Semantic Search Systems based on Triple Stores

Semantic search systems are going to play a crucialrole in the evolution from the current Web to theSemantic Web by providing new complexfunctionalities for retrieving aggregating andcombining information. To successfully accomplishthis migration three main directions shall beconsidered: different media formats differentinformation sources and different search strategies.The more systems will be able to seamlessly operateon all dimensions the more will be the chances toprovide an unprecedented level of versatility toknowledge management systems.This dissertation aims at building the representationbasis for systems able to operate on all these threeaxis. A new resource model is proposed which is ableto support different media formats to hostresource-specific data such as metadata descriptionstext explanations etc. and to provide a containerfor semantic interpretations of suchresource-specific data (i.e. for semantic knowledgeextracted from it). A real-world case study is discussed where anadapted version of a Semantic Indexing and Retrievalplatform is used to index and search severalheterogeneous resources.