Systems and methods for improving collaborative filtering

• US 7,630,916 B2
• Filed: 06/25/2003
• Issued: 12/08/2009
• Est. Priority Date: 06/25/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Meek, Christopher A., Heckerman, David E., Kadie, Carl M., Lind, Jesper B.
• Primary Examiner(s): DeCady; Albert
• Assistant Examiner(s): ROBERTSON, DAVID
• Application Number:
  US10/603,541
• Publication Number:
  US 20040267596A1
• Time in Patent Office: 2,358 Days
• Field of Search: None
• US Class Current: 705/7.29
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions

Abstract

The present invention provides collaborative filtering systems and methods employing statistical smoothing to provide quickly creatable models that can efficiently predict probability that a user likes an item and/or similarities between items. Smoothing is accomplished by utilizing statistical methods such as support cutoff, single and multiple prior on counts, and prior on measure of association and the like. By improving model-based collaborative filtering with such techniques, performance is increased with regard to product-to-product recommendations. The present invention also provides improvements over systems based on dependency nets (DN) in both areas of quality of recommendations and speed of model creation. It can also be complementary to DN to improve the value of an existing collaborative filtering system'"'"'s overall efficiency. It is also employable with low frequency user preference data.

Claims

1. A method of data analysis, employing collaborative filtering, implemented as instructions executed on a processor operatively coupled to memory, the method comprising:

receiving an item set containing at least one item of input data;

selecting an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and the item set;

scoring at least one item of the item set by employing the selected measure of association;

selecting at least one additional measure of association based on the item set;

scoring at least one item of the item set by employing the at least one additional measure of association;

smoothing at least one item of the item set via a selected smoother; and

employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item;

wherein the instructions executed on the processor operatively coupled to memory facilitate the receiving, selecting, scoring, smoothing and employing.

2. The method of claim 1,

wherein the collaborative filtering employs Lift, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set.

3. The method of claim 1,

wherein the collaborative filtering employs Lift, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

4. The method of claim 1,

wherein the collaborative filtering employs Lift, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

5. The method of claim 1,

wherein the collaborative filtering employs Weight of Evidence, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

6. The method of claim 1,

wherein the collaborative filtering employs Weight of Evidence, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

7. The method of claim 1,

wherein the collaborative filtering employs Weight of Evidence, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

8. The method of claim 1,

wherein the collaborative filtering employs Weight of Evidence, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

9. The method of claim 1,

wherein the collaborative filtering employs Yule'"'"'s Q, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

10. The method of claim 1,

wherein the collaborative filtering employs Yule'"'"'s Q, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

11. The method of claim 1,

wherein the collaborative filtering employs Yule'"'"'s Q, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

12. A method of claim 1,

wherein the collaborative filtering employs Yule'"'"'s Q, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

13. The method of claim 1,

wherein the collaborative filtering employs tau measures, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

14. The method of claim 1,

wherein the collaborative filtering employs tau measures, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

15. The method of claim 1,

wherein the collaborative filtering employs tau measures, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

16. The method of claim 1,

wherein the collaborative filtering employs tau measures, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

17. The method of claim 1,

wherein the collaborative filtering employs Phi, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

18. The method of claim 1,

wherein the collaborative filtering employs Phi, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

19. The method of claim 1,

wherein the collaborative filtering employs Phi, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

20. The method of claim 1,

wherein the collaborative filtering employs Phi, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set;

and

employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item.

21. The method of claim 1,

wherein the collaborative filtering employs cross-product, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

22. The method of claim 1,

wherein the collaborative filtering employs cross-product, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

23. The method of claim 1,

wherein the collaborative filtering employs cross-product, smoothed via informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

24. The method of claim 1,

wherein the collaborative filtering employs cross-product, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

25. The method of claim 1,

wherein the collaborative filtering employs log of cross-product, smoothed via cutoff smoothing techniques, as a measure of association for scoring at least one item of the item set.

