Working Papers
Rethinking the Income-Emissions Relationship: Evidence from Trade-Embodied CO₂ Emissions Networks (working paper, 2026)
- Developed reproducible computational workflows for integrating and processing large-scale international datasets on embodied carbon emissions, trade, and macroeconomic indicators.
- Constructed weighted global trade networks and implemented longitudinal network analyses to characterize the evolution of global embodied carbon trade networks.
- Integrated network analysis and panel econometrics to examine how economic development and network position shape domestic and trade-embodied CO₂ emissions across advanced and developing economies.
Global migration network and the geography of development (working paper, 2026)
- Created a longitudinal panel of directed, weighted bilateral migration networks spanning 23 years to examine global migration dynamics.
- Developed reproducible computational workflows for network construction, visualization, and longitudinal analysis using R.
- Contributed to network model development, core-periphery detection, interpretation of network statistics, and manuscript preparation.
Product Reproducibility and International Trade Patterns: Evidence from Cultural Goods (under review, 2026)
- Constructed computational pipelines for harmonizing, processing, and analyzing large-scale international trade data covering 190 countries.
- Combined structural gravity modelling (PPML) with dynamic network modelling (STERGM) to examine the formation and persistence of international trade relationships.
- Developed reproducible research workflows integrating data processing, statistical modelling, visualization, version control, and documentation using Python, R, and Git.
- Prepared and submitted the manuscript as the corresponding author.
Gendered Cultural Hierarchies in Social Media Audience Discourse: Evidence from Bangladeshi Artists (working paper, 2026)
- Developed scalable Python pipelines using the YouTube Data API v3 for automated collection, preprocessing, and analysis of over 8 million multilingual social media comments.
- Applied computational text mining and natural language processing techniques, including sentiment analysis (VADER) and structural topic modelling (quanteda/STM), to identify patterns in audience discourse and sentiment across artists, genders, and artistic categories.
- Developed a human-in-the-loop classification workflow for a large-scale YouTube video dataset, using the Gemini API for preliminary classification of artist gender and artistic category from video titles, followed by manual validation and correction of approximately 48,000 videos.
- Built reproducible computational pipelines integrating automated data collection, preprocessing, statistical modelling, visualization, and documentation using Python and R.