 |
 |
 |
 |
| |
|||||
|
|
|
Home > News & Events > MITRE Publications > The MITRE Digest > | |||||||||||||||||||
Share and Share Alike: Building an Information Interoperability Toolkit February 2010
A captain at the helm of a container ship traveling along the Atlantic coastline can easily discern the locations of other container ships, passenger vessels, and tankers in the region on a computer screen. But what if he wants to add the locations of U.S. nuclear plants to his display? Or airports? More often than not, the information needed to complete this type of combined picture simply isn't in the correct format, so it can't be done in a timely manner. "Existing data sources often don't mesh with the latest technological standards," explains Peter Mork, a principal software engineer at MITRE. "But there is great value to combining information from many different sources, so knowledge can be shared and extended." Information interoperability is a huge challenge for both industry and government, whether it's a large corporation that must quickly consolidate systems that have overlapping product lines, or a counterterrorism alert that must be rapidly disseminated across federal, state, and local agencies. A team of MITRE engineers is striving to create a fast-turnaround data sharing capability, by addressing the differences in modeling languages, data structures, and semantics. They are working in conjunction with industry and academic leaders to define, design, and build an open source information interoperability toolkit. "Our sponsors are in need of affordable, vendor-neutral interoperability tools that can be expanded on by software developers," confirms Len Seligman, a co-principal investigator on the project, known as Open Information Interoperability (OpenII, pronounced "Open Eye-Eye"). "Our particular challenge is how to express a data model in a way that is neutral enough so that everyone can understand what's going on. Ultimately, we're asking, 'what is the lingua franca for interoperability tools?'" In Harmony "OpenII addresses the fact that independently developed systems often use different terms to mean the same thing, or the same terms to refer to different things," Mork points out. Similarly, the context in which certain terms or measurements are used might be subtly different.
"A simple example is distance stated in meters versus feet, but in other cases, there are much more complex transformations that must occur to make data usable," Seligman says. "Choices of modeling language, like the use of XML versus UML, and data structures and semantics must be addressed." (For a quick look at these terms, see "XML versus UML, Explained," at left.) Several years ago, MITRE technical experts published a paper on how interoperability tools might be harmonized. "In the process, rather than just develop a one-off connection of Product A to B, we tried to take a larger view, examining the fundamental building blocks of a complete data information interoperability tool suite," Mork explains. When the paper was read by researchers in the data integration field, it triggered fervent industry and academic collaboration. A senior researcher at Google liked the framework that the MITRE paper put forward and approached the team to see if there was an opportunity to join forces. The result was a conference of some of the most astute minds in the field at Google's California campus, co-organized by MITRE in 2008. Researchers from Microsoft, IBM, Yahoo, University of California at Irvine, University of Wisconsin, University of Pennsylvania, and others arrived to explore the possibilities of a vendor-neutral, open source information interoperability toolkit. Cross-Country Collaboration Since then, the collaboration among MITRE, industry, and academic leaders has continued to steam ahead, towards the ultimate goal of the first release of a prototype toolkit. While tools developed at MITRE form the core infrastructure, several parties are taking an active role in the project's development. Google and the University of California at Berkeley, for example, are working on high-performance schema searches, while IBM contributed two tools to the OpenII infrastructure. One of the biggest challenges is taking into account a multitude of scenarios when unanticipated data sharing might suddenly become critical. For example, if there is a natural disaster—or the threat of one—federal agencies have an urgent need to exchange information about the national infrastructure, from water supplies and the power grid to nuclear plants. Getting Real While information interoperability has been a subject of study in the data management field for more than two decades, it's only in the last several years that research is beginning to make a significant impact. "We're at an exciting point, because the application of the technology is just emerging," Seligman says. The next phase for the project is to incorporate an XML query capability, which is being developed at UC-Irvine. "Adding that feature will complete the pipeline, providing the very end-to-end product that we envisioned at the beginning of OpenII," Mork adds. Meanwhile, there have already been some positive impacts for MITRE's sponsors. MITRE engineers produced a flexible data management system for the U.S. Army's Warfighter Development Office. That analysis capability allows the system to rapidly configure a variety of simulations, concluding with results that can be evaluated before next steps are taken. "This removed the need for the sponsor to manually design one-time-use databases and write custom XML programs," Seligman says. "The project has proven to be major step towards the goal of creating fast-turnaround analytic capabilities that can be transitioned to MITRE's other sponsors. "We're a networked company. We have close working relationships with leaders in this field, and we've nurtured them over the years so we have an awareness of our customers' most pressing problems." "OpenII isn't just about information-sharing," Mork concludes. "It's also about being able to make sense of it, with meaningful results coming out at the other end so users get the true benefit of combining diverse data sources." —by Cheryl Scaparrotta Related Information Articles and News
Technical Papers and Presentations
Websites |
||||||||||||||||||||
Page last updated: February 26, 2010 | Top of page |
Solutions That Make a Difference.® |
|
|
