An extensible and scalable Content Adaptation Pipeline architecture to support heterogeneous clients

The importance of middleware and content adaptation has previously been demonstrated for pervasive use of Web-based applications. In this paper we propose a modular extensible, and scalable middleware component called the Content Adaptation Pipeline that performs content adaptation on arbitrarily complex data types not limited to text and graphic images. Furthermore, the architecture can be used as part of many client-server applications, not just Web browsers. In our work we leverage the XML language as a uniform means to describe all the elements in our architecture, including the client device and user profiles, the data characteristics, the transcoding operations performed on the data, and the resultant adapted data. We illustrate the flexibility of our architecture to support new data types and adaptation operations by first showing its use with data from a real-world medical application and then extending its capabilities to handle animated graphics and also real-time streaming RTP data. Finally, we demonstrate scalability in our architecture by executing the Content Adaptation Pipeline over a distributed set of servers running an efficient protocol.