METHOD AND SYSTEM FOR MINING WEBSITES

• US 20100161785A1
• Filed: 12/22/2008
• Published: 06/24/2010
• Est. Priority Date: 12/22/2008
• Status: Active Grant

• Current Assignee: NBC Universal Media LLC (Comcast Corporation)
• Original Assignee: General Electric Company
• Inventors: Xue, Feng, Gustafson, Steven Matt, Moitra, Abha
• Granted Patent: US 9,230,030 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

One website mining embodiment is for characterizing first time users of a website, collecting user session data of the users visiting the website and identifying first time visitors, determining features of the first time visitors utilizing the user session data, determining rules utilizing the features of the first time visitors, monitoring actions of the first time visitors on the website, updating the rules utilizing the monitored actions of the first time visitors and recommending web content utilizing the rules to the first time visitor.

Claims

1. A method of website mining, comprising:

collecting user session data of users visiting the website and identifying first time visitors;

determining features of the first time visitors utilizing the user session data;

determining rules utilizing the features of the first time visitors;

monitoring actions of the first time visitors on the website;

recommending web content utilizing the rules to the first time visitor; and

updating the rules utilizing the monitored actions of the first time visitors.

2. The method of claim 1 wherein the first time visitors are tagged for identification of the first time visitors.

3. The method of claim 1, wherein the first time visitors are the users not visiting the website for a predefined time period T and/or not downloading data less than a predetermined size from the website.

4. The method of claim 1, wherein the features of the first time visitors include date of visit, time of visit, client IP address, user name, service name, client server name, client server IP address, server port, method, protocol status, Win32 status, bytes sent, bytes received, protocol version, host, user agent, cookie, referrer, website pages viewed, operating system, city of the first time visitor, country of the first time visitor and session id.

5. The method of claim 1, wherein the features of the first time visitors are stored in a hash table.

6. The method of claim 1, wherein the features of the first time visitor are stored in a logfile database.

7. The method of claim 1, wherein the rules of the first time visitors are derived using Classification and Regression Trees.

8. The method of claim 1, wherein the monitoring of actions of the first time visitors include acceptance of the recommendations, rejection of the recommendations, time spent to review recommended data and time spent on the website after recommendations.

9. The method of claim 1, wherein the determination and updating the rules is done offline.

10. The method of claim 1, wherein the recommendation of web content includes recommending content on the basis of the rules corresponding to the features of the first time visitor.

11. The method of claim 1, wherein the recommendation of web content to the first time visitor is done online.

12. The method of claim 1, wherein the recommendations are made utilizing the most highly ranked rules in the rules database.

13. The method of claim 1, wherein the rules determined have a ranking of preference.

14. The method of claim 1, wherein updating the rules include updating rankings of the existing rules.

15. The method of claim 1, wherein updating the rules include comparision of a rule determined with an existing rule.

16. The method of claim 15 wherein the determined rule is stored in the database if the determined rule does not exist in the existing rules.

17. A system of website mining comprising:

a webserver for storing web usage data of a website in a storage device; and

a webserver engine connected to the webserver for;

processing the user session data to determine features of first time visitors of the website and determining rules based on the features of the first time users;

recommending web content to the first time visitors using the rules;

monitoring actions of the first time visitor to update the rules; and

updating the rules based on the actions of the first time visitors.

18. The system of claim 17 wherein the website has a finite number of hierarchical levels with mapping from each website page into the hierarchical levels, and wherein each user session is mapped to the hierarchical levels of the website.