文本数据建模的政治语言学应用研究——以日本安全和经济政策偏好分析为例

Political Linguistic Research Based on Text Data Modeling——Take Japan's Security and Economy Policy Preference as an Example

随着大数据时代的到来,基于文本数据挖掘的政治语言学研究越来越受到关注。本文采用日本首相国会演讲的关键词数据进行相关性分析,发现日本的安全政策调整会提前反映在首相演讲的意识形态变化之中,且二战后至今日本的政策偏好结构具有在经济手段和军事手段之间权衡的特征。基于文本数据建立的回归模型表明,日美同盟框架下的贸易自由化是日本经济的主要增长点,安全防卫等级提升则阻碍总体经济增长。深入分析日本政策偏好结构有利于对日本的安全政策动向做出预判。

Big data era arouses more attention on political linguistic research based on text data digging. Taking the data from the Japanese Prime Minister’s speech as analyzing object, the correlation analysis shows that Japan’s security policy adjustments will be reflected in the ideological changes of the prime minister’s speech in advance. Since World War II, Japan’s policy preference structure has an obvious feature that Governments often transform their economic and military instruments. The regression model based on text data shows that the trade liberalization under the framework of the Japan-US alliance is the main growth point of the Japanese economy, and the promotion of the security and defense level hinders the macroeconomic growth. Further analysis of the Japanese policy preference structure is conducive to make a pre judgment on Japan’s security policy trends.