Character Interaction Network Analysis of Chinese Literary Work-A Preliminary Study

摘要

Considering the shortages of character interaction network analysis from literary work by literary theorists and computer scientists, this paper combined their methods and conducted qualitative and quantitative approach to extract and analyze character interaction network from literary text. Firstly, it proposed the analysis framework of character interaction network from literary work: one-mode and two-mode network analysis. One-mode network analysis was applied in macro-level (chapters), meso-level (social events) and micro-level (character dialogues). Two-mode network analysis was applied in meso-level (social events). Then two literary scholars manually extracted social events and character dialogues from the fiction “Journey to the West Prequel” and manually recognized characters in different levels. Then it constructed the one-mode and two-mode matrices based on the proposed standard formula which was based on the idea of co-occurrence. Different matrices were computed and visualized by SNA software. The results validated by literary scholars showed that the proposed analysis framework was applicable in interpreting and revealing the contents of the fiction. Finally, the discussions and further researches were summarized.