Research And Realize On Pronominal Anaphora Resolution System In Chinese Text

Anaphora is a common phenomenon in the research on NLP (Natural Language Processing), it appears a lot in the discourses or the dialogues. The use of anaphoric words makes discourse looks brief. Anaphora resolution plays an important role in text information processing. With the increasing development of dealing with the discourses, anaphora resolution shows the unprecedented importance, and becomes a hot research on processing the information of text. It is very important in Machine Translating, Automatic Summarization, Question Answer, Information Extraction and other natural language processing area, and it becomes an important evaluating task of MUC and ACE.In this paper, based on the in-depth analysis of anaphoric features of pronoun in the paroxysmal Chinese texts, we present an approach of anaphora resolution, which is based on corpus adopting the statistical machine learning arithmetic and combining with the preference selection strategy. The method takes into account all kinds of anaphoric features, and uses the decision tree arithmetic to construct the filter. It is a tool reducing the noise of the system, which can decrease the number of waiting resolution words. The preference selection strategy can resolve other anaphoric phenomena, which cannot be resolved by the previous method. These two methods cooperate well.The features of this model are shown as follow:(1) The model of machine automatic learning. This method is an anaphora resolution system trained automatically in large-scale corpus. It only needs few interventions of people. All of the related features of antecedent can be gained in training.(2) Decreasing the noise of non-coreference words. This method uses the decision tree arithmetic to construct the filter, which reduces the noise of the system, eliminating many non-coreference words and increases the...