Disambiguating data using contextual and historical information

• US 9,218,568 B2
• Filed: 03/15/2013
• Issued: 12/22/2015
• Est. Priority Date: 03/15/2013
• Status: Active Grant

• Current Assignee: Business Objects Incorporated (SAP SE)
• Original Assignee: Business Objects Incorporated (SAP SE)
• Inventors: Burbank, JohnPaul, Sam, Jennifer, Tsumura, Michael, Sun, George, Wang, Yuru
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): Misir, Dave
• Application Number:
  US13/843,513
• Publication Number:
  US 20140279803A1
• Time in Patent Office: 1,012 Days
• Field of Search: 706 46- 47
• US Class Current: 1/1
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

Techniques for disambiguating data using contextual and historical information include determining that data is potentially associated with two or more classifications of a plurality of classifications; obtaining contextual information associated with the data; obtaining historical information associated with a user of the computing system; and determining that the data is more likely associated with one classification of the two or more classifications than other classifications of the two or more classifications based on the contextual information and the historical information.

Claims

1. A method performed with a computing system for disambiguating data, the method comprising:

determining that data is potentially associated with two or more classifications of a plurality of classifications;

obtaining contextual information associated with the data;

obtaining historical information for a user associated with the data;

determining at least one accuracy modifier associated with an information type of the contextual information and the historical information, the at least one accuracy modifier indicating an accuracy of at least one previous use of the information type to resolve a data classification ambiguity;

employing the at least one accuracy modifier as at least one weight for the contextual information and the historical information;

determining at least one weighted score based on the at least one accuracy modifier; and

determining that the data is more likely associated with one classification of the two or more classifications than other classifications of the two or more classifications, the determining based at least partly on a maximum of the at least one weighted score.

2. The method of claim 1, wherein determining that data is potentially associated with two or more classifications of a plurality of classifications comprises:

determining that the data is potentially associated with a first classification of the two or more classifications based on one or more rules associated with the first classification; and

determining that the data is potentially associated with a second classification of the two or more classifications based on one or more rules associated with the second classification.

3. The method of claim 1, wherein:

obtaining contextual information associated with the data comprises;

obtaining an image containing text,

performing optical character recognition on the image to identify the text, and

identifying one or more terms of the text as describing at least one of a location, a date, a time, a monetary value, a product, or a service associated with the data; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the one or more terms of the text is associated with the one classification.

4. The method of claim 1, wherein:

obtaining contextual information associated with the data comprises;

determining geographical coordinates of a computing device, and

obtaining information identifying a location associated with the data based on the geographical coordinates; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the location is associated with the one classification.

5. The method of claim 1, wherein:

obtaining historical information associated with the user of the computing system comprises, for each classification of the two or more classifications, obtaining a set of one or more terms previously identified by the user as being associated with the classification; and

determining that the data is more likely associated with one classification than other classifications comprises determining that one or more terms of the data is included in a set of one or more terms previously identified by the user as being associated with the one classification.

6. The method of claim 1, wherein:

obtaining historical information associated with the user of the computing system comprises obtaining information identifying at least one of people, places, or things that the user has interacted with in the past; and

determining that the data is more likely associated with the one classification than other classifications comprises determining that the information identifying at least one of people, places, or things that the user has interacted with in the past is associated with the one classification.

7. A non-transitory computer readable storage medium encoded with a computer program, the program comprising instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

determining that data is potentially associated with two or more classifications of a plurality of classifications;

obtaining contextual information associated with the data;

obtaining historical information associated with a user;

determining at least one accuracy modifier associated with an information type of the contextual information and the historical information, the at least one accuracy modifier indicating an accuracy of at least one previous use of the information type to resolve a data classification ambiguity;

employing the at least one accuracy modifier as at least one weight for the contextual information and the historical information;

determining at least one weighted score based on the at least one accuracy modifier; and

determining that the data is more likely associated with one classification of the two or more classifications than other classifications of the two or more classifications, the determining based at least partly on a maximum of the at least one weighted score.

