Leveraging our existing expertise in open-source software research, we plan to better understand Ethereum’s open-source ecosystem based on data-driven analysis. Specifically, we will formally test multiple hypotheses inspired by studies on the history of other open-source software, reveal patterns that are relevant for Ethereum, and hopefully provide guidance for Ethereum ecosystem’s long-term development. Check the latest updates on SSRN.

Our web platform provides visualizations for understanding Ethereum open-source activity dynamics. Navigate through interactive representations, examining real-time trends, contributions, and collaborations. Easily utilize data-driven insights for academic exploration, engaging in a neutral visual discourse on the dynamics of open-source endeavors.

We use Gen AI text analytics approach to better understand of GitHub repositories content. We use a generative pre-trained trasformer (GPT) model from OpenAI, i.e., GPT-3/3.5/4 for topic modeling and classification. The results are superior to traditional topic modeling tools. 

We develop an agentic AI approach to classify the funding model of OSS projects using ChatGPT, information from their GitHub accounts, and broader web data. Our goal is to prompt the agent to learn to identify the funding model with high accuracy while optimizing the learning process. We have surpassed the 51% accuracy level on sample data and continue improving the learning strategy.