It's no secret that the amount of open-source data has skyrocketed over the past two decades. Methods to parse through that data have improved, and the creation of events datasets such as GDELT and ACLED showcase their utility. Open-source intelligence is a key component of decisionmaking today.
Recognizing this value, the Scientific Collection of Open-Source Policy Evidence (SCOPE) is a data-driven initiative which aims to boost transparency and awareness of development, security, and diplomatic activities around the world. How? SCOPE seeks to build on existing work by offering the capability to construct events datasets which simultaneously benefit from the high-speed collecting capabilities of machine learning as well as the discerning eye of trained researchers. We are designing a suite of free-to-use, data-driven methods which we believe will go beyond our homefield of international relations and contribute to the collection of massive text-based datasets in a variety of other fields. To do so, SCOPE leverages technologies in natural language processing, GIS, and machine learning to triangulate and parse information from across the web.
In the short term, you can expect to see a breadth of content on international relations topics and information extraction methodologies as well as the original datasets we are constructing. In the long term, SCOPE will house a variety of projects linked by their use of our interchangeable human and AI-based workflows. Stay tuned for our official launch in 2021.