dssg/hitchhikers-guide

What it solves

This guide provides a comprehensive framework for training data scientists to apply machine learning and AI to social impact projects. It addresses the challenge of bridging the gap between technical data skills and the practical, ethical, and collaborative requirements of working with non-profits and government agencies.

How it works

The repository acts as a central knowledge base for the Data Science for Social Good (DSSG) fellowship. It is divided into two main sections: a manual for program logistics and a curriculum covering technical and social science skills. The curriculum includes tutorials on data acquisition, machine learning, programming best practices, and ethical considerations, all designed to help teams move from vague project goals to actionable, reproducible solutions.

Who it’s for

While primarily designed for incoming DSSG fellows, the guide is open to anyone interested in learning how to perform responsible, socially impactful data science. It is particularly useful for students, researchers, or practitioners looking to understand the intersection of data analysis, public policy, and ethical project management.

Highlights

• Comprehensive curriculum covering the full data science lifecycle, from web scraping and database management to causal inference and machine learning.
• Dedicated focus on the "soft" skills required for social impact, including project scoping, communication, and team management.
• Deep integration of ethics, privacy, and fairness into technical workflows to ensure data-driven projects do not cause harm.
• Practical resources for reproducible research, including testing, version control, and coding best practices.
• A collaborative wiki containing technical tips for common infrastructure tasks like cloud computing and database migration.