Automatic Composition of Geospatial Web Services using Intelligent Agents

Geospatial web services have been utilized to disseminate geospatial data and functionalities over the internet. Combining geospatial web services and generating composite web services out of atomic services can provide users with a higher level of functionality and therefore supply more complex needs. These new composite functionalities can also be published over the network as new web services.
Considering the increasing amount of geospatial web services that are published over the web and the dynamic nature of the web, searching and combining geospatial web services manually is a complex and time-consuming task. Therefore, combining multiple web services in an automated manner has received much attention during recent years in geospatial communities.
This thesis deals with automatic composition of geospatial web services. It proposes a solution by which intelligent planner agents are utilized to auomaticlly generate composite services out of semantically described atomic geospatial web services. This thesis presents a novel approach for semantic description of geospatial web services based on SAWSDL and WSMO-Lite technologies. Semantic describtion of geospatial web services are exploited by intelligent agents to generate required composition that can satisfies the required goal. The intelligent agents uses Artificial Intelligence (AI) planning algorithm to combine proper services and synthesis the composition. This research developed an AI planning dialect based on a state transition system in which states are described as RDF graphs, operations of geospatial web services are considered actions, the conditions and effects of each operation are defined as SPARQL ask queries, and to generate a subsequent state, the condition and effect must be applied to the RDF graph of the current state using SPARQL update operations.
In case that an angent is not able to generate proper composition, this theis presented a distributed, multi-agent AI planning algorithm by which multiple agents can cooperate to generate composite geospatial web services.