AN OVERBROAD PATENT ON providing recommendations based on relative preferences across items for a customer. - This application from MIT seeks to patent the idea of...identifying a relative preference list of all users, comparing a user's relative preference list with the canonical preference list, and inferring preferences of the user on the basis of comparison with other users. 10 minutes of your time can help narrow US patent applications before they become patents. Follow @askpatents on twitter to help.
QUESTION - Have you seen anything that was published before 8/29/2011 that discusses:
- Providing a personalized recommendation to a user regarding items that the user has not ranked.
If so, please submit evidence of prior art as an answer to this question. We welcome multiple answers from the same individual.
EXTRA CREDIT - A user's preference is identified based on pair-wise preferences of items.
TITLE: Determining preferences of a user based on comparative rating across items.
Summary: [Translated from Legalese into English] Comparing preference of a user with identified preferences of all users and inferring preference of the user based on a correlation factor between the user preference and each of the identified preferences.
- Publication Number: US 20130054498 A1
- Application Number: US 13/598,124
- Assignee: MIT
- Prior Art Date: Seeking prior Art predating 8/29/2011
- Link to Google Prior Art Search - "Find Prior Art"
Claim 1 requires each and every step below:
A method for inferring the preferences of a user in relation to a population, comprising the steps of:
identifying a preference distribution of said population comprising a plurality of ranked lists and a weight of each of said plurality of ranked lists;
comparing a revealed preference of said user to one of said plurality of ranked lists;
assigning an affinity weight between said user and each of said plurality of ranked lists; and
taking a weighted average of each of said affinity weights.
In English this means:
A method for inferring preferences of a user, the method comprises the following steps:
identifying preferences of all users over a number of choices;
comparing preferences of the user with each of the identified preferences;
assigning a correlation factor between the user preferences and each of the identified preferences; and
averaging the assigned correlation factors to infer preferences of the user.
Good prior art would be evidence of a system that did each and every one of these steps prior to 8/29/2011
You're probably aware of ten pieces of art that meet this criteria already... separately, the applicant is claiming A preference list providing comparisons between every pair of items for a user is represented by a comparison matrix indexed by i and j, where entry (i, j) is set to 1 if item i is preferred to item j, and 0 otherwise.
"Providing recommendations based on pair-wise comparison of items."
What is good prior art? Please see our FAQ.
Want to help? Please vote or comment on submissions below. We welcome you to post your own request for prior art on other questionable US Patent Applications.