During the recent Semantic Technology and Business Conference in San Francisco, a motley crew of expert presenters got up in front of a packed room, took a deep breath, and spoke passionately about the semantic projects nearest and dearest to their hearts while the unforgiving clock ticked their five precious minutes away. At the conference I shared highlights from some of those aptly named Lightning Sessions. Here are a few more snappy sessions that captivated the room that day:
Semantic Technology to Shed Light on Big Dark Data with Ben Zamanzadeh, DataPop
DataPop is a startup in the field of semantic advertising. The company seeks to create actionable insights for clients with semantics. As Ben put it, "Ad data is still dark data. Consumer actions are very hard to understand and even harder to predict." The talk description explains DataPop's approach: "DataPop's Semantic Advertising Technology uses Machine Learned Semantic Models to build and analyze advertising campaigns that surpasses conventional advertising capabilities. Composite Semantic Data Models are used to translate Big piles of Data into meaningful entities, then Inference Engines transcribe information such that decisions and strategies can be formed. Semantic Methods has made it possible for us to explain the reasoning behind 'why' things happen."
DataPop utilizes their own taxonomies and ontologies, but they also rely on the human element to analyze data with the best results. Ben insisted that with semantics, DataPop is able to model data better, and this leads to clearer insights. "The human component is key," Ben said, "but we can learn more when we can model data better. Semantic models allow for actionable insights. They bring together seller intent, consumer behavior, ad perception, and more." Ben's five minutes ended abruptly, but you can learn more about what he's doing at DataPop.com.
Why Schema.org with Richard Wallis, OCLC
Richard Wallis is well regarded throughout the semantic technology community for his speaking prowess, and the five minutes he was allotted for his lightning talk didn't limit Richard in the slightest. Giving the talk that garnered by far the most laughter and enthusiasm from the crowd, Richard discussed why schema.org is a good (dare he say the best?) way to publish structured information on the web. He spoke specifically about library data, but Richard opined that schema.org is the right choice for all types of structured data.
Richard started by pointing out that people looking for books don't turn to library catalogs anymore. They don't even turn to Google Books or Google Scholar. They just go to Google. In fact, Richard noted, "80% of searchers who try to find books in catalogues are coming from outside search engines." He continued, "Google doesn't have the data. But Google does understand schema.org." Schema.org is a broad vocabulary for describing things on the web, and it is being adopted at a rapid pace. Just one year after it came into the public sphere, seven to ten percent of searches have schema.org mark-up. Libraries should use schema.org at the same time as other standards and vocabularies so that book and media information can be found by the rest of the world, not just local communities, Richard insisted. Learn more about what Richard is doing at the OCLC website.
Image: Courtesy Angela Guess