Expanding concept types in conceptual graphs

• US 8,463,808 B2
• Filed: 11/07/2008
• Issued: 06/11/2013
• Est. Priority Date: 11/07/2008
• Status: Active Grant

• Current Assignee: Forcepoint Federal Holdings LLC (Forcepoint LLC)
• Original Assignee: Raytheon Company (Rtx Corporation)
• Inventors: Peoples, Bruce E., Johnson, Michael R., Smith, Jonathon P.
• Primary Examiner(s): Lee, Wilson
• Assistant Examiner(s): BUI, THUY T
• Application Number:
  US12/266,671
• Publication Number:
  US 20100287179A1
• Time in Patent Office: 1,677 Days
• Field of Search: 707/706, 707/736, 707/999.003, 707/999.01, 707/769
• US Class Current: 707/769
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence

Abstract

In one embodiment, searching documents includes generating a query conceptual graph for a search query. The query conceptual graph has graph terms representing concept types. A set of conceptually similar terms is identified for each graph term, and the set is mapped to the concept type represented by the graph term. Documents are searched in accordance with the set of conceptually similar terms. In another embodiment, selecting a document includes determining documents that satisfy a search query expressed as a query conceptual graph. A document conceptual graph that has graph terms representing concept types is generated for each document. A set of conceptually similar terms is identified for each graph term, where the set is mapped to the concept type represented by the graph term. A document is selected if terms of the sets associated with the document match terms associated with the query conceptual graph.

Claims

1. A method comprising:

generating a query conceptual graph for a search query, the query conceptual graph comprising;

a plurality of graph terms, a graph term representing a concept type;

a plurality of conceptual relation nodes that each comprise text describing a relationship between two or more of the plurality of graph terms; and

a plurality of arcs that each indicate a direction of the relationship described by the text of a corresponding conceptual relation node of the plurality of conceptual relation nodes;

identifying at least one set of conceptually similar terms for at least one graph term of the query conceptual graph, the at least one set of conceptually similar terms mapped to the corresponding concept type represented by the respective at least one graph term;

searching a plurality of documents in accordance with a search query based at least in part on the query conceptual graph and on the set of conceptually similar terms;

selecting a document from the plurality of documents as a potential match to the search query if, when the query conceptual graph is compared to a document conceptual graph for the document, there is at least one corresponding term associated with the query conceptual graph that matches at least one corresponding term associated with the document conceptual graph, wherein the corresponding term associated with each respective conceptual graph comprises at least one or more of;

terms representing a concept type, and terms that are conceptually similar to the terms representing concept types;

accessing one or more onomasticons, each onomasticon comprising a list of one or more mappings, each mapping comprising a mapping of a predetermined concept type to a corresponding set of conceptually similar terms; and

validating that the potential match is a valid match if a set of conceptually similar terms of the query conceptual graph, and a set of conceptually similar terms of the document conceptual graph that is associated with the potentially matching document, map to the same concept type in one or more onomasticons.

2. The method of claim 1, the generating the query conceptual graph for the search query further comprising:

generating the query conceptual graph in accordance with one or more query terms of the search query.

3. The method of claim 1, the identifying the at least one set of conceptually similar terms for the at least one graph term further comprising:

determining a semantic sense for a search query concept of the search query; and

identifying the at least one set of conceptually similar terms in accordance with the semantic sense.

4. The method of claim 1, the identifying the at least one set of conceptually similar terms for the at least one graph term further comprising:

determining a plurality of conceptually similar term options in accordance with a semantic sense for a search query concept of the search query; and

identifying the at least one set of conceptually similar terms from the plurality of conceptually similar term options.

5. The method of claim 1:

the at least one set of conceptually similar terms comprising at least one foreign language term comprising a foreign language translation of a native language term conceptually similar to the search query.

6. A system comprising:

a computer system comprising a processor and a memory, the memory configured to store a query conceptual graph for a search query, the query conceptual graph comprising;

a plurality of graph terms, a graph term representing a concept type;

a plurality of conceptual relation nodes that each comprise text describing a relationship between two or more of the plurality of graph terms; and

a plurality of arcs that each indicate a direction of the relationship described by the text of a corresponding conceptual relation node of the plurality of conceptual relation nodes; and

logic stored in one or more tangible media and when executed by the computer system, the logic is configured to;

generate the query conceptual graph;

identify at least one set of conceptually similar terms for at least one graph term of the query conceptual graph, the at least one set of conceptually similar terms mapped to the corresponding concept type represented by the respective at least one graph term; and

search a plurality of documents in accordance with a search query based at least in part on the query conceptual graph and on the set of conceptually similar terms;

select a document from the plurality of documents as a potential match to the search query if, when the query conceptual graph is compared to a document conceptual graph for the document, there is at least one corresponding term associated with the query conceptual graph that matches at least one corresponding term associated with the document conceptual graph, wherein the corresponding term associated with each respective conceptual graph comprises at least one or more of;

terms representing a concept type, and terms that are conceptually similar to the terms representing concept types;

access one or more onomasticons, each onomasticon comprising a list of one or more mappings, each mapping comprising a mapping of a predetermined concept type to a corresponding set of conceptually similar terms; and

validate that the potential match is a valid match if a set of conceptually similar terms of the query conceptual graph, and a set of conceptually similar terms of the document conceptual graph that is associated with the potentially marching document, maps to the same concept type in one or more onomasticons.

