Method of automatically verifying document content

• US 9,177,014 B2
• Filed: 07/01/2014
• Issued: 11/03/2015
• Est. Priority Date: 08/14/2007
• Status: Active Grant

• Current Assignee: John Nicholas and Kristin Gross Trust
• Original Assignee: John Nicholas and Kristin Gross Trust
• Inventors: Gross, John Nicholas
• Primary Examiner(s): Beausoliel, Jr., Robert
• Assistant Examiner(s): ALLEN, NICHOLAS E
• Application Number:
  US14/321,686
• Publication Number:
  US 20140316911A1
• Time in Patent Office: 490 Days
• Field of Search: 707/709
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/95 Retrieval from the web
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/0243 Comparative campaigns
• CPC Class Codes: G06Q 30/0251 Targeted advertisements

Abstract

A news and/or search engine system predicts events and stories based on existing events, stories, etc., extracted from a knowledge domain. These predictions can be used for a number of different purposes, including locating new content and actual event outcomes.

Claims

1. A method of verifying content of a news story with a computing system, comprising:

a) monitoring published items with the computing system in a knowledge domain including one or more social network domains, news content domains, blog domains and/or message board domains;

b) processing news content with the computing system in a first set of items published in said knowledge domain to extract at least a first set of events referenced in such news content;

c) identifying at least a first event in said first set of events described in a first content item published in said knowledge domain;

d) automatically verifying with the computing system if said first event in said first content item has occurred in fact;

wherein said automatically verifying is performed automatically in real-time by checking one or more verification sources by constructing and retrieving results from an automated query based on first content keywords extracted from said first content item; and

e) generating a verification rating for said first event indicating a result of said verification step (d).

2. The method of claim 1 wherein an automated check is performed during step (d) to determine if a news story identified in said first content item is legitimate.

3. The method of claim 1 wherein a news story is identified in said first content item and a timeliness of said first event relative to other events for said news story is automatically determined.

4. The method of claim 1, further including a step:

presenting said first content item as part of a news feed only when said verification rating exceeds a predetermined threshold.

5. The method of claim 1 further including a step:

automatically determining names of persons in said first content item.

6. The method of claim 5 further including a step:

determining a time based sequence of events for persons in said news content based on temporal decoding of said news content other than timestamps for such first set of items.

7. The method of claim 1 wherein step (d) is performed by automatically reviewing verification sources in said knowledge domain.

8. The method of claim 1 wherein step (d) is performed by automatically reviewing verification sources separate and outside said knowledge domain.

9. The method of claim 1 wherein said verification sources are processed and given a reliability rating for accuracy, and are selected at step (d) based on such reliability rating.

10. The method of claim 1 wherein said verification rating is provided in part by selected human sources that are also consulted during step (d).

11. The method of claim 9 wherein said selected human sources are determined by automatically detecting an event site identified in said first content item, and such that only selected human sources proximate to event site are consulted.

12. The method of claim 11 wherein speech input from said selected human sources is recognized and decoded for said verification.

13. The method of claim 11 wherein fixed length limited text messages provided by said selected human sources is used for said verification.

14. The method of claim 1 wherein step (d) is performed periodically by the computing system to determine a change in prevalence in said verification sources.

15. The method of claim 1 further including a step:

selecting and presenting an advertisement to a user reviewing said news content in an electronic interface, which advertisement is based on first content automatically detected for said first item and a site of said first event automatically detected.

16. The method of claim 1 further including a step:

selecting and presenting an advertisement to a user reviewing said news content in an electronic interface, which advertisement is based on first content automatically detected for said first item and a prediction of a mood for said user also determined automatically from said first content.

17. The method of claim 14 wherein said mood prediction is also based on characteristics of the user, including an identified affinity with entities identified in said first content.

18. The method of claim 1 further including a step:

predicting future content of future unpublished stories in said knowledge domain based on automatically processing said first event.

19. The method of claim 16 wherein said future content is determined by automatically identifying prior news stories containing said first event or similar events and further temporal sequenced content relating to later developments of such news stories.

20. The method of claim 16 wherein said future content is determined by a parser classifying said first event, and by referencing a template specifying a predetermined predictable change in content for said first event.

21. The method of claim 1 wherein said news content is in a news broadcast.

22. The method of claim 1 wherein said automated query also includes predicted keywords determined from phrases determined to be correlated to said first content keywords.

23. The method of claim 1 wherein said automated query also includes predicted keywords determined from phrases determined to be temporally related to said first content keywords.

24. The method of claim 1 wherein said automated query is repeated later with a second set of verification sources predicted to be relevant based on a progress predicted for said first set of events.

25. A method of verifying content of a news story with a computing system, comprising:

a) monitoring published items with the computing system in a knowledge domain including one or more social network domains, news content domains, blog domains and/or message board domains;

b) processing news content with the computing system in a first set of items published in said knowledge domain to extract at least a first set of events referenced in such news content;

c) identifying at least a first event in said first set of events described in a first content item published in said knowledge domain;

d) automatically verifying with the computing system if said first event in said first content item has occurred in fact;

wherein said automatically verifying is performed automatically in real-time by checking one or more verification sources;

e) generating a verification rating for said first event indicating a result of said verification step (d); and

f) automatically determining names of persons in said first content item and determining a time based sequence of events for the persons in said news content based on temporal decoding of said news content other than timestamps for such first set of items.

26. The method of claim 25 wherein an automated check is performed during step (d) to determine if a news story identified in said first content item is legitimate.

27. The method of claim 25 wherein a news story is identified in said first content item and a timeliness of said first event relative to other events for said news story is automatically determined.

28. The method of claim 25, further including a step:

presenting said first content item as part of a news feed only when said verification rating exceeds a predetermined threshold.