Electrical computers and digital processing systems: multicomput – Remote data accessing
Reexamination Certificate
1998-12-05
2001-12-11
Lim, Krisna (Department: 2153)
Electrical computers and digital processing systems: multicomput
Remote data accessing
C709S203000, C711S118000, C707S793000
Reexamination Certificate
active
06330592
ABSTRACT:
TECHNICAL FIELD
This invention relates generally to the usage of a computer network by a user as more specifically to the techniques of providing specialized information to a network user based on accumulated user data.
BACKGROUND
The World Wide Web (WWW) of computers is a large collection of computers operated under a client-server computer network model. In a client-server computer network, a client computer requests information from a server computer. In response to the request, the server computer passes the requested information to the client computer. Server computers are typically operated by large information providers, such as commercial organizations, governmental units, and universities, and are typically referred to as “web sites”. Client computers are typically operated by individuals.
To ensure interoperability in a client-server computer network, various protocols are observed. For example, a protocol known as the Hypertext Transport Protocol (HTTP) is used to move hypertext files across the WWW. In addition, the WWW observes several protocols for organizing and presenting information, two examples being the Hypertext Markup Language (HTML) and the Extensible Markup Language (XML). The information delivered by the server computer is typically referred to as a “web page”.
A server computer can use a technique known as “dynamically-generated customized pages” to create a web page in response to a request for information from a client computer. A dynamically-generated customized page results in a set of information in a particular format. For example, a first client computer may support the ability to represent information in a number of columns, while a second client computer may support the ability to represent information in a table. Thus, a server computer receiving a request from the first client computer can dynamically generate the requested information in a format with columns. It can respond to a request from the second client computer by dynamically generating the requested information in table format. In this example, two customized pages are created to represent the same information.
It is not unusual for a server computer on the WWW to contain thousands or even tens of thousands of web pages. This large quantity of e makes it difficult for a person, i.e., a “web site visitor”, operating a client computer to locate the information of most interest to them. In much the same way that dynamically-generated customized pages can be used to present the same information in a different presentation format for each client computer, dynamically-generated customized pages can be used to select the information to be displayed so that each web site visitor may see information customized to their specific interests. This process is known in the art as personalization.
Personalization can be achieved through current technology using survey questions to ascertain the visitor's interests, and using dynamically-generated customized pages compute customized pages for each visitor. There are two disadvantages to this approach. First, web site visitors frequently prefer to not fill out questionnaires when visiting a web site, making it difficult for a site to gather the necessary visitor preference data. Second, dynamic generation of every page on a server computer does not scale well for large numbers of requests. In other words, existing methods provide a relatively slow response when a large number of requests are made for personalized pages. This slow response time is attributable to the fact that in existing systems a computer program must be executed to completely generate each dynamic page on every single request.
In view of the foregoing, it would be highly desirable to provide a technique to unobtrusively gather web site visitor preference data and efficiently respond to a large number of requests for personalized pages.
SUMMARY OF THE INVENTION
The invention is a method and apparatus for learning in what a visitor is interested and what demographics the visitor may demonstrate so as to deliver personalized information to the visitor based upon accumulated data, and to do so without requiring dynamic page generation for each individual visitor.
For example, a visitor may demonstrate interest in football and, in particular, his favorite football team. The present invention learns this by observing the behavior of the visitor, i.e., which sports articles he reads and if such articles are focused even further. If a tendency is observed, the learned knowledge is then used to deliver more information about that team to the visitor. Such preferred articles can be recycled by having the invention deliver the same information to other visitors who have the same favorite team.
Visitor interests can be tracked by including “keyword directives” in content contained within the web site. These keyword directives specify a keyword indicating the type of category of information represented by the content. As the content is delivered to the visitor in the form of a web page, the number of keyword directives attached to the content is accumulated into a specified visitor profile. Over time, this visitor profile can represent the types of information the visitor has viewed and serve as an indicator of his or her preferences. In this way, the invention can accumulate a visitor profile unobtrusively, without requiring the visitors to fill out a survey or questionnaire. The profile may also be augmented with explicit information the visitor provides over time, such as a name or address provided when ordering a product from the site.
The present invention then delivers personalized pages to the visitor by examining such visitor's profile. Another directive, called a personalization directive, may be placed into web pages that are to be customized by the invention. These directives cause a personalization function to be applied to the visitor's profile data. The result of the personalization function defines an attribute to be used for locating personalized page fragments, called “page components”, that the invention then assembles into a customized page for the visitor. In this manner, each visitor may receive a page containing three different classes of data: common data received by all visitors, personalized data received by a similar group of visitors, and individual data received only by this one visitor. The present invention assembles all of this data and delivers a “personalized” page to the visitor.
The present invention stores personalized page components in a cache. Subsequent delivery of the same page components is satisfied by retrieving the information from the cache, rather than by dynamically generating it each time. The present invention can therefore take advantage of a common situation where large groups of visitors share similar interests and should receive the same data. Since previously generated personalized page components need not be re-generated for every visitor, computational overhead is reduced tremendously by supplying such pre-generated page components.
For example, a home page for a large web site might include a personalization directive describing the inclusion of an article related to a visitor's favorite NFL team. The personalization directive function examines the visitor profile, determines the favorite team, and includes the appropriate page with information about that team. In this way, each visitor to the web site might receive a different introductory web page, customized for their preferences. Even though every visitor receives a page that appears to be customized for them, since, in fact, there are only 30 or so NFL teams; the caching mechanism of the invention ensures that the dynamic page generation only occurs at most 30 or so times. If one million visitors come to the site, most of the visitors simply receive a web page that was already dynamically generated for a previous visitor. In essence, the invention allows “personalized” pages to be constructed by choosing from a set of previously computed pages, rather than by
Makuch Michael K.
Webber Neil
Gray Care Ware & Freidenrich LLP
Lim Krisna
Vignette Corporation
LandOfFree
Method, memory, product, and code for displaying... 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, memory, product, and code for displaying..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method, memory, product, and code for displaying... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2573516