Data mining of user activity data to identify sequential item acquisition patterns

Abstract

A data mining component collectively analyzes item acquisition histories of users of an electronic catalog of items and identifies pairs of catalog items that tend to be acquired sequentially. The data mining component may also generate data regarding such sequential item acquisition patterns. For example, the data mining component may determine whether user acquisitions of the two items tend to be spaced apart in time by a characterizing time interval, and/or may determine percentages of users who have followed particular sequential acquisition patterns. Information regarding the detected sequential item acquisition patterns may be exposed to users on electronic catalog pages, and/or may be used to select the timing with which particular items are recommended to users.

1 . A data mining method, comprising:
 storing, in computer storage, item acquisition data of users of an electronic catalog of items, said item acquisition data including information reflective of timings of item acquisition events, said electronic catalog including item detail pages that correspond to particular catalog items;   detecting, based on an analysis of the item acquisition data by a computer system, a sequential item acquisition pattern in which users who acquire a first catalog item tend to subsequently acquire a second catalog item; and   causing an indication of the sequential item acquisition pattern to be incorporated into an item detail page for the first catalog item, to thereby expose an existence of the sequential item acquisition pattern to users of the electronic catalog.   
     
     
         2 . The method of  claim 1 , further comprising generating, by the computer system, based on the stored item acquisition data, statistical data regarding the sequential item acquisition pattern, and causing said statistical data to be incorporated into the item detail page with the indication of the sequential item acquisition pattern. 
     
     
         3 . The method of  claim 2 , wherein the statistical data comprises a representation of an amount of time users typically wait to acquire the second catalog item after acquiring the first catalog item. 
     
     
         4 . The method of  claim 2 , wherein the statistical data comprises data regarding what percentage of users who acquire the first catalog item subsequently acquire the second catalog item. 
     
     
         5 . The method of  claim 4 , wherein the percentage is tied to a bounded time interval range. 
     
     
         6 . The method of  claim 2 , wherein generating the statistical data comprises determining, based on an analysis of time intervals between user acquisitions of the first and second catalog items, whether a characterizing time interval exists that represents a typical amount of time users wait to acquire the second catalog item after acquiring the first catalog item. 
     
     
         7 . The method of  claim 1 , wherein the item acquisitions are item purchases. 
     
     
         8 . Non-transitory computer storage that stores executable program code that directs a computer system comprising one or more computers to perform a process that comprises:
 storing, in computer storage, data regarding item acquisitions of users of an electronic catalog of items, said data including information reflective of timings of item acquisition events;   detecting, based on an analysis of the stored data regarding item acquisitions, a sequential item acquisition pattern in which users who acquire a first catalog item subsequently acquire a second catalog item; and   causing an indication of the sequential item acquisition pattern to be incorporated into an electronic catalog page associated with the first catalog item, to thereby expose an existence of the sequential item acquisition pattern to users of the electronic catalog.   
     
     
         9 . The non-transitory computer storage of  claim 8 , wherein the process further comprises generating, based on the stored data regarding item acquisitions, statistical data regarding the sequential item acquisition pattern, and causing said statistical data to be incorporated into the electronic catalog page in association with the indication of the sequential item acquisition pattern. 
     
     
         10 . The non-transitory computer storage of  claim 9 , wherein the statistical data comprises a representation of an amount of time users typically wait to acquire the second catalog item after acquiring the first catalog item. 
     
     
         11 . The non-transitory computer storage of  claim 9 , wherein the statistical data comprises data regarding what percentage of users who have acquired the first catalog item have subsequently acquired the second catalog item. 
     
     
         12 . The non-transitory computer storage of  claim 9 , wherein generating the statistical data comprises determining, based on an analysis of time intervals between user acquisitions of the first and second catalog items, whether a characterizing time interval exists that represents a typical amount of time users wait to acquire the second catalog item after acquiring the first catalog item. 
     
     
         13 . The non-transitory computer storage of  claim 8 , wherein the item acquisitions are item purchases. 
     
     
         14 . The non-transitory computer storage of  claim 8 , in combination with the computer system, wherein the computer system is programmed with said executable program code to perform said process. 
     
     
         15 . A data mining method, comprising:
 storing, in computer storage, item acquisition data of users of an electronic catalog of items, said item acquisition data including information reflective of timings of item acquisition events;   identifying a pair of catalog items, item A and item B, that, based on said item acquisition data, have been acquired in the sequence item A followed by item B by each of a plurality of said users; and   determining, based on time intervals between user acquisitions of item A and item B among said plurality of users, an amount of time users typically wait to acquire item B after acquiring item A;   said method performed by a computer system that comprises one or more computers.   
     
     
         16 . The data mining method of  claim 15 , wherein the amount of time is determined as a range of time intervals. 
     
     
         17 . The data mining method of  claim 15 , wherein determining the amount of time users typically wait comprises determining whether a characterizing time interval exists. 
     
     
         18 . The data mining method of  claim 15 , further comprising programmatically using the determined amount of time to select a timing with which to recommend item B to a user who has acquired item A. 
     
     
         19 . The data mining method of  claim 15 , further comprising causing an electronic catalog page associated with item A to be supplemented with an indication of said amount of time users typically wait to acquire item B after acquiring item A. 
     
     
         20 . The data mining method of  claim 19 , further comprising calculating, by the computer system, what percentage of users who have acquired item A have acquired item B after waiting said amount of time, and causing said electronic catalog page to be supplemented with an indication of said percentage.