With the web now into its second decade, leading lights in the web world want to turn it from a phenomenon into a science.
Representatives from the University of Southampton and the Massachusetts Institute of Technology (MIT) on Thursday announced the Web Science Research Initiative (WSRI), a multidisciplinary project to study the social and technological implications of growing web adoption.
The universities intend to raise money from corporations — the WSRI already has financial backing from Google and IBM —to establish a research centre that will sponsor PhD students and ultimately create undergraduate curricula in web science.
"The web is basically a web of people. It's a way that social people interact," Tim Berners-Lee, who invented the basic software of the web and is director of the World Wide Web Consortium (W3C) standards group, said. "Because it's something we created, we have a duty to make it better."
Berners-Lee, who is also a senior research scientist at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), detailed the initiative with other organisers at MIT on Thursday.
The universities intend to combine several disciplines, including social sciences, psychology and life sciences, with technology development.
The social aspect of the web — and the web's huge impact on society — demands that a field separate from computer science be explored, organisers said. For example, eBay is interesting because it relies on the involvement of millions of people. Similarly, Google used a mathematical algorithm that examines how millions of individuals link to other pages to improve search results.
"We want to throw some light on forecasting what these new technologies might lead to in the human sense, in the community sense — and in the business," said Wendy Hall, a professor of computer science at the University of Southampton.
Social scientists can help analyse online communities, and experts in life sciences can help web scientists understand how complex systems like the human body — or the web — operate, she said.
Researchers would like systems that can better reflect the social relationships between people, said Daniel Weitzner, principle research scientist at CSAIL.
For example, finding out basic information on meeting participants, such as phone numbers or professions, from an online calendar entry would entail a lot of manual work. But socially aware web applications could make the task much easier
"The web fails to capture the nature of social relationships. We want the eb to be more responsive to the existing relationships people actually have," said Weitzner.
Researchers intend to keep "lanes open" between the universities and the W3C to standardise suitable research, Berners-Lee said. The plan is to make technology available without royalties to encourage adoption, said Weitzner.
As more and more information goes online — submitted by individuals or organisations — new technical and policy problems get introduced, project backers said.
Berners-Lee gave the example of email, which worked well until it was adopted on a mass scale. That tipping point enticed spammers to start abusing the system.
To prevent problems such as these, web scientists can study the incentives for individuals to do harm as well as the technology that can be misused, such as viruses, Berners-Lee said. Similarly, researchers can analyse the human psychology as it relates to web-based interactions and the legal implications.
The WSRI plans to find ways to embed the rules and policies on data use into software, such as taking into account existing laws around privacy.
For example, there are European Union directives that require that government information be made available in a general way for reuse.
But at present there is no easy way to take into account the policies that govern the use of information, some of which could be sensitive health data, said Nigel Shadbolt, professor of artificial intelligence at the University of Southampton and the incoming president of the British Computer Society.
A PhD student could design a system that describes the provenance of shared data and details the rules on how data could be used for certain purposes, Shadbolt explained.
Adoption of the Semantic Web — a W3C standardisation effort — should aid researchers in developing mathematical models and creating systems to analyse data, organisers said.
Semantic Web standards are now being developed and will be built into products, albeit slower than HTML — a key web-coding format — because the Semantic Web is more complicated and developed in a more rigorous fashion, said Berners-Lee.
Organisers said that the creation of a field of study for web science does not eliminate the need for continued research in existing fields, such as engineering and computer science. Rather, they envisioned web science as a field that straddles both the technical and social nature of the web and complements other fields.
"We want to see the web as an object of scientific study from the perspective of various different disciplines," said Weitzner. "What we are looking for is to direct scholarly attention and research attention to this particular new subject."
