Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
1998-09-25
2002-01-08
Homere, Jean R. (Department: 2177)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C707S793000, C709S203000, C709S217000, C709S223000
Reexamination Certificate
active
06338066
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates in general to computer behavior prediction, and in particular predicting a given web surfer's behavior based on the behavior of past surfers.
2. Description of Related Art
As the popularity and usefulness of the Internet grows, and more of the general public are able to access the Internet both at home and at work, marketing of Internet services and other items has become more popular. However, mass marketing of certain items to everyone becomes wasteful and expensive, because each Internet user (surfer) visits different Uniform Resource Locator (URL) sites. Further, each Internet user (surfer) has various susceptibilities to different products, marketing techniques, and presentation styles.
Each URL has an associated web server log, which records each surfer's visit, which pages were downloaded, and whether a purchase was made from that URL. However, the web server log has not been utilized by advertisers to determine or predict behavior of a given surfer based on prior users.
Currently, advertisers are mass marketing their advertisements to all users. This approach, although simple, is less effective than other advertising techniques, e.g., time slotting of certain advertisements, demographic studies, etc. However, most Internet users are anonymous, making demographic targeting difficult. Further, no modeling of Internet user behavior, nor any studies of Internet advertisement success, have been performed to predict behavior.
One potential way to distinguish surfers is through a unique identifier that a site may provide to a customer's web browser software for use in identification during future visits to the site, also called a “cookie.” Unfortunately this method has the potential to violate consumer privacy. Further, current web browsers, such as Netscape, make it easy for even novice users to disable this identification feature, making identification difficult. Further, this tool is unlikely to be an effective user identification in the future.
A second way to distinguish surfers is to question them directly as a requirement for entering the site. The answers to such questions may or may not be associated with a given user id and password for use in the future. This method has two problems. First, users may be unwilling to supply such information and therefore may forego visiting the site. Second, even when the information is supplied, its veracity is subject to question.
A third way to distinguish surfers is to analyze their surfing behavior while visiting the site and deduce from this behavior how they might best be influenced to increase their buying behavior. This method of distinguishing surfers has been tried in the past by other companies including Aptex® and FireFly® using neural nets and Nearest Neighbor techniques. However, these methods have not shown much success.
It can be seen, then, that there is a need for a way to predict a given web surfer's behavior based on past surfer behavior. It can also be seen that there is a need for advertisers to target advertisements to certain surfers or groups of surfers on the Internet.
SUMMARY OF THE INVENTION
To overcome the limitations in the prior art described above, and to overcome other limitations that will become apparent upon reading and understanding the present specification, the present invention discloses a method for predicting a given web surfer's behavior based on the behavior of past surfers.
A method in accordance with the teachings of the invention comprises storing a log of visits to a website and categorizing a plurality of pages of the website into a group. The log and group are processed into at least three classifications, e.g., features, actions, and behaviors, and the features and behaviors are sorted by the actions associated with the features and behaviors. The features and behaviors are then grouped by the associated action, models are generated for each action; and a predictive model for each action and behavior is generated.
An object of the present invention is to predict a given web surfer's behavior based on past surfer behavior. Another object of the present invention is to target advertisements to certain demographic groups on the Internet.
REFERENCES:
patent: 5664115 (1997-09-01), Fraser
patent: 5678041 (1997-10-01), Baker et al.
patent: 5712979 (1998-01-01), Graber et al.
patent: 5802292 (1998-09-01), Mogul
patent: 5892917 (1999-04-01), Myerson
patent: 5917891 (1999-06-01), Will
patent: 5974572 (1999-10-01), Weinberg et al.
patent: 5977964 (1999-11-01), Williams et al.
patent: 6020884 (2000-02-01), MacNaughton et al.
patent: 6052730 (2000-04-01), Felciano et al.
patent: 6055569 (2000-04-01), O'Brien et al.
patent: 6055573 (2000-04-01), Gardenswartz et al.
patent: 6085226 (2000-07-01), Horvitz
patent: 6167441 (2000-12-01), Himmel
patent: 6182133 (2001-01-01), Horvitz
Brachmann, R. et al., “The Process of Knowledge Discovery in Databases”, Section 2, pp. 37-57 (Date Unknown).
Duda, R. et al., “Pattern Classification and Scene Analysis”,Stanford Research Institute, Menlo Park, California, a Wiley-Interscience Publication, John Wiley & Sons, 8 pages (Date Unknown).
Monthly Newsletter “Intelligent Software From Cutter Information Corp. Strategies”, XII(12):1-16 (Dec. 1996).
Press W. et al., “Numerical Recipes in C”,The Art of Scientific Computing Second Edition, Cambridge University Press, 2 pages (Date Unknown).
Strang, G., “Linear Algebra And Its Applications”, Second Edition, Academic Press, Cover Sheet and pp. 136-145 (Date Unknown).
Martin David Charles
Miranda Hansel Joseph
Plutowski Mark Paul
Spangler William Scott
Vaithyanathan Shivakumar
Altera Law Group LLC
Homere Jean R.
LandOfFree
Surfaid predictor: web-based system for predicting surfer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Surfaid predictor: web-based system for predicting surfer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Surfaid predictor: web-based system for predicting surfer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2823572