7. The system of claim 6, the logic configured to generate the query conceptual graph for the search query by:

generating the query conceptual graph in accordance with one or more query terms of the search query.

8. The system of claim 6, the logic configured to identify the at least one set of conceptually similar terms for the at least one graph term by:

determining a semantic sense for a search query concept of the search query; and

identifying the at least one set of conceptually similar terms in accordance with the semantic sense.

9. The system of claim 6, the logic configured to identify the at least one set of conceptually similar terms for the at least one graph term by:

determining a plurality of conceptually similar term options in accordance with a semantic sense for a search query concept of the search query; and

identifying the at least one set of conceptually similar terms from the plurality of conceptually similar term options.

10. The system of claim 6:

the at least one set of conceptually similar terms comprising at least one foreign language term comprising a foreign language translation of a native language term conceptually similar to the search query.

11. A method comprising:

determining a plurality of documents that satisfy a search query expressed as a query conceptual graph;

performing the following for each document of the plurality of documents to identify one or more sets of conceptually similar terms for the each document;

generating a document conceptual graph for each document, the document conceptual graph comprising;

a plurality of graph terms, a graph term representing a concept type;

a plurality of conceptual relation nodes that each comprise text describing a relationship between two or more of the plurality of graph terms; and

a plurality of arcs that each indicate a direction of the relationship described by the text of a corresponding conceptual relation node of the plurality of conceptual relation nodes; and

identifying one or more sets of conceptually similar terms for at least one graph term of the document conceptual graph, the one or more sets of conceptually similar terms mapped to the concept type represented by the at least one graph term;

selecting at least one document from the plurality of documents as a potential match to the search query if, when the query conceptual graph is compared to a document conceptual graph for the document, there is at least one corresponding term associated with the query conceptual graph that matches at least one corresponding term associated with the document conceptual graph, wherein the corresponding term associated with each respective conceptual graph comprises at least one or more of;

terms representing a concept type, and terms that are conceptually similar to the terms representing concept types;

accessing one or more onomasticons, each onomasticon comprising a list of one or more mappings, each mapping comprising a mapping of a predetermined concept type to a corresponding set of conceptually similar terms; and

validating that the potential match is a valid match if a set of conceptually similar terms of the query conceptual graph, and a set of conceptually similar terms of the document conceptual graph that is associated with the potentially matching document, maps to the same concept type in one or more onomasticons.

12. The method of claim 11, further comprising:

validating a match between a subset of the one or more sets of conceptually similar terms of the document and the one or more terms associated with the query conceptual graph.

13. The method of claim 11, further comprising:

initiating translation of at least one document from a foreign language to a native language of the search query.

14. The method of claim 11, the identifying the one or more sets of conceptually similar terms for the each graph term further comprising:

determining a semantic sense for a search query concept of the search query; and

identifying the one or more sets of conceptually similar terms in accordance with the semantic sense.

15. The method of claim 11, the identifying of the one or more sets of conceptually similar terms for the at least one graph term further comprising:

identifying the one or more sets of conceptually similar terms using the one or more onomasticons.

16. A system comprising:

a computer system comprising a processor and memory, the memory configured to store a search query expressed as a query conceptual graph; and

logic stored in one or more tangible media and when executed by the computer system, the logic is configured to;

determine a plurality of documents that satisfy the search query;

perform the following for each document of the plurality of documents to identify one or more sets of conceptually similar terms for each document;

generate a document conceptual graph for each document, the document conceptual graph comprising;

a plurality of graph terms, a graph term representing a concept type;

a plurality of conceptual relation nodes that each comprise text describing a relationship between two or more of the plurality of graph terms; and

a plurality of arcs that each indicate a direction of the relationship described by the text of a corresponding conceptual relation node of the plurality of conceptual relation nodes;

identify one or more sets of conceptually similar terms for each graph term of the document conceptual graph, the one or more sets of conceptually similar terms mapped to the concept type represented by each graph term;

select a document as a potential match if one or more terms of the one or more sets of conceptually similar terms for the document match one or more terms associated with the query conceptual graph;

access one or more onomasticons, each onomasticon comprising a list of one or more mappings, each mapping comprising a mapping of a predetermined concept type to a corresponding set of conceptually similar terms; and

validate that the potential match is a valid match if one or more of the sets of conceptually similar terms a set of conceptually similar terms of the query conceptual graph, and a set of conceptually similar terms of the document conceptual graph that is associated with the potentially matching document, maps to the same concept type in one or more onomasticons.

17. The system of claim 16, the logic further configured to:

select the validate the potential match between a subset of the one or more sets of conceptually similar terms of the document and the one or more terms associated with the query conceptual graph.

18. The system of claim 16, the logic further configured to:

initiate translation of at least one document from a foreign language to a native language of the search query.

19. The system of claim 16, the logic configured to identify the one or more sets of conceptually similar terms for the each graph term by:

determining a semantic sense for a search query concept of the search query; and

identifying the one or more sets of conceptually similar terms in accordance with the semantic sense.

20. The system of claim 16, the logic further configured to:

identify the one or more sets of conceptually similar terms for the each graph term by identifying the one or more sets of conceptually similar terms using the one or more onomasticons.