English Noun Phrase Coreference Resolution via a Maximum Entropy Model

In this paper, a novel corpus based learning approach to noun phrase coreference resolution is presented This approach aims to solve not only pronoun anaphora problem, but also a more general noun phrase coreference one, which is introduced by MUC By applying the maximum entropy （M E ） model and utilizing a flexible object based architecture, the system is able to make use of a range of knowledge sources in training the classifier and achieves an F measure of 60 2%, which is very close to the state of art result （61 8%）, on the MUC 7 coreference resolution task corpus