Submitted by raquel on Mon, 07/26/2010 - 19:27
The UMIREC corpus, ellaborated in collaboration with the Genesis research group at MIT, is publicly available. This corpus comprises 62 files in English (30 folktales and 32 Wall Street Journal articles) annotated with the location of referring expressions, coreference relations between the referring expressions, and the descriptive and distinctive information associated to the referring expressions.
The corpus can be downloaded in http://dspace.mit.edu/handle/1721.1/54766.