Concept Rollups for OMOP Data

Phenotyping in OMOP data means working with tens of thousands of granular SNOMED CT concepts, most of them too sparse to model directly. This project developed a method that aggregates these concepts into interpretable groups using TF-IDF-style weighting, reducing dimensionality while preserving clinical interpretability. I presented this work with Pfaff and NC TraCS colleagues at the 2025 AMIA Informatics Summit: “From Complex to Comprehensible: A TF-IDF Approach for Hierarchical Aggregation of Clinical Conditions in SNOMED CT.”