, wherein each of the predicates, instances, types of said instances, and literal values is found in at least one of the Semantic Web Structured resources; and

an object viewer for summarizing said obtained predicates, instances, types of said instances, and literal values for presentation to said user as summarized search results, receiving user input selecting a summarized subject of interest, and displaying a list of predicates associated with the selected summarized subject of interest.

6. The system of claim 5, wherein said object viewer comprises:

means for arranging said predicates, instances, types, and literal values into one or more graphical representations; and

means for grouping said one or more graphical representation according to at least one of said types and said literal.

7. The system of claim 5, comprising:

means for identifying Semantic Web structured resources to obtain identified Semantic Web structured resources;

means for gathering statements from said identified Semantic Web structured resources to obtain gathered statements;

means for presenting said gathered statements for parsing of said gathered statements;

means for iteratively invoking said means for identifying, said means for gathering and said means for presenting, and

wherein said processor comprises means for updating said index based on the parsing of said gathered statements.

8. The system of claim 7, wherein said object viewer comprises:

means for arranging said predicates, instances, types, and literal values into one or more graphical representations; and

means for grouping said graphical representations according to at least one of said types and said literal values.

9. A computer program, disposed on a computer readable medium, for enabling searching of and presentation of search results from Semantic Web structured resources, said computer program including instructions for causing a processor to:

identifying tagged statements in at least one Semantic Web structured resource, wherein the tagged statements each include subject/object/predicate triples;

parse said tagged statements from the at least one Semantic Web structured resource to identify component words;

construct an index from said component words, said index f said component words to said tagged statements;

compare said component words to a search term to identify matching words;

identify related ones of said tagged statements for said matching words based on said index;

obtain predicates, instances, types of said instances, and literal values of said related ones of said tagged statements, wherein each of the predicates, instances, types of said instances, and literal values is found in at least one of the at least one Semantic Web Structured resources; and

summarize said obtained predicates, instances, types of said instances, and literal values for presentation to a user as summarized search results;

receiving user input selecting a summarized subject of interest; and

displaying a list of predicates associated with the selected summarized subject of interest.

10. The computer program of claim 9, wherein said instructions for causing a processor to summarize further comprise instructions for causing a processor to:

arrange said predicates, instances, types, and literal values into one or more graphical representations; and

group said one or more graphical representations according to at least one of said types and said literal values.

11. The computer program of claim 9, wherein said instructions further comprise instructions for causing a processor to:

identify Semantic Web structured resources to obtain identified Semantic Web structured resources;

gather statements from said identified Semantic Web structured resources to obtain gathered statements according to tags associated with each statement, wherein the tagged statements each include subject/object/predicate triples;

present said gathered statements for parsing of said gathered statements;

wherein said instructions for causing a processor to construct an index comprise instructions for causing a processor to update said index based on the parsing of said gathered statements; and

wherein said instructions for causing a processor to identify, gather and present comprise instructions for causing a processor to iteratively identify, gather and present.

12. The computer program of claim 11, wherein said instructions for causing a processor to summarize further comprise instructions for causing a processor to:

arrange said predicates, instances, types, and literal values into one or more graphical representations; and

group said graphical representations according to at least one of said types and said literal values.

13. A method, comprising:

identifying, in a computing device that includes a processor and a memory, tagged statements in a plurality of Semantic Web structured resources, wherein the tagged statements each include subject/object/predicate triples;

parsing the tagged statements from the Semantic Web structured resources to identify component words, wherein, for each Semantic Web structured resource, there are a plurality of tagged statements, and for each tagged statement there are a plurality of component words;

constructing a non-hierarchical index from said component words, said index relating said component words to said tagged statements;

comparing said component words to a search term to identify matching words;

identifying related ones of said tagged statements for said matching words based on said index;

obtaining predicates, instances, types of said instances, and literal values of said related ones of said statements, wherein each of the predicates, instances, types of said instances, and literal values is found in at least one of the at least one Semantic Web Structured resources; and

summarizing said obtained predicates, instances, types of said instances, and literal values for presentation to a user as summarized search results;

receiving user input selecting a summarized subject of interest; and

displaying a list of predicates associated with the selected summarized subject of interest.

14. The method of claim 13, wherein at least one of the Semantic Web structured resources has been converted from non-web data.

15. The method of claim 14, wherein the non-web data includes at least one of a database table and a database extract.