Method and system for site path evaluation using web session clustering

• US 8,572,233 B2
• Filed: 07/15/2004
• Issued: 10/29/2013
• Est. Priority Date: 07/15/2004
• Status: Expired due to Fees

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Yu, Qingfeng, Lakshminarayan, Choudur K., Benson, Alan
• Primary Examiner(s): Kim, Hee Soo
• Application Number:
  US10/891,989
• Publication Number:
  US 20060015504A1
• Time in Patent Office: 3,393 Days
• Field of Search: 709/203, 709/223, 709/224, 709/217, 709/219, 709/227, 709/228, 709/229, 707/6, 707/7, 707/10
• US Class Current: 709/224
• CPC Class Codes: G06F 16/906 Clustering; Classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A method and system for site path evaluation using web session clustering is provided. The method identifies properties of a plurality of web site traversal paths. Data is acquired from a plurality of sessions corresponding to at least a portion of the plurality of web page traversal paths. Portions of the web site traversal paths are grouped into a unified web page category. The plurality of sessions is clustered into a plurality of web session clusters according to a similarity measure. One of the plurality of web session clusters most closely exhibiting at least one predefined metric is selected for analysis of the propertied of a web page traversal path contained therein. A system includes a plurality of web pages, a monitoring program and a computational process configured to receive data and identify properties of the plurality of traversal paths.

Claims

1. A method for identifying properties of a plurality of web page traversal paths, comprising:

acquiring data from a plurality of sessions corresponding to said plurality of web page traversal paths;

grouping, by a computer, said web page traversal paths into web page categories, wherein grouping said web page traversal paths into said web page categories includes correlating traffic amount relationships between specific ones of a plurality of web pages in said web page traversal paths to identify said web page categories, wherein correlating the traffic amount relationships between the specific web pages comprises correlating the specific web pages based on an amount of traffic between the specific web pages;

using, by the computer, the web page categories to map the plurality of sessions to new sessions;

clustering, by the computer, said new sessions according to a similarity measure into a plurality of web session clusters, wherein clustering said new sessions comprises defining the similarity measure between sessions, and partitioning said new sessions into said plurality of web session clusters according to said similarity measure; and

selecting, by the computer, one of said plurality of web session clusters most closely exhibiting at least one predefined metric from said plurality of web session clusters for analysis of properties of a web page traversal path contained therein.

2. The method of claim 1, wherein said acquiring data from the plurality of sessions, includes:

acquiring data from a plurality of user logs corresponding to said plurality of web page traversal paths during the corresponding plurality of sessions.

3. The method of claim 2, wherein said acquiring data includes:

tagging each web page of said plurality of web page traversal paths;

for each of said plurality of sessions,

forwarding a URL to a monitoring program to store as said data;

said monitoring program compiling a user log for each of said plurality of sessions;

upon completion of one of said plurality of sessions, forwarding said user log for compiling into said data from said one of said plurality of sessions.

4. The method of claim 3, wherein said tagging includes embedding a Java script fragment into each of said plurality of web pages.

5. The method of claim 1, wherein each of said plurality of sessions comprises at least one URL uniquely identifying a specific one of the plurality of web pages.

6. The method of claim 1, wherein grouping said web page traversal paths into said web page categories further includes parsing URL text strings identifying specific ones of the plurality of web pages.

7. The method of claim 1, wherein said partitioning comprises clustering said new sessions according to a pair wise sequence alignment algorithm.

8. The method of claim 1, wherein selecting the one of said plurality of web session clusters comprises:

extracting an open sequence as a representative sequence from each of said web session clusters; and

analyzing properties of a web page traversal path of said open sequence of each of said web session clusters.

9. The method of claim 8, wherein said open sequence is identified according to an APRIORI algorithm.

10. A non-transitory computer-readable medium having computer-executable instructions for identifying properties of a plurality of web page traversal paths, said computer-executable instructions executable for:

acquiring data from a plurality of sessions corresponding to said plurality of web page traversal paths;

grouping said web page traversal paths into web page categories, wherein grouping said web page traversal paths into said web page categories includes correlating traffic amount relationships between specific ones of a plurality of web pages in said web page traversal paths to identify said web page categories, wherein correlating the traffic amount relationships between the specific web pages comprises correlating the specific web pages based on an amount of traffic between the specific web pages;

using the web page categories to map the plurality of sessions to new sessions;

clustering said new sessions according to a similarity measure into a plurality of web session clusters, wherein clustering said new sessions comprises defining the similarity measure between sessions, and partitioning said new sessions into said plurality of web session clusters according to said similarity measure; and

selecting one of said plurality of web session clusters most closely exhibiting at least one predefined metric from said plurality of web session clusters for analysis of properties of a web page traversal path contained therein.

11. The non-transitory computer-readable medium of claim 10, wherein grouping said web page traversal paths includes parsing URL text strings identifying specific ones of the plurality of web pages.

12. The non-transitory computer-readable medium of claim 10, wherein selecting one of said plurality of web session clusters comprises:

extracting an open sequence as a representative sequence from each of said web session clusters; and

analyzing properties of a web page traversal path of said open sequence of each of said web session clusters.

13. A method comprising:

in plural sessions, sending requests from a browser to a web site for web pages of the web site, wherein the web pages include corresponding tags;

in response to receiving a particular one of the web pages, the browser using the tag of the particular web page to request a monitoring service by a monitoring program separate from the web site;

grouping, by a computer, URLs of the web pages into page categories, wherein grouping the URLs includes correlating traffic amount relationships between specific ones of the web pages to identify the page categories, wherein correlating the traffic amount relationships between the specific web pages comprises correlating the specific web pages based on an amount of traffic between the specific web pages;

using, by the computer, the page categories to map the plural sessions to new sessions having reduced attribute information; and

clustering, by the computer, the new sessions into clusters, wherein clustering the new sessions includes defining a similarity measure between sessions, and partitioning the new sessions into the clusters according to the similarity measure.

14. The method of claim 13, wherein mapping the plural sessions to new sessions includes performing feature selection to select features for characterizing the new sessions.