System and method for forecasting information using collective intelligence from diverse sources

• US 7,155,510 B1
• Filed: 03/26/2002
• Issued: 12/26/2006
• Est. Priority Date: 03/28/2001
• Status: Active Grant

• Current Assignee: Predictwallstreet LLC
• Original Assignee: Predictwallstreet LLC
• Inventors: Kaplan, Craig Andrew
• Primary Examiner(s): DINH, KHANH Q
• Application Number:
  US10/107,420
• Time in Patent Office: 1,736 Days
• Field of Search: 709/203, 709/223, 709/228, 709/245, 709/224, 709/227, 709/238, 713/1, 364/146, 705/8, 705/10, 715/507
• US Class Current: 709/224
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

System, method, computer program product, and business operating model. Information processing system comprising: server computer having processor for executing computer instructions and memory coupled to processor for storing instructions and data; communications link for communicating between server and information appliance client entities external to server, communication link enabling receipt of first information from a plurality of external entities and for transmitting second information to second entities, first and second entities may be same entities or different entities; data store defining database for storing first and second information; computer program having plurality of executable command instructions and executing in processor and memory for generating second information from at least some of plurality of first information; wherein first information comprising plurality of individual information items and second information comprising processed collective information generated from plurality of individual information items and having increased value to second entities by virtue of collective contribution and processing applied thereto.

Claims

1. An information processing system comprising:

a server computer having a processor for executing computer instructions and a memory coupled to said processor for storing instructions and data during processing;

a communications port for coupling with an external communications link or network adapted to support communications between said server computer and devices external to said server computer, said communication link or network enabling receipt of first raw contributed information from a plurality of external first entities and for transmitting second consensus collective information to any of said external first entities or to external second entities;

a data store defining a database for storing said first raw contributed information and said second consensus collective information; and

a computer program having a plurality of executable command instructions and executing in said processor and said memory for generating said second consensus collective information from at least some of said plurality of first raw contributed information;

said first raw contributed information comprising a plurality of individual raw contributed information items each generated by an external human contributor and received by the server over said communications link or network; and

said second consensus collective information generated from said plurality of individual raw contributed information items by weighting each of each said individual raw contributed information items according to an objective accuracy measure and aggregating said weighted raw contributed individual information items, and having an increased value by virtue of the accuracy provided by the objective accuracy-based weighted consensus collective aggregation of the plurality of individual raw contributed information items.

2. The system in claim 1, wherein the first entities comprise individual persons who provide the raw contributed information items and said second entities comprise individual persons who are consumers of the processed collective information.

3. The system in claim 1, wherein the first entities comprise individual persons that provide the raw contributed information and are different from the second entities who comprise individual persons, or where the second entities request different consensus collective information that is not based on the raw contributed information items they submitted to the server.

4. The system in claim 1, wherein the raw contributed information comprises numeric future event prediction information.

5. The system in claim 1, wherein the raw contributed information comprises non-numeric future event prediction information.

6. The system in claim 1, wherein the communications port for coupling with a communications link or network of for coupling with an Internet network communication link, and said external devices comprise client computers or information appliances coupled to said server computer via said Internet network communication link.

7. The system in claim 6, wherein at least one of said computer server and said client computers or information appliances are coupled to said server computer via said Internet network communication link and interface for communication with said server computer using an Internet web-based browser.

8. The system in claim 1, wherein the communications link or network comprises a wireless communications link, and said external devices comprise wireless client computers or information appliances selected from the set consisting of a notebook computer, a PDA, and a cellular or mobile telephone, coupled to said server computer via said wireless communications link.

9. The system in claim 8, wherein the external device include software of firmware applets or other downloadable client software that executes on wireless or wired devices.

10. The system in claim 1, wherein said server computer further comprise an Internet website interface accessible over said communications link or network that can be accessed by wireless or non-wireless browsers executing on said external devices.

11. The system in claim 1, wherein the first raw contributed information is selected from the set of information types consisting of:

stock prices, bond prices, security prices, real estate prices, interest rates, inventory figures, sales figures, consumer confidence estimates, other financial information, vacation plans, traffic routes, sports scores, election results, predictions of actions by individuals, predictions of actions by groups, predictions of actions by companies, predictions of actions by nations, battlefield intelligence, military intelligence, national security intelligence, other non-financial information, and combinations thereof.

