The idea behind the Linguistic Watermark package is to preserve a logical independence between wrapped linguistic resources, and the different, possible embedding applications which may exploit their content.

As we can observe from the Class Diagram on the right, we sketched a sort of ontology of linguistic resources, with the addition of operational aspects.

The core of the package is composed of an Abstract Class, named LinguisticInterface, which is both the locus for a formal description of a given linguistic resource and a service-provider for exposing the resource specific methods for accessing its content. The other abstract classes and interfaces in the package, which can be implemented or not, depending on the profile of the resource being wrapped, provide instead the signatures for known interface methods.

Linguistic Watermark class diagram (click to enlarge)