Ontologies and Knowledge Graphs

Overview

Standardized vocabularies and structured data models giving bioprocessing data shared meaning across organizations, enabling AI to generate insights and identify gaps in upstream, downstream, and manufacturing contexts. Building ontologies and knowledge graphs is described as critically underestimated and as the key intermediate step between raw data and AI-generated insights — once built, large language models can identify which data are missing and what insights can be derived. Merck built bioprocessing ontologies in collaboration with external industry consortia to standardize data language across upstream, downstream, and beyond, enabling the creation of its first digital twins and large language model integrations. (Segment 6 (25:00-30:00))