Class Activity: From the List to the Semantic Web

Objective: To evaluate lists of sites or links for meaning and rhetoric, and to also use the model of Tim Berners-Lee's Semantic Web (a markup protocol that provides rich meaning contexts to information) in order to see how designers are teaching computers to "think".

The inability of computer to think, per se, opens up a lot of possibilities for artists projects. Lists or topic headings that don't offer the best organization of their topics provide discomfort, but also humor. When computers leave out important information, meaning nevertheless enters into the context of the data. This may be a condition of only the early internet, given the vision of Tim Berners Lee, for meaning-rich descriptions of data. More on that momentarily.

There is also an effect of large lists: they equalize their items. We have to bring qualitative judgment to the items in the list, since a list is usually not a hierarchy if not a part of a hierarchy. Encyclopedism, surrounding the creation of encyclopedias of knowledge in the 18th Century provides a background to the cross-references of the web (another term for linking), as well as the scholarly and not so scholarly practices behind collections and indexes.

If you read Anthony Grafton's work on encyclopedists or the development of historical footnotes, you will see that cataloguers and indexers--list-makers-- have bordered on eccentric. For the connection between collecting, cataloguing, the pursuit of knowledge, and list-making read the article on Flaubert.

The current web is a web of lists and links. I've outlined one of the meanings of everything being in a list. This is only the initial effect of our perception of organized data. What other meanings due to sequence and due to qualities that we bring to the list can be seen in the following lists?

Great Books

Consumer Technology

Science and the Humanities

Russian Constructivism

Aside: The work at The World Wide Web Consortium

The researchers at the W3 Consortium are working toward a vision of "smarter" data. Through a sophisticated system for labeling data, or RDF (Resource Definition Framework), the web of the future will create rich "meaning contexts".

Say you have a document that is an interview between a famous musician and a TV personality. In this markup, the statements of the musician, can be marked up as <YoYoMa> while the statements of the TV Personality can be marked up as <LarryKing>. The creator of the web-based document of this interview will supply basic information that defines both of these, such as their functions, musician and TV personality.

The computer connected to the internet can simply ask for a definition or ask for an occupation and the interview document could supply Musician or TV personality. What is the difference between this scenario and today's web? Imagine going to a search engine and typing in "author of Crime and Punishment". The search engine would know that this is a work of Russian Literature by Dostoyevsky and would know that it is a book. It would explicitly retrieve Dostoyevsky from "author of Crime and Punishment."

For this exercise you will generate lists of categories to describe the content for normal search engine terms. Choose from the following list:

For instance, one possible category of Of Mice and Men: <JohnSteinbeckNovel> or <au=Steinbeck> where au is author.

Jacques Chirac: one possible category/semantic term: <FrenchNationalPresident>

Compact Disc:one possible: <storageMedia>

We will engage in a discussion for the list and semantic web exercise.