AML: a modeling language for designing adaptive web applications

Abstract

Modern WWW methodologies focus on developing context-aware data delivery applications on the Web, but very few proposals provide a conceptual modeling of their adaptation specifications. In this paper, we present AML—Adaptation Modeling Language—to specify complex Adaptive Web Applications at the conceptual level. AML models the adaptation process by using a special workflow involving modeling primitives that execute elementary manipulations. The adaptation considers different and orthogonal components of the context and operates over the selection of the most suitable contents, the organization of the hypertext structure of the response, and the layout of the final pages. The language and its design process are fully implemented in a CASE tool, called FAWIS. Specifically, FAWIS is able to access a Web data source and automatically filter and adapt information delivery on the basis of the context of the client. The tool is able to manage heterogeneous contexts, is easy extensible to new coordinates of adaptation, and can integrate different adaptation requirements solving possible conflicts between them. It allows the designer to define the response that better satisfies the requirements of adaptation for a context.