A Study on the Distribution Differences of Sentence Group Coherence in Chinese Texts of Different Genres

摘要

Chinese sentence group plays an important role in text coherence analysis. Because of the complexity and the diversity of Chinese language, the sentence groups of different genres often show different coherence distribution characteristics. This paper analyzed the coherence of four different corpora in news, application, prose and encyclopedia based on the different statistical features of two independent annotators. In this work, the coherence distribution characteristics of sentence groups in four corpora are analyzed, and the differences of sentence group coherence in different genres are compared in detail. The study lays a good foundation for the automatic segmentation of the boundary of sentence group and automatic analysis of the relation between sentences in the future.