This resource is no longer available

Personalization is increasingly an expected part of the American news cycle, as more and more diverse audiences demand a curated feed in a world with more content than ever.

In order to better recommend news to their audience, the Sohu app turned to Milvus, an open-source vector database, and built a semantic vector search inside their app which allowed the company to fine tune recommendations to user profiles over time.

Read this article to learn the challenges that come with using semantic search as personalization vectors and discover how Sohu engineered an app that beat these challenges and provided real-time news recommendations.