Yesterday we said a fond farewell to 2013. Today, we look ahead to the New Year, with the help, once again, of our panel of experts:
Phil Archer, Data Activity Lead, W3C:
For me the new Working Groups (WG) are the focus. I think the CSV on the Web WG is going to be an important step in making more data interoperable with Sem Web.
I’d also like to draw attention to the upcoming Linking Geospatial Data workshop in London in March. There have been lots of attempts to use Geospatial data with Linked Data, notably GeoSPARQL of course. But it’s not always easy. We need to make it easier to publish and use data that includes geocoding in some fashion along with the power and functionality of Geospatial Information systems. The workshop brings together W3C, OGC, the UK government [Linked Data Working Group], Ordnance Survey and the geospatial department at Google. It’s going to be big!
[And about] JSON-LD: It’s JSON so Web developers love it, and it’s RDF. I am hopeful that more and more JSON will actually be JSON-LD. Then everyone should be happy.
Christine Connors, founder and information strategist, TriviumRLG:
Much of my attention this past year has been on technologies for automating categorizing and classifying information. More of the Fortune 2000 wants these capabilities, and the players in the space want to expand their offerings with taxonomy management capabilities and the “NoSQL” glamour of a “triplestore.”
I think there are going to be some very interesting joint product offerings in 2014, and likely more M&A activity because of it. More blending of the what and the how, as it were. I think we’re going to see datastore companies buy up NLP/auto-categorizing tools and metadata management tools to beef up their offerings.
We’ve seen it before, I think the cycle is almost come back around. Time will tell!
Seth Grimes, industry analyst, consultant and organizer of the Sentiment Analysis Symposium:
I expect continued strong growth in content analytics – content content = text, images, audio, and video – powered by emerging deep learning methods, with increasing adoption of semantics-reliant synthesis and sense-making methods and technologies, accompanied by a strong dose of visual analytics.
If you look at the message boards and discussion groups on schema.org and the W3C one of the most interesting is about coming up with a vocabulary to describe actions. The web is mostly about nouns today and we need to get verbs, and schema.org is part of a much larger ecosystem of players required to make that happen. We are going to start to get there.
Also, the transition to mobile is [huge]. The modern mode of [using the] web is access from mobile [devices], and you don’t have the option of the same flexibility and dexterity as on the desktop where you have a monitor, keyboard and mouse. Here the system needs to have a greater understanding of what a user is up to [and] having structured data enables the system to develop this deeper level of understanding.
James Hendler, Tetherless World Senior Constellation Professor and
Director, Rensselaer Institute for Data Exploration and Applications,
Department of Computer Science and Cognitive Science Department
Rensselaer Polytechnic Institute (RPI):
I expect to see the continued growth of the work that is starting from search and social networking to spread. I hope we’ll also see some more companies that are using these technologies start to make it bigger – we’ve seen a number of successes in the small, and of course the big kids are using the technologies in various ways. What I don’t think we’ve seen yet, and I keep waiting for, is a breakout company where this is a critical component — we’ve seen some good attempts, and there’s some interesting things I’ve seen in stealth, so I’m optimistic we’ll see increasing use of these technologies.
The other thing I think we will see is an increasing growth in the relationship between semantic technologies and Big Data. I’m being asked to speak on that more and more these days, and it is interesting that many of these are coming from traditional data conferences and people. More and more people are realizing that discovery, integration, validation and other such critical aspects of Big Data need semantics. I think we’ll see more and more of this over the next few years.
Elisa Kendall, principal, Thematix Partners:
I think [the standards initiatives related to infrastructure, such as the Linked Data Platform, that were kicked off or received more attention this year (see Good-Bye 2013)] is an area to watch for next year, and that we’ll see more and more products that build on this kind of infrastructure over the coming months. Included in this mix will be increasing evidence that triple stores and graph databases are part of the standard web, n-tiered architecture; that the “Big Data” conversation turns more towards content classification, clustering, and visualization rather than the nature of the back-end; increasing talk of and actual use of large scale recommender systems, and so on.
We’ve [also] seen increasing numbers and kinds of tools for search in particular over the last year – Google Now and Hummingbird, Microsoft Satori, Facebook’s Graph Search, among them. I think this is the tip of the proverbial iceberg, and that we’ll see more and more tooling become available, particularly around semantically-enabled content search and management. These will have to include:
● better tools for mark-up using schema.org, rNews, Prov-O, and other vocabularies for Microsoft, Google, Adobe, and others;
● tools for richer content organization and provenance that enable “normal people” to use and build out vocabularies and ontologies to support their internal corporate mark-up requirements;
● core content management systems (CMS) that follow Drupal’s lead in integrating RDFa; and
● more/better front-end tools to support question answering and decision support that leverage semantics for use by “business users”, including authoring as well as visualization capabilities, among others.
[Also], it’s taken us a very long time to get to the point where people really want to standardize domain vocabularies outside of healthcare. Dublin Core and SKOS have been available as standards, providing vocabulary infrastructure that others can build on, for almost 10 years. I think in 2014 we will see increasing roll-out of more vertical ontology development activities, not only for healthcare but in finance, insurance, law, retail, and other domains where there is increasing need from an information classification perspective, and also with respect to data quality management and governance, now that industry is beginning to see the value proposition.
Finally, and as our infrastructure, tools, and ontologies begin maturing, I think we will see increasing adoption of semantic technologies as a component of larger enterprise architectures — for content management, decision support, question answering and recommendation systems, information harmonization and consolidation, master data management, search, and so forth.
But, for this to happen, we as a community need to move our messaging and education efforts from being semantics-centric to semantics-as-a-piece-of-a-larger-enterprise-strategy centric, which is only just beginning to happen as well.
The year has only just begun, and so too have our experts’ predictions. Join us tomorrow for Part 2 of what to look forward to in 2014!