8. The non-transitory computer readable storage medium of claim 7, wherein determining that data is potentially associated with two or more classifications of a plurality of classifications comprises:

determining that the data is potentially associated with a first classification of the two or more classifications based on one or more rules associated with the first classification; and

determining that the data is potentially associated with a second classification of the two or more classifications based on one or more rules associated with the second classification.

9. The non-transitory computer readable storage medium of claim 7, wherein:

obtaining contextual information associated with the data comprises;

obtaining an image containing text,

performing optical character recognition on the image to identify the text, and

identifying one or more terms of the text as describing at least one of a location, a date, a time, a monetary value, a product, or a service associated with the data; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the one or more terms of the text is associated with the one classification.

10. The non-transitory computer readable storage medium of claim 7, wherein:

obtaining contextual information associated with the data comprises;

determining geographical coordinates of a computing device, and

obtaining information identifying a location associated with the data based on the geographical coordinates; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the location is associated with the one classification.

11. The non-transitory computer readable storage medium of claim 7, wherein:

obtaining historical information associated with the user comprises, for each classification of the two or more classifications, obtaining a set of one or more terms previously identified by the user as being associated with the classification; and

determining that the data is more likely associated with one classification than other classifications comprises determining that one or more terms of the data is included in a set of one or more terms previously identified by the user as being associated with the one classification.

12. The non-transitory computer readable storage medium of claim 7, wherein:

obtaining historical information associated with the user comprises obtaining information identifying at least one of people, places, or things that the user has interacted with in the past; and

determining that the data is more likely associated with the one classification than other classifications comprises determining that the information identifying at least one of people, places, or things that the user has interacted with in the past is associated with the one classification.

13. A system of one or more computers configured to perform operations comprising:

determining that data is potentially associated with two or more classifications of a plurality of classifications;

obtaining contextual information associated with the data;

obtaining historical information associated with a user;

determining at least one accuracy modifier associated with an information type of the contextual information and the historical information, the at least one accuracy modifier indicating an accuracy of at least one previous use of the information type to resolve a data classification ambiguity;

employing the at least one accuracy modifier as at least one weight for the contextual information and the historical information;

determining at least one weighted score based on the at least one accuracy modifier; and

determining that the data is more likely associated with one classification of the two or more classifications than other classifications of the two or more classifications, the determining based at least partly on a maximum of the at least one weighted score.

14. The system of claim 13, wherein determining that data is potentially associated with two or more classifications of a plurality of classifications comprises:

determining that the data is potentially associated with a first classification of the two or more classifications based on one or more rules associated with the first classification; and

determining that the data is potentially associated with a second classification of the two or more classifications based on one or more rules associated with the second classification.

15. The system of claim 13, wherein:

obtaining contextual information associated with the data comprises;

obtaining an image containing text,

performing optical character recognition on the image to identify the text, and

identifying one or more terms of the text as describing at least one of a location, a date, a time, a monetary value, a product, or a service associated with the data; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the one or more terms of the text is associated with the one classification.

16. The system of claim 13, wherein:

obtaining contextual information associated with the data comprises;

determining geographical coordinates of a computing device, and

obtaining information identifying a location associated with the data based on the geographical coordinates; and

determining that the data is more likely associated with one classification than other classifications comprises determining that the location is associated with the one classification.

17. The system of claim 13, wherein:

obtaining historical information associated with the user comprises, for each classification of the two or more classifications, obtaining a set of one or more terms previously identified by the user as being associated with the classification; and

determining that the data is more likely associated with one classification than other classifications comprises determining that one or more terms of the data is included in a set of one or more terms previously identified by the user as being associated with the one classification.

18. The system of claim 13, wherein:

obtaining historical information associated with the user comprises obtaining information identifying at least one of people, places, or things that the user has interacted with in the past; and

determining that the data is more likely associated with the one classification than other classifications comprises determining that the information identifying at least one of people, places, or things that the user has interacted with in the past is associated with the one classification.