ASEAN launched negotiations for the Digital Economic Framework Agreement (DEFA) in 2023 with the goal of concluding negotiations by 2025. The region seeks to establish “a coherent, harmonized, collaborative, and rules-based approach to ASEAN’s cooperation in the digital ecosystem” through DEFA. To meet this stated objective, DEFA would have to contend with the reality that ASEAN and ASEAN member states are already parties to a growing body of other Digital Economy Agreements and digitally related trade agreements.
To help disentangle this “spaghetti-bowl” typical of the ASEAN policy environment, we develop a machine-readable text database of digitally-related trade provisions in trade agreements. This session will address questions such as how does one even begin creating such a database? What are the challenges? How can it be used for policy recommendations?
By applying machine learning on the textual data, we are able to discover meaningful patterns underlying the unstructured data with the advantage of computational superiority to answer these questions. The session will discuss how text-as-data techniques like topic modelling and textual distance measurement are used to identify thematic variations as well as similarities and dissimilarities between digital trade provisions.