Collections of linked databases

• US 8,635,217 B2
• Filed: 03/15/2007
• Issued: 01/21/2014
• Est. Priority Date: 09/15/2004
• Status: Active Grant

• Current Assignee: 3degrees LLC
• Original Assignee: 3degrees LLC
• Inventors: Markus, Michael J., McGuire, Heather A., Kionga-Kamau, Peter M., Smith, Brian N.
• Primary Examiner(s): Hasan, Syed
• Assistant Examiner(s): MUELLER, KURT A
• Application Number:
  US11/686,416
• Publication Number:
  US 20080228947A1
• Time in Patent Office: 2,504 Days
• Field of Search: 707/734
• US Class Current: 707/734
• CPC Class Codes: G06F 16/906 Clustering; Classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering

Abstract

This invention is generally directed to one or more methods or systems relating to inferential networks. This invention is further directed to methods and systems for making and using inferential networks. Still further, this invention is generally directed to making one or more inferential networks by using data generated from directed searches of collections of linked databases.

Claims

1. A method for ordering search results using inferential network data, said method comprising:

searching a first network of linked nodes for a search term, each of the nodes connected to another one of the nodes by a link, wherein the search begins at a node in the first network specified in the search query;

retrieving stored inferential network data describing inferences made about nodes in the first network, the inferences derived from electronic search data collected from previously conducted searches of the first network, the electronic search data including a) search terms used in the previously conducted searches and b) corresponding search results, wherein the inferential network data describes the inferences as a second, inferential network of linked inferential nodes, each of the inferential nodes corresponding to a node of the first network and connected to another one of the inferential nodes by an inferential link, the inferential link including a strength, the strength increasing with the number of previously conducted searches that result in an inference corresponding to the inferential link and decreasing as a function of time, wherein a previously conducted search results in an inference corresponding to the inferential link when a node of the first network associated with the inferential link is designated as an anchor database or a node of the first network associated with the inferential link is designated as an anchor database in a search for an identical or related search term;

for each of the results of the search;

identifying an inferential node of the inferential network that represents the result; and

assigning a weight to the result based on the combined strengths of all inferential links leading to the identified node from other inferential nodes; and

weighing the results of the search based on the assigned weights of the results; and

displaying the results of the search ordered according to the weighing.

2. The method of claim 1, wherein inferential network data is selected from the group consisting of search data from searches conducted using an anchor database, broadcast-search data, and a combination of broadcast-search search data and search data from searches conducted using an anchor database.

3. The method of claim 1, wherein the inferential network nodes are members of a social network.

4. The method of claim 1, wherein the inferential network nodes are internet or intranet websites.

5. The method of claim 1, wherein the inferential link is a subject matter based qualification derived from a search term used in a search that included a designated anchor database.