Understand Basics
Explanations about the common terms we use when discussing how to use Zilliz Cloud.
AUTOINDEX Explained [READ MORE]
Zilliz Cloud offers Performance-optimized and Capacity-optimized clusters. Because of their different purposes, building indexes on these clusters requires different approaches. To save users the trouble of tuning and tweaking index parameters, AUTOINDEX comes into play.
Cluster, Collection & Entities [READ MORE]
A Zilliz Cloud cluster is a managed Milvus instance associated with certain computing resources. You can create collections in the cluster and insert entities into them. In comparison to a relational database, a collection in a cluster is similar to a table in the database, and an entity in a collection is similar to a record in the table.
Schema Explained [READ MORE]
Defining a data schema is the process of creating a clear and organized data structure for storing information in a cluster. A data schema also illustrates how the data is connected. The objective of a data schema is to recognize all the data components in a dataset, demonstrate how they are linked, and determine the most effective ways to represent these relationships.
ANN Search Explained [READ MORE]
A k-nearest neighbor (kNN) search finds the k-nearest vectors to a query vector. Specifically, it compares a query vector to every vector in a vector space until k exact matches appear. Although kNN searches guarantee perfect accuracy, they are time-consuming, especially for large datasets comprising high-dimensional vectors.
Similarity Metrics Explained [READ MORE]
Similarity metrics are used to measure similarities among vectors. Choosing an appropriate distance metric helps improve classification and clustering performance significantly.
Consistency Level [READ MORE]
In distributed databases, consistency ensures that every node or replica offers the same data view during read or write operations. Zilliz Cloud provides three consistency levels Strong, Bounded Staleness, and Eventually, with Bounded Staleness set as the default.
Data Security [READ MORE]
Data security is a crucial aspect of any cloud platform, and Zilliz Cloud is no exception. To safeguard data, Zilliz Cloud provides robust measures in various aspects, including authorization and authentication, network isolation, encryption, and backup and restoration.
Reranking [READ MORE]
Zilliz Cloud enables hybrid search capabilities using the hybridsearch() API, incorporating sophisticated reranking strategies to refine search results from multiple `AnnSearchRequest` instances. This topic covers the reranking process, explaining its significance and implementation of different reranking strategies in Zilliz Cloud.