As we close out 2012, we’ve asked some semantic tech experts to give us their take on the year that was. Was Big Data a boon for the semantic web, or is the opportunity to capitalize on the connection still pending? Is structured data on the web not just the future but the present? What sector is taking a strong lead in the semantic web space?
We begin with Part 1, with our experts listed in alphabetical order:
John Breslin, lecturer at NUI Galway, researcher and unit leader at DERI, creator of SIOC, and co-founder of Technology Voice and StreamGlider:
I think the schema.org initiative really gaining community support and a broader range of terms has been fantastic. It’s been great to see an easily understandable set of terms for describing the objects in web pages, but also leveraging the experience of work like GoodRelations rather than ignoring what has gone before. It’s also been encouraging to see the growth of Drupal 7 (which produces RDFa data) in the government sector: Estimates are that 24 percent of .gov CMS sites are now powered by Drupal.
Martin Böhringer, CEO & Co-Founder Hojoki:
For us it was very important to see Jena, our Semantic Web framework, becoming an Apache top-level project in April 2012. We see a lot of development pace in this project recently and see a chance to build an open source Semantic Web foundation which can handle cutting-edge requirements.
Still disappointing is the missing link between Semantic Web and the “cool” technologies and buzzwords. From what we see Semantic Web gives answers to some of the industry’s most challenging problems, but it still doesn’t seem to really find its place in relation to the cloud or big data (Hadoop).
Christine Connors, Chief Ontologist, Knowledgent:
One trend that I have seen is increased interest in the broader spectrum of semantic technologies in the enterprise. Graph stores, NoSQL, schema-less and more flexible systems, ontologies (& ontologists!) and integration with legacy systems. I believe the Big Data movement has had a positive impact on this field. We are hearing more and more about “Big Data Analytics” from our clients, partners and friends. The analytical power brought to bear by the semantic technology stack is sparking curiosity – what is it really? How can these models help me mitigate risk, more accurately predict outcomes, identify hidden intellectual assets, and streamline business processes? Real questions, tough questions: fun challenges!
One thing I had hoped to see more of in 2012 was a tighter integration of NLP and ontologies – it IS happening, via vendor partnerships, open source integrations and academic research. I had thought the close of 2012 would have seen more successful production deployments. Ask me again at SemTechBizSF!
Seth Grimes, industry analyst, consultant and organizer of the Sentiment Analysis Symposium:
- The advance of pragramatic approaches to semantics for real-world business and technology challenges that include search and social-media analysis. Think the Google Knowledge Graph, Facebook’s Open Graph protocol, rNews for news-specific metadata, and RDFa and microformats. Let the idealists and theorists continue to pursue unrealizable Semantic Web aspirations. Pragmatic approaches get the semantics job done.
- Growing sentiment analysis adoption, making sense of opinions and attitudes in social, online, and enterprise sources, and growing sophistication of commercially available sentiment technologies.
Leading-edge technologies encompass aspect-oriented sentiment, resolved at a granular, feature level and not just at the document level, and beyond-polarity classification according to emotion, genre, and intent.
Dr. Michael Hausenblas, Research Fellow and Research Center coordinator at DERI:
- Uptake of structured Web data via schema.org (Yandex joined the consortium, a number of extensions such as GoodRelations have been taken on board, Dan Brickley appointed as Developer Advocate at Google who’s taking care of it).
- More and more of the big players join the semantic technologies arena: IBM and EMC co-chairing W3C Linked Data Platform WG; Oracle co-chairing RDB2RDF (and now the standard is available).
- People are getting serious about industry standard benchmarks, see LDBC.
- I’m still missing Linked Data applications (which is something that is also true for apps that heavily use/depend on Open Data).
See us tomorrow for part 2!