What if War in Taiwan–Applying Machine Learning to Analyze the “Resist Will”

• Item type: Symposium(1)
• 公開日: 2023-12-02
• タイトル: What if War in Taiwan–Applying Machine Learning to Analyze the “Resist Will”
• タイトル / 言語: en
• タイトル: What if War in Taiwan–Applying Machine Learning to Analyze the “Resist Will”
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: machine learning; sentiment analysis; social media; computational social science; conflict studies; War in Taiwan
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_5794
• 資源タイプ: conference paper
• 著者所属: Department of East Asian Studies ,National Taiwan Normal University
• 著者所属(英): en
• 著者所属(英): Department of East Asian Studies ,National Taiwan Normal University
• 著者名: Shao, Hsuan-Lei
• 著者名(英): Hsuan-Lei, Shao
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: This study aims to explore the challenges of the "Taiwan issue" for Xi Jinping's third term. Besides the international structural factors of geopolitical dynamics, the 'resist will' within Taiwanese society is identified as a decisive element. Consequently, this research proposes a novel methodology and process to measure and understand the variations in Taiwan's 'resist will' across different scenarios. Traditional survey methods have limitations in capturing audience perceptions and reactions to various contexts. Therefore, this study employs text mining to analyze online forum discourse, designing scenario-specific queries to more accurately gauge the resist will. Compared to traditional approaches, this method facilitates a more comprehensive understanding of audience emotions and attitudes, achieving a standardized process. The findings indicate significant variations in the resist will across different contexts. For instance, scenarios requiring personal sacrifice tend to elicit more negative emotions from the audience, whereas those positively impacting national capabilities are likely to provoke positive emotions. By analyzing the changes in the “resist will” across various scenarios, as well as the stance and sentiment of social media posts, the study aims to delve into the factors influencing Taiwanese people's display of resistance. Specifically, the research calculates the positive and negative sentiment values of articles to determine their stance; in conjunction with 'article attention' and 'comment support' metrics, it analyzes social media responses to further understand reader emotions and positions. This approach not only breaks through the constraints of traditional survey methods but also achieves real-time responsiveness and systematic data processing. Moreover, the flexibility of the research method makes it applicable to various contexts and needs, offering insights for related studies and policy formulation.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: This study aims to explore the challenges of the "Taiwan issue" for Xi Jinping's third term. Besides the international structural factors of geopolitical dynamics, the 'resist will' within Taiwanese society is identified as a decisive element. Consequently, this research proposes a novel methodology and process to measure and understand the variations in Taiwan's 'resist will' across different scenarios. Traditional survey methods have limitations in capturing audience perceptions and reactions to various contexts. Therefore, this study employs text mining to analyze online forum discourse, designing scenario-specific queries to more accurately gauge the resist will. Compared to traditional approaches, this method facilitates a more comprehensive understanding of audience emotions and attitudes, achieving a standardized process. The findings indicate significant variations in the resist will across different contexts. For instance, scenarios requiring personal sacrifice tend to elicit more negative emotions from the audience, whereas those positively impacting national capabilities are likely to provoke positive emotions. By analyzing the changes in the “resist will” across various scenarios, as well as the stance and sentiment of social media posts, the study aims to delve into the factors influencing Taiwanese people's display of resistance. Specifically, the research calculates the positive and negative sentiment values of articles to determine their stance; in conjunction with 'article attention' and 'comment support' metrics, it analyzes social media responses to further understand reader emotions and positions. This approach not only breaks through the constraints of traditional survey methods but also achieves real-time responsiveness and systematic data processing. Moreover, the flexibility of the research method makes it applicable to various contexts and needs, offering insights for related studies and policy formulation.
• 書誌情報: じんもんこん2023論文集; 巻 2023, p. 31-36, 発行日 2023-12-02
• 出版者 / 言語: ja
• 出版者: 情報処理学会