Semantics

We are convinced that Semantics, Semantic Technology, and the Semantic Web can now play a major role in reshaping enterprise information systems. We recognized early on the importance of semantic technology, and for the past ten years have actively promoted and popularized its adoption and application.

Designing and Building Business Ontologies

This is one of the consulting services we offer–we are recognized as expert at designing and building business ontologies. In building your enterprise ontology, we bring the benefits of speed, an enduring foundation, elegance of design, and unsurpassed visualization methods. Forward-looking firms see the time as NOW to move ahead with building their business ontologies. We can jump start the process and save you time and cost.

Designing and Building Business Ontologies (DBBO) is also the name of a four-day hands-on course we offer that covers the key technologies, how they work and what they can be used for. This link has a detailed description and the dates and location of the next public version.

We also conduct this course in-house, and have done so for half a dozen public and private organizations.

If you’re interested in knowing more, contact Mark at (970) 402-1474 or marka@semanticarts.com

gist

gist is our minimalist upper ontology. It is available free of charge as an OWL ontology at http://ontologies.semanticarts.com/gist/gist.owl You can read more about it here.

Semantic Conference

We co-produce and chair the largest conference in this sector, The Semantic Technology Conference.

Now in its sixth year, it attracts well over 1,000 attendees, and is widely hailed by the participants as a high energy, high impact place to be.

In 2010 we’ll be in San Francisco, June 21- 25.

Our Applied Research Program for Semantics in the Enterprise

We are attempting to bridge these two observations:

One of the greatest contributors to the runaway complexity in current enterprise environments is the unchecked proliferation of subtle distinctions between similar bits of information. The drawback with current technology is that these slight distinctions in meaning must be baked into the structure of your databases. Taming the beast will require large doses of semantic discipline and the application of semantic technologies.
At the same time, semantic technologies have not yet been deployed widely in traditional corporate environments, leaving open questions about its robustness and ability to scale.

We are pleased to report that there are implementations with over 10 billion triples that perform many tasks as well as or better than equivalent relational systems. Many of the technologies can be delivered at design time instead of run time, reducing the risk of performance issues.

We have now built enterprise ontologies for a half-dozen clients, and have tackled a number of issues around performance and scaling. Further, we have built our ontologies based on gist (our minimalist upper ontology) which results in a much more elegant model with less ambiguity.

In 2010 we have built tools that automate the process of deriving a message model from an ontology. Note that this is similar to the practice of creating “canonical messages” but done with more rigor.

Further Research on Semantics in the Enterprise

We have been working on a number of additional initiatives and experiments that are about ready to be applied. We are looking for visionaries and early adopters. Understand that these projects are not completely proven; these are not projects that all your peers have already done and you just need to play catch up. At the same time they are not pure R&D: most of the basic research has been done and what’s needed now is some discipline to apply that research to real world problems.

Our list of “shovel ready” semantic projects (in approximate order from least experimental to most):

Internal LOD Cloud – The linked open data cloud is pretty well established (although harder to use than it seems like it should be). We believe all the pieces are in place to create an in-house, easier to use and more tightly controlled cloud. Major advantages would be finally linking up structured with unstructed data, and driving this through wiki-like UIs.
RDF-based SOA – SOA is currently based on static XML (typically WSDL Web Services). One of the better ways of getting to a more dynamic payload will involve encoding the message in rdf. There are several possible variations on this theme, depending on the needs of the particular situation.
Semantically Driven MDA – We built and patented some pretty sophisticated Model Driven Architecture in the late 90′s. It was pure run-time data-driven MDA, with no code generation. It was based on an implicit central database model. We now know a great deal more about topology and semantics. It is possible to build a distributed, ontologically driven MDA.
Scalable complete inferencing – One of the biggest hurdles to the adoption of Semantics is the poor performance of complex inferencing. We have an approach and a set of algorithms, looking for a sponsor, that we believe will blow away everything out there and make possible a whole new generation of applications.

If you’re interested in pursuing any of these, Dave is more than eager to talk: (970) 490-2224 or mccomb@semanticarts.com.