A Deep Neural Network for Chinese Zero Pronoun Resolution

This paper investigates the problem of Chinese zero pronoun resolution. Most existing approaches are based on machine learning algorithms, using hand-crafted features, which is labor-intensive. Moreover, semantic information that is essential in the resolution of noun phrases has not been addressed enough by previous approaches on zero pronoun resolution. This is because that zero pronouns have no descriptive information, which makes it almost impossible to calculate semantic similarity between the zero pronoun and its candidate antecedents. To deal with these problems, we aim at exploring learning algorithms that are capable of generating semantic representations for zero pronouns, capturing the intricate relatedness between zero pronouns and candidate antecedents, and meanwhile less dependent on extensive feature engineering. To achieve this goal, in this paper, we propose a zero pronoun-specific neural network for Chinese zero pronoun resolution task. Experimental results show that our approach significantly outperforms the state-of-theart method.