Research on Automatic Metonymy Recognition in the Japanese Economic Corpus of China
Zongnan Liu, Le Chang · 2025
Corpus automatic recognition is an important topic in artificial intelligence language research. Compared to metaphor automatic recognition, research on metonymy in Japanese corpora, is especially insufficient. This study takes Japan’s economic discourse towards China as research object, combines cognitive linguistics conceptual metonymy theory and corpus linguistics methods to select a self-built closed corpus of Japan’s economic discourse towards China from 2013 to 2023, and develops a complete and effective annotation system to automatically identify metonymies in the corpus. Machine learning algorithms are used to build recognition models for experimentation. The results show that metonymy is widely used in Japanese economic discourse, and the types are diverse. The corpus can serve as a fundamental database for Japanese language research, providing strong support for automatic recognition of metonymies and other concepts in the future.