Make no mistake about it: The semantic web has been a success and that’s not about to stop now. That was essentially the message delivered by W3C Data Activity Lead Phil Archer, during his keynote address celebrating the semantic web’s ten years of achievement at last month’s Semantic Technology & Business Conference in San Jose.
After acknowledging that he’s heard it all about the semantic web being a failure, about it being rebranded as Linked Data and that being a failure too, he summed up those impressions in one distinctly British word: “Bollocks.” The list of successes ranged across the spectrum, from the use in federated data portals of the Data Catalog Vocabulary from the W3C’s Government Linked Data Working Group, to the 47-million triples strong Open Phacts pharmacology discovery platform, to all the job postings that come up on The Semantic Web Blog – including one recently for the J. Craig Venter Institute, named for the pioneer genomic researcher who sequenced the human genome, which is looking for a bioinformatics analyst with OWL expertise in his or her resume.
That’s just a taste of the many citations he offered of the semantic web’s successes to date, not least among them his own strong familiarity with Linked Data’s use in government, where the technologies, he said, are used “to make our government more efficient” by having one organization make an authoritative data set the others can link to to support data-sharing across agencies. (For the full view into Archer’s take on the semantic web’s successes, you can view the entire keynote here.)
2024, Here We Come
But it’s where things are beginning to head – and where they should be further promoted – that’s going to have the semantic web celebrating what Archer said he hopes will be 20 years of success by 2024. Some further notes of encouragement, for example, come from other communities taking closer note of the semantic web, like the XML world which Archer said is recognizing the power of using HTTP URIs as identifiers “and looking at the semantics that we can offer them.” The XML world, he said, “is looking at what we are doing.”
In the geospatial realm, large-scale spatial data infrastructure systems based on the Open Geospatial Consortium’s (OGC) OpenGIS Geography Markup Language (GML), an XML grammar for expressing geographical features, is spending more time look at semantics and is beginning to use HTTP URIs as identifiers, he said. “They gave us GeoSPARQL,” he said. “They have more experiments going on in that world.”
W3C and OGC, in fact, are working together on the subject of how to do geospatial data on web, and it’s expected that they will form a joint working group that would include among its efforts finishing work on the OWL Time Ontology, he said.
Manufacturing and supply chain sectors also are beginning to see the value in the idea that all their components all have data about them that could be made available throughout the manufacturing process right up to the front end customer site. While there are some cultural obstacles to overcome there, Archer said things are starting to happen and, “more importantly, it’s not just us working on that. Other people want that to happen,” he said, referencing the GS1 standards body that is responsible for barcode and identification standards like GTIN (Global Trade Item Numbers) and GLNs (Global Location Numbers) in the supply chain.
“They have a project to make this stuff available and to use the web and persistent identifiers to help improve that efficiency through manufacturing up through the supply chain until the end,” he said (See our discussion with Bernie Hogan, GS1 Senior Vice President, Emerging Capabilities and Industries, on extending the standards used for the identification of goods in the brick and mortar retail world into the web realm, including how its systems and standards can fit in with the semantic web.)
Bring Others Into the Fold
Semantic web technologies also are making their way into the hearts and minds of those “who don’t know what they’ve got and probably don’t care,” Archer said, courtesy in large part of JSON-LD, the JSON-based serialization for Linked Data. Indeed, these very same people may actively lambast RDF while embracing JSON-LD, but Archer urged tolerance. “Those people are our potential customers,” he said.
And even if some of them treat JSON-LD as JSON only, which he noted is explicitly possible (although it may result in untidy RDF), “it gives us an interface between the semantic web world and everybody else,” Archer said. “Let’s not knock that.”
As Manu Sporny, one of the primary creators of JSON-LD has written, “JSON-LD was created for web developers that are working with data that is important to other people and must interoperate across the Web.” And it is the regular web developer, Archer said during his keynote—the ones that don’t care about SPARQL or OWL but love things like the idea of web components or using new Web APIs to take second-screen interactions to new levels – that must be seen as part of the semantic web customer base, as much as commercial customers are.
“We need to make our stuff available to people who never write never need to write a SPARQL query,” he said. “I hope that probably in 2016 we can look at standardizing something around that – how you provide a NoSPARQL interface to the semantic web.”
Despite criticism and nay-saying, a lot of success has been achieved so far, Archer summed things up, and a lot more is to come. “You should be proud of what this community has achieved,” he said. “We have a stable, broad, deep, mature technology stack and we at the W3C will continue to support you and your efforts to try and fill in gaps.”
And in the long term, he concluded, “we have to use our technology, skills, and data analysis ability so that everyone…can access this immensely successful and powerful technology stack that we have.”