Package Ecosystem
| Capture | Store | Analyze |
|---|---|---|
| provR | prov-cpl | Rclean |
| CamFlow | Dataverse | encapsulator |
| provViz |
- Science is a process that is works better when its shared
- Software should not be a barrier to data science
- Computation should be leveraged to make sharing easier
- Tools are best when they have clear utility and simple API
Purpose
ProvTools is creating an open process scientific culture through tools that capture, store and analyze data provenance.
Big Goal
Broad incorporation of provenande collection and utilization across many scientific disciplines. Fully transparent, end-to-end publication of scientific research. 50% reduction in the rate of irreproducible computational studies and code-rot.
Vision
The crucial challenge at the frontier of data science is now the integration of information and the develompment of synthetic knowledge. The scientific landscape is rapidly developing software for data collection and analysis. Software creation is no longer just in the hands of engineers. Tools to aid data science best practices is a limiting factor to a majority of scientists, especially those with no formal data science or software engineering training. In the future, scientific data will be fully shared. Code for data processing and analysis will also be freely accessible. Data provenance will facilitate the legibility across studies and disciplines in the distributed information processing network of formal and informal investigators. Data and code will not only be more available they will also be clearer because production of well curated information about the processes behind them will be aided by automatic meta-data capture tools. The gains made from this increased efficiency will enable benefaction across socio-econimic and cultural barriers, inreasing the the diversity of the community of participants and stakeholders.