We’re all waiting to see what’s next for the Web of Data – that is, the Semantic Web – after a year that had numerous highlights including: the official publication of HTML5, which boasts the ability to describe the structure of a web document with standard semantics; a flurry of activity for schema.org, including a new actions vocabulary, new types such as roles, and several community- and partner-led improvements in areas from bibliographies to sports to events; and, the continuing progression of Knowledge Graphs from the likes of Google, Facebook, Microsoft, and Yahoo.
Read on for some insight into how the Semantic Web’s promises and challenges will play out in 2015, from some of the experts who study and aim to resolve them every day.
Bringing Worlds Together: What’s ahead in the way of collaboration?
The W3C’s data activity lead Phil Archer says to expect announcements in January related to collaboration between the W3C, which has under membership review the draft charter for the Spatial Data on the Web Working Group, and the Open Geospatial Consortium (OGC). The outcome of a workshop Archer ran in March on Linking Geospatial Data, he says, made it clear that there was a desire for the two standards developing organizations (SDOs) to work together “so that geospatial and Web can work together, probably, although not exclusively, through Linked Data.” At its most basic, he says, that boils down to the concept of using URIs as identifiers in spatial data:
“2015 is when we see those two worlds come together. That’s going to be important for obvious areas like environmental data, but it also applies to the automotive industry, media (news stories are usually tied to a place where whatever it was happened), government statistics and more. Interestingly, it also applies to life sciences. Researchers can describe the spread of a pathogen across a tissue sample using points, lines and polygons with a suitable coordinate reference system – just like any other spatial data.”
Linked Data Lineup: What other milestones to expect there?
The Linked Data Platform 1.0 – the editor’s draft is available here – actually should become a W3C Recommendation before 2014 is out, says Arnaud Le Hors, co-chair of the W3C’s Linked Data Platform (LDP) Working Group and Linked Data Standards Lead at IBM. The mission of the LDP Working Group has been to produce a W3C Recommendation for HTTP-based (RESTful) application integration patterns using read/write Linked Data. This advance creating a standard set of principles and patterns for interacting with Linked Data over HTTP should be welcome news for developers and enterprises, opening wider the door that will make it easier for them to use Linked Data more effectively and more broadly in 2015.
As the Working Group continues its activities, including polishing the Linked Data Patch Format Working Draft that defines a language for updating Linked Data resources incrementally, it’s also contemplating what it will tackle next, says Le Hors. “All along the development of LDP 1.0 we’ve had to put aside features that Working Group members were interested in developing but didn’t have the time for,” he says. They’ll be revisiting those features – including access control and optimizations of different kinds requiring fewer roundtrips between the client and the server than what LDP 1.0 does – and prioritizing them to develop a new charter that will be presented to the W3C membership for approval.
“We’re expecting to charter an LDP 2 group to look at, among other things, access control,” confirms Archer, who also comments that the LDP is going to be important in 2015 in areas including manufacturing supply chain data. As Archer explains, it generally takes a whole chain of manufacturers and assemblers to produce the goods we buy, and in many cases, a company in a chain will only know the other companies immediately before and after them in the chain. They may not even know where the product they’re contributing to actually ends up. While there may be legitimate concerns and commercial sensitivities for the secrecy, many times there are not, he says:
“What we need to do it so make it easier for manufacturers to share their product specifications, food ingredients, etc. as data, and be able to control access to different parts of the dataset,” Archer says. “Data that can be made public can be generated at source and made available across the whole chain up to and including the end consumer.”
Turning the spotlight onto the lightweight Linked Data format JSON-LD – which is primarily intended to be a way to use Linked Data in Web-based programming environments, to build interoperable Web services, and to store Linked Data in JSON-based storage engines – keep an eye on work underway at the W3C’s Social Web Working Group. It is developing Activity Streams, which Archer says is how developers encode things like “Martin Smith posted an article to the blog ‘Martin’s Blog’ at 3:04 PM GMT on February 2, 2011.” Activity Streams, Archer says “are serialized in JSON-LD but have the notion of a default @context file so that a [person conversant with the Semantic Web Stack’s Resource Description Framework (RDF), a standard model for data interchange on the Web] can interpret the JSON as JSON-LD even if no context declaration is made.”
Tackling New Demands: Where might the Semantic Web make further headway in 2015?
Scientific researchers and publishers are on Archer’s list, as he sees the potential for developing better interfaces between the Semantic Web world and the HTML, CSS, JavaScript, and JSON worlds. Researchers are now obliged to publish the data that supports their papers and the publishers are responding to this, says Archer. “A lot of the focus is on discovery metadata, which is obviously important, but I believe there’s a lot of potential for Sem Web in making sense of the data itself, making links and enabling new insights – that’s what the Semantic Web is for after all!” he notes. Archer hope to run a workshop in the first half of 2015 around that topic as a guide to helping scientific publishers better curate, enrich and add value to the data they’re now starting to publish alongside the research.
The enterprise architect community is taking heed, as well. John Schlesinger, Chief Enterprise Architect at Temenos, which develops software for retail banking companies, including solutions that will leverage triple stores, says that enterprise architects whose role requires understanding the business and its information systems increasingly believe that they need to build ontologies of the business in order to succeed. Ontologies are a formal and explicit specification of a shared conceptualization, and can be created using the Web Ontological Language (OWL.) “I think we will start to see much more enterprise architecture that’s ontology-driven,” he says.
Customer support communities should also look for semantic technologies to make a growing impact. Tony Sarris, founder and principal at semantic technology consultancy N2Semantics, says that he’s impressed to see some small companies building on successes they’ve had with natural language processing-based Q&A systems for practical applications like enhanced customer support. Consider, for example, Artificial Solutions, which is “trying to build a framework for linking applications at the level of customer support to lower-level IoT-based devices, and for NLP-based assistants in specialized ‘vertical’ domains to coordinate where and when needed through a more centralized ‘horizontal’ assistant,” he says.
Business Buy-In: Is the time at hand?
Interest in the concepts of the Semantic Web will grow among the business community at large in the upcoming year, says Chris Nelligan, VP and CTO of Dow Jones Professional Information Business, which includes offerings such as Factiva, a product that applies semantic tools to its content set to surface relevant information. Companies are finding greater value in how the technologies within the Semantic Web stack – such as RDF, SKOS, OWL, and SPARQL – can be applied to their specific information needs across content internal and external to their organizations, he says. With the advent of Big Data technologies, “organizations have learned that there is a better way to getting answers to their questions than just searching and reading headlines, and that technologies such as natural language processing, graph databases, and inference engines are important tools for extracting the gems of information in their mountains of content,” says Nelligan.
Cloud Computing, which makes it possible to process extremely large sets of structured and unstructured content quickly and with reasonable cost, is a further catalyst to advance Semantic Web adoption in the enterprise, he notes. “Not only have the architectures and platforms needed to construct a semantic processing engine advanced, so have the methodologies, standards and tools (open source and premium) for semantically enriching and interrogating content,” Nelligan says.
Others see similar pickup of general business interest for similar reasons and ends. Paweł Zarzycki, CEO of semantic platform vendor Cognitum, says his company has seen interest pick up across a broad variety of industries. He expects that in the next year, semantically enhanced searching across large volumes of documents they’re managing will be a high item on many companies’ agendas. “This probably will be a vivid market,” he says. Combining semantic technology with a model-driven approach that enables businesses to formalize organizational knowledge, he says, “will deliver great support for end users to perform daily tasks.”
