Item selection by prediction and refinement

• US 6,078,740 A
• Filed: 11/04/1996
• Issued: 06/20/2000
• Est. Priority Date: 11/04/1996
• Status: Expired due to Term

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: DeTreville, John D.
• Primary Examiner(s): Teska, Kevin J.
• Assistant Examiner(s): CHOI, KYLE JAEHUN
• Application Number:
  US08/740,737
• Time in Patent Office: 1,324 Days
• Field of Search: 395/500, 395/200.48, 395/200.49, 705/26, 705/10
• US Class Current: 703/22
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

In a computerized method for predicting a particular user preference for an item based on observations made about the item by other users, client computers are used to enter the observations about the items. The observations are forwarded to a server computer via a network. The observations are collected in a database of the server computer. Using factor analysis, a solver module of the server computer analyzes the observations to generate a model of the observations. The models are distributed to the client computers via the network. The client computer makes predictions of preferences of the particular user using the models.

Claims

1. A computerized method for predicting a particular person'"'"'s preferences for items based on observations made about the items by other persons, comprising:

at a plurality of client computers, collecting observations of the items by the other persons, and passing observation data corresponding to the observations to a server computer;

at the server computer, analyzing the observation data from the client computers to generate a model of the observations, and distributing copies of the model to the plurality of the client computers; and

at a particular one of the client computers associated with the particular person, receiving the model from the server computer, and predicting the preference of the particular person for a particular one of the items using the model.

2. The method of claim 1, including

at the server computer, repeatedly refining the model as additional observation data is received from the client computers, and distributing the refined model to the client computers.

3. The method of claim 2, including

the client computers, associated with various ones of the persons, receiving the model from the server computer, and predicting the preference of the persons associated with the client computers for each of a plurality of the items using the model.

4. The method of claim 1, wherein the observation data includes information identifying the persons associated with a particular observations about the items, information identifying the items, and scores indicating relative preferences for the items by the persons.

5. A distributed computer system, for predicting a particular person'"'"'s preferences for items based on observations made about the items by other persons, comprising:

a server computer;

a plurality of client computers for collecting observations of the items by the other persons and for passing observation data corresponding to the observations to the server computer;

the server computer including a database for storing the observation data from the client computers and apparatus for analyzing the observation data from the client computers to generate a model of the observations and for distributing copies of the model to the plurality of the client computers; and

the client computers including prediction means for receiving the model from the server computer for using the model to predict the preference of the particular person for a particular one of the items using the model.

6. The system of claim 5, wherein the server computer apparatus includes apparatus for repeatedly refining the model as additional observation data is received from the client computers, and for distributing the refined model to the client computers.

7. The system of claim 6, wherein the prediction means in the client computers is configured to be predicting the preferences of a plurality of different ones of the persons for each of a plurality of the items using the model.

8. The system of claim 5, wherein the observation data includes information identifying the persons associated with a particular observations about the items, information identifying the items, and scores indicating relative preferences for the items by the persons.

9. A computer program product for use in conjunction with a plurality of computers in a distributed computer system, the computer program product comprising a computer readable storage medium and a computer program mechanism embedded therein, the computer program mechanism comprising:

an observation collection module, for execution by each of a plurality of client computers in the distributed computer system, for collecting observations of the items by the other persons, and passing observation data corresponding to the observations to a server computer in the distributed computer system;

a server module, for execution by the server computer, for analyzing the observation data from the client computers to generate a model of the observations, and for distributing copies of the model to the plurality of the client computers; and

a prediction module, for use at a particular one of the client computers associated with the particular person, for receiving the model from the server computer, and for predicting the preference of the particular person for a particular one of the items using the model.

10. The computer program product of claim 9, wherein the server module includes instructions for repeatedly refining the model as additional observation data is received from the client computers and for distributing the refined model to the client computers.

11. The computer program product of claim 10, wherein the prediction module is for use the client computers, associated with various ones of the persons, and including instructions for predicting the preference of the persons associated with the client computers for each of a plurality of the items using the model.

12. The computer program product of claim 9, wherein the observation data includes information identifying the persons associated with a particular observations about the items, information identifying the items, and scores indicating relative preferences for the items by the persons.