12. The system in claim 1, wherein the computer program generating said second consensus collective information from at least some of said plurality of said first raw contributed information items includes computer program instructions computing an average to generate said second consensus collective information from said plurality of first raw contributed information items.

13. The system in claim 1, wherein said plurality of first raw contributed items are contributed by human contributors via said communications port;

and the computer program generating said second consensus collective information from at least some of said plurality of said first raw contributed information items includes computer program instructions computing a weighted average by applying a weighing amount to each of said plurality of first raw contributed information items to generate said second consensus collective information from said plurality of first raw contributed information items.

14. The system in claim 13, wherein the weighting applied to each of said first raw contributed information items is based on the historical accuracy of the particular first entity supplying the particular first raw information item to the server.

15. The system in claim 13, wherein the weighting uses other methods for assigning and applying weights to the individual raw information items selected from the set of weighting methods consisting of a weighted majority algorithm, a shifting expert algorithm, or any other weighing algorithms known in the art.

16. The system in claim 13, wherein the weighting applied to each of said first information items is determined by linear regression or other statistical techniques of regression.

17. The system in claim 13, wherein the weighting applied to each of said first information items incorporates a time factor such that more recent performance counts more heavily than older performance.

18. The system in claim 13, wherein the method of weighting changes as the system itself learns better ways of weighting in order to minimize the error between a forecast second information value and an actual value for the forecast value.

19. The system in claim 13, wherein the algorithms for generating the second information are selected from the set of algorithms consisting of:

computing statistics other than the average, and using statistical methods that attempt to determine how likely or unlikely a given event is.

20. The system in claim 19, wherein the statistics other than the average are selected from the set consisting of:

computing a median, computing a mode, computing a range, computing a variance, and computing a standard deviation; and

the statistical methods that attempt to determine how likely or unlikely a given event is includes using a Bayesian analyses to determine how likely or unlikely a given event is.

21. The system in claim 1, wherein the generation includes actively and dynamically sampling a population of human contributors to obtain current sentiment with respect to a variable of interest including an investor current sentiment variable.

22. The system in claim 1, wherein the system further includes means for making an investment decision and the investment decision is automatically driven by a future value forecasts generated by the system.

23. The system in claim 1, wherein input raw information is filtered to remove information from entities determined to be attempting to provide information intended to game the system.

24. The system in claim 1, wherein the second information is an output forecast information and the display of the output forecast information takes the form of a scrolling ticker display output on an output device that displays both current values for a given variable and future values for the same variable.

25. The system in claim 24, wherein the variables include stock prices.

26. The system in claim 24, wherein there are two or more lines in the scrolling ticker, and one of the two or more lines displays the current values of a given variable and the other of the two or more lines displays the future values for the same given variable.

27. The system in claim 24, wherein the user may customize what appears on the ticker based on user preferences.

28. The system in claim 24, wherein the ticker is resizable.

29. The system in claim 24, wherein the ticker can be configured to display horizontally or vertically.

30. The system in claim 24, wherein the ticker is capable of displaying advertisements interspersed with the scrolling content.

31. The system in claim 30, wherein the ad display mechanism is intelligent and recognizes the content of the ticker and displays ads that are deemed most relevant to the content that is currently scrolling.

32. The system in claim 31, wherein the ticker displays ads in a non-scrolling frame portion of the ticker.

33. The system in claim 32, wherein the ad display mechanism is intelligent and recognizes the content of the ticker and displays ads that are deemed most relevant to the content that is currently scrolling.

34. The system in claim 24, wherein the ticker can display both numeric, alphanumeric, and symbol information, which is color-coded.

35. The system in claim 24, wherein the ticker can be downloaded to a PC web browser, PDA, cell phone, or other wireless or non-wireless device as a standalone applet or piece of client code that can communicate with a server to display information dynamically.

36. The system in claim 35, wherein the PDA, cell phone, or other wireless device can also periodically load the latest forecast information into the ticker applet and operate independently without continuous connectivity or with only occasional connections to the network.

37. The system in claim 1, wherein the system further includes means for processing not only information submitted by individuals but also other information available from the network in order to generate value-added information.

38. The system in claim 1, wherein first raw contributed information providing entities get credit for providing information to the system.

