LOGO: Semantic Technology & Business Conference; June 2-5, 2013, San Francisco, CaliforniaWhy should enterprise IT leaders start steeping themselves in semantic technologies? The answer to that question will become apparent to anyone attending the June Semantic Technology & Business conference in San Francisco, where many sessions will explore the value CIOs and their staffs can gain from going semantic. (You can register for SemTechBiz here.)

Let’s start with the problem of forcing enterprise knowledge workers into rigid procedures to accomplish their activities, the result of having to adhere to flow-charted business processes whose silo’d components are pieced together via fixed integration points. Dave Duggal, co-founder and managing director of EnterpriseWeb LLC, will paint a picture at this session instead of a world of smart, connected business processes to stand up a team of empowered and interactive knowledge workers. Once accorded certain rules-enabled permissions and information access rights, those employees can put their smarts to work “to do their jobs in a goal-oriented way to meet the objectives of the organization,” as Duggal explains it.

A business, he says, is a network – one of people, capabilities, information and rules, “and in an ideal situation and universe you would seamlessly interoperate among those,” vs. having to work according to the confines of a flow chart. The various pieces of this ideal network would come together on demand in the service of real-time business applications and processes, and getting to that state, he says, means stepping into the domain of semantic enterprise application integration. “There is EAI and there is real-time semantic EAI, where at millisecond speed we find the resources, connect them, transform them, process and orchestrate them, and push back a personalized response to the business user or customer every time,” Duggal says.

That’s the outcome of a system that indexes every event, activity, and piece of history, and makes each available to every execution of every process all the time via write-once-use-many-times adapters. “That makes the network available. It creates an abstraction, a virtualization layer that makes it look like a unified semantic layer where all the processes can be interconnected, even if some data lives in silos or third-party systems, which it will,” he says. “All those connections now live in that beautiful semantic layer, and it facilities something very powerful.”

What that powerful thing is is innovation, thanks to the system optimizing for the knowledge worker, around an actual present interaction vs. around some procedural, preconceived notion of what the process should be. “You want system automation to analyze your current situation and return to you a best response as well as proactive guidance and next best actions, so you can move through the enterprise as efficiently as possible,” he says.

Climb Aboard For Data Agility

In another take on the topic, Blue Slate Solutions’ CTO David S. Read and senior consulting software engineer Michael Delaney will add up when to consider semantic technology at this session – and when not to. “If they have a need for a flexible data architecture, if you will – if the layout of their data changes regularly, then semantics are going to get the point,” says Delaney. “If they have large influxes of new kinds of data, semantics is something to look at.” At the same time, there’s no sense in force-fitting it where it doesn’t belong: “A high-volume system like web ordering is a problem that has been solved, so semantics doesn’t fit that,” he says.

So, back to the issue of data agility, which plays out in ways like simplifying analytics work. “A lot of time gets spent on the process of creating a variety of analytic models, on pre-processing data and bringing it together,” says Read. With semantic technology in the picture, resources don’t have to be dedicated to figuring out if two columns in different tables are the same – over and over and over again. “That’s done for you. You’ve already decided what something means and the fact that it’s in five places is okay.” If you don’t have to finagle with data in various pre-processing steps, you can more quickly get to leveraging analytic techniques. Not only that, but from there it’s easier to tie results to the original meaning of the data, “because you started with an ontology of what stuff means” and to what it relates.

Things get even cooler when the enterprise can take advantage of semantic OWL reasoners to provide a forward-chaining environment, to go beyond the data and relationships that already are understood. “The real power is for them to start to combine other information, and those relationships, and let the environment infer some of the interesting connections for them,” Read says.

As an example of how all this can serve, Blue Slate plans to highlight its experience with a client in the Medicare payer business that was dealing with lengthy and costly claims review processes, including for the majority of cases where there were no defects or concerns with the claim. “They were paying people just to look and say that it’s right,” Read says.

What the client needed was a system that could start to learn from the data so that it could figure out what claims actually should be extracted for review. “Semantics fits well there. You can create concepts on the fly, and manipulate the ontology in a straightforward way. You’re also dealing with different types of information, so you need the semantic paradigm to give you that flexibility,” he says. The system that was deployed integrates with the Medicare payer’s legacy databases but adds the semantic tier and reasoner on top of it. Says Read, “They’re excited, and now are starting to use the reasoner to find patterns within the data.”

Tune in tomorrow for part 2.