System for communication and collaboration

• US 8,688,673 B2
• Filed: 09/26/2006
• Issued: 04/01/2014
• Est. Priority Date: 09/27/2005
• Status: Active Grant

• Current Assignee: Devajyoti Sarkar
• Original Assignee: Sarkar PTE. LTD.
• Inventors: Sarkar, Devajyoti
• Primary Examiner(s): LEWIS, CHERYL RENEA
• Application Number:
  US11/992,665
• Publication Number:
  US 20100223261A1
• Time in Patent Office: 2,744 Days
• Field of Search: 707/705, 707/706, 707/708, 707/723, 707/728, 707/732, 707/737, 707/748, 707/726
• US Class Current: 707/706
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A system and method for communication and collaboration are disclosed. A generalized annotation based mechanism using an emergent self organization characteristic of the natural language of the annotations allows users to search for relevant items, users to search for relevant users, and items to search for relevant users based on aggregation of stored annotations having associations between keywords, items and users in a context space. Aggregations based on correlations between users, items and keywords, are used to form a collaborative content relevance that allows users to be directed to items or other users, and items to be directed to users.

Claims

1. A computer system for using an emergent self-organization characteristic in a natural language for establishing a collaborative content relevance between users and items in a context space, the computer system comprising:

at least one processing unit and

at least one computer-readable media for storing instructions for implementation by the at least one processing unit and data used by the at least one processing unit in implementing instructions,

wherein the at least one processing unit includes;

means for identifying a plurality of items in the context space with unique item identifiers, one of the plurality of items capable of being independently annotated in the context space by one or more of a plurality of users operating independently of one another without knowledge of each other'"'"'s activities or existence, the annotating being done in the context space with unique user identifiers using an annotation of at least one keyword from the natural language during an annotation event, the annotation including an association between the at least one keyword, one of the unique item identifiers associated with the one of the plurality of items being independently annotated, and one of the unique user identifiers associated with the independently annotating one of the plurality of users;

means for copying each of the annotations generated during ones of the annotation event to the at least one computer-readable media for storage therein, the stored annotations containing the ones of the unique user identifier of the independently annotating ones of the plurality of users, the ones of the unique item identifier of the ones of the items being independently annotated, and the at least one keyword, ones of the plurality of users capable of using a different keyword to annotate the same one of the plurality of items;

means for aggregating the stored annotations based on a correlation between users, items and keywords to form the collaborative content relevance between the plurality of items and the plurality of users; and

means for associating, using the collaborative content relevance, one of;

ones of the plurality of users with other ones of the plurality of users; and

ones of the plurality of items with ones of the plurality of users such that the ones of the plurality of users are capable of discovering the other ones of the plurality of users, and the ones of the plurality of items are capable of discovering the ones of the plurality of users, based on the associated collaborative content relevance.

2. A method for using an emergent self-organization characteristic in a natural language for establishing a collaborative content relevance between users and items in a context space, implemented using the computer system of claim 1, the method comprising the steps of:

identifying a plurality of items in the context space with unique item identifiers, one of the plurality of items capable of being independently annotated in the context space by one or more of a plurality of users operating independently of one another without knowledge of each other'"'"'s activities or existence, the annotating being done in the context space with unique user identifiers using an annotation of at least one keyword from the natural language during an annotation event, the annotation including an association between the at least one keyword, one of the unique item identifiers associated with the one of the plurality of items being independently annotated, and one of the unique user identifiers associated with the independently annotating one of the plurality of users, performed using the means for identifying;

copying each of the annotations generated during ones of the annotation event to the at least one computer-readable media for storage therein, performed using the means for copying, the stored annotations containing the ones of the unique user identifier of the independently annotating ones of the plurality of users, the ones of the unique item identifier of the ones of the items being independently annotated, and the at least one keyword, ones of the plurality of users capable of using a different keyword to annotate the same one of the plurality of items;

aggregating the stored annotations based on a correlation between users, items and keywords, the correlation associated with the emergent self organization characteristics to form the collaborative content relevance between the plurality of items and the plurality of users, performed using the means for aggregating the stored annotations; and

associating, using the collaborative content relevance, one of;

ones of the plurality of users with other ones of the plurality of users; and

ones of the plurality of items with ones of the plurality of users such that the ones of the plurality of users are capable of discovering the other ones of the plurality of users, and the ones of the plurality of items are capable of discovering the ones of the plurality of users, based on the associated collaborative content relevance, performed using the means for associating.