26. The method of claim 1,

wherein the collaborative filtering employs log of cross- product, smoothed via prior on counts smoothing techniques, as a measure of association for scoring at least one item of the item set, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

27. The method of claim 1,

wherein the collaborative filtering employs log of cross-product, smoothed vi informative priors on measures of association smoothing techniques, as a measure of association for scoring at least one item of the item set.

28. The method of claim 1,

wherein the collaborative filtering employs log of cross-product, smoothed via nonuniform prior smoothing techniques, as a measure of association for scoring at least one item of the item set.

29. The method of claim 1, further comprising receiving a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

30. The method of claim 1, further comprising utilizing a highest value score of scores applicable to an item as the single score of the multiple-score collaborative filtering evaluation method.

31. The method of claim 30, further comprising receiving a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

32. The method of claim 1, the selected smoother is selected, based on the item set, from among known smoothers.

33. A collaborative filtering system, the system comprising:

a processor;

a memory;

instructions stored in the memory and executed by the processor, the instructions comprising;

a measure of association selection component that selects an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and an item set containing at least one item of input data;

a smoothing component that smoothes at least one item of the item set via a selected smoother;

a measure of association computing component that scores at least one item of the item set by employing the selected measure of association;

wherein the measure of association computing component additionally employs at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item; and

a filtering component that employs Lift as a measure of association for scoring at least one item of the item set.

34. A kiosk employing the system of claim 33.

35. A search engine employing the system of claim 33.

36. A set-top box employing the system of claim 33.

37. A television guide employing the system of claim 33.

38. A video guide employing the system of claim 33.

39. A media guide employing the system of claim 33.

40. A music guide employing the system of claim 33.

41. A merchandising system employing the system of claim 33.

42. A targeted advertising system employing the system of claim 33.

43. A recommendation system employing the system of claim 33.

44. A commerce server employing the system of claim 33.

45. A computer readable medium having stored thereon computer executable components of the system of claim 33.

46. A device employing the system of claim 33 comprising at least one selected from the group consisting of a computer, a server, and a hand held electronic device.

47. A collaborative filtering system, the system comprising:

a processor;

a memory;

instructions stored in the memory and executed by the processor, the instructions comprising;

a measure of association selection component that selects an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and an item set containing at least one item of input data;

a smoothing component that smoothes at least one item of the item set via a selected smoother;

a measure of association computing component that scores at least one item of the item set by employing the selected measure of association;

wherein the measure of association computing component additionally employs at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item; and

a filtering component that employs informative priors on a measure of association for smoothing the measure of association utilized in collaborative filtering.

48. A method of data analysis, employing collaborative filtering, implemented as instructions executed by a processor operatively coupled to memory, the method comprising:

receiving an item set containing at least one item of input data;

selecting an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and the item set;

scoring at least one item of the item set by employing the selected measure of association;

selecting at least one additional measure of association based on the item set;

scoring at least one item of the item set by employing the at least one additional measure of association;

smoothing at least one item of the item set via a selected smoother; and

employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item, wherein Lift is employed as a measure of association in the collaborative filtering.

51. A device employing the method of claim 48 comprising at least one selected from the group consisting of a computer, a server, and a hand held electronic device.

52. A computer readable medium having stored thereon computer executable instructions for performing the method of claim 48.

49. A method of data analysis, employing collaborative filtering, implemented as instructions executed by a processor operatively coupled to memory, the method comprising:

receiving an item set containing at least one item of input data;

selecting an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and the item set;

scoring at least one item of the item set by employing the selected measure of association;

selecting at least one additional measure of association based on the item set;

scoring at least one item of the item set by employing the at least one additional measure of association;

smoothing at least one item of the item set via a selected smoother;

employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item; and

employing informative priors on a measure of association for smoothing the measure of association utilized in the collaborative filtering.

