Jennifer Zaino
SemanticWeb.com Contributor

American Tower, an owner and operator of communications sites for the wireless and broadcast industries, announced this week that it is moving from using Microsoft SQL Server Integration Services for its data warehousing application to use Expressor Software’s semantic data integration system.

What is semantic data integration? In Expressor’s case, it’s a patent-pending solution that leverages what it calls smart semantics for data mapping. The traditional path to data integration—combining data from different sources into some sort of unified view of the information that the business requires—relies on mapping sources to targets against technical or physical metadata, and then writing business or transformation rules that are more or less tied to these physical metadata constructs. The result is that there’s too much time spent on mapping over and over again to each and every application a business needs, and as the volume of data grows exponentially the challenge does as well.

“We looked at our space and said how, for so many years people did data integration over and over the same way, mapping sources to targets, physical metadata to physical metadata, and it’s not the right way of doing it because there’s no abstraction,” says Michael Waclawiczek, VP of Marketing at Expressor.

What Expressor does is called semantic rationalization, employing algorithms that automatically propose common business definitions that most likely resemble the business term a company would like to use, based on the metadata it reads in from sources and targets. Once that metadata is mapped to common definitions, any business rule or transformation rule is written against those common definitions, which enables those rules to be reusable across projects.

“There are several roles in our system and several user interfaces where you are totally abstracted out from the physical metadata, we’re not exposing it to you. So your whole way of interacting is based on those common business definitions,” he says.

A benefit there is that both the business and the technical community that are involved in these projects can speak the same language rather than get lost in complicated conversations. “The conversation in our system is more based on those abstracted terms and that’s the communication those sites would use,” he says.

As for its use of the term semantics to define the software, “the reason we are using semantics is because we are reconciling the differences between bits and pieces of metadata,” says Waclawiczek.

While its notion of a semantic rationalization process does not adhere to W3C Semantic Web definitions, the company’s overall strategy is to build over time more of an ontology into the system, collecting other pieces of metadata around subject areas—domains such as finance, for instance—and exposing them as defined metatypes in highly graphical ways to users. Also on the table are plans to launch a community metadata repository where companies or individuals can submit rules and metadata, check them in and out and use them within their own domains.

“So over time this could be pretty cool,” Waclawiczek says. “If I am building a new application and interested in some sort of common definitions and business rules for this domain, then I could go to the community repository, see what’s there, and check stuff out and make contributions.”

“We have assumed over time that customers will build so many rich sets of metadata that there will be, I think, lots of interesting things we can do on the semantic and the ontology side of things,” he says.

With its solution Expressor is aiming to target both global and mid-sized companies, the latter of which have often been left out of these opportunities because of what typically are high costs.

“There’s definitely the notion out there in market that data integration software is too expensive,” Waclawiczek says.

Expressor says it offers a usage based pricing model that can let users get their job done at up to 80 percent less expense than some of the leading data integration solutions on the market.