Schemalet: A Modeling Language for Data and Schema Translation

K. Kuramitsu (Japan)



A translation is a promising approach to resolve semantic heterogeneity. The problem with that is an exponentially increasing number of translation rules. To reduce the specification task of them, we design a conceptual modeling language, called Schemalet. The Schemalet specifies the concept of data by the interrelations of data elements. The comparison of two interrelations enables us to detect the difference of two data concepts. If both are of the same concept, the schemalet automates to derive structural mappings from the interrelation. The strength of our approach is that we can derive a new translation rule with preserving semantic consistency.

Important Links:

Go Back