Non-profit and public-sector organizations acquire, store, and use data
in ever-increasing volumes and varieties, but
they typically lack the in-house technical expertise needed to
develop improved ways to use their existing data holdings, enhance
their data collection and analysis methods, or create entirely new
applications of their data.

CDS is leading efforts to provide education and research pathways for
aspiring data scientists to apply their knowledge and skills to benefit
society. Students do not have many opportunities to be exposed to the
vital and compelling work of community non-profits and public-sector
agencies, yet are increasingly motivated by the desire to direct their
talents towards making the world a better place.

DS4CG harnesses growing student interest in social-good
causes and connects it to partner organizations that stand to benefit
from the students’ growing data science expertise.

DS4CG is a summer program conducted over 12 weeks from May to August.
Participating students, called DS4CG fellows, perform their work and are
supervised by CDS faculty and staff in paid internships comparable to
private-sector industry internships. No fee is charged to the
partner organizations; however, to fully realize the benefits of
participation, partners are expected to allocate staff time to engage
regularly with our student teams. In addition, fellows are supported by
volunteer professional data science advisors drawn from CDS’s pool of
corporate affiliates.

DS4CG serves organizations that lack staff capacity to exploit
state-of-the-art analysis and modeling of their business data. Ideal
DS4CG projects incorporate analysis and integration of very large
volumes of data, with different types and formats. DS4CG
fellows compile, organize and clean these datasets,
mathematically explore their statistical properties, build predictive
models and forecasts, and visualize the results. Project deliverables
can include actionable data-driven insights, presentations, reports, and
proof-of-concept software tools.