Scientific progress today requires multi-institutional and cross-disciplinary sharing and analysis of data. Many disciplines, such as social and health-related sciences, face a web of policies and technological constraints on data due to privacy concern over, for example, Personal Health Information (PHI) or Personally Identifiable Information (PII). Issues of privacy, safety, competition, and ownership have led to regulations controlling data location, availability, movement, and access. Compliance poses obstacles to traditional data-processing practices and slows research; yet, increasingly, pressing scientific and societal problems demand collaborative efforts involving data from multiple stakeholders.
ImPACT (Infrastructure for Privacy-Assured CompuTations) will free researchers to focus more fully on science by supporting the analysis of multi-institutional data while satisfying relevant regulations and interests. It is designed to facilitate secure cooperative analysis, meeting a pressing need in the research community. ImPACT seeks to develop an integrative model for management of trust, deploying a collection of supportive mechanisms at scale into a model cyber-infrastructure. The project will develop methodologies with best practices in networking, data management, security, and privacy preservation to fit a variety of use cases.