1,721,016 research outputs found
Sulla Classificazione delle Sedi di Pubblicazione nella Valutazione della Produzione Scientifica
No full textLarge-Scale Deployment of Middleware-Oriented Government Interoperability Frameworks
No full textWe discuss deployment solutions for e-Government Interoperability Frameworks (GIFs). We concentrate on middleware-oriented GIFs, i.e., those in which middleware modules act as intermediaries among information systems that need to exchange data and services. A prominent example is the Italian SPCoop interoperability framework. We review the SPCoop architecture, and two popular open-source implementations of its core modules, called OpenSPCoop and freESBee. We argue that the comparison of these two solutions is relevant since they obey to radically different philosophies in terms of their internal architec- tures, which we call “container-bound” and “container-independent”. Then, we discuss one of the main problems in large-scale deployment of SPCoop-like GIFs, namely the need to quickly deploy a large number of middleware instances over a relatively small number of servers. We report a number of experiments that show how container-bound solutions guarantee better scalability in single-instance de- ployments, while container-independent solutions represent a better solution in multiple-instance deployments
Government interoperability frameworks: Middleware tools for the Italian SPCoop
Full text linkWe discuss deployment solutions for e-Government Interoperability Frameworks (GIFs). We concentrate on the italian SPCoop, a prominent example of a middleware-oriented architectures, where middleware modules act as intermediaries among information systems. We review the SPCoop architecture, and two popular open-source implementations, called OpenSPCoop and freESBee. The two solutions follow radically different philosophies in terms of their internal architectures, which we call "container-bound" and "container-independent". We compare the two systems with respect to one of the main problems in largescale deployment of SPCooop-like GIFs, namely the need to quickly deploy a large number of middleware instances over a relatively small number of servers. We report a number of experiments to discuss how the different design choices impact performance
Middleware-oriented government interoperability frameworks: A comparison
No full textWe discuss deployment solutions for e-Government Interoperability Frameworks (GIFs). We concentrate on middleware-oriented GIFs, i.e., those in which middleware modules act as intermediaries among information systems that need to exchange data and services. A prominent example is the Italian SPCoop interoperability framework. We review the SPCoop architecture, and two popular open-source implementations of its core modules, called OpenSPCoop and freESBee. We argue that the comparison of these two solutions is relevant since they obey to radically different philosophies, both in terms of the relationship to the underlying J2EE container, and of their internal module organization. Then, we discuss one of the main problems in large-scale deployment of SPCoop-like GIFs, namely the need to quickly deploy a large number of middleware instances over a relatively small number of servers. We report a number of experiments to discuss how the different design choices impact performance. To the best of our knowledge, this is the first large-scale test of the framework, from which a number of important lessons can be learned
Noodles: A Clustering Engine for the Web
No full textThe paper describes the Noodles system, a clustering engine for Web and desktop searches. By employing a new algorithm for document clustering, based on Latent Semantic Indexing, Noodles provides good classification power to simplify browsing of search results by casual users. In the paper, we provide some background about the problem of clustering search results, give an overview of the novel techniques implemented in the system, and present its architecture and main features
BUNNI: Learning Repair Actions in Rule-driven Data Cleaning
Full text linkIn this work, we address the challenging and open problem of involving non-expert users in the data-repairing problem as
irst-class citizens. Despite a large number of proposals that have been devoted to cleaning data from the point of view of
expert users (IT staf and data scientists), there is a lack of studies from the perspective of non-expert ones. Given a set of
available data quality rules, we exploit machine learning techniques to guide the user to identify the dirty values for each
violation and repair them. We show that with a low user efort, it is possible to identify the values in tuples that can be trusted
and the ones that are most likely errors. We show experimentally how this machine-learning approach leads to a unique
clean solution with high quality in scenarios where other approaches fail
- …
