=Paper= {{Paper |id=Vol-500/paper-3 |storemode=property |title=A Parametric Design Approach to Scalability Management of Semantic Middleware |pdfUrl=https://ceur-ws.org/Vol-500/paper3.pdf |volume=Vol-500 }} ==A Parametric Design Approach to Scalability Management of Semantic Middleware== https://ceur-ws.org/Vol-500/paper3.pdf
         A Parametric Design Approach to Scalability
            Management of Semantic Middleware

                                  Reto Krummenacher
                          reto.krummenacher@sti-innsbruck.at

                        Forschungsinstitut STI, Universität Innsbruck


     The World Wide Web is one of the ground-breaking achievements of modern com-
puter science. The core technology, so simple to use, allows billions of users to publish,
share and consume information from all over the world. With the emergence of the
Web2.0 paradigm, Web services and the Semantic Web, the World Wide Web has trans-
formed to an open interactive market place. Heterogeneous and distributed producers
and consumers – or rather, in the current context of the Web, so-called prosumers –
have to integrate data and to coordinate their activities. Bringing this all together and
lifting it to the level of large-scale Web computing triggers increased scalability, hetero-
geneity and dynamism challenges. In response, recent work in the field of middleware
technology proposes semantics-aware tuplespaces that enable the integration, commu-
nication and coordination of billions of autonomous, distributed and heterogeneous data
and service prosumers. Sources are no longer addressed directly by their endpoint lo-
cator, but are abstracted by the semantic space middleware that provides a unified and
virtually global view of the data published in diverse databases. Such middleware is the
fundamental building block for the realization of distributed application in the true Web
sense: scalable, open and very simple to use.

    In this work we discuss in-depth the requirements and challenges arising in the
context of large-scale Web computing. As main contribution, we elaborate on scalability
implications, and present a parametric design-based user requirements analysis tool that
supports architects during the design phase of semantic space realizations. The tool
aims at facilitating the configuration of scalable and dependable space installations that
match the application needs in terms of functional and non-functional properties. As
proof of concept, we present a Web-scale semantic space implementation that ensures
scalability and decentralization by means of established P2P technology.