A favorite component of the Semantic Technology and Business Conference is the Lightning Round: an hour of five-minute talks given by excited executives and entrepreneurs fighting the clock to get across their ideas in just five minutes. If the participants went over the time limit (which was boldly displayed for all of us in the packed audience to monitor) they were very politely clapped off the stage. Surprisingly, many of the presenters had their speeches timed to the second, but others would have killed for fifteen more seconds.
At the end of the hour, the crowd of attendees left with brains stuffed to the brim with new ideas about a wide range of topics. The following are just a few highlights from the action-packed hour.
Language Meets Knowledge - How to Achieve Cross-Border Interoperability with Jochen Hummel, ESTeam AB -- In the first five-minute round, Jochen elegantly described how the European Union's Office for Harmonization in the Internal Market (OHIM) has, with ESTeam's help, reduced the amount of time it takes to register a trademark across the 27 countries of the European Union from over a year to just ten weeks. Issues involved in the trademark process include checking that the new trademark doesn't conflict with any existing trademarks in any of the 22 languages represented in the union.
To do this in a timely manner, OHIM could not rely solely on translation. Instead, they turned to linked data, mapping key concepts from each of the 22 languages to an ontology. In essence, they created a map between concept systems in different languages, allowing for a much more efficient examination process than translation alone could offer. The applications for such a system go far beyond trademarks, but Jochen, alas, reached the end of his five minutes before he could delve too deeply into the obvious potential for international businesses.
Ontology Evaluation across the Life Cycle: Key findings from a four-month collaborative exploration with Amanda J Vizedom, Independent Ontologist -- Amanda spoke more passionately than any of her fellow speakers about the need for semantic professionals to come to a consensus regarding how ontologies should be evaluated. She commented that there is currently quite a bit of adoption of ontologies, but very little communication between the creators of those ontologies, and that lack of discussion means that best practices are not being formed and time is being wasted.
At the recent Ontology Summit 2013, attendees worked on ways to fix this problem and came up with ideas about how to talk meaningfully about ontology life cycle management. The group found that while no two ontologies go through the same sequence, there are patterns in the dependencies that must be met between various phases of the ontology's life cycle. "Evaluation against phase appropriate requirements is key to creating the most effective ontology," Amanda commented, and she urged her listeners to pay more attention to this entirely too-overlooked topic.
Semantic Reading with Eric Freese, codeMantra -- As a lover of good books, the five-minute session that piqued my interest more than any other was Eric's presentation of the cloudshelf enhanced eReading system. Eric started by pointing out, "The knowledge locked in a book is inaccessible as soon as the ink hits the page, and the same is true of eBooks." All that you get is what you see. With cloudshelf, deeper levels of knowledge contained within the pages of a book are unlocked.
Using RDFa, the system generates links within the text of eBooks to relevant supplementary information, definitions of key words, images, etc. The system also allows you to ask questions of the book to find information you know is in there somewhere or to find all of the passages about a certain topic. For example, Eric showed that a typical search of the Steve Jobs biography for "mammals" won't return any results, but an enhanced semantic search through cloudshelf will prove that Steve was, in fact, a human being.
Image: Courtesy Angela Guess