Method for implementing search engine
First Claim
Patent Images
1. A method of influencing behavior of an electronic World Wide Web (WWW) Internet search engine, to identify search items in response to a query from an online network user, comprising:
- compiling a set of N popular items {X1, X2, . . . , Xi, . . . , XN},wherein for each popular item Xi amongst the set of N popular items, a plurality of M early adopter web pages {Y[Xi]1. . . Y[Xi]M} is determined resulting in N separate sets of M early adopter web pages, said M early adopter web pages being determined based on a time of adoption of said popular item Xi at said M early adopter web pages;
correlating said N separate sets of M early adopter web pages to identify a trend predictor value of each of the early adopter web pages and to determine which of the early adopter web pages occur most frequently in the set {Y[Xi]j}, where 1≦
i≦
N and 1≦
j≦
M,wherein said most frequently occurring early adopter web pages are qualified as trend predictors;
measuring an adoption rating for a first search item identified in a set of web pages, said set of web pages including one or more web pages in said N separate sets of at least M early adopter web pages; and
responding to a query from a user for search items at least in part based on said adoption rating for said first search item.
2 Assignments
0 Petitions
Accused Products
Abstract
A system and method for implementing/influencing a search engine which provides search results to users based on characteristics of certain trendsetter web pages identified on the Internet. The trendsetter web page are determined by studying historical adoption behavior of a group within the universe of websites, or by reference to known indicia.
-
Citations
22 Claims
-
1. A method of influencing behavior of an electronic World Wide Web (WWW) Internet search engine, to identify search items in response to a query from an online network user, comprising:
-
compiling a set of N popular items {X1, X2, . . . , Xi, . . . , XN}, wherein for each popular item Xi amongst the set of N popular items, a plurality of M early adopter web pages {Y[Xi]1. . . Y[Xi]M} is determined resulting in N separate sets of M early adopter web pages, said M early adopter web pages being determined based on a time of adoption of said popular item Xi at said M early adopter web pages; correlating said N separate sets of M early adopter web pages to identify a trend predictor value of each of the early adopter web pages and to determine which of the early adopter web pages occur most frequently in the set {Y[Xi]j}, where 1≦
i≦
N and 1≦
j≦
M,wherein said most frequently occurring early adopter web pages are qualified as trend predictors; measuring an adoption rating for a first search item identified in a set of web pages, said set of web pages including one or more web pages in said N separate sets of at least M early adopter web pages; and responding to a query from a user for search items at least in part based on said adoption rating for said first search item. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
-
-
12. A method of classifying web pages, comprising:
-
using a computer to identify a set of web pages; extracting, electronically, a plurality of M early adopter web pages from said set of web pages, which plurality of M early adopter web pages are determined by ranking said set of web pages in accordance with a time in which they first make a reference to a predetermined set of N popular items {X1, X2, . . . , Xi, . . . , XN}, wherein for each popular item Xi amongst the set of N popular items, the set of M early adopter web pages {Y[Xi]1 . . . Y[Xi]M} is determined, resulting in N separate sets of M early adopter web pages; correlating, electronically, said N separate sets of M early adopter web pages to identify a trend predictor value of each of the early adopter web pages and to determine which of the early adopter web pages occur most frequently in the set {Y[Xi]j}, where 1≦
i≦
N and 1≦
j≦
M,wherein said most frequently occurring early adopter web pages are qualified as trend predictors; and increasing a ranking for a web page in response to a search request posed to a search engine when said web page is from said set of M early adopter web pages. - View Dependent Claims (13, 14, 15, 16, 17)
-
-
18. A method of influencing behavior of an electronic search engine system, to provide results to queries initiated by a community of online users of the World Wide Web, comprising:
-
identifying N popular items {X1, X2, . . . , Xi, . . . , XN}, as measured by interest demonstrated by online users for such items, wherein for each popular item Xi amongst the N popular items, a plurality of M early adopter web pages {Y[Xi]1. . . Y[Xi]M} is determined resulting in N separate sets of M early adopter web pages; correlating said N separate sets of M early adopter web pages to identify a trend predictor value of each of the early adopter web pages and to determine which of the early adopter web pages occur most frequently in the set {Y[Xi]j}, where 1≦
i≦
N and 1≦
j≦
M,wherein said most frequently occurring early adopter web pages are designated as trend predictor web pages; measuring a trend predictor web page adoption rating for a particular item; modifying a search result provided by the search engine to an online user based on said adoption rating, wherein said adoption rating is used, at least during certain time periods, to influence which search results are presented to a user query; eliminating statistically said trend predictor web pages that have adopted over a certain threshold of popular items, to avoid biasing of the search result; and presenting online advertising to users of trend predictor web pages based on said trendsetter web page adoption rating. - View Dependent Claims (19, 20, 21, 22)
-
Specification