News aggregator and search engine using temporal decoding

• US 9,405,792 B2
• Filed: 08/14/2008
• Issued: 08/02/2016
• 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:
  US12/191,896
• Publication Number:
  US 20090055359A1
• Time in Patent Office: 2,910 Days
• Field of Search: 707/787, 707/725
• 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. 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. providing a set of temporal ordered electronic documents related to an event, which event is associated by the computing system with a plurality of corresponding event states defined for a progress template for such event;

wherein said event states are associated by the computing system with content distinctive to different temporal periods within said progress template;

b. wherein each of said set of temporal ordered electronic documents is associated with a temporal value established by the computing system, and wherein said temporal value is not any of a creation time alone, a publication time alone, or a combination of a creation time and a publication time alone;

c. comparing the contents of a separate first electronic document with the computing system to the contents of one or more electronic documents in said set of temporal ordered electronic documents to determine a closest content match document;

d. with the computing system, determining if said separate first electronic document and said closest content match document based are substantial duplicates;

filtering one of said separate first electronic document and said closest content match document when they are determined to be substantial duplicates;

comparing said first electronic document to said closest content match document when they are determined not to be duplicates to identify temporal differences with the computing system, including by comparing first content identifying a first state for the event in said first electronic document to second content identifying a second state for said event in said closest content match document;

e. generating a temporal related value for said first electronic document based on steps (c)-(d) with the computing system, which temporal value is derived from comparing said first state and said second state for said event and identifying their respective temporal periods within said progress template; and

f. adjusting a temporal order of said set of temporal ordered documents as needed with the computing system based on said temporal related value determined for said separate first electronic document.

2. The method of claim 1, wherein said closest content match is a subset or superset of content found in said first electronic document.

3. The method of claim 1 wherein during step (c) the computing system searches for and identifies subject/predicate pair counterparts of temporal components of said first electronic document located in said closest content match.

4. The method of claim 3 wherein said subject/predicate pair counterparts include semantic variants.

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

generating an ongoing master set of reference temporal components, which master set contains all distinctive temporal related content snippets from said set of ordered electronic documents.

6. The method of claim 1 further including steps:

g) moving said separate first electronic document into a temporal position held by said closest content match within said order of said set of temporal ordered documents; and

h) repeating steps (c) through (f) using one or more next highest ranked temporal documents until said separate first electronic document fails to achieve a temporal value greater than a next highest ranked temporal document.

7. The method of claim 1, wherein source data for said separate electronic document is also determined.

8. The method of claim 7, wherein said temporal value is adjusted based on said source data.

9. The method of claim 8 wherein said temporal value is adjusted based on a timeliness factor associated with said source.

10. The method of claim 8 wherein said temporal value is adjusted based on an accuracy factor associated with said source.

11. The method of claim 1, wherein site data relating to a location of said event is also determined automatically from said first electronic document by the computing system.

12. The method of claim 11, wherein said site data is used by the computing system to determine one or more sources of content for said event.

13. The method of claim 11 wherein said site data is used by the computing system to filter data from sources reporting on said event based on their distance to such site.

14. The method of claim 11 wherein said site data is used by the computing system to weight data from sources reporting on said event based on their distance to such site.

15. The method of claim 11 wherein said site data is used by the computing system to identify sources in a geographic region near said site suitable for verifying a temporal value of said separate first electronic document.

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

automatically soliciting input on content for said event by the computing system from persons located near said event as determined by said site data.

17. The method of claim 16 wherein said temporal value is adjusted by the computing system based on contributions computed from said persons located near said event.

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

verifying a temporal value of said separate first electronic document with the computing system from one or more verification sources separate from a source identified for said separate first electronic document.

19. The method of claim 18, wherein said verification sources include one of a search engine, a news source, an RSS feed, a web log, a social networking site, and/or an electronic message board.

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

21. 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 a predicted subsequent state for said event.

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

verifying said second state for the event by computing a prevalence rate for content identifying such second state in a set of benchmark sites and confirming said prevalence rate is within a particular range.

23. The method of claim 3 wherein said subject/predicate pairs are weighted by the computing system in accordance with their temporal significance to determine said temporal related value.

24. The method of claim 3 wherein subject/predicate pairs are linked to other subject/predicate pairs by the computing system to determine said temporal related value.

25. A method of causing a computing system to automatically present electronic documents in accordance with their temporal characteristics comprising:

a. automatically retrieving a set of one or more documents pertaining to a first event identified with the computing system;

selecting an event template for said first event, which event template includes a plurality of corresponding event states for said first event correlated and assigned to different temporal periods of such first event;

b. identifying first content embodied in a first document which characterizes a first state of such event using the computing system;

correlating said first state to one of said plurality of corresponding event states;

c. presenting said first electronic document within a presentation portion of a web page in human perceptible form using the computing system;

d. it analyzing a second electronic document with second content characterizing a second state of said first event using the computing system;

correlating said second state to one of said plurality of corresponding event states;

e. determining any temporal differences between said first document and said second document with respect to said first content and said second content using the computing system, including differences between said first state and second state within said event template; and

f. wherein after determining any temporal differences said computing system;

i. sorts said first electronic document and said second electronic document in a temporal order based on said event template with respect to their content describing said first event,

wherein said temporal order is based on content associated with said documents but is not any of creation time alone, publication time alone, or a combination of creation time and publication time alone;

ii. determines how such first electronic document and said second electronic document should be arranged to denote a temporal order within said presentation portion of said web page;

and further wherein duplicate versions of electronic documents describing a duplicate state of said first event can be selectively omitted from said presentation portion of a web page.

