When Text Mining Meets WTO's Trade Policy Review: The Cases of South Korea, Japan, and China

Among the key institutional pillars of the World Trade Organization (WTO), the Dispute Settlement Mechanism (DSM) has attracted the lion's share of scholarly attention in contemporary research relating to the trade organization. Yet this study notes that the other institutional ballast of the WTO, namely the Trade Policy Review Mechanism (TPRM), has remained the least studied, despite its significance as a safeguard against protectionism. The main reason for the lack of attention by trade scholars is the semantic complexity of review reports. In an effort to navigate through the complex, subtle, and often ambivalent text, this study uses information extraction (IE) techniques for both a horizontal and vertical analysis of TPR reports. A total of 18 TPR reports on the three representative trading states in East Asia—South Korea, China, and Japan—have been analyzed by the Rapid Automation Keyword Extraction (RAKE) and TextRank algorithms. The findings of this study show important policy patterns and their changes.