39. The system in claim 38, wherein such credit is determined in part by the quality of the information provided.

40. The system in claim 39, wherein quality is determined in part by one or more of (i) the accuracy of the information as compared to the actual information, (ii) the timeliness of the information, and (iii) the period of time in advance that an accurate forecast is made when the information is a future forecast.

41. The system in claim 38, wherein first entities can use the received credit to buy the right to see value-added information including the right to see the second consensus collective information.

42. The system in claim 38, wherein entities can exchange credit for prizes, cash, or other compensation.

43. The system in claim 1, further comprising means of combining information from many individuals and processing it to produce value-added information is incorporated as part of the mechanism for a larger collective intelligence or global intelligence system.

44. The system in claim 43, wherein such a system is a military intelligence system.

45. The system in claim 1, wherein:

the first entities comprise individual persons who provide the raw information are said second entities who comprise individual persons who are also the consumers of the processed collective information;

the communications link comprises an Internet network communication link, and said external devices comprise client computers or information appliances coupled to said server computer via said Internet network and at least one of said computer server and said client computer or information appliance interface for communication using an Internet web-based browser;

the first raw contributed information is selected from the set of information types consisting of;

stock prices, bond prices, security prices, real estate prices, interest rates, inventory figures, sales figures, consumer confidence estimates, other financial information, vacation plans, traffic routes, sports scores, election results, predictions of actions by individuals, predictions of actions by groups, predictions of actions by companies, predictions of actions by nations, battlefield intelligence, military intelligence, national security intelligence, other non-financial information, and combinations thereof;

the computer program generating said second information from at least some of said plurality of said first information includes computer program instructions computing a weighted average by applying a weighing amount to each of said plurality of first information items to generate said processing second information from said plurality of first information inputs, and the weighting applied to each of said first information items is based on the historical accuracy of the particular first entity supplying the information; and

the second information is a forecast information and the display of the forecast information takes the form of a scrolling ticker that displays both current values for a given variable and future values for the same variable.

46. The system in claim 1, wherein said first and second entities are the same entities.

47. The system in claim 1, wherein said first and second entities are different entities.

48. The system in claim 1, wherein the first raw contributed information is selected from the set of information types consisting of:

predictions of stock prices, predictions of bond prices, predictions of security prices, predictions of real estate prices, predictions of interest rates, predictions of inventory figures, predictions of sales figures, predictions of consumer confidence, predictions of other financial information, predictions of vacation plans, predictions of traffic routes, predictions of sports scores, predictions of election results, predictions of actions by individuals, predictions of actions by groups, predictions of actions by companies, predictions of actions by nations, predictions of battlefield intelligence, predictions of military intelligence, predictions of national security intelligence, predictions of other non-financial information, and combinations thereof.

49. The system in claim 48, wherein each of the predictions is a newly made current prediction made by a human predictor just prior to communicating it to the server.

50. The system in claim 1, wherein the objective accuracy measure is a past-performance based objective historical accuracy performance measure.

51. The system in claim 1, wherein entities that provide the information are also the entities that consume the information.

52. The system in claim 1, wherein entities provide the first information so that they have the option of receiving the second information.

53. The system in claim 1, wherein the first information includes a first individual information and the second information includes a weighted aggregate consensus information.

54. The system in claim 1, wherein the first information includes a first individual prediction information and the second information includes a weighted aggregate consensus prediction information.

55. The system in claim 1, wherein the weighted aggregate consensus information having a value that is related to its accuracy relative to a future actual information.

56. The system in claim 1, wherein the first information is provided in response to a query to human contributors for information not contained in the database prior to the query and submitted to the system and entered into the database only after the query.

57. The system in claim 1, wherein the processed consensus collective information comprises a prediction of a future value.

58. The system in claim 1, wherein the weighting of new information created by humans whereas Krishnan'"'"'s involves weighting relevance of pre-existing objects in a database.

59. The system in claim 1, wherein the weighting is a weighting according to historical accuracy not preferences.

60. The system in claim 1, wherein the weights are selected or adjusted according to an absolute accuracy metric and are not weighted according to like, dislike, or relevancy.

61. The system in claim 1, wherein the first information includes a first individual prediction information and the second information includes a weighted aggregate consensus prediction information, and the weighted aggregate consensus prediction information is generated at least in part by applying a different non-unity weight to different ones of the first individual prediction information items.

