In one decade the World Wide Web has exploded into 14 billion pages that touch almost all aspects of modern life. The network has grown in a grassroots way, based on a handful of pervasive protocols and aloof guidance from the World Wide Web Consortium, a forum based at M.I.T. for Web developers. Essentially, millions of devotees have spent countless hours advancing the Web bit by bit. Although forceful, the effort has also been piecemeal and inefficient.
Furthermore, vast emergent properties are beginning to arise on the Web, and no one is studying how they have blossomed or what they may mean for society. E-mail led to instant messaging, which grew into social networks such as MySpace. The transfer of documents led to file sharing sites such as Napster, which led to user-generated portals like YouTube. Tagging documents with identifying labels is prompting the emergence of a Semantic Web, a global effort to allow computers to recognize not just what online documents are, but what kinds of information they contain and what it might mean. The Semantic Web promises to bring all sorts of useful data to users, not just text and imagery.
According to Tim Berners-Lee, who invented the Web and leads the consortium, it is now time to analyze the Web in a scientific way. By modeling its structure, articulating the architectural principles that have fueled its phenomenal growth, and discovering how social conventions drive online human interactions, Web science hopes to find mechanisms that will ensure the network continues to grow productively and in ways that support the basic social values of trustworthiness and privacy. As transformational as the Web has been, far more advancement could be realized with deeper scientific investigation of decentralized information systems.
This work will occur through a Web Science Research Initiative, headed by Berners-Lee at M.I.T. and Wendy Hall, who directs Southampton's School of Electronics and Computer Science.
The first step will be to define a cogent research agenda. That will be a challenge that could require patience. "We're setting this up so it will exist for a long time," Hall says. She notes that no one knows exactly what Web science will look like. The same situation faced computer science, which developed only after computers existed. Fundamentally, Hall says, "we want to answer the question: What is this amazing thing that is evolving? Some people liken the Web to a living system. But its evolution is different. Living cells know what they're supposed to be; the Web, however, changes as people change it. The evolutionary pattern itself is constantly changing."
The initiative hopes to tackle a number of key questions: What features of current Web protocols make the system work? How do Web users represent the meaning, or semantics, of Web content? Can developers exploit the statistical patterns and distribution of content to understand meaning and relevance? What properties of the Web result in social effects? How do users address online privacy protection, intellectual property rights and security? What trends could fragment the Web?
Two other experts will steer the work: Nigel Shadbolt, professor of artificial intelligence at Southampton, and Daniel Weitzner, principal research scientist at M.I.T.'s Computer Science and Artificial Intelligence Laboratory. The group, which just launched its site at www.webscience.org, anticipates start-up costs of $19 million, which they are raising through government and corporate grants and donations from foundations and individuals.