The increasing openness of data, methods, and collaboration networks has
created new opportunities for research, citizen science, and industry.
Whereas openly licensed scientific, governmental, and institutional data
sets can now be accessed through programmatic interfaces, compressed
archives, and downloadable spreadsheets, realizing the full potential of
open data streams depends critically on the availability of targeted
data analytical methods, and on user communities that can derive value
from these digital resources. Interoperable software libraries have
become a central element in modern statistical data analysis, bridging
the gap between theory and practice, while open developer communities
have emerged as a powerful driver of research software development.
Drawing insights from a decade of community engagement, I propose the
concept of open data science, which refers to the new forms of research enabled by open data, open methods, and open collaboration.