Scholarlink

<p>This study introduces the Geopolitics–Media–Sentiment (GMS) framework to deconstruct the international image of the Hainan Free Trade Port (HFTP). Using a multilingual corpus extracted from LexisNexis (January 2018 to December 2024), this study implemented a rigorous multistage pipeline – from an initial collection of 91,196 raw records to a refined analytical dataset of 23,723 validated news reports. By deploying language-specific transformer models (RoBERTa, BETO, CamemBERT, and German-BERT) alongside BERTopic modelling, this study reveals a profound macro–micro narrative asymmetry. While macro-level event signals indicate policy-driven cooperation, micro-level textual analysis reveals a systematic ‘refraction’ into cautious and critical sentiment. Distinct linguistic framings have emerged: English media navigate a binary of opportunity versus institutional risk, German reporting maintains high technical neutrality, Spanish outlets focus on social-benefit narratives, and French media polarise around geopolitical themes. This disconnect demonstrates how geopolitical structures are re-coded into affect-laden narratives through professional media practices, offering a robust framework for understanding place branding in a multipolar information era.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Global media representations of China’s Hainan Free Trade Port (2018–2024): a cross-lingual computational study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月23日

查看全文
  • Jumabayi Ayixiren,
  • Xiu Quan He,
  • Qiqi Zhang

摘要

This study introduces the Geopolitics–Media–Sentiment (GMS) framework to deconstruct the international image of the Hainan Free Trade Port (HFTP). Using a multilingual corpus extracted from LexisNexis (January 2018 to December 2024), this study implemented a rigorous multistage pipeline – from an initial collection of 91,196 raw records to a refined analytical dataset of 23,723 validated news reports. By deploying language-specific transformer models (RoBERTa, BETO, CamemBERT, and German-BERT) alongside BERTopic modelling, this study reveals a profound macro–micro narrative asymmetry. While macro-level event signals indicate policy-driven cooperation, micro-level textual analysis reveals a systematic ‘refraction’ into cautious and critical sentiment. Distinct linguistic framings have emerged: English media navigate a binary of opportunity versus institutional risk, German reporting maintains high technical neutrality, Spanish outlets focus on social-benefit narratives, and French media polarise around geopolitical themes. This disconnect demonstrates how geopolitical structures are re-coded into affect-laden narratives through professional media practices, offering a robust framework for understanding place branding in a multipolar information era.