Omar Boucelma

web services customization: a composition-based approach

In this paper, we propose a framework for dynamic Web services customization. The basic idea consists to automatically transform Web services configurations, published in services directories, into other ones better suited for client needs. The goal of this automatic transformation, based mainly on dynamic Web services composition, is twofold. On the one hand, it makes it easier for Web services providers to perform services customization. On the other hand, it allows services requesters to obtain a Web service with the configurations that satisfy their preferences, even when those preferences are not explicitly published in Web services directories.

The virGIS WFS-based spatial mediation system

The proliferation of spatial data on the Internet is beginning to allow a much wider access to data currently available in various Geographic Information Systems (GIS). In order to provide effective spatial database integration, we need to provide flexible and powerful GIS data integration solutions. Indeed, GIS are highly heterogeneous: not only they differ by their data representations, but they also offer radically different query languages. A GIS mediation approach should provide an integrated view of the data supplied by all sources, and a geographical query language to access and manipulate integrated data.

Schema matching in GIS

With the proliferation of spatial data on the Internet, there is an increasing need for flexible and powerful GIS data integration solutions. Recall that a data integration system provides users with a uniform access to a multitude of (local) data sources. The user poses his query against a virtual (global) schema, which is rewritten into queries against the local sources. A key issue in this context is to provide (semi)automatic schema mappings between schemas. In this paper we describe a machine-learning based approach for GIS schema matching. Our approach extends existing machine-learning approaches for (traditional) data mapping but departs from them due to the nature of geographic data. Our solution reduces the complex mappings by identifying different values of a determining property.

Dynamic Web Services Personalization

In order to fulfil current customers requirements, companies and services providers need to supply a large panel of their products and services. This situation has led recently to the mass customizing paradigm, meaning that products and services should be designed in such a way that makes it possible to deliver and adapt different configurations. The increasing number of services available on the Web, together with the heterogeneity of Web audiences, are among the main reasons that motivate the adoption of this paradigm to Web services technology. In this paper we describe a solution that allows automatic customization of Web services: a supplier configuration, published in a services repository, is automatically translated into another configuration better suited for fulfilling customers needs

Customizable-Resources description, selection, and composition: a feature logic based approach

Users preferences heterogeneity in distributed systems often forces resources suppliers to offer customizable-resources in order to fulfill different customer needs We present in this paper a Feature Logic based approach to customizable-resources description, selection, and composition In our approach, resources and requests are both specified in a logical framework by feature terms The feature terms unification technique allows reasoning on these specifications in order to select and possibly compose the resources that are candidate to satisfy a client request.

Managing web GIS quality

Data quality descriptions play a key role in Geographic Information Systems (GIS). With the availability of geographic data sources over the internet, there is a real issue to be addressed, that is how to ensure quality of web-based heterogeneous GIS. The issue is threefold: first, we have to perform data integration, second we have to manage the quality information if any, and third, we have to provide a mechanism that mixes two technologies, which, to the best of our knowledge has not yet been fully addressed.