Leveraging Big Data and Machine Learning to Identify and Forecast Factors that Influence the War in Ukraine
Abstract
Military analysts often rely on publicly available information from traditional news platforms and social media to gain insight on the development of international events of interest. Relying too heavily on English language sources can be detrimental to the quality of analysis, however, particularly if focusing on non-Western actors. Examining the ongoing war between Russia and Ukraine, it is hard to quickly aggregate publicly available information in Ukrainian, Russian, Persian, and other non-English languages. The Global Database of Events, Language, and Tone (GDELT) offers a potential solution, as it monitors the worlds news media in over 100 languages from every country in the world, with 65 languages being automatically translated into English. While the GDELT data collection is openly available, its size and complexity present a significant challenge in collecting, parsing, and scrubbing relevant data. This study aims to leverage the GDELT dataset to forecast important factors and actors influencing the war in Ukraine and will use extracted data and machine learning techniques to develop predictive models.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2023
- Accession Number
- AD1213604
Entities
People
- Benjamin Polzin
Organizations
- Naval Postgraduate School