CU Record Linkage (CURL) System
PI: Toan Ong, PhD
A platform to perform record linkage operations including linkage configuration and management, data normalization, data encryption and hashing, data linkage, data de-duplication, and linked data dissemination. CURL platform supports both centralized and distributed record linkage. CURL implements modular software architecture which allows each module to be updated, added or removed without the need to change the source code of the core platform.
END TO END LINKAGE PROCESS
• Comprehensive linkage data preparations including data normalization, customization and hashing
• Customizable salts for hashing
• Honest broker-centric clear-text and privacy preserving record linkage (PPRL)
• Central control of hashing and linkage job configurations via the CURL Center
SIMPLIFIED RECORD LINKAGE
• User-friendly graphical web user interface
• Simple and flexible software deploying process via Docker or Java applications
• Work directly on data in flat file format (no direct access to backend database is needed)
CUSTOMIZABLE DATA PREPARATION AND LINKAGE METHODS
• Built-in data preparation and data linkage methods
• Modular architecture for easy method customization and extension
• Flexible linkage method configurations including mixing deterministic and probabilistic linkage methods
SECURE AND EFFICIENT LINKAGE OPERATIONS
• SHA2 and Bloom filter-based hashing methods to support deterministic and probabilistic data blocking and linkage
• Parallel processing to utilize the power of computing hardware
• Independent software components running behind secure network firewall
Want to try out CURL or need more info? Please contact curl.support@ucdenver.edu
Contact:
Toan Ong, PhD is an Assistant Professor in the Department of Pediatrics. He is a data scientist with research concentrations in data quality, data reporting, and data visualization.