62. A system as in claim 1, wherein said second collective consensus information includes a component given by an expression of the form:

CPI=K×

((WPIE₁×

P₁)+(WPIE₂×

P₂)+ . . . +(WPIE_n×

P_n))/Σ

WPIE_{1 . . . n},

where K is a proportionality constant.

63. A system as in claim 1, wherein said second collective consensus information includes a component that is a normalized sum of a plurality of individual inputs that are each weighted by a weighting factor.

64. An information processing system comprising:

a server computer having a processor for executing computer instructions and a memory coupled to said processor for storing instructions and data during said processing;

a communications link for communicating between said server computer and information appliance client entities external to said server computer, said communication link enabling receipt of first information from a plurality of external entities and for transmitting second information to second entities, said first and second entities may be the same entities or different entities;

a data store defining a database for storing said first information and said second information;

a computer program having a plurality of executable command instructions and executing in said processor and said memory for generating said second information from at least some of said plurality of said first information;

said first information comprising a plurality of individual information items and said second information comprising a processed collective information generated from said plurality of individual information items and having an increased value to said second entities by virtue of the collective contribution and processing applied thereto;

the computer program generating said second information from at least some of said plurality of said first information and includes computer program instructions computing a weighted average by applying a weighing amount to each of said plurality of first information items to generate said processing second information from said plurality of first information inputs; and

said second information is a consensus collective prediction given by an expression of the form;

CPI=K×

((WPIE₁×

P₁)+(WPIE₂×

P₂)+ . . . +(WPIE_n×

P_n))/Σ

WPIE_{1 . . . n},

where K is a proportionality constant.

65. A method of generating a consensus collective information for a particular item from a plurality of raw information inputs contributed by a plurality of human contributors for that same item, the method comprising:

soliciting raw information contribution inputs for a particular item over an interactive electronic network;

gathering a plurality of raw information contribution inputs in response to said solicitation from a plurality of first entities for the particular item, each raw information contribution input being generated by an external human contributor and received over said interactive electronic network;

storing said plurality of gathered raw information contribution inputs into a database;

processing said plurality of raw information contribution inputs for said particular item to generate a processed consensus collective information for said item, said processing including weighting each of each said plurality of raw information inputs by a weighting factor determined according to an objective accuracy measure and aggregating said weighted raw information inputs to generate the processed consensus collective information as an output and having an increased value by virtue of the accuracy provided by the objective accuracy-based weighted consensus collective aggregation of the plurality of individual raw contributed information inputs;

storing said processed consensus collective information into said database; and

communicating said processed consensus collective information to second entities over said interactive electronic network.

66. A method as in claim 65, wherein said raw information input comprises an individual raw prediction and said processed collective information comprises a processed consensus collective prediction based on a plurality of weighted individual raw prediction inputs.

67. A method as in claim 65, wherein said raw information input comprise a non-numeric information and said processed consensus collective information comprises a processed non-numeric information based on a plurality of weighted individual raw non-numeric information inputs.

68. A method as in claim 67, wherein said non-numeric information comprises a geographic location.

69. A method as in claim 67, wherein said non-numeric information comprises a travel route.

70. A method as in claim 65, wherein said raw information comprises a travel time.

71. A method as in claim 65, wherein said raw information comprise a quantitative prediction and said processed collective information comprises a processed consensus collective prediction.

72. A method as in claim 65, wherein said raw information comprises a quantitative prediction value and said processed collective information comprises a processed collective prediction value.

73. A method as in claim 65, wherein said raw information comprises a quantitative numeric prediction value and said processed collective information comprises a processed collective numeric prediction value.

74. A method as in claim 65, wherein at least some of the first entities from which the plurality of individual prediction values are gathered are the same as the second entities.

75. A method as in claim 65, wherein at least some of the first entities from which the plurality of individual prediction values are gathered are different entities from the second entities.

76. A method as in claim 65, wherein the first entities from which the plurality of individual prediction values are gathered are different entities from the second entities.

77. A method as in claim 65, wherein the information comprises a numeric information.

78. A method as in claim 65, wherein the numeric information comprises a security price.

79. A method as in claim 65, wherein the gathering of raw information comprises gathering the raw information on a web-server and storing the gathered information in a database within or communicatively coupled with said web server.

80. A method as in claim 65, wherein the web-server interacts with at least one of said entities through a web-server client web-browser interface.

81. A method as in claim 65, wherein the web-server interacts with at least one of said entities through a web-server client web-browser interface on a personal computer.

