Import.io: Semantic Web Is Dead, Long Live Structured Web!

We need to turn the Web into a big database, says Andrew Fogg, co-founder and Chief Data Officer of the UK start-up Import.io. So in order to make online data more accessible, his company turns any website into a spreadsheet or an Application Programming Interface (API), for free and without the need to write any code.

“The Semantic Web has had its day – no one has adopted any of the standards,” he told us in an interview with TechWeek. “If you look at the statistics, if you search for the Semantic Web technologies like RDF and SPARQL, the adoption stands at something like 0.01 percent. It’s a great philosophy, but in some respect, we as an industry need to move on – there’s just too much baggage around this term.”

He does not make this claim lightly, having worked with data his entire career for organisations including Microsoft Research, Barclays Capital, RBS, Cambridge University and the Wellcome Trust. Fogg says we need to replace the Semantic Web with the easier-to-grasp idea of the ‘Structured Web’ – the Internet squeezed into rows and columns.

Eventually, Import.io hopes to build a marketplace where any business can sell access to its data, similarly to how Twitter offers access to its ‘firehose’. The key word here is “access” – Import.io doesn’t trade in databases, instead it builds API connectors to link information on the Web, making it more useful.

Structuring the Web

The term Semantic Web, coined by the ‘father of the Internet’ Sir Tim Berners-Lee in 2001 and still promoted by the World Wide Web Consortium (W3C), describes a network consisting of interlinked pieces of data posted in common formats and accompanied by machine-readable metadata.

A component of ‘Web 3.0’, the Semantic Web would enable people to share and reuse information beyond the boundaries of applications and websites. More importantly, it would allow online content to be read and interpreted by machines. However, it has long been criticised as an unworkable concept by people like Cory Doctorow, who says that Semantic Web is unfeasible due to the faults of human nature, such as laziness and dishonesty.

“The vision for a web of data has been around for a while. It’s a valid vision, one we firmly believe in, but the Semantic Web is just not being adopted,” Fogg told us.

Import.io launched in 2012 with a mission to build a system that brings together data users and data owners. Fogg says that the Web was designed for documents, and not for data – but thanks to the power of APIs, data producers can defeat the limitations of the system which recently celebrated its 25^th birthday. Such an approach to information is already used in recruitment, retail and mobile development.

“Traditionally, if you are a company and you want an API for your website, it requires you commissioning a not insignificant technology project, making it available, documenting the API, it’s a big piece of work,” explained the data scientist. In contrast, building an API using tools from Import.io takes just a few minutes.

“Some people want to do it on their own websites, but very often it’s about people wanting to get lots of data from the Web inside their organisation.”

For example, the British Red Cross recently wanted to create an app that would feature information about all the hospitals in the UK, provided by the NHS. But even though there’s an NHS website that contains this data, there were no APIs to get it out. Using Import.io, the Red Cross team was able to build a relevant API in just a few minutes, and the app is currently available on the iTunes store.

Meanwhile, HP has used Import.io to build a system that automatically monitors whether its channel partners are observing the minimum retail prices for its products, to ensure fair competition. Finally, the start-up has helped a recruitment company to automatically monitor client websites and immediately add listings for new job vacancies. “That’s something they wanted to do forever, but they haven’t been able to, because it wasn’t technically feasible,” says Fogg.

According to ProgrammableWeb, a directory of all freely available Web APIs, there are around 11,000 currently in existence. Meanwhile, users of Import.io have so far created nearly 20,000.

The movement

But why are services like Import.io emerging now? “I think it’s the volume of data. There is more data on the Web than ever before, and pressure on companies to incorporate this data into their decision-making is also greater than ever before. Companies that are not making data-based decisions are going to be beaten by the competition,” suggested Fogg.

“What we’re tapping into is the fact that the Web is the biggest repository of data available to a business.”

In the next few years, Fogg wants to see Import.io creating not just a niche for itself, but a whole movement dedicated to describing, interlinking and structuring the Web. He says there are plenty of start-ups that share this vision, but the market for such services is still very young.

“Much like cloud emerged as a concept from a couple of companies, Salesforce in particular, running software on their own servers instead of your servers, I want Import.io leading the way in the discussion about the structured Web.”

Fogg says that due to the convenience of APIs, in the nearest future Big Data applications will begin to integrate a lot more information that’s being offered in real-time, rather than static data prevalent in today’s analytics projects.

TechWeek has previously met Fogg at Le Web 2013 conference in Paris, to talk about the importance of APIs. You can watch the interview here.

What do you know about Big Data analytics? Take our quiz!