Beyond information extraction: The role of ontology in military report processing

Abstract

Information extraction tools like SMES transform natural language into formal representation, e.g. into feature structures. Doing so, these tools exploit and apply linguistic knowledge about the syntactic and morphological regularities of the language used. However, these tools apply semantic as well as pragmatic knowledge only partially at best. Automatic processing of military reports has to result in a visualization of the reports content by map as well as in an actualization of the underlying database in order to allow for the actualization of the common operational picture. Normally, however, the information provided by the result of the information extraction is not explicit enough for visualization processes and database insertions. This originates from the reports themselves that are elliptical, ambiguous, and vague. In order to overcome this obstacle, the situational context and thus semantic and pragmatic aspects have to be taken into account. In the paper at hand, we present a system that uses an ontological module to integrate semantic and pragmatic knowledge. The result of the completion contains all the specifications to allow for a visualization of the report’s content on a map as well as for a database actualization.