The end of the month should see the release of an update to Cognitum’s Fluent Editor 2014 ontology editor, which will bring with it new capabilities to further drive its usage not only in academia but also in industrial segments such as energy, pharmaceuticals and more.
The company is including among its additions the ability to run analytical computations over ontologies with the open source language R and its Controlled Natural Language. What has been lacking when it comes to performing computations over Big Data sets, says CEO Paweł Zarzycki, is some shortcut to easily combine the semantic and numerical worlds. R is great for doing statistical analysis over huge sets of numerical data, he says, but more knowledge opens up when Cognitum’s language is leveraged for analysis in a qualitative way.
As an example, using a quantitative statistical approach, a business could discover numerical properties related to concepts, classes and other taxonomy items representing its inventory – such as how many of various types of items are in its warehouse, how many of these have been shipped in a certain timeframe, how many are due to ship, and so on. This data could be used for procurement, sales and production processes. But the findings can be enriched even further when semantic analysis of the taxonomy enters the picture – how the numbers that were come up with in the statistical analysis relate to specific values for each item (parts, customer segmentations and so on) for all the classes of managed items. “Semantics opens the way to instantly ask many different questions in one place at one time,” he says.
The upcoming version also now supports direct integration with the Protégé open source ontology editor. “Protégé focuses on the structure of taxonomies and ontologies while Fluent Editor focuses mainly on actual meaning,” of OWL ontologies, says Zarzycki, and community feedback Cognitum received was that knowledge engineers would like to have both views. Rather than trying to bring Protégé-like featuers into Fluent Editor, Cognitum thought it would be easier to support collaboration between the tools in one place.
“With the new version there actually is an instant way to transfer ontologies between the windows between the two solutions, so it’s very convenient for the end user to work with both,” he says, to efficiently dig deeper into structure while also modeling rules and ontologies with its Controlled Natural Language, asking questions via SPARQL and natural language, and using plug-ins to visualize it all. This version also responds to users’ requests to efficiently manage industrial vocabularies and ontologies. It implements free-text natural-language annotations for defined terms in SKOS in support of collaborative knowledge editing. “Collaborative knowledge editing becomes a real problem for several organizations since there are more and more engineers and analysts involved, and this process itself also requires proper orchestration,” says Zarzycki.
Additionally, the Poland-based company is ramping up its research partner program, including helping the Maria Curie Skłodowska University by supplying it with its software as it opens a new curriculum for students focused on data processing with semantic technology. Cognitum’s Ontorion Server for building large, scalable solutions for the semantic will be part of that infrastructure. Ontorion Server also has proven successful at creating a clinical decision support application for a research project to improve cancer treatment with the Warsaw Cancer Care Center, which is being evaluated by medical doctors now. “The first goal was to prove this kind of application and complex system can be really described using semantic technology, and that is done,” says Zarzycki.