Automatic identification of unreliable user ratings

  • US 7,519,562 B1
  • Filed: 03/31/2005
  • Issued: 04/14/2009
  • Est. Priority Date: 03/31/2005
  • Status: Active Grant
First Claim
Patent Images

1. A method for a computing system to automatically detect unreliability of a rating from a user based at least in part on a pattern identified using prior ratings of the user, the method comprising:

  • receiving an indication of a rating that reflects an evaluation by a user related to a target item that is available for acquisition, the indicated rating being a rating by the user that is of a review of the target item provided by another user;

    obtaining one or more indications of multiple other ratings by the user that reflect evaluations related to other items;

    automatically analyzing by the computing system the indicated rating and the other ratings to determine whether one or more patterns in the indicated rating and other ratings identify an unreliability of the indicated rating, wherein the computing system is configured with instructions to perform the automatic analyzing of the indicated rating and the other ratings to determine whether one or more patterns in the indicated rating and other ratings identify an unreliability of the indicated rating; and

    if the unreliability of the indicated rating is identified based on the one or more patterns, providing an indication that the indicated rating is unreliable so as to exclude use of the indicated rating as part of an aggregate rating related to the target item, and otherwise including the use of the indicated rating as part of the aggregate rating related to the target item.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×