Combining Different NLP Methods for HUMINT Report Analysis.

摘要

In this paper we present a combined approach to the automatic (pre-) analysis of intelligence reports. The combination encompasses information extraction (IE) and information enrichment by means of ontologies. The combined approach proves to yield superior results compared to standalone IE. For our work we mainly use open standards and open source software. For the purpose of IE, for instance, we use the GATE system, whereas our ontology work is based on the W3C OWL standard and the Protege ontology editor.