The workshop on the Second HAREM, a satellite meeting of PROPOR 2008, took place at the University of Aveiro, in the afternoon of 7th September, 2008 (14h-18h30).
HAREM is an evaluation contest for named entity recognition in Portuguese, organized by Linguateca, whose first edition took place in 2005-2006, and had international participants from Spain and Mexico as well.
The call for participation for the Second HAREM (issued in September, 2007) confirmed international interest in this venue. Not only we got a considerable number of participants, but geographical spread was considerably increased.
This edition features, in addition to a similar setup as the First HAREM, two new tracks: extended classification and normalization of temporal expressions, proposed by Hagège, Baptista & Mamede, 2007 (guidelines, in Portuguese) and a new pilot track on the recognition (detection and classification) of semantic relations among named entities (guidelines, in Portuguese).
In the workshop the general methodology underlying the Second HAREM was discussed, and detailed evaluation results was presented as well. The workshop worked as a forum for discussing future initiatives on this respect.The final program, in Portuguese, can be found in the Portuguese page.