Fuzzy ontologies for multilingual document exploitation

Multilingual document exploitation (MDE) involves assessing the relevance of individual foreign language documents to the course of a military mission. The current approach to relevance assessment (RA) in FALCon, an MDE system, runs a machine translation (MT) program to convert the documents into English and then provides a simple keyword search with a frequency count of the matched keywords. This paper explores the potential that fuzzy mathematics and ontologies have for improving performance in MDE. Research on information retrieval and filtering is examined and fuzzy extensions to these applications are presented for inclusion in RAVEN, an alternate MDE system design to FALCon.