3. The method of claim 2, further comprising:

in the aggregating step, aggregating the stored annotations for each of the plurality of items such that the keywords associated with the item are aggregated from all the annotations of the item;

searching for one of the plurality of users based on the collaborative content relevance of the annotations associated with the one of the plurality of users and at least one search keyword associated with the searching;

the collaborative content relevance established by one of;

determining items whose keywords match the at least one search keyword; and

determining users that annotated the determined items; and

returning the determined users as the search result.

4. The method according to claim 3, wherein the collaborative content relevance is further established by ranking the determined items on the basis of a Information Retrieval ranking algorithm and ranking the determined users based on a correspondence between ones of the ranked determined items and ones of the determined users that annotated the items;

and

returning the determined users according to rank as the search result.

5. The method according to claim 4, further comprising:

associating an advertisement with the at least one search keyword; and

delivering the advertisement to at least one of a subset of the highest-ranked determined relevant users.

6. The method according to claim 5, wherein the advertisement is selected from at least one from a group consisting of text, audio and video, and classified advertisement.

7. The method according to claim 3, further comprising:

considering each annotation as a synthetic hyperlink between the annotating user and the annotated item;

treating the determined users as hubs and determined items as authorities; and

using a link analysis ranking algorithm as the basis to compute the ranks for determined items and using a link analysis ranking algorithm to compute the ranks for the determined users.

8. The method according to claim 7, wherein the Link Analysis Ranking algorithm includes one of an Indegree algorithm, a HITS algorithm, a Randomized HITS algorithm, a Subspace HITS algorithm, a SALSA algorithm, a HUBAVG algorithm, an Authority Threshold family of algorithms, a MAX algorithm, a BFS algorithm, a BAYESIAN algorithm, a Simplified BAYESIAN algorithm, a PageRank algorithm, a Personalized PageRank algorithm, a TrafficRank algorithm, a Tensor decomposition algorithm, a TOPHITS algorithm, a CubeSVD algorithm, a PHITS algorithm, and a PLSA+PHITS algorithm.

9. The method according to claim 3, further comprising:

associating an advertisement with the at least one search keyword; and

delivering the advertisement to at least one of the determined users.

10. The method according to claim 9, wherein the advertisement is selected from at least one from a group consisting of text, audio and video, and classified advertisement.

11. The method according to claim 3, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

restricting the determined items to items where the at least one search keyword matches at least one of the defining keywords of the item.

12. The method of claim 2 further comprising:

matching annotations that have the at least one search keyword;

in the aggregating step, aggregating the unique keywords of the matched annotations after removing the at least one searched keyword;

ranking the unique keywords based on the number of users or items;

sorting the unique keywords by rank;

presenting each such unique keyword as a sub-category that a user may drill-down; and

responding to the user'"'"'s drill-down by appending the drill-down keyword to the at least one search keyword and re-issuing the search.

13. The method according to claim 12, wherein the annotations are generated from the clickstream of a search engine.

14. The method of claim 12 further comprising ranking drilldowns based on number of users and items.

15. The method according to claim 2, further comprising:

in the aggregating step, aggregating annotations by category contexts such that an annotation corresponds to a category context if it contains the keywords of the category context;

determining category contexts that have a predefined minimum number of unique user identifiers and a predefined minimum number of unique item identifiers in their aggregated annotations; and

clustering items and users simultaneously by category context such that the unique user identifiers for a determined category context is identified as a cluster of users and the unique item identifiers for the same determined category context is identified as a cluster of items.

16. The method according to claim 15, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

restricting the aggregation of annotations by category contexts to annotations where at least one of the category context'"'"'s keywords matches at least one of the defining keywords of the item in the annotation.

17. The method according to claim 15, further comprising:

associating an advertisement with a determined category context; and

delivering the advertisement to at least one of the relevant users.

18. The method according to claim 17, wherein the advertisement is selected from at least one from a group consisting of text, audio and video, and classified advertisement.

19. The method according to claim 15, further comprising:

associating an item with a determined one of the category contexts; and

publishing the item to the determined category context such that it is made available to any of the cluster of users associated with the category context.

20. The method according to claim 15, further comprising:

subscribing to items by a determined category context by a user; and

delivering items corresponding to the determined category context'"'"'s cluster of items to the user.

