System and method for determining affinity using objective and subjective data

• US 6,697,800 B1
• Filed: 05/19/2000
• Issued: 02/24/2004
• Est. Priority Date: 05/19/2000
• Status: Expired due to Term

• Current Assignee: Moon Glow, Series 82 of Allied Security Trust I (Allied Security Trust)
• Original Assignee: ROXIO Incorporated (MelodyVR Group Plc)
• Inventors: Jannink, Jan F., Shivakumar, Narayanan, Schirmer, Thomas E.
• Primary Examiner(s): Robinson, Greta
• Assistant Examiner(s): Pannala, Sathyanaraya R
• Application Number:
  US09/574,108
• Time in Patent Office: 1,376 Days
• Field of Search: 707/1-10,100-104.1 704/9 395/760
• US Class Current: 1/1
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/957 Browsing optimisation, e.g. caching or content distillation
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein each objective item in the plurality includes a plurality of objective item properties and the search item includes one or more search item properties, the method comprising: grouping a subset of objective items based on an objective relationship between one or more of the plurality of objective item properties and the one or more search item properties; modifying the subset based on a subjective relationship, wherein the subjective relationship indicates an association between objective items in the subset and the search item, thereby calculating an affinity value between each objective item in the subset and the search item; and ranking the objective items in the modified subset based on the affinity value calculated.

Claims

1. A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein each objective item in the plurality includes a plurality of objective item properties and the search item includes one or more search item properties, the method comprising:

a. grouping a subset of objective items based on an objective relationship between one or more of the plurality of objective item properties and the one or more search item properties;

b. modifying the subset based on a subjective relationship, wherein the subjective relationship indicates an association between objective items in the subset and the search item, thereby calculating an affinity value between each objective item in the subset and the search item, and wherein subjective data of the subjective relationship is stored in a subjective database;

c. ranking the objective items in the modified subset based on the affinity value calculated; and

d. scaling the affinity value computed by one or more predetermined normalization weights, wherein the one or more predetermined normalization weights are initially set by the user.

2. The method according to claim 1 further comprising computing one or more subjective property values based on subjective data collected, wherein the one or more subjective property values indicate a strength of the association between the plurality of objective items and the search item.

3. The method according to claim 2 further comprising adjusting the affinity value based on the one or more subjective property values computed.

4. A method according to claim 1 further comprising providing the user with the objective items in the modified first subset.

5. The method according to claim 4 wherein the ranked order of objective items provided to the user is displayed via a user interface.

6. The method according to claim 1 wherein the plurality of objective items are stored in an objective database.

7. The method according to claim 6 further comprising updating the subset of objective items as a new objective item is added to the objective database.

8. The method according to claim 2 wherein the subjective data is collected based on user browsing activity over a predetermined period of time.

9. The method according to claim 2 wherein the subjective data is provided directly by the user.

10. A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein each objective item in the plurality includes a plurality of objective item properties and the search item includes one or more search item properties, the method comprising:

a. grouping a subset of objective items based on an objective relationship between one or more of the plurality of objective item properties and the one or more search item properties;

b. modifying the subset based on a subjective relationship, wherein the subjective relationship indicates an association between objective items in the subset and the search item, thereby calculating an affinity value between each objective item in the subset and the search item, and wherein the subset of objective items is modified by a cache aging technique;

c. ranking the objective items in the modified subset based on the affinity value calculated; and

d. scaling the affinity value computed by one or more predetermined normalization weights, wherein the one or more predetermined normalization weights are initially set by the user.

11. A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein each objective item in the plurality includes a plurality of objective item properties and the search item includes one or more search item properties, the method comprising:

a. grouping a subset of objective items based on an objective relationship between one or more of the plurality of objective item properties and the one or more search item properties;

b. modifying the subset based on a subjective relationship, wherein the subjective relationship indicates an association between objective items in the subset and the search item, thereby calculating an affinity value between each objective item in the subset and the search item, the affinity value being s calculated by taking a difference between the one or more search item properties and each of the plurality of objective item properties;

c. computing one or more subjective property values based on subjective data collected, wherein the one or more subjective property values indicate a strength of the association between the plurality of objective items and the search item;

d. ranking the objective items in the modified subset based on the affinity value calculated; and

e. scaling the affinity value computed by one or more predetermined normalization weights, wherein the one or more predetermined normalization weights are initially set by the user.

