Tsundoku: A Python toolkit for social network analysis

Eduardo Graells-Garrido, Nicolás García, Andrés Carvallo

Abstract: Tsundoku is a Python toolkit for analyzing social media data, focusing on text and network analysis. It offers user classification, bot detection, community identification, and topic modeling, with an active learning component to improve model accuracy. Tsundoku generates detailed reports with visualizations, making it accessible to researchers across disciplines. By streamlining the analysis pipeline from data collection to insight generation, Tsundoku helps researchers tackle the challenges of large-scale social media data analysis.