×

Methods and systems for selecting and presenting content based on a comparison of preference signatures from multiple users

  • US 8,380,726 B2
  • Filed: 03/06/2007
  • Issued: 02/19/2013
  • Est. Priority Date: 03/06/2006
  • Status: Active Grant
First Claim
Patent Images

1. A method of selecting and presenting a subset of content items to a user based on content item selections by other users having similar preferences as determined at least in part based on analyzing the weights of relative preferences and based on analyzing intervals of time occurring between user selection of similar content items, the method comprising:

  • providing access to a set of content items, each content item having at least one associated descriptive term to describe the content item;

    receiving a plurality of incremental inputs, each incremental input being entered by a corresponding one of a plurality of users of an input device, wherein the incremental input includes at least one character in a series of characters;

    in response to each character of the incremental input, presenting a corresponding subset of content items;

    receiving selection actions of content items from the user who entered the incremental input for which the subset of content items was presented;

    determining a geographic location of the user at the time of at least one of the selection actions;

    analyzing the descriptive terms associated with the selected content items to identify sets of actions resulting in the selection of similar content items, wherein similarity is determined by comparing the descriptive terms associated with any one of the selected content items with any of the previously selected content items;

    analyzing the date, day, and time of at least three of the individual selection actions of the sets of actions to learn arbitrary periodicities corresponding to a particular set of actions for selecting similar content items, wherein the arbitrary periodicities each comprise repeated similar intervals of time between user actions resulting in the selections of similar content items;

    for each user of the plurality of users, learning the preferences of each user by associating a weight and arbitrary periodicity with each descriptive term associated with the selection actions of the user, said weight being indicative of the amount of use by the user of the selected content items associated with each of the descriptive terms and said arbitrary periodicity being the learned arbitrary periodicity of the descriptive terms associated with the set of similar content items, so that the user preferences are specified as a set of descriptive terms, each with associated weights and arbitrary periodicities, the learning the preferences of each user further including associating the geographic location with each descriptive term associated with the corresponding selection actions of the user;

    identifying users with similar preferences by comparing sets of learned user preferences to determine if the sets of preferences of the users match within a specified threshold; and

    presenting content items to at least one user of the plurality of users by identifying for presentation content items selected by users identified as having learned user preferences that are similar to the learned user preferences of said at least one user;

    wherein at least one of receiving a plurality of incremental inputs, presenting the corresponding subset of content items, receiving selection actions, learning the preferences of each user, identifying users with similar preferences, and presenting content items is performed on a server system remote from the user.

View all claims
  • 11 Assignments
Timeline View
Assignment View
    ×
    ×