Découverte des services dans les systèmes d'information géographiques répartis sous réseau ad-hoc

REZEG, khaled (2011) Découverte des services dans les systèmes d'information géographiques répartis sous réseau ad-hoc. ["eprint_fieldopt_thesis_type_phd" not defined] thesis, Université Mohamed Khider – Biskra.

[img]
Preview
Text
Découverte des services dans les systèmes d'information géographiques répartis sous réseau ad-hoc.pdf

Download (2MB) | Preview

Abstract

The actual geographic information systems today provide access to a large number of heterogeneous and distributed sources. The multiplication of resources and increased data requires the use of Web services that appear as a solution to ensure interoperability between different resources. They are able to collaborate and share control of process data and information between applications on different platforms. The use of geographic data through Web services shows the geospatial Web Services (GWS). A growing number of GWS designed to interoperate spatial information over the network have emerged. GWS are changing the way in which spatial information systems and applications are designed, developed, and deployed. When GWS is increasing, the difficulty of the service discovery increases too. The service discovery represents one of the interactions between Web service components (supplier, customer and UDDI). It is unfortunately that it is based only on the syntax. The semantic aspect necessary to satisfy the user is absent. The use of semantic Web technology in the Web service discovery phase gives rise to the Web service semantic discovery. The latter is based on semantic reasoning. It enhances the accuracy of search results compared to traditional techniques of Web service discovery, the additional matching accuracy in terms of computing power. In this paper we are interested to the distributed GWS semantic discovery based on geographical metadata held in geo-catalogs. The metadata used in our case is designed according to the ISO 19119 standard reinforced by the quality criteria. The suggested approach is Geospatial Web services semantic discovery using metadata and multi-agents system. It is based on the co-operation of a set of agents using the semantic capacities of the geospatial Web service through the catalogues based on the metadata according to standard ISO19119. The agents of our architecture take in account the principal properties of the agent technology (Portability, Autonomy, Stability, Persistence, Flexibility, and Sociability). The based idea of our approach is to classify to geospatial Web services by functionalities by using metadata of ISO 19119 standards. Optimization of the recent approach is done by using quality and personalization. Quality management plays a vital role to satisfy the user needs (explicit or implicit). Personalization aims to facilitate the expression of user needs and make information intelligible to the selected user and exploitable. Our discovery approach based on semantic metadata organized in catalogs, proposed in the publication phase of GWS two levels of description. The first level concerns the GWS classes catalog based on the ISO19119 standards and quality criteria; the second level concerns the UDDI registry through the WSDL. In the discovery phase, this solution allows to restrict the search space and increase the number of relevant services through the application of the algorithm of "matching". In the metadata level, the matching algorithm “LARKS” is adjusted by applying the first three filters (context, profile, similarity) on the quality criteria contained in the request and description of classes available in GWS metadata catalog. This solution enables us to have a result like "relaxed match”. In the UDDI registry level, the application of “Matching” algorithm is based on the interrogation mechanism of the UDDI registry using the different pages (White pages, yellow pages, green pages). A list of GWS classified by order of correspondence (quality aspect, functional aspect) is the result of the discovery. For the implementation of our approach, we propose the tourism field as application domain; we use a prototype ontology which includes only the part of the terminology field of tourism. We develop a Web application, an adaptation of a traditional Web architecture. Our implementation is based on the use of metadata and the algorithm of "matching". To make these persistent metadata, we chose the pair XML / XML Schema, based on the principle of markup languages SGML family. A set of interface is available to the user (Provider /Customer / User), the interface is made using PHP that provides ease of interfacing with databases. These interfaces are: 1. Interfaces for acquiring knowledge of the user profile; 2. GWS publication interfaces; 3. GWS discovery interfaces.

Item Type: Thesis (["eprint_fieldopt_thesis_type_phd" not defined])
Uncontrolled Keywords: Catalog, Semantic discovery, ISO 19119, Metadata, Ontology, User profile, quality, Geospatial Web Service, Multi-agents systems.
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76 Computer software
Divisions: Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie > Département d'informatique
Depositing User: Admin02 TMLBiskra
Date Deposited: 16 Jun 2014 10:01
Last Modified: 16 Jun 2014 10:01
URI: http://thesis.univ-biskra.dz/id/eprint/33

Actions (login required)

View Item View Item