Highly Multilingual Coreference Resolution Exploiting a Mature Entity Repository

In this paper we present an approach to large-scale coreference resolution for an ample set of human languages, with a particular emphasis on time performance and precision. One of the distinctive features of our approach is the use of a mature multilingual named entity repository (persons and organizations) gradually compiled over the past few years. Our experiments show promising results – an overall precision of 94% tested on seven different languages. We also present an extrinsic evaluation on seven languages in the context of summarization where we gauge the contribution of the coreference resolver towards the end summarization performance.JRC.G.2-Global security and crisis managemen