System and method for obtaining user preferences and providing user recommendations for unseen physical and information goods and services

• US 6,801,909 B2
• Filed: 07/23/2001
• Issued: 10/05/2004
• Est. Priority Date: 07/21/2000
• Status: Expired due to Term

• Current Assignee: Triplehop Technologies (Oracle Corporation)
• Original Assignee: Triplehop Technologies (Oracle Corporation)
• Inventors: Delgado, Joaquin, LaPlanche, Renaud, Turck, Mathias
• Primary Examiner(s): Coby, Frantz
• Application Number:
  US09/909,997
• Publication Number:
  US 20020052873A1
• Time in Patent Office: 1,170 Days
• Field of Search: 707/104.1, 707/2, 707/10, 707/3, 707/7, 705/27, 705/10, 705/26, 709/203, 345/745, 700/83, 706/45
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9532 Query formulation
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A system and method for searching data repositories and for obtaining user preferences and providing user recommendations for unseen physical and information goods and services. The system includes, for example, four levels of filtering: content-based, collaborative, event-based and context-based. Filtering is designed to understand and anticipate a user'"'"'s physical and information goods and services needs by learning about the user'"'"'s preferences and the preferences of users similar to the user.

Claims

1. A method for obtaining user preferences, comprising:

providing input search criteria;

matching information about a user to information stored in a database, the database having a multi-level structure enabling grouping of the information for scalability, the information about the user modified by an event filter;

selecting a target attribute and calculating a long-term memory value for the user, and setting a short-term memory value to 1 each time the target attribute is selected;

ranking and displaying attributes according to a predicted relevance to the user based on the matched and selected calculations; and

updating the information in the database to reflect a similarity between users in a group and to better predict an individual user'"'"'s interest in one of the attributes.

2. The method according to claim 1, further comprising:

assigning a plurality of two-value pairs to each visited attribute, each two-value pair comprising a short-term memory (STM) value and a long term-memory (LTM) value, the LTM being calculated as the ratio of the number of times one of the users has selected the attribute divided by the total number of searches performed by the user, the STM decaying at β

∈

[0,1] based on d(STM^(t+1))=STM^(t)×

β

; and

averaging the LTM and STM to create an attribute-interest ratio, where an attribute interest level corresponds to the attribute-interest ratio.

3. The method according to claim 1, further comprising:

determining one of the user'"'"'s interest in an attribute using a collaborative filter by selecting the most visited attribute among the attributes common to the group, wherein a similarity between users in the group is used to predict the interest, a similarity calculated by comparing the attribute values in a compared users'"'"' profile tree with any overlap between the values using the following equation;

$\mathrm{sim}<br><br>\left(U,i\right)=\frac{\underset{j=1}{\overset{M}{\sum <br><br>}}<br><br>\text{\hspace{1em}<br><br>}<br><br>{\mathrm{\hslash <br><br>}}^2*v_{U,j}*v_{i,j}}{\sqrt{\underset{j=1}{\overset{M}{\sum <br><br>}}<br><br>\text{\hspace{1em}<br><br>}<br><br>{\left(\mathrm{\hslash <br><br>}*v_{U,j}\right)}^2<br><br>\underset{j=1}{\overset{N}{\sum <br><br>}}<br><br>\text{\hspace{1em}<br><br>}<br><br>{\left(\mathrm{\hslash <br><br>}*v_{i,j}\right)}^2}}.$

4. A system for obtaining information, comprising:

a content-based filter to acquire and maintain at least one user profile, and comparing instances being filtered using a similarity measure to evaluate relevance;

an attribute based collaborative filter to monitor shared impressions of a group of user profiles, the impressions rated such that unknown impressions can be selected and shown to specific users in the group based on a rating;

an event-based filter to track movements of users; and

a context-based filter to determine the a relevance of contextual information embedded in searches conducted by the users.

5. The system according to claim 4, further comprising:

an automated collaborative filter to determine users in the group having similar impressions, and to analyze and facilitate the creation of interest groups for the purpose of targeting specific users.