In this paper, we describe K-Extractor, a powerful NLP framework that provides integrated and seamless access to structured and unstructured information with minimal effort. The K-Extractor converts natural language documents into a rich set of semantic triples that, not only, can be stored within an RDF semantic index, but also, can be queried using natural language questions, thus eliminating the need to manually formulate SPARQL queries. The K-Extractor greatly outperforms a free text search index-based question answering system.