70 likes | 218 Vues
Semantics-based Interoperability. Gopal Gupta The University of Texas at Dallas. Interoperability. The IEEE defines interoperability as: The ability of two or more systems or components to exchange information and to use the information that has been exchanged
E N D
Semantics-based Interoperability Gopal Gupta The University of Texas at Dallas
Interoperability • The IEEE defines interoperability as: The ability of two or more systems or components to exchange information and to use the information that has been exchanged • Challenge: The 2 systems may speak “different language” • Solution: Translate between the various languages • Requirement: Translator rapidly developable for this approach to be practical
The Translation Problem • Given a sentence S (a program, a file, etc.), written in a notation Ls, obtain an semantically equivalent sentence T in another notation Lt. • Of course, It should be possible to map the semantics of the source notation Ls to (a subset of) the target notation Lt. • Structure of the notation to be translated has to be inferred via parsing and semantically processed to obtain the encoding in target notation. • Common approach: Build a compiler that compiles programs/documents in notation Ls to those in Lt. Use the compiler to translate S.
Semantics-based Approach • Translators can be rapidly built by writing semantics of Ls in terms of Lt using logic programming (Prolog) • Prolog is a declarative, high level language which allows rapid construction of software • Parsers for even context sensitive languages can be built rapidly (e.g., Nexus) translate(S, T) :- Ps(S, Rs), Mst(Rs, Rt), Pt(T, Rt).
CIF based Approach • Find a common intermediate format (e.g., NEXUS) • Build back-and-forth translators between all other notations and NEXUS • Past & Current Work • Translator between Nemeth Math and Latex; subsequently CIF approach used with Latex as the CIF (spun into a company) • Translators between Architectural Description Languages • Translatorsbetween OO DB and Relational DB • Translators between various XMLs (HTML -> VoiceXML) • Translators between GUI test scripting languages
Other Work & References • Work also done on service discovery & composition as well as semantically describing web-services • References: • G. Gupta Horn Logic Denotations and Their Applications," The Logic Prog. Paradigm: A 25 year perspective Springer Verlag. pp. 127-160. • G. Gupta, E. Pontelli, et al. Semantic-Based Filtering: LP's Killer Approach? 4th PADL LNCS 2257, Springer Verlag, pp. 82-100. • S. Kona, A. Bansal, G. Gupta. Semantics-based Efficient Service Discovery and Composition. In Proc.International Conference on Web Services. 2007. • E. Pontelli, D. Ranjan, G. Gupta, B. Milligan. Design & Implementation of a Domain Specific Language for Phylogenetic Inference. Journal of Bioinformatics and Computational Biology, 1(2):2003. pp. 201-230. • + Others