21. The method according to claim 15, further comprising:

identifying a subset of the determined category contexts such that the category contexts of the identified subset have more people in its cluster of users relative to the determined category contexts as a whole and have fewer items in its cluster of items relative to the determined category contexts as a whole; and

presenting the identified subset of category contexts as hot topics.

22. The method according to claim 21, further comprising:

allocating resources based on the hot topics.

23. The method according to claim 15, further comprising:

offering drilldowns to a search based on the determined category contexts.

24. The method of claim 2, further comprising:

each item having at least one publisher identifier;

in the aggregating step, aggregating the annotations for each publisher identifier such that the keywords and the user identifiers associated with the publisher identifier are aggregated from all the annotations corresponding to the publisher identifier;

searching based on at least one keyword;

determining publisher identifiers that match the at least one searched keyword from the aggregated annotations corresponding to the publisher identifier; and

ranking the determined publisher identifiers.

25. The method according to claim 24, wherein the ranking of publisher identifiers is done using an information retrieval ranking algorithm.

26. The method according to claim 24, further comprising:

considering each annotation as a synthetic hyperlink between the annotating user and the publisher identifier;

treating the determined users as hubs and determined publisher identifiers as authorities; and

using a link analysis ranking algorithm as the basis to compute the ranks for determined publisher identifiers and using a link analysis ranking algorithm to compute the ranks for the determined users.

27. The method according to claim 26, wherein the Link Analysis Ranking algorithm includes one of an Indegree algorithm, a HITS algorithm, a Randomized HITS algorithm, a Subspace HITS algorithm, a SALSA algorithm, a HUBAVG algorithm, an Authority Threshold family of algorithms, a MAX algorithm, a BFS algorithm, a BAYESIAN algorithm, a Simplified BAYESIAN algorithm, a PageRank algorithm, a Personalized PageRank algorithm, a TrafficRank algorithm, a Tensor decomposition algorithm, a TOPHITS algorithm, a CubeSVD algorithm, a PHITS algorithm, and a PLSA+PHITS algorithm.

28. The method of claim 24 further comprising:

searching for items based on at least one keyword;

determining items whose keywords match the at least one search keyword;

and ranking the determined items based on the ranking of their publisher identifiers.

29. The method according to claim 28, further comprising:

personalizing the results for a user wherein the publisher identifiers are re-ranked based on the user'"'"'s past annotations; and

re-ranking the items based on the new publisher identifier ranks; and

presenting the re-ranked items to the user.

30. The method according to claim 28, further comprising searching in a periodic fashion automatically based on a user profile.

31. The method according to claim 30 further comprising delivering an advertisement based on the user profile.

32. The method according to claim 31, wherein the advertisement is selected from at least one from a group consisting of text, audio and video, and classified advertisement.

33. The method according to claim 24, wherein the at least one search keyword is restricted to category contexts.

34. The method of claim 24 further comprising each item having at least one publisher identifier, and ranking the publisher identifiers in a category context.

35. The method according to claim 24, wherein the published item is an advertisement.

36. The method according to claim 24, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

restricting the determined items to items where the at least one search keyword matches at least one of the defining keywords of the item.

37. The method as claimed in claim 2, further comprising:

publishing an item when a user annotates an item with a set of keywords;

in the aggregating step, aggregating the annotations for each item such that the keywords associated with the item are aggregated from all the annotations of the item;

determining items whose keywords match the at least one publishing keyword;

determining users that annotated the determined items; and

ranking the published item highly in searches for a subset of the determined users until a predetermined condition is fulfilled.

38. The method according to claim 37, further comprising:

ranking the determined users using an information retrieval algorithm; and

selecting the subset of the determined users from highly ranked users.

39. The method according to claim 38, wherein the information retrieval ranking algorithm is a link analysis ranking algorithm.

40. The method according to claim 39, wherein the Link Analysis Ranking algorithm includes one of an Indegree algorithm, a HITS algorithm, a Randomized HITS algorithm, a Subspace HITS algorithm, a SALSA algorithm, a HUBAVG algorithm, an Authority Threshold family of algorithms, a MAX algorithm, a BFS algorithm, a BAYESIAN algorithm, a Simplified BAYESIAN algorithm, a PageRank algorithm, a Personalized PageRank algorithm, a TrafficRank algorithm, a Tensor decomposition algorithm, a TOPHITS algorithm, a CubeSVD algorithm, a PHITS algorithm, and a PLSA+PHITS algorithm.

