Content area
Abstract
Automatic Service Discovery and Selection are crucial phases in the workflow of geospatial web services. The traditional Web services architecture use a keyword based approach to match a query to one or more service providers. Because of the rich diversity of geospatial terms and their usage in geospatial domain, strict keyword search leads to an inefficient service discovery mechanism. This results in human interventions in these phases. The issue of semantic heterogeneities also has to be taken care during the service discovery phase. In this thesis, we have proposed, designed and implemented DAGIS – Automatic Discovery of Geospatial Information Services Framework which provides an interactive web based user interface for automatic discovery, selection and dynamic compositions of the geospatial services on the fly to solve the user queries. The emerging standards for Semantic Web and Semantic Web Services have brought a new dimensionality to this research challenge. In this thesis, the major issue of semantic interoperability between different geospatial data sources is being addressed by employing higher-level service ontology specifications for the web services handling geospatial data using OWL-S and through domain specific Ontologies. The issue of selecting the best service from set of functionally similar services for the user query is also handled in DAGIS framework. QoS based service selection algorithm through semantic matching using QoS Ontologies have been developed for this framework. We have also proposed a set of geospatial criteria that can be used alongside the regular QoS parameters in service discovery and invocation. The framework is capable of performing dynamic compositions on the fly through recursive inference based back chaining algorithm. The framework is evaluated by solving queries posed by users in various geospatial decision making scenarios. This thesis provides the phases involved in building this framework, challenges faced, and the future work.
