Facility for the intelligent selection of information objects (persona)

• US 5,854,923 A
• Filed: 06/21/1994
• Issued: 12/29/1998
• Est. Priority Date: 06/21/1994
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Farber, Joel F., Dockter, Michael J., Seppi, Kevin D.
• Primary Examiner(s): Black, Thomas G.
• Assistant Examiner(s): ALAM, HOSAIN T
• Application Number:
  US08/262,834
• Time in Patent Office: 1,652 Days
• Field of Search: 395/600, 395/159, 395/161, 395/700, 395/148-157, 395/50-60, 395/605, 395/160
• US Class Current: 707/694
• CPC Class Codes: G06F 16/907 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A system to intelligently link a computer user to enormous databases of multi-media information. The system provides an automated facility to determine the value of information based on each user'"'"'s subjective preferences. An execution is created by comparing these preferences with prior characterizations of information objects. Once the comparisons are made, an ordered list is compiled. The list provides a set of data sources starting with the most relevant and ending with the least relevant. By selecting and reviewing only the most significant references, the probability of locating a pertinent reference is increased dramatically while increasing the effectiveness of time spent reviewing results.

Claims

1. A computer-based method for ordering information objects for presentation to a requester, comprising the steps of:

receiving a set of information objects, connections between the information objects, and attributes associated with the connections;

receiving a first profile descriptor representing a first user'"'"'s information preferences and predetermined characteristics;

said first profile descriptor including a preference field and a confidence field associated with said preference field;

determining a subset of the information objects, connections, and attributes based on said first profile descriptor;

computing at least first utility values for the connections in the determined subset as a function of the connection attributes and the first profile descriptor;

normalizing the utility values, and

wherein the preferences of the information objects identified by the connections are ordered according to the normalized utility values computed for the connections.

2. The method as recited in claim 1, further comprising the steps of:

receiving a second profile descriptor representing a second user'"'"'s information preferences and prior characterizations;

determining a second subset of the information objects, connections, and attributes based on the second profile descriptor;

computing utility values for the connections in the second determined subset as a function of the connection attributes and the profile descriptor for the second user;

combining the normalized utility values for the first user with normalized utility values for the second user to form meta utility values, and

ordering information objects identified by the connections according to the meta utility values computed for the connections.

3. The method as recited in claim 1, wherein the received first user profile descriptor represents user information preferences and prior characterizations for a computer user other than the first computer user receiving the user profile descriptor.

4. The method as recited in claim 1, wherein the received user profile descriptor includes combined information preferences for a plurality of computer users.

5. The method as recited in claim 1, wherein the connection attributes are valued meta data.

6. The method as recited in claim 1, wherein the function used to compute the utility values is a neural net.

7. The method as recited in claim 1, wherein the function used to compute the utility values is rule-driven.

8. The method as recited in claim 1, wherein the function used to compute the utility values is a regression.

9. The method as recited in claim 1, wherein the function used to compute the utility value is the combination of two or more mathematical functions which produce utility values.

10. The method as recited in claim 1, further comprising the steps of:

computing second utility values for the connections in the determined subset as a second function for the connection attributes and the first profile descriptor;

normalizing the second utility values, and

ordering information objects identified by the connections according to the normalized second utility values computed for the connections.

11. The method as recited in claim 1, further comprising the steps:

comparing the normalized utility values to a pre-determined critical utility level; and

highlighting or coloring the presentation of information objects associated with normalized utility values above the predetermined critical utility level.

12. The method as recited in claim 1, wherein multiple user preferences may be layered.