82. A method as in claim 65, wherein the web-server interacts with at least one of said entities through a web-server client web-browser interface on a personal data assistant (PDA).

83. The method in claim 65, wherein the raw information input is selected from the set of information input types consisting of:

predictions of stock prices, predictions of bond prices, predictions of security prices, predictions of real estate prices, predictions of interest rates, predictions of inventory figures, predictions of sales figures, predictions of consumer confidence, predictions of other financial information, predictions of vacation plans, predictions of traffic routes, predictions of sports scores, predictions of election results, predictions of actions by individuals, predictions of actions by groups, predictions of actions by companies, predictions of actions by nations, predictions of battlefield intelligence, predictions of military intelligence, predictions of national security intelligence, predictions of other non-financial information, and combinations thereof.

84. A server adapted to receive a plurality of raw prediction information contributions over a communications link or network and to generate a consensus collective weighted forecast information, said server comprising:

an interface to said communications link or network for receiving said plurality of raw prediction information contributions and for sending said consensus collective weighted forecast information to external entities;

a processor for executing computer program instructions and a memory coupled to said processor for storing instructions and data during said execution;

a data store adapted to store said received plurality of first raw contributed prediction information and said generated consensus collective weighted forecast information;

said first raw contributed prediction information comprising a plurality of individual raw contributed prediction information items each generated by an external human contributor and received by said server over said communications link or network; and

a computer program having a plurality of executable command instructions for execution in said processor and said memory for generating a second consensus collective weighted forecast information from at least some of said plurality of first raw contributed prediction information from at least some of said plurality of said first raw contributed prediction information, said computer program including instructions for receiving said plurality of first raw contributed prediction information items and for applying weighting factors determined according to an objective historical contributor accuracy measure for each of said raw prediction information contribution items, and for aggregating said weighted raw prediction information contribution items to generate said consensus collective weighted forecast information.

85. A server as in claim 84, wherein said raw contributed prediction information comprises a stock market financial numerical or non-numerical prediction for a future period, and said consensus collective weighted forecast information comprises a consensus collective weighted forecast information for the same stock market financial numerical or non-numerical prediction for the same future period.

86. A server as in claim 84, wherein said consensus collective weighted forecast information includes a component given by an expression of the form:

CPI=K×

((WPIE₁×

P₁)+(WPIE₂×

P₂)+ . . . +(WPIE_n×

P_n))/Σ

WPIE_{1 . . . n},

where K is a proportionality constant.

87. A business method for providing a consensus collective weighted forecast information, said method comprising:

soliciting a plurality of raw prediction information contribution inputs for a particular topic over an interactive electronic network;

receiving said plurality of raw prediction information contribution inputs in response to said solicitation from a plurality of contributors over said interactive electronic network, each raw prediction information contribution input being generated by an external human contributor and received over said interactive electronic network;

storing said plurality of received raw prediction information contribution inputs into a database;

processing said plurality of raw prediction information contribution inputs for said particular topic to generate said consensus collective weighted forecast information for said topic by applying weighting factors to each said plurality of raw prediction information contribution inputs determined according to an objective historical contributor accuracy measure and aggregating said weighted raw contributed prediction information inputs to generate said processed consensus collective weighted forecast information as an output and having an increased value by virtue of the accuracy provided by the objective accuracy-based weighted consensus collective aggregation of the plurality of individual raw prediction information contribution inputs;

storing said consensus collective weighted forecast information into said database; and

communicating said consensus collective weighted forecast information for said particular topic to a recipient in exchange for a payment over said interactive electronic network.

88. A business method as in claim 87, further comprising providing a credit to each of said contributors for a raw prediction contribution;

and wherein the payment may be made in the form of a monetary payment or in exchange for said credit.

89. A business method as in claim 87, wherein said raw prediction information comprises a stock market financial numerical or non-numerical prediction for a future period, and said consensus collective weighted forecast information comprises a consensus collective weighted forecast information for the same stock market financial numerical or non-numerical prediction for the same future period.

90. A business method as in claim 87, wherein said consensus collective weighted forecast information includes a component given by an expression of the form:

CPI=K×

((WPIE₁×

P₁)+(WPIE₂×

P₂)+ . . . +(WPIE_n×

P_n))/Σ

WPIE_{1 . . . n},

where K is a proportionality constant.