国际中文阅读文本自动简化: 局部简化、整体简化与可控简化

The task of text simplification aims to adjust the language and content of texts to make them easier to read and comprehend. Given the unique characteristics of Chinese reading materials for international readers and the growing global audience, research on the simplification of Chinese texts has become increasingly important. Recent advancements in artificial intelligence (AI) and natural language processing (NLP) have provided technical support for the automatic simplification of such texts, opening new possibilities for their pedagogical applications. Based on the varying demands and strategies of experts involved in manual text simplification, automatic simplification of Chinese texts for international readers can be categorized into three main paradigms: local adjustments, which employs linguistic rules to identify and replace difficult points; global modifications, which leverages parallel corpora to train models for content simplification; and controllable simplification, which utilizes prompting strategies to achieve personalized interaction and generation through human-computer interaction. This paper systematically elaborates on the core characteristics of these three paradigms, analyzes their respective advantages and limitations, explores the driving factors and current challenges in automatic simplification of Chinese texts for international readers, and offers perspectives on future research directions and potential pedagogical applications in this field.