Goal of iDocument is a semantic answering of queries on non-structured document corpora(e.g., news, large document collections). Based on existing background knowledge and an unstructured - and on the basis of size and fluctuation open - document body, an intelligent process is developed which filters out relevant documents from the document corpus with the help of a query issued by a user.
The contents of the relevant documents are transformed to a semantic network by methods of information extraction and by linkage with the background knowledge. A semantic search then examines this network with regard to the original query and delivers resulting relations and concepts to the user. With this ad hoc processing of documents for the semantic search, previously non-structured (resp. semantically enriched) sources of information are exploited to semantic search. With it another step is provided to bring the Semantic Web to life.