Temporal Document Sorter and Method Using Semantic Decoding and Prediction

• US 20090049018A1
• Filed: 08/14/2008
• Published: 02/19/2009
• Est. Priority Date: 08/14/2007
• Status: Abandoned Application

• Current Assignee: John Nicholas and Kristin Gross Trust U/A/D April 13 2010
• Original Assignee: John Nicholas and Kristin Gross Trust U/A/D April 13 2010
• Inventors: Gross, John Nicholas
• Application Number:
  US12/191,973
• Publication Number:
  US 20090049018A1
• Time in Patent Office: Days
• 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

Electronic documents are classified and compared according to their temporal qualities. The content of a document relating to an event is analyzed to identify temporal components, including semantic counterparts. These components can be compared with corresponding counterparts in other documents to identify a relative temporal order, and/or to generate predictions of future news stories. The invention can be used in environments such as automated news aggregators, search engines, and other electronic systems which compile information having temporal qualities.

Claims

1. A method of automatically sorting electronic documents by their temporal characteristics with a computing system comprising:

a) identifying at least a first event having first content embodied in a first electronic document with the computing system;

b) analyzing said first electronic document with the computing system to identify first temporal components;

wherein said first temporal components relate to a state of said first event;

c) determining first semantic variants of said first temporal components with the computing system;

d) analyzing a second electronic document with the computing system to identify second content related to said at least first event;

e) comparing said second content to identify counterparts of said first temporal components and/or said first semantic variants;

f) generating an output identifying which of said first and second electronic documents contains content describing a more current state of said first event.

2. The method of claim 1 wherein said first semantic variants include semantic variations of words and phrases.

3. The method of claim 1 wherein said first semantic variants are derived from either both WORDNET and/or FRAMENET based semantic tools.

4. The method of claim 1 wherein additional semantic variants are generated based on said second content, and said additional semantic variants are also used for comparison with said first temporal components during step (e).

5. The method of claim 1 wherein said methods are implemented as one or more computer software routines embodied in a tangible media and adapted to cause a computing system to perform the operations recited therein.

6. A method of automatically sorting electronic documents by their temporal characteristics with a computing system comprising:

a) identifying at least a first event having first content embodied in a first electronic document with the computing system;

b) analyzing said first electronic document with the computing system to identify a first set of temporal components;

wherein said first set of temporal components relate to a first state of said first event;

c) generating a second predicted set of temporal components relating to a different predicted state of said first event with the computing system;

d) generating a search query with the computing system to locate additional electronic documents with content relating to said different state of said first event.

7. The method of claim 6 wherein said different predicted state is a future state of said event.

8. The method of claim 6 wherein said additional electronic documents are used to train an automatic document classifier.

9. The method of claim 6 wherein said second predicted set of temporal components are used to derive a target set of query sources for locating said additional electronic documents.

10. The method of claim 6 wherein a predicted locale of said different predicted state for said event is also calculated.

11. The method of claim 10 wherein said predicted locale is used to filter and select a set of local content sources for locating said additional electronic documents.

12. The method of claim 6 wherein said methods are implemented as one or more computer software routines embodied in a tangible media and adapted to cause a computing system to perform the operations recited therein.

18. The method of claim 10 wherein said predicted locale is used to filter and select a set of local content sources for locating said additional electronic documents.

13. A method of identifying electronic news stories with a computing system comprising:

a) identifying at least a first event having first content embodied in a first news story with the computing system;

b) analyzing said first news story with the computing system to identify a first state of said first event and a first locale associated with said event;

c) generating a second different predicted state of said first event with the computing system;

d) collecting content from news sources located geographically proximate to said first locale to identify additional news stories related to said second different predicted state of said first event.

14. The method of claim 13 wherein a predicted locale of said different predicted state for said event is also calculated.

15. The method of claim 13 wherein said second different predicted state is a future state of said event.

16. The method of claim 13 wherein said second predicted state of said first event is used to derive a target set of query sources for locating additional electronic documents related to said first event.

17. The method of claim 13 wherein a predicted locale of said different predicted state for said event is also calculated.

19. The method of claim 13 wherein said methods are implemented as one or more computer software routines embodied in a tangible media and adapted to cause a computing system to perform the operations recited therein.