Research Data that Can Be Trusted
Published in SpringerBriefs in Computer Science, Springer, 2026
About the book
The book argues for a new approach to data provenance and explains how recent advances in data-processing workflow automation make it possible. It introduces descriptive dataflow operators, an approach that integrates descriptive workflow languages with a data-modeling domain-specific language (DSL), and reviews current workflow automation in that light.
Modern healthcare relies on complex data pipelines over diagnostic information, clinical records, and research data. That complexity, together with AI and ML applications and stricter regulatory oversight, demands disciplined data preparation, documentation, and validation. The book presents the Dorieh Data Platform as an exemplar implementation of such a DSL: a framework for reproducible research whose infrastructure supports data-lineage documentation, validation, and error logging, so that data processes are transparent, auditable, and conformant.
The framework is applied to healthcare claims data quality, revealing inconsistencies and deficiencies, and the book makes the case that precise, reproducible data-transformation methods are a precondition for reliable research outcomes.
Co-authored with colleagues from the Harvard Research Computing community. Also available on Amazon.
Recommended citation: Bouzinier, M., Etin, D., Khoshnevis, N., Shad, M., & Yockel, S. (2026). Research Data that Can Be Trusted. SpringerBriefs in Computer Science. Springer. https://link.springer.com/book/9783032210319
Download Paper