Innovative Operating Software

This effort examines the merit of using Semantic Web technology, as developed by the World Wide Web Consortium and DARPA, to reduce the time and money required by MDA to acquire and maintain software. MDA has a large collection of legacy software andalgorithms in a variety of programming languages. These same algorithms are useful throughout MDA, yet the cost of rewriting code is enormous. We propose to investigate the use of eXtensible Markup Language (XML) and ontologies to develop a framework forlanguage translation. This framework will reduce the resources required to build new software with tools like MATLAB and to reuse legacy software by translating from original languages and MATLAB to a variety of programming languages. The framework isrobust, as it extends to new input and output languages easily. The heart of the framework is the definition of an abstract language, CodeML. By translating legacy or MATLAB languages into CodeML as an intermediate step, the languages can then be analyzedand optimized. MDA can then share algorithms and tools across platforms. We propose a library to store source code and its CodeML representation. This library will have an intelligent search capability using Semantic Web technology. The results of thiseffort will benefit MDA and other Government agencies by fielding and maintaining quality software faster and cheaper than is required today. MDA will be able to share algorithms across systems, not just programs. It will be less costly to maintain overthe life cycle of the system. There are many commercial applications, wherever language translation will reduce software development costs and time.