41. The method according to claim 37, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

restricting the determined items to items where the at least one publishing keyword matches at least one of the defining keywords of the item.

42. The method according to claim 37, wherein the at least one publishing keyword is a category context.

43. The method according to claim 37, wherein the published item is an advertisement.

44. The method according to claim 43, wherein the advertisement consist of at least one from a group of text, audio and video, and classified advertisement.

45. The method according to claim 37, wherein the act of publishing an item corresponds to the completion of a step in a workflow process.

46. The method according to claim 45, wherein the step is one of a group comprising authorization, requesting, assigning, termination and notification.

47. The method according to claim 37, wherein the subset of users includes one of highly ranked publishers, highly ranked users and random selected users.

48. The method as claimed in claim 2, further comprising:

in the aggregating step, aggregating the annotations for each item such that the keywords associated with the item are aggregated from all the annotations of the item;

determining items whose keywords match the at least one search keyword;

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

restricting the determined items to items where the at least one search keyword matches at least one of the defining keywords of the item;

determining users that annotated the determined items;

considering annotations as synthetic hyperlinks between the annotating user and the annotated item;

treating the determined users as hubs and determined items as authorities;

using a link analysis ranking algorithm as the basis to compute the ranks for determined items; and

returning the determined items sorted by rank as the search result.

49. The method according to claim 48, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item; and

restricting the determined items to items where the at least one search keyword matches at least one of the defining keywords of the item.

50. The method according to claim 48, wherein the Link Analysis Ranking algorithm includes one of an Indegree algorithm, a HITS algorithm, a Randomized HITS algorithm, a Subspace HITS algorithm, a SALSA algorithm, a HUBAVG algorithm, an Authority Threshold family of algorithms, a MAX algorithm, a BFS algorithm, a BAYESIAN algorithm, a Simplified BAYESIAN algorithm, a PageRank algorithm, a Personalized PageRank algorithm, a TrafficRank algorithm, a Tensor decomposition algorithm, a TOPHITS algorithm, a CubeSVD algorithm, a PHITS algorithm, and a PLSA+PHITS algorithm.

51. The method according to claim 2, wherein the item includes one of a digital asset, a physical asset, a person, an animate entity, a text advertisement, a video advertisement, an event, a location, a state, a process, an act, a group, a concept, a file, an email, an instant message, a blog posting, a podcast, a web page, a web site, a web service, a data structure, a software module, a software object, an application, an operating system, a row in a table in a relational database, XML data and a resource represented in RDF.

52. The method according to claim 2, wherein the unique identifier includes one of a hash value, a URL, a URI, a URN, a UNC, a bar code, a RFID, a fiducial marker, an email address, a social security number, a vehicle registration number, and a telephone number.

53. The method according to claim 2, wherein the identifiers are globally unique.

54. The method according to claim 2, wherein each annotation event used to compute the result set contains all the keywords of the query.

55. The method according to claim 2, wherein the annotations are restricted to annotations that have taken place in a predetermined time interval.

56. The method according to claim 2, further comprising:

determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item; and

restricting the aggregated annotations to annotations wherein at least one keyword of the annotation matches at least one of the defining keywords of the annotated item.

57. The method according to claim 2, wherein the event source includes one of:

remembering a dialog, publishing an item, tagging an item, collaboratively tagging an item, examining a clickstream of a search, underlining words in an item, the underlined words forming keywords for the item, saving a file to a file system, acquiring keywords from text of emails, acquiring keywords from text of documents, acquiring keywords from text of subject lines of mail messages, acquiring keywords from text of mail messages, and performing link text analysis.

58. The method according to claim 2, wherein the annotation event is digitally signed by the user.

59. The method according to claim 2, wherein the computer system comprises at least one server and the at least one processing unit is included in the at least one server, and wherein the annotation events are sent via a network to the at least one server for aggregation.

60. The method according to claim 59, wherein the network is the Internet.

61. The method according to claim 2, wherein at least one keyword is a machine representation of meaning.

62. The method according to claim 2, wherein at least one keyword is a semantic metadata.

63. The method according to claim 62, further describing the semantic metadata by at least one Lexicon and disambiguating the keywords of the search as well as the aggregated keywords of items and users based on the Lexicon.

