A cloud-based platform for semantic enrichment, linked data publishing and search technologies is underway now at startup Redlink, which bills it as the world’s first project of its kind.
The company has a heritage in the European Commission-funded IKS (Interactive Knowledge Stack) Open Source Project, created to provide a stack of semantic features for use in content management systems and the birthplace of Apache Stanbol, as well as in the Linked Media Framework project, from which Apache Marmotta derived. The founding developers of those open source projects are founders of Redlink, and Apache Stanbol, which provides a set of reusable components for semantic content management (including adding semantic information to “non-semantic” pieces of content), and Apache Marmotta, which provides Linked Data platform capabilities, are core to the platform, as is Apache Solr for enterprise search.
“We have taken core open source technologies and now we are going to help enterprise integrators leverage them so they can help their customers make sense of their data,” says John Pereira, cofounder and managing director of Redlink. These technologies come together to enable a workflow to enrich textual contents, index them with custom vocabularies, link them to the Linked Open Data cloud, and then help users find what they need via semantic search.
The idea is to make it possible for integrators to deliver these capabilities hat Stanbol, Marmotta and Solr provide to enterprise customers and build value on top of them, by giving them the ability to access via the cloud customized APIs and SDKs connectors into major enterprise applications and content management frameworks. The platform opens the door to enterprises having a common, uniform representation of the data in their company, and having it correctly marked up so that all systems benefit from the high-quality metadata. News and media organizations can gain value, too, with high-quality metadata at the entity level that helps efforts to reuse content with minimal manual curation, Pereira says.
“Integrators create value in terms of feeding the enhanced and linked data into business workflow processes,” Pereira says. And plug-ins for capabilities ranging from natural language processing engines to content analytics apps means delivering clients the extended functionality they want without themselves having to become too deeply embedded into semantic technology. “We’re out to democratize semantic technology,” he says.
Redlink sees these highly configurable, easy-access plug-ins being available on the Redlink Marketplace, which should debut next year, courtesy of premium data providers and other third-part application providers. It’s already testing capabilities with sources including an NLP provider that performs entity disambiguation across multiple languages, as well as providers working specifically on Alfresco, WordPress and Drupal environment plug-ins.
NLP definitely seems to be a hot spot, he notes, mentioning that in Europe that are a lot of NLP providers in the commercial and academic sector who really could benefit from a marketplace to score with enterprises, who he says are beginning to demand being able to interact with data using a more natural-language approach.
Redlink plans to go live with APIs and SDKs in the first quarter of 2014, and in the meantime is looking to conduct additional beta testing of the platform with selected large enterprise integrators. In practice, those integrators will buy its services on a processing-unit basis, and charge those fees back to their clients, Pereira notes.
“We want a long-term commitment to open source and to innovation in this field,” he says, “and the value is in wrapping these technologies and providing the industry with easy connectors into different kinds of platforms and selling that on a CPU basis.”