Research Assistant, Environmental Research Center, Duke Kunshan University (Aug 2024 - May 2025)
Built an NLP-based computational framework in Python to quantify psychological antecedents of collective action in environmental movements. Processed and analyzed social-media text to study how group identity and emotional contagion evolve over time. Applied topic modeling and sentiment scoring to map environmental grievances to mobilization signals and interpret implications for Just Transition policy work. • Ingested and analyzed 3,000 social media posts for longitudinal identity/emotion tracking • Used NLP Transformers and GPT-API to support feature extraction and modeling workflows • Applied LDA topic modeling to identify themes related to grievances (e.g., critical mineral impacts) • Used sentiment scoring to connect online discourse patterns with online/offline mobilization