26. The method of claim 25, wherein said duplicate versions can be presented or omitted based on user selectable options presented within said web page.

27. The method of claim 25 wherein said steps are implemented as one or more computer software routines stored in a tangible media and adapted to cause one or more computing systems to perform the operations recited therein.

28. A method of processing a search query with an online computing system comprising:

a) receiving a search query relating to a first event and/or a first topic;

determining if said search query includes temporal parameters related to said first event and/or said first topic;

b) generating an initial set of search results, including an initial set of electronic documents, which are responsive to the search query with the online computing system;

c) based on steps (a) and (b) processing documents, including said initial set of electronic documents, to determine temporal characteristics of such documents with respect to said first event and/or said first topic;

wherein said temporal characteristics are associated with content in said documents pertaining to {object, object status} pairs but are not any of creation time alone, publication time alone, or a combination of creation time and publication time alone;

further wherein said processing determines said temporal characteristics in said documents by analyzing said {object, object status} pairs and assigning an object status for each object to a distinct temporal period; and

d) altering a composition and/or ordering of said initial set of search results based on said step (c) and returning such as a final set of search results responsive to said search query.

29. The method of claim 28 wherein said processing documents step (c) also is performed prior to said query with a corpus of reference documents to identify predetermined content keywords correlated to said distinct temporal periods, where said temporal characteristics are stored in an electronic file index associated with said electronic documents, and where said altering step (d) is performed according to said stored temporal characteristics.

30. The method of claim 28 wherein said temporal characteristics of said electronic documents are computed dynamically during said query.

31. The method of claim 30 wherein said temporal characteristics are calculated by the computing system only for a top N set of said initial set of search results which are determined at an initial evaluation by the computing system to contain content representative of a temporal period that is later in time than that of others of said set of initial set of search results.

32. The method of claim 30 further including a step:

generating one or more content based electronic ads relating to temporal aspects of such electronic documents, including predetermined content keywords matching said {object, status} pairs.

33. The method of claim 30 wherein said final set of search results are presented within at least one of a search engine, a web portal, a web page, a web log, and/or a television related interface.

34. The method of claim 28 wherein advertising is presented to a user presenting said search query based on a current state identified for said first event or topic determined from said {object, status} pairs.

35. The method of claim 28 wherein advertising is presented to a user presenting said search query based on a temporal sequence identified for an electronic document as it is reviewed by said user.

36. The method of claim 28 wherein said steps are implemented as one or more computer software routines stored in a tangible media and adapted to cause one or more computing systems to perform the operations recited therein.

37. A method of processing a search query with an online computing system comprising:

a) receiving a search query relating to a first event;

determining if said search query includes temporal parameters related to said first event;

b) generating an initial set of search results, including an initial set of electronic documents, which are responsive to the search query with the online computing system;

c) based on steps (a) and (b) processing said initial set of electronic documents to determine temporal characteristics of such documents with respect to one or more events and/or topics, wherein said temporal characteristics are associated with content in said documents pertaining to {event, event state} content snippets but are not any of creation time alone, publication time alone, or a combination of creation time and publication time alone;

further wherein said processing determines said temporal characteristics in said documents by analyzing said {event, event state} content snippets and assigning an event state for each event to a distinct temporal period;

d) generating a search result for said search query to contain a subset of said initial set of electronic documents selected based on said determined temporal characteristics; and

e) presenting said search result to said user responsive to said search query.

38. The method of claim 37, wherein said search result represents documents containing more current event state information for said first event referenced in said search query compared to other information in other documents in said initial set of search results.

39. The method of claim 37, wherein said search result represents documents from a search index, web logs and news sources.

40. The method of claim 37, further including a step:

repeating steps (b) through (e) to compile a chronological report on said first event, said chronological report being broken down into discrete temporal categories.

41. The method of claim 37, wherein said search result is presented in a ranked list within a web page presented to said user, said ranked list being based on a timeliness factor associated with each data source associated with an entry on said list.

42. The method of claim 37, wherein said temporal characteristics are determined in part based on search behavior observed for other users.

43. The method of claim 37 further including a step:

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

44. The method of claim 42 further including a step:

performing a verification step by querying members of a social networking site located near a site of such event to confirm a current event state determined for said first event.

45. The method of claim 42 further including a step:

performing a verification step to confirm a current event state determined for said first event.