Federal CIO Vivek Kundra identified ten principles for improving federal transparency in his testimony before a House Committee on Oversight and Government Reform subcommittee today. They're worth repeating here:
- Build end-to-end digital processes – Automate transfer of data between systems to increase productivity, protect data integrity, and speed data dissemination. Capitalize on game-changing technologies to increase transparency.
- Build once, use often – Architect systems for reuse and share platforms to reduce costs, streamline systems and processes, reduce errors, and foster collaboration.
- Tap into golden sources of data – Pull data directly from authoritative sources to improve data quality, shorten processes and protect data integrity.
- Release machine-readable data and encourage 3rd party applications – Make data machine-readable to allow the public to easily analyze, visualize and use government information.
- Use common data standards – Develop and use uniform, unique identifiers and data standards to ease the flow of data and reduce system complexity.
- Validate data up front – Correct errors during collection and at the point of entry to block bad data from ever entering the system.
- Release data in real time and preserve for future use – Release data as quickly as feasible to enhance its relevance and utility while maintaining future accessibility.
- Reduce burden – Collect data once and use it repeatedly. Pull from existing data sets to reduce costs and burden and to increase productivity and uniformity.
- Protect privacy and security – Safeguard the release of information to increase public trust, participation, preserve privacy, and protect national security. Open Government doesn’t mean vulnerable government.
- Provide equal access and incorporate user feedback – Provide a common view of data to all stakeholders to foster collaboration. Incorporate user feedback to help identify high-value, meaningful data sets, set priorities, to continuously drive and improve future planning and processes.