64. The method of claim 2 further comprising:

identifying a plurality of items with unique item identifiers and a plurality of users with unique user identifiers at a search engine;

allowing a user to search based on keywords and displaying items that correspond to the search as a results,

monitoring items that the user clicks from among the results; and

tagging each clicked item with the keywords of the search and the user'"'"'s unique identifier.

65. The method according to claim 2, wherein at least one keyword for the annotations is obtained from text included in one of emails, documents, mail messages, and subject lines.

66. The method according to claim 2, wherein the ranking is computed using an algorithm including one of a Tensor decomposition algorithm, a TOPHITS algorithm, a CubeSVD algorithm, and a PLSA+PHITS algorithm.

67. The computer system of claim 1, wherein the means for aggregating performs the function of aggregating the annotations for each item such that the keywords associated with the item are aggregated from all the annotations of the item, and wherein the at least one processing unit further includes:

means for searching for users based on at least one search keyword;

means for determining items whose keywords match the at least one search keyword;

means for determining users that annotated the determined items; and

means for returning the determined users as the search result.

68. The computer system of claim 1, wherein the means for aggregating performs the function of aggregating the unique keywords of the matched annotations after removing the at least one searched keyword, and wherein the at least one processing unit further includes:

means for matching annotations that have the at least one search keyword;

means for ranking the unique keywords based on the number of users or items;

means for sorting the unique keywords by rank;

means for presenting each such unique keyword as a sub-category that a user may drill-down; and

means for responding to the user'"'"'s drill-down by appending the drill-down keyword to the at least one search keyword and re-issuing the search.

69. The computer system of claim 1, wherein the means for aggregating performs the function of aggregating annotations by contexts such that an annotation corresponds to a context if it contains the keywords of the context, and wherein the at least one processing unit further includes:

means for determining contexts that have a predefined minimum number of unique user identifiers and a predefined minimum number of unique item identifiers in their aggregated annotations; and

means for clustering items and users simultaneously by context such that the unique user identifiers for a determined context is identified as a cluster of users and the unique item identifiers for the same determined context is identified as a cluster of items.

70. The computer system of claim 1,

wherein each item has at least one publisher identifier, and the means for aggregating performs the function of aggregating the annotations for each publisher identifier such that the keywords and the user identifiers associated with the publisher identifier are aggregated from all the annotations corresponding to the publisher identifier;

and

wherein the at least one processing unit further includes

means for searching for items based on at least one keyword;

means for determining publisher identifiers that match the at least one searched keyword from the aggregated annotations corresponding to the publisher identifier; and

means for ranking the determined publisher identifiers.

71. The computer system of claim 1, wherein the means for aggregating performs the function of aggregating the annotations for each item such that the keywords associated with the item are aggregated from all the annotations of the item, and wherein the at least one processing unit further includes:

means for publishing an item when a user annotates an item with a set of keywords;

means for determining items whose keywords match the at least one publishing keyword;

means for determining users that annotated the determined items; and

means for ranking the published item highly in searches for a subset of the determined users until a predetermined condition is fulfilled.

72. The computer system of claim 1, wherein the means for aggregating performs the function of aggregating the annotations for each item such that the keywords associated with the item are aggregated from all the annotations of the item, and wherein the at least one processing unit further includes:

means for determining items whose keywords match the at least one search keyword;

means for determining defining keywords for items where each defining keyword of an item is a keyword that has been used by more than a predetermined percentage of the users annotating the item;

means for restricting the determined items to items where the at least one search keyword matches at least one of the defining keywords of the item;

means for determining users that annotated the determined items;

means for considering annotations as synthetic hyperlinks between the annotating user and the annotated item;

means for treating the determined users as hubs and determined items as authorities;

means for using a link analysis ranking algorithm as the basis to compute the ranks for determined items; and

means for returning the determined items sorted by rank as the search result.

73. The computer system of claim 1, wherein the at least one processing unit further includes:

means for identifying a plurality of items with unique item identifiers and a plurality of users with unique user identifiers at a search engine;

means for allowing a user to search based on keywords and displaying items that correspond to the search as a results,

means for monitoring items that the user clicks from among the results; and

means for tagging each clicked item with the keywords of the search and the user'"'"'s unique identifier.