50. A data analysis system employing collaborative filtering, the system comprising:

means for receiving an item set containing at least one item of input data;

means for selecting, based on the item set, an appropriate measure of association from among known measures of association;

means for scoring at least one item of the item set by employing the selected measure of association;

means for selecting at least one additional measure of association based on the item set;

means for scoring at least one item of the item set by employing the at least one additional measure of association;

means for smoothing at least one item of the item set via a selected smoother; and

means for employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item, the collaborative filtering based, at least in part, on employing Lift as a measure of association.

53. A collaborative filtering system embodied on a computer readable medium, comprising:

means for receiving an item set containing at least one item of input data;

means for selecting, based on the item set, an appropriate measure of association from among known measures of association;

means for scoring at least one item of the item set by employing the selected measure of association;

means for selecting at least one additional measure of association based on the item set;

means for scoring at least one item of the item set by employing the at least one additional measure of association;

means for smoothing at least one item of the item set via a selected smoother; and

means for employing at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item.

54. A collaborative filtering system embodied on a computer readable medium, comprising:

a measure of association selection component that selects an appropriate measure of association from among known measures of association, the selection is based on the known measures of association and an item set containing at least one item of input data;

a smoothing component that smoothes at least one item of the item set via a selected smoother; and

a measure of association computing component that scores at least one item of the item set by employing the selected measure of association;

wherein the measure of association computing component additionally employs at least one multiple-score collaborative filtering evaluation method to obtain a single score for an item when more than one measure of association score applies to that item.

55. The system of claim 54, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

56. The system of claim 54, the multiple-score collaborative filtering evaluation method utilizing a highest value score of scores applicable to an item as the single score.

57. The system of claim 56, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

58. The system of claim 54, further comprising:

a filtering component that employs Lift, smoothed via prior on counts smoothing techniques.

59. The system of claim 58, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

60. The system of claim 58, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

61. The system of claim 60, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

62. The system of claim 54, further comprising:

a filtering component that employs Lift, smoothed via informative priors on measures of association smoothing techniques.

63. The system of claim 62, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

64. The system of claim 62, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

65. The system of claim 64, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

66. The system of claim 54, further comprising:

a filtering component that employs Lift, smoothed via nonuniform prior smoothing techniques.

67. The system of claim 66, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

68. The system of claim 66, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

69. The system of claim 68, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

70. The system of claim 54, further comprising:

a filtering component that employs Weight of Evidence, smoothed via cutoff smoothing techniques.

71. The system of claim 70, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

72. The system of claim 70, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

73. The system of claim 72, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

74. The system of claim 54, further comprising:

a filtering component that employs Weight of Evidence, smoothed via prior on counts smoothing techniques, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

75. The system of claim 74, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

76. The system of claim 54, further comprising:

a filtering component that employs Weight of Evidence, smoothed via informative priors on measures of association smoothing techniques.

77. The system of claim 76, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

78. The system of claim 76, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

79. The system of claim 78, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

80. The system of claim 54, further comprising:

a filtering component that employs Weight of Evidence, smoothed via nonuniform prior smoothing techniques.

81. The system of claim 80, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

82. The system of claim 80, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

83. The system of claim 82, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

84. The system of claim 54, further comprising:

a filtering component that employs Yule'"'"'s Q, smoothed via cutoff smoothing techniques.

85. The system of claim 84, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

86. The system of claim 84, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

87. The system of claim 86, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

88. The system of claim 54, further comprising:

a filtering component that employs Yule'"'"'s Q, smoothed via prior on counts smoothing techniques.

89. The system of claim 88, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

90. The system of claim 54, further comprising:

a filtering component that employs Yule'"'"'s Q, smoothed via informative priors on measures of association smoothing techniques.

91. The system of claim 90, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

92. The system of claim 90, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

93. The system of claim 92, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

94. The system of claim 54, further comprising:

a filtering component that employs Yule'"'"'s Q, smoothed via nonuniform prior smoothing techniques.

