US 6,584,468 B1Granted

Method and apparatus to retrieve information from a network

AssigneeNineSigma, Inc.

InventorsKaigham J. Gabriel, Evan M. Indianer, Christopher M. Umbel, Joel Lenhart

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method and apparatus to index network information is described. A network is searched for files of information relevant to people and resources in a particular field using a search list of weighted links to said files. The information is parsed into content and additional links to additional files. The content is weighted and copied to memory (such as a database). A determination is made as to whether the additional links are relevant to the people and resources in the given technical field. Those additional links that are relevant are weighted using a predetermined weighting algorithm. The relevant additional weighted links are copied to the search list. This process continues until an ending condition occurs.

Claims28

Dependent claims

A method to index network information, comprising:

(a) establishing a list of parameters to be searched over a network;
(b) establishing a search list of weighted links;
(c) assigning a predetermined initial weight to each parent link in said search list;
(d) searching said network for files of information containing one or more of said parameters using said search list of weighted links to said files;
(e) parsing said information into content and additional links to additional files;
(f) weighting said content;
(g) copying said weighted content into a memory;
(h) comparing each of said additional links to a locally-stored excluded domain file, wherein said excluded domain file contains a list of irrelevant links;
(i) identifying those of said additional links that are found relevant in accordance with said comparison using said excluded domain file;
(j) assigning a predetermined link weight to each said relevant additional link, wherein each said relevant additional link is initially assigned an identical link weight that is different from said initial weight of a corresponding parent link;
(k) adjusting said link weight of each said relevant additional link to be more than, less than or equal to said initial weight of said corresponding parent link depending on at least one of the following:
whether said relevant additional link has been previously processed,
whether said relevant additional link has been previously unprocessed, and
a number of said parameters present in said content corresponding to said relevant additional link;
(l) copying said relevant additional weighted links to said search list;
and
(m) performing steps (d)-(l) until an ending condition occurs.

14 more elements

A machine-readable medium whose contents cause a computer system to index network information by performing the following:

(a) establishing a list of parameters to be searched over a network;
(b) establishing a search list of weighted links;
(c) assigning a predetermined initial weight to each parent link in said search list;
(d) searching said network for files of information containing one or more of said parameters using said search list of weighted links to said files;
(e) parsing said information into content and additional links to additional files;
(f) weighting said content;
(g) copying said weighted content into a memory;
(h) comparing each of said additional links to a locally-stored excluded domain file, wherein said excluded domain file contains a list of irrelevant links;
(i) identifying those of said additional links that are found relevant in accordance with said comparison using said excluded domain file;
(j) assigning a predetermined link weight to each said relevant additional link, wherein each said relevant additional link is initially assigned an identical link weight that is different from said initial weight of a corresponding parent link;
(k) adjusting said link weight of each said relevant additional link to be more than, less than or equal to said initial weight of said corresponding parent link depending on at least one of the following:
whether said relevant additional link has been previously processed,
whether said relevant additional link has been previously unprocessed, and
a number of said parameters present in said content corresponding to said relevant additional link;
(l) copying said relevant additional weighted links to said search list;
and
(m) performing steps (d)-(l) until an ending condition occurs.

14 more elements

An apparatus to index network information, comprising:

(a) means for establishing a list of parameters to be searched over a network;
(b) means for establishing a search list of weighted links;
(c) means for assigning a predetermined initial weight to each parent link in said search list;
(d) means for searching said network for files of information containing one or more of said parameters using said search list of weighted links to said files;
(e) means for parsing said information into content and additional links to additional files;
(f) means for weighting said content;
(g) means for copying said weighted content into a memory;
(h) means for comparing each of said additional links to a locally-stored excluded domain file, wherein said excluded domain file contains a list of irrelevant links;
(i) means for identifying those of said additional links that are found relevant in accordance with said comparison using said excluded domain file;
(j) means for assigning a predetermined link weight to each said relevant additional link, wherein each said relevant additional link is initially assigned an identical link weight that is different from said initial weight of a corresponding parent link;
(k) means for adjusting said link weight of each said relevant additional link to be more than, less than or equal to said initial weight of said corresponding parent link depending on at least one of the following:
whether said relevant additional link has been previously processed,
whether said relevant additional link has been previously unprocessed, and
a number of said parameters present in said content corresponding to said relevant additional link;
(l) means for copying said relevant additional weighted links to said search list;
and
(m) means for performing functions in (d)-(l) until an ending condition occurs.

14 more elements

Dependencies are inferred from the claim text.