Integrated semantic search in the product development phase

Finding relevant and needed information within company's specific repositories or internet resources has become a time consuming part of the development process. In order to optimally exploit the large amounts of engineering information one of the key challenges is, that an engineer has to be aware of proper keywords and all kind of relevant repositories in order to start a successful search. Especially with respect to unstructured repositories such as fileservers an explicit search process becomes complicated, time consuming and sometimes impossible, due to the lack of appropriate metadata. By combining existing technologies from the areas of natural language processing, search engines and desktop integration frameworks a concept for an integrated semantic search tool is proposed, where a user has not to be aware of the repositories and their specifics. As an intermediate result, we present a prototypical implementation, where the knowledge retrieval process is initiated in the background on basis of an automatic identification of a selected text paragraph and its semantic similarity to existing documents in an unstructured repository. This way the system delivers valuable context specific documents just by a simple click.