Where is semantic technology better poised to be better commercialized – the U.S. or Europe? With The Semantic Business and Technology conference heading across the ocean to the U.K. next month, it seems a good time to provide some perspective on the question.
At the last SemTech conference in San Francisco, 3RoundStones took first place in the Startup competition with Callimachus Enterprise. During a discussion of some of the product’s winning features, talk turned to some of the differences between how semantic technology has progressed in the States and overseas.
“I finally see the difference between Europe and the U.S.,” said 3RoundStones CTO David Wood. “For years we here in the U.S. have been looking across the Atlantic Ocean at Europe and saying semantic technology is doing so much better there,” with a lot more R&D dollars put its way.
“Frankly, we’ve been jealous. The work done in R&D there was very important and the large government projects were very impressive,” he said. “But now we are starting to see an interesting thing happening. In the U.S. we don’t spend a lot of money in top-down, federally funded R&D, and not in big central government data projects.” But what is here is a commercially-focused culture, and now large enterprises based here, seeing big vendors like IBM and Oracle talking up semantics, are buying into the idea of semantic tech in a bigger way. “I heard people at SemTech saying now that Europe is becoming jealous of America because our projects, because they are economic in nature, are more sustainable,” rather than being dependent on government or institutional funding that can run out.
Weighing in on the issue, Antonia Bradford of the U.K.’s AB Computing, a consultancy that has developed the Entity Query Language, has this to add: “There has been a lot of EU funding for semantic web projects and for semantic technology projects in general. There’s been funding at the government level and support in the university, but it does need to move out – it will never get anywhere if it stays in university labs,…or continues to receive government funding,” Bradford notes. She adds that tightening purse strings will have an impact there. Still, when projects live on government or university funding, she thinks, there isn’t as much vested interest as when the risk lies with the owner. “I, as the owner of AB Computing, I want to make money. I have invested my time, my money, my emotions – I need to make it work.”
Not, she adds, that research projects don’t have a place, but they have to be undertaken with the right ends in mind – and that is, “to provide actually physical benefits for business, or it won’t get anywhere.” The American model is commerce-driven at its core, she says, so semtech outfits in the States come to the table very well prepared to show how their semantic database solutions, for instance, make life easier for workers, and why they should be chosen over relational databases for certain use cases.
“The American model needs to take over and turn it into something that makes money for people,” adds Martin Bradford, primary developer at AB Computing.
Things can – and do – go right for some semantic projects born with funding from sources like the EU Framework. “Very little of what you see in Sindice would be there if we had not had this funding, so I am grateful for that,” says Giovanni Tummarello, who led the efforts around the Sindice search and analysis engine and is founder and CEO of SindiceTech. “This funding let us reach a certain level of maturity in the way we see things. At the same time, the companies that themselves carved a market in the U.S. did it on their own and they are very pragmatic,” he says, while European efforts reliant upon funding don’t always transition well to the commercial space, he says. Sometimes the fundamentals don’t get the right focus in projects that are academic efforts or government-funded – applying large-scale data processing, sound principle of information management, comparisons of what costs are, including costs of training people, and how they stand up to what was there before, he says, are important. “In the semantic web world a lot of people got a lot of money but don’t have the fundamentals,” he says. It’s “the common sense you need that altogether makes something that can be realistic and useful using semantic technologies.”
For his part, Craig Edmonds, the co founder of U.K.-based Whisk, which is applying semantics to bringing online recipes and online ingredients shopping together, and who is speaking at London SemTech, would like to see more investment opportunities from the private sector into commercially-minded startups. The problem in the U.K., he says, is “a lack of available funds for most types of Internet startup businesses,” he says. “You look at the funding environment here compared to what happens in Silicon Valley, where you can relatively easily raise a couple million based on a proof of concept, and compare that here, where you have to prove traction and ideally revenue before anyone will think about writing a couple hundred-thousand pound check,” he says.