95. The system of claim 94, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

96. The system of claim 94, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

97. The system of claim 96, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

98. The system of claim 54, further comprising:

a filtering component that employs tau measures, smoothed via cutoff smoothing techniques.

99. The system of claim 98, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

100. The system of claim 98, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

101. The system of claim 100, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

102. The system of claim 54, further comprising:

a filtering component that employs tau measures, smoothed via prior on counts smoothing techniques.

103. The system of claim 102, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

104. The system of claim 54, further comprising:

a filtering component that employs tau measures, smoothed via informative priors on measures of association smoothing techniques.

105. The system of claim 104, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

106. The system of claim 104, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

107. The system of claim 106, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

108. The system of claim 54, further comprising:

a filtering component that employs tau measures, smoothed via nonuniform prior smoothing techniques.

109. The system of claim 108, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

110. The system of claim 108, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

111. The system of claim 110, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

112. The system of claim 54, further comprising:

a filtering component that employs Phi, smoothed via cutoff smoothing techniques.

113. The system of claim 112, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

114. The system of claim 112, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

115. The system of claim 114, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

116. The system of claim 54, further comprising:

a filtering component that employs Phi, smoothed via prior on counts smoothing techniques, wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

117. The system of claim 116, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

118. The system of claim 54, further comprising:

a filtering component that employs Phi, smoothed via informative priors on measures of association smoothing techniques.

119. The system of claim 118, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

120. The system of claim 118, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

121. The system of claim 120, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

122. The system of claim 54, further comprising:

a filtering component that employs Phi, smoothed via nonuniform prior smoothing techniques.

123. The system of claim 122, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

124. The system of claim 122, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

125. The system of claim 124, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

126. The system of claim 54, further comprising:

a filtering component that employs cross-product, smoothed via cutoff smoothing techniques.

127. The system of claim 126, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

128. The system of claim 126, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

129. The system of claim 128, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

130. The system of claim 54, further comprising:

a filtering component that employs cross-product, smoothed via prior on counts smoothing techniques, wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

131. The system of claim 130, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

132. The system of claim 54, further comprising:

a filtering component that employs cross-product, smoothed via informative priors on measures of association smoothing techniques.

133. The system of claim 132, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

134. The system of claim 132, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

135. The system of claim 134, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

136. The system of claim 54, further comprising:

a filtering component that employs cross-product, smoothed via nonuniform prior smoothing techniques.

137. The system of claim 136, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

138. The system of claim 136, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

139. The system of claim 138, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

140. The system of claim 54, further comprising:

a filtering component that employs log of cross-product, smoothed via cutoff smoothing techniques.

141. The system of claim 140, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

142. The system of the claim 140, the multi-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

143. The system of claim 142, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

144. The system of claim 54, further comprising:

a filtering component that employs log of cross-product, smoothed via prior on counts smoothing techniques, wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

145. The system of claim 144, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

146. The system of claim 54, further comprising:

a filtering component that employs log of cross-product, smoothed via informative priors on measures of association smoothing techniques.

147. The system of claim 146, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

148. The system of claim 146, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

149. The system of claim 148, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

150. The system of claim 54, further comprising:

a filtering component that employs log of cross-product, smoothed via nonuniform prior smoothing techniques.

151. The system of claim 150, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

152. The system of claim 150, the multiple-score collaborative filtering evaluation method comprising utilizing a highest value score of scores applicable to an item as the single score.

153. The system of claim 152, the item set comprising a higher-order item set wherein more than one item is represented on a left-hand side of an association rule applicable to at least one item in the item set.

154. The system of claim 54, the smoothing component selects, based on the item set, the selected smoother from among known smoothers.

155. The system of claim 54, the measure of association selection component selects at least one additional measure of association based on the item set, and a filtering component scores at least one item of the item set by employing the at least one additional measure of association.