Posted On: Mar 24, 2020
Amazon Elasticsearch Service now offers k-Nearest Neighbor (k-NN) search which can enhance search by similarity use cases like product recommendations, fraud detection, and image, video and semantic document retrieval. Built using the lightweight and efficient Non-Metric Space Library (NMSLIB), k-NN enables high scale, low latency nearest neighbor search on billions of documents across thousands of dimensions with the same ease as running any regular Elasticsearch query.
Given a space of data points, the k-NN plugin finds the number (k) of data points at closest distance to a query data point. A new field type for k-NN, enables you to seamlessly integrate k-NN search with Elasticsearch’s extensive features such as aggregations and filtering to further improve the precision of the search results. Elasticsearch’s distributed architecture enables the k-NN plugin to ingest and process large datasets, support incremental updates, thereby delivering you a highly performant similarity search engine with fast inference.
k-NN similarity search is powered by Open Distro for Elasticsearch. To learn more about Open Distro for Elasticsearch visit the website.
k-NN similarity search is available on domains running Elasticsearch 7.1 and higher. To learn more, see the documentation.