Temporal document sorter and method

• US 10,762,080 B2
• Filed: 08/14/2008
• Issued: 09/01/2020
• Est. Priority Date: 08/14/2007
• Status: Active Grant

• Current Assignee: John Nicholas and Kristin Gross Trust U/A/D April 13 2010
• Original Assignee: John Nicholas and Kristin Gross Trust
• Inventors: Gross, John Nicholas
• Primary Examiner(s): Mamillapalli, Pavan
• Application Number:
  US12/191,809
• Publication Number:
  US 20090049037A1
• Time in Patent Office: 4,401 Days
• Field of Search: 707787, 707725
• 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. These components can be compared with corresponding counterparts in other documents to identify a relative temporal order. 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;

wherein said first event is assigned by the computing system to a plurality of different event states each represented by a different temporal value within a temporal sequence defined for such first event;

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

wherein said first temporal components relate to a first state of said first event and include first temporal related data for said first state that is not any of a creation time alone, a publication time alone, or a combination of a creation time and publication time alone;

c) repeating step (b) to analyze a second electronic document with second content related to said at least first event with the computing system to identify second temporal components relating to a second state of said first event and a second temporal value;

d) comparing said first temporal value and said second temporal value to generate an output with the computing system identifying which of said first and second electronic documents contains content describing a more current state of said first event within said defined temporal sequence.

2. The method of claim 1 wherein said temporal components consist of selected text words and/or phrases determined by a natural language engine as identifying subjects and predicates for said first event in said first content and said second content.

3. The method of claim 1 further including the steps:

automatically locating and compiling a set of documents relating to said at least first event including from social network site postings; and

sorting said documents according to their temporal characteristics relating to said first event.

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

generating an alert in human detectable form in response to said output of step (d) to inform a human operator of potential new content including a new event state relating to said first event.

5. The method of claim 1, further including steps:

determining a first source of first electronic document and a second source of said second electronic document; and

attributing respective confidence rankings to each of said first source and said second source, which confidence rankings are used in determining which of said electronic documents contains content relating to a more current state of said first event.

6. The method of claim 1 wherein said current state relates to an absolute time associated with said event.

7. The method of claim 1 wherein said content corresponds to a first electronic incarnation by the computing system of previously unpublished factual information relating to said event.

8. The method of claim 1 wherein said content corresponds to a first electronic incarnation by the computing system of previously unpublished factual information relating to a current state of said event.

9. The method of claim 1, further including steps:

determining a geographic site where said first event is occurring; and

at least in part prioritizing collection of electronic documents relating to said first event from one or more selected sources geographically associated with said site.

10. The method of claim 9, further including a step:

weighting electronic documents from said one or more selected sources with a higher score relative to other sources reporting on said first event.

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

validating an existence of said more current state by comparing to a search first event state returned by one or more online electronic searches automatically generated and referencing said first event including from news outlet sources.

12. The method of claim 11 further including a step:

determining an awareness rate of said current state by a calculating a change in presence of such event state in content over time.

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

validating an existence of said more current state automatically with the computing system by comparing rankings provided by one or more online human reviewers of temporal values of said first and second electronic documents to temporal values determined by the computing system.

14. The method of claim 13, wherein said online human reviewers are drawn at least in part from social network site users located within a geographical region associated with said event.

15. The method of claim 13, wherein said online human reviewers are drawn primarily from mobile phone users within a controllable sized geographical region associated with said event.

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

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

predicting a subsequent event state for said first event with the computing system from said plurality of event states based on determining said current state.

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

adjusting an advertising auction pricing for keywords with a computing system based on a prediction of content associated with said predicted subsequent state for said first event.

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

selecting one of said first and second electronic documents with the computing system to present in a time based electronic data feed based on identifying which describes a more current state of said first event.

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

selecting one of said first and second electronic documents with the computing system in response to search engine query based on identifying which describes a more current state of said first event.

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

automatically generating a collection of reference documents with distinct temporal values spanning said temporal sequence for said first event using said first electronic document and said second electronic document.

22. The method of claim 1 wherein temporal components for said first event are generated to include a data construct with text snippets specifying an entity, event and location.

23. The method of claim 22 wherein said data construct for said temporal components also includes an age field and weighting value as a temporal predictor.

24. The method of claim 1 including a step:

wherein said temporal sequence is determined from a predictable progress pattern identified for events similar to said first event.

25. The method of claim 1, including a step of presenting said first electronic document within a web browser to a user with content referencing said first event, and automatically updating temporal data for said first event in response to said user selecting said first event.

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

a. with the computing system;

[i] generating a temporal sequence for a first event, said temporal sequence including a plurality of different event states for such first event having a corresponding temporal interpretation, based on processing content for a reference set of documents relating to said first event;

wherein the computing system further processes said reference set of documents to create a comparison set of documents representing distinct temporal interpretations related to said first event;

[ii] generating a first approximate temporal interpretation of a first electronic document relative to a state of said first event;

b. wherein said first approximate temporal interpretation is based on temporal characteristics from of a set of electronic training documents and include temporal characteristics for said state that are not any of a creation time alone, a publication time alone, or a combination of a creation time and publication time alone;

c. using said first approximate temporal interpretation to identify a comparison set of documents for said first electronic document;

d. generating a second temporal interpretation of said first electronic document based on said first approximate temporal interpretation and correlating such document with one or more second electronic documents having a third temporal interpretation value and taken from said comparison set.

27. The method of claim 26 wherein said steps are implemented as one or more computer software routines embodied in a tangible media and adapted to cause one or more computing systems to perform the operations recited therein.

28. The method of claim 26 wherein said comparison set of documents are analyzed by natural language processing to identify selected subject-predicate pairs corresponding to distinct temporal interpretations.

29. The method of claim 26 wherein said comparison set of documents are analyzed by natural language processing to identify text constructs that change in a predictable sequence in time.

30. The method of claim 26 wherein said comparison set of documents are analyzed by natural language processing to select and use text phrases having a form in which only a subset of words change dynamically in time.

31. The method of claim 26 including a step generating a confidence score for said second temporal interpretation indicating whether it exceeds a predetermined threshold.

32. The method of claim 26 including a step:

generating a plurality of comparison sets of documents with the computing system from a document corpus, which plurality of comparison sets of documents are assigned to a spectrum of temporal bins covering multiple topics.

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

a. providing a set of reference temporal interpretation values for a first reference event based on processing a set of reference documents;

wherein said first reference event is associated by the computing system with a plurality of different event reference states each represented by and corresponding to a different reference temporal interpretation value defined for such first reference event;

further wherein each of said reference temporal interpretation values is based on temporal components of content identified in the documents that are not any of a creation time alone, a publication time alone, or a combination of a creation time and publication time alone;

b. processing a first electronic document with the computing system to identify if content therein describes a second event matching said first reference event, and to identify a corresponding first initial coarse temporal interpretation value of said second event;

wherein said first initial coarse temporal interpretation value maps said first electronic document to a first temporal bin;

c. comparing said first electronic document with a second electronic document in the first temporal bin within the set of reference documents and having a known reference temporal interpretation value to determine a second reference temporal interpretation value of said second event;

d. generating a final temporal related value for said first electronic document within said first temporal bin based on said first initial coarse temporal interpretation value and said second reference interpretation value .

34. The method of claim 33 wherein said first electronic document is part of a batch of documents which are sorted into temporal order with respect to a particular second event based on a comparison to a temporal sequence previously determined for a first reference event.

35. The method of claim 33 wherein said steps are implemented as one or more computer software routines embodied in a tangible media and adapted to cause one or more computing systems to perform the operations recited therein.