12. A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein each objective item in the plurality includes a plurality of objective item properties and the search item includes one or more search item properties, the method comprising:

a. grouping a subset of objective items based on an objective relationship between one or more of the plurality of objective item properties and the one or more search item properties;

b. modifying the subset based on a subjective relationship, wherein the subjective relationship indicates an association between objective items in the subset and the search item, thereby calculating an affinity value between each objective item in the subset and the search item;

c. ranking the objective items in the modified subset based on the affinity value calculated; and

d. scaling the affinity value computed by one or more predetermined normalization weights, wherein the one or more predetermined normalization weights are initially set by the user.

wherein the subset of objective items is updated using an equation;

13. A method of determining a relationship between a search item provided by a user and a plurality of objective items, wherein the search item includes one or more search item properties and each objective item in the plurality includes one or more objective item properties, the method comprising:

a. computing an affinity value between each of the one or more search item properties and each of the one or more objective item properties;

b. grouping a number of objective items into an appropriate one or more clusters, wherein the number of objective items grouped have the affinity value within a predetermined distance with respect to the search item;

c. partitioning a plurality of subjective data into one or more bins, wherein each of the one or more bins corresponds to an appropriate one or more clusters;

d. assigning a rule within each of the one or more bins, each rule representing a user defined relationship between the objective items and the search item;

e. computing a subjective property value for each rule, wherein the subjective property indicates a strength of the user-defined relationship;

f. adjusting the affinity value based on the subjective property value computed, thereby generating an adjusted affinity value; and

g. updating the one or more clusters, wherein the plurality of objective items are rearranged between the one or more clusters as a result of the adjusted affinity value.

14. The method according to claim 13 further comprising scaling the affinity value computed by one or more predetermined normalization weights.

15. The method according to claim 14 wherein the one or more predetermined normalization weights are initially set by the user.

16. The method according to claim 13 further comprising providing the plurality of objective items in the updated one or more clusters to the user in a ranked order based on the adjusted affinity value of each objective item.

17. The method according to claim 16 wherein the ranked order provided to the user is displayed via a user interface.

18. The method according to claim 13 wherein the plurality of objective items are stored in an objective database.

19. The method according to claim 13 wherein the plurality of subjective data are stored in a subjective database.

20. The method according to claim 18 further comprising updating the one or more clusters as a new objective item is added to the objective database.

21. The method according to claim 13 wherein the affinity value is calculated by taking a difference between the one or more search item properties and the one or more objective item properties.

22. The method according to claim 13 wherein the affinity value is adjusted using an equation:

23. A system for retrieving information in response to a search item, the system comprising:

a. an objective database including a plurality of objective items;

b. a subjective database having one or more rules, wherein each of the one or more rules represents a subjective relationship between one or more of the plurality of objective items and the search item;

c. a collaborative filtering module for assigning a subjective property value to each of the one or more rules, wherein the subjective property value indicates a strength of the subjective relationship; and

d. an affinity calculation module for determining an affinity value between the search item and each of the plurality of objective items in the objective database, wherein one or more objective items having an affinity value within a predetermined distance are initially grouped together in one or more clusters, whereby the affinity calculation module rearranges the one or more objective items in the one or more clusters upon receiving the subjective property value from the collaborative filtering module.

24. The system according to claim 23 wherein the affinity calculation module determines the affinity value between the search item and each of the objective items by taking a difference between one or more search item properties and one or more objective item properties.

25. The system according to claim 23 wherein the affinity calculation module rearranges the one or more objective items by computing an equation: