TAG Newsletter

[Previous] [Next]

<TAG> Newsletter (http://tagnewsletter.com) has been published since 1987 as a technical newsletter for the information management business. I have been the editor of the newsletter since 1991. In 1992, <TAG> Newsletter became a monthly newsletter, and in 1999, we made it available online as a database-driven, interactive site that uses XML and XSL to create personalized information for our subscribers.

The mission of <TAG> is to provide XML implementers with the tools and information they need. <TAG> is mostly technical in nature, providing tutorials with code samples and technical tips. We also try to publish as many case studies and book reviews as possible to keep the publication well rounded.

In 1999, we re-engineered the site to provide our readers with a personalized view of the publication. We converted more than 1000 articles to XML and stored them in a database. Using the database, we are able to provide many different views of each article based on the profiles of each user. For example, a user visiting the site by using Netscape Navigator will get a page that has a different structure than a user accessing the site by using Microsoft Internet Explorer. The site does this by recognizing the difference between Web browsers and other access methods (offline subscriptions and cell phones, for example) and creates information responses tailored to those technologies.

One of the benefits of this personalized system is that it gives paid subscribers instant access to all articles on the site, up to and including the current edition. Nonsubscribers have access to everything on the site, but they only get the title, author, and abstract of articles that are less than a year old.

We perform personalization and browser-specific rendering using XML, XSLT, and middle-tier scripting in a holistic way. As we developed the site, we discovered some great ways to present our information assets. For example, when an article mentions an acronym, we tag it as such. This gives us the ability to render the acronym for the print publication as smaller capitals, but also allows us to "hide" the definition as a pop-up when we render for electronic delivery. In other places as well, we found information that could be rendered many ways for different outputs and readers. We wanted to share this information with our users, so we made a multipart case study/tutorial on the topic, starting in June 1999. Our goals were to create an intelligent content management and delivery environment and to spend as little money as possible on it. We achieved both. Our only expense was our time.