Coreference resolution systems usually attempt to find a suitable antecedent for (almost) every noun phrase. Recent studies, however, show that many definite NPs are not anaphoric. The same claim, obviously, holds for the indefinites as well. In this study we try to learn automatically two classifications, +-discourse_new and +-unique, relevant for this problem. We use a small training corpus (MUC-7), but also acquire some data from the Internet. Combining our classifiers sequentially, we achieve 88.9% precision and 84.6% recall for discourse new entities. We expect our classifiers to provide a good prefiltering for coreference resolution systems, improving both their speed and performance.
展开▼
