Intelligent data agent for a knowledge management system

• US 9,020,879 B2
• Filed: 10/22/2012
• Issued: 04/28/2015
• Est. Priority Date: 10/22/2012
• Status: Active Grant

• Current Assignee: Bank of America Corp.
• Original Assignee: Bank of America Corp.
• Inventors: McClung, Susan, Hofmeister, Michael K.
• Primary Examiner(s): Gaffin, Jeffrey
• Application Number:
  US13/656,927
• Publication Number:
  US 20140114902A1
• Time in Patent Office: 918 Days
• Field of Search: 706/47, 706/50, 706/59
• US Class Current: 706/50
• CPC Class Codes: G06F 16/367 Ontology
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Abstract

A system includes a memory operable to store one or more classification rules. The system also includes a processor communicatively coupled to the memory. The processor is operable to retrieve one or more data elements from a data source. The processor is further operable to identify a structured data element among the one or more data elements. The processor is further operable to parse the structured data element using one or more filter processes to produce a plurality of tokens. The processor is further operable to classify the plurality of tokens based at least in part on the one or more classification rules and an ontology, which includes a plurality of concepts and a plurality of relationships between the concepts. The processor is further operable to generate a knowledge assertion, which includes the plurality of classified tokens and one or more relationships between the classified tokens.

Claims

1. A system, comprising:

a memory operable to store one or more classification rules; and

a processor communicatively coupled to the memory and operable to;

retrieve one or more data elements from a data source;

identify a structured data element among the one or more data elements;

parse the structured data element using one or more filter processes to produce a plurality of tokens;

classify the plurality of tokens based at least in part on the one or more classification rules and an ontology, the ontology comprising a plurality of concepts and a plurality of relationships between the concepts;

identify a conflict between a first classified token and a second classified token;

resolve the conflict by evaluating the first and second classified tokens based at least in part on the ontology; and

generate a knowledge assertion comprising the plurality of classified tokens and one or more relationships between the classified tokens.

2. The system of claim 1, wherein:

the ontology further comprises concept statistics associated with the plurality of concepts; and

the processor is further operable to resolve the conflict by evaluating the first and second classified tokens based at least in part on the concept statistics.

3. The system of claim 1, wherein the processor is further operable to:

generate data source statistics for the data source; and

retrieve one or more updated data elements from the data source according to an update schedule, the update schedule based at least in part on the data source statistics.

4. The system of claim 1, wherein the processor is further operable to:

identify an unstructured data element among the one or more data elements;

parse the unstructured data element to produce a second plurality of tokens; and

return the second plurality of tokens and source information for inclusion in a search index.

5. The system of claim 4 wherein the source information comprises one or more of:

a pointer to the data source; and

a pointer to the unstructured data element.

6. The system of claim 1 wherein the processor is further operable to:

receive a query comprising a particular classified token; and

generate a result comprising;

one or more additional classified tokens generated based on information retrieved from the data source; and

one or more relationships between the additional classified tokens, wherein each additional classified token has at least one relationship with the particular classified token.

7. A method, comprising:

retrieving one or more data elements from a data source;

identifying a structured data element among the one or more data elements;

parsing the structured data element using one or more filter processes to produce a plurality of tokens;

classifying, by a processor, the plurality of tokens based at least in part on one or more classification rules and an ontology, the ontology comprising a plurality of concepts and a plurality of relationships between the concepts;

identifying a conflict between a first classified token and a second classified token;

resolving the conflict by evaluating the first and second classified tokens based at least in part on the ontology; and

generating a knowledge assertion comprising the plurality of classified tokens and one or more relationships between the classified tokens.

8. The method of claim 7, wherein:

the ontology further comprises concept statistics associated with the plurality of concepts; and

resolving the conflict by evaluating the first and second classified tokens is based at least in part on the concept statistics.

9. The method of claim 8, further comprising:

generating data source statistics for the data source; and

retrieving one or more updated data elements from the data source according to an update schedule, the update schedule based at least in part on the data source statistics.

10. The method of claim 7, further comprising:

identifying an unstructured data element among the one or more data elements;

parsing the unstructured data element to produce a second plurality of tokens; and

returning the second plurality of tokens and source information for inclusion in a search index.

11. The method of claim 10, wherein the source information comprises one or more of:

a pointer to the data source; and

a pointer to the unstructured data element.

12. The method of claim 7, further comprising:

receiving a query comprising a particular classified token; and

generating a result comprising;

one or more additional classified tokens generated based on information retrieved from the data source; and

one or more relationships between the additional classified tokens, wherein each additional classified token has at least one relationship with the particular classified token.

13. One or more non-transitory computer readable storage media embodying logic that is operable when executed by one or more processors to:

retrieve one or more data elements from a data source;

identify a structured data element among the one or more data elements;

parse the structured data element using one or more filter processes to produce a plurality of tokens;

classify the plurality of tokens based at least in part on one or more classification rules and an ontology, the ontology comprising a plurality of concepts and a plurality of relationships between the concepts;

identify a conflict between a first classified token and a second classified token;

resolve the conflict by evaluating the first and second classified tokens based at least in part on the ontology; and

generate a knowledge assertion comprising the plurality of classified tokens and one or more relationships between the classified tokens.

14. The storage media of claim 13, wherein:

the ontology further comprises concept statistics associated with the plurality of concepts; and

the logic is further operable when executed to resolve the conflict by evaluating the first and second classified tokens based at least in part on the concept statistics.

15. The storage media of claim 13, wherein the logic is further operable when executed to:

generate data source statistics for the data source; and

retrieve one or more updated data elements from the data source according to an update schedule, the update schedule based at least in part on the data source statistics.

16. The storage media of claim 13, wherein the logic is further operable when executed to:

identify an unstructured data element among the one or more data elements;

parse the unstructured data element to produce a second plurality of tokens; and

return the second plurality of tokens and source information for inclusion in a search index.

17. The storage media of claim 16, wherein the source information comprises one or more of:

a pointer to the data source; and

a pointer to the unstructured data element.

18. The storage media of claim 13, wherein the logic is further operable when executed to:

receive a query comprising a particular classified token; and

generate a result comprising;

one or more additional classified tokens generated based on information retrieved from the data source; and

one or more relationships between the additional classified tokens, wherein each additional classified token has at least one relationship with the particular classified token.