Electrical computers and digital processing systems: multicomput – Computer network managing – Computer network monitoring
Reexamination Certificate
1997-03-31
2004-03-30
Geckil, Mehmet B. (Department: 2152)
Electrical computers and digital processing systems: multicomput
Computer network managing
Computer network monitoring
C709S217000, C709S218000, C709S219000, C705S014270
Reexamination Certificate
active
06714975
ABSTRACT:
SPECIFICATION
The present invention relates to advertisements on pages of a World Wide Web site. Specifically, it is directed to a method for dynamically assigning advertisements to web pages according to self-learned user information.
BACKGROUND OF THE INVENTION
The internet is a network of networks and gateways that use the TCP/IP suite of protocols. A client is a computer accessed by a user or viewer which issues commands to another computer called a server. The server performs a task associated with the client's command. The World Wide Web (WWW or Web) is the internet's application which displays information on the internet in a user-friendly graphical user interface format called a Web page. A Web server typically supports one or more clients. The Web allows users (at a client computer) who seek information on the internet to switch from server to server and database to database by viewing objects (images or text) and clicking (with a pointing device or keystroke) on corresponding highlighted words or phrases of interest (hyperlinks).
The Web can be considered as the internet with all of the resources addressed or identified as Universal Resource Locators (URLs) and which displays the information corresponding to URLs and provides a point-and-click interface to other URLs. A URL can be thought of as a Web document version of an e-mail address. Part of a URL is termed the Internet Protocol (IP) address.
An internet browser or Web browser is a graphical interface tool that runs internet protocols and displays results on the user's screen. The browser can act as an internet tour guide, complete with pictorial desktops, directories and search tools used when a user “surfs the net.”
With the recent explosion of web-related sites and services, the internet has become a great opportunity for web-related advertising. For site administrators who maintain a large number of pages, it has become almost necessary to use some process of optimizing placement of the advertisements and scheduling of the web pages.
Typically, advertisements occur as objects in the form of (text or inline) links on Web pages. Advertisers measure the effectiveness of contracting with a particular Web server (web site) by analyzing the number of times a viewer clicks (“hits”) on an advertisement. The cost of such a contract may be either directly or indirectly dependent upon this click rate. Consequently, from the point of view of the site administrator, it is desirable to maximize the total number of hits to the advertisements on the server.
Each page on a server may have a certain number of predefined slots which have standard sizes containing the object (inline image or text) links to the actual advertisement pages. The object may be an advertisement.
An advertisement is exposed when a page which contains the slot with the advertisement is served to a client accessing the page. Since a page may typically contain more than one slot, more than one advertisement may be exposed at a single time. This exposure of an advertisement is also called an impression.
An advertisement is clicked when a client decides to choose (with a pointing device or keystroke) the link corresponding to an exposed advertisement. Thus, the number of clicks for an advertisement is always a certain fraction of the number of exposures. Since the advertisement agencies measure the effectiveness of an advertisement by the number of clicks that an ad receives, and since the sum of the total number of exposures received over all slots by a site is defined by the traffic to that site, it is advantageous for a server to assign a particular advertisement to a slot in a manner such that the advertisement's click/exposure ratio is maximized.
Presently, placement of advertisements on web pages is executed by using different variations of static assignment. Static assignment produces web pages with advertisements which generally do not change unless and until the site administrator adjusts them according to some historic information. This static method of placing advertisements does not take into consideration various real-time characteristics of the web which can improve placement optimization.
For instance, certain web pages in a web site are more likely than others to be accessed by users. Without knowing which pages are the present “hot spots,” static assignment is likely to result in underexposure of certain advertisements and overexposure of others. Furthermore, some advertisements are much more likely than others to be accessed at a certain time of the day. Static assignment does not take this time-dependency into consideration. Finally, static assignment provides no way of taking into consideration the individual tendencies of a particular user to choose a specific advertisement based on appearance, size, shape or location on a page.
It is a primary object of the present invention to provide a method for dynamically assigning advertisements to appropriate slots on appropriate web pages based on a characteristic of the requesting client or user, depending upon self-learning data obtained from historical user behavior.
A further object of the present invention is to provide a method of scheduling advertisements which exploits popular web pages or “hot spots.”
A further object of the present invention is to provide a method of placing advertisements which avoids the problem of over-exposure and under-exposure of advertisements.
A further object of the present invention is to provide a method of placing advertisements which takes into consideration the time of day.
SUMMARY OF THE INVENTION
The present invention overcomes the prior art limitations by providing a method for placing advertisements web pages which is capable of making placement decisions “on the fly” depending on characteristics of the client or user from whom the request is received.
According to the present invention, the method for dynamically placing objects in slots on a web page in response to a current client request for the web page includes the steps of classifying users into two or more user groups based on at least one user characteristic, accumulating self-learning data based on user click behavior for each user group, matching the current request with a corresponding user group and scheduling real-time selection of the slots for the objects on the web page based on at least the self-learning data of the corresponding user group. The objects are preferably advertisements.
Preferably, a group click/exposure ratio is accumulated for each user group and the scheduling of real-time selection of the slots for the objects on the web page is based on the group click/exposure ratio for the corresponding group.
The method preferably includes the step of generating probabilistic assignment data for each user group based on a contract requirement of the objects wherein the scheduling is further based on the assignment data.
Preferably, the method also includes the step of generating probabilistic assignment data for each user group based on an exposure requirement of the object for a period of time.
It is preferable that the method further includes the step of generating probabilistic assignment data for each user group based on a popularity characteristic of the web page, such as a number of clicks on the web page for a period of time.
Because processing resources are limited in a web server, the classifying step of the present invention preferably includes the steps of probabilistically choosing a fraction of previous requests and classifying users of the fraction of previous requests into two or more user groups. In web sites which have a registration procedure (for example the New York Times), explicit demographic information is available, and this may be used in order to perform advertisement placement. Thus, according to a preferred embodiment of the invention, classification of users is executed based on user demographic information.
According to a second and third embodiment of the present invention, classification of users is executed based on user cl
Aggarwal Charu Chandra
Wolf Joel Leonard
Yu Philip Shi-lung
F. Chau & Associates LLP
Geckil Mehmet B.
Prieto Beatriz
LandOfFree
Method for targeted advertising on the web based on... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for targeted advertising on the web based on..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for targeted advertising on the web based on... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3287026