In a recent acquisition WebMD Ignite inherited a large patient education content collection. A unified content corpus that could be delivered across all systems on its platform.
Combination and deduplication were required to support multiple use cases.
To support this primary business need, methods to speed up annotation of all content were developed.
Outcomes achieved:
- A consumer knowledge graph with an associated taxonomy that is a good fit for describing all content in two collections.
- An optimized semantic layer to feed downstream applications and support a unified relevance-based search experience across the platform
This session is an exploration of how LLMs with consumer-based semantic layer were used for content integration.
