Aligning IATE Criminal Terminology to SUMO

Abstract

In this paper we apply an ontology matching system in the context of an information extraction project regarding criminal data. Our data comes from social network, in order to make better sense and analysis of the information provided we consider the IATE (InterActive Terminology for Europe) regarding its crime related sub-domain. The alignment of this terminology to an ontology is a further step towards enriching the semantics of the data. We evaluate a recently proposed domain top ontology matcher (based on Wordnet-SUMO previous alignment) to the task of aligning this IATE sub-domain to SUMO, a general purpose top ontology. Another aspect to explore is the use of multi-linguality in the disambiguation problem itself, as in the case of “alvo” (aim as target and not goal) which becomes more clear in the Portuguese than the English version.