Web pioneers call for new 'web science' discipline

2019-03-02 04:14:00

By Celeste Biever The social interactions that glue the World Wide Web together are now so complex it has outgrown the relatively narrow field of computer science, say pioneering internet researchers. “You can’t tell how something will spread just by looking at how the html code works,” says Tim Berners-Lee, who invented the hypertext code that connected the first web pages back in 1989, and who now works as a researcher at MIT in the US. To fully harness the web’s potential, Berners-Lee says, “web science” must include elements of social science, psychology, economics and law, along with computer science and engineering. “The web as a structure has got so large that we need new forms of analysis to understand its properties,” adds Nigel Shadbolt, who heads the Intelligence, Agents, Multimedia research group at the University of Southampton in the UK. On 2 November, scientists from that university and MIT launched a collaboration aimed at promoting such analysis, called the Web Science Research Initiative (WSRI), pronounced “wiz-ree”. The initiative will, for example, study the ways social networks, blogs, e-commerce and government regulation all contribute to the web’s overall structure. This should help academics and businesses alike, say the initiative’s founders. “We are doing this partly out of excitement, partly out of duty,” says Berners-Lee. “We have a duty to understand [the web] because if we can make it better, we should make it better.” Scientists have long studied the web’s network topology and come up with various hypotheses to explain its structure. For example, the “small world” hypothesis says that any person can be connected to any other person through just six online acquaintances. The “scale-free” proposal concludes that some parts of the network are more vulnerable than others because they have points that connect a huge number of other points – so destroying a these “hubs” can disrupt the entire network. But these hypotheses are just the tip of the iceberg, says Shadbolt. Further studying the intersection between computing and social science will shine light on how and why these connections form in the first place. Google famously harnessed one aspect of the web’s structure to create its search engine. The company’s search algorithms examine the frequency and quality of web links to determine how relevant a web page is to a particular search query. Wendy Hall, another researcher at the University of Southampton, says Google, Amazon and eBay have all exploited the complex interplay between social science and computer science to build their businesses. “They only work because people use them,” she says. “And the more people that use them, the better they become.” A better understanding of the way the web works could also help to protect people online, says Daniel Weitzner, a lawyer from MIT and another a founding member of WSRI. Currently, he says, “the web is particularly bad at helping people to replicate the social expectations and legal expectations that we all function with in society”. Students who chose to study web science may, for example, examine ways to make public information available through the web without violating personal privacy. Or they might study how data distributed over multiple servers can best be searched. “One thing that is difficult for computer science is querying in a distributed fashion,” says Shadbolt. “How do you do that in an efficient way?” Ultimately, Hall says, the aim is to create a new breed of scientist: