https://blog.vespa.ai/approximate-nearest-neighbor-search-in-vespa-part-1/
Approximate Nearest Neighbor Search in Vespa – Part 1
Geir Storli
Principle Vespa Engineer
“Searching for the nearest neighbors of a data point in a high dimensional vector space is an important problem…. we’ll explain why we selected HNSW (Hierarchical Navigable Small World Graphs) as the baseline algorithm… Requirements included supporting real-time updates, integration with the Vespa query tree, and being flexible enough to fit a wide range of use cases.”