Vector Similarity Search (VSS) search allows developers to retrieve information based on audio, natural language, images, video clips, voice recordings, and many more data types. Searching over unstructured data makes VSS a foundational technology for building advanced similarity search experiences. With advances in AI, data scientists can build models that can transform almost any data “entity” into its vector representation. An entity here could be a transaction, a user profile, an image, a sound, a long piece of text (sentence or paragraph), a time series, or a graph. Any of these can be turned into its “feature vector,” also known as “embedding.” AI/ML practitioners are familiar with generating “dense” feature representations (a.k.a embeddings) for their data entities. They can now store these feature vectors in Redis and perform similarity searches. From a visual search on an e-commerce website to automated chatbots / Q&A systems and multiple types of recommendation systems. VSS is generally helpful on any app where spotting similarity in real-time is essential to unlocking value. Common applications are E-commerce recommendations, Semantic similarity, Similarity in user profiles or products, Similarity in time-series data, graph data, and transactions.
In this talk, we’ll learn how to implement VSS in Spring applications using Redis Stack enhanced search capabilities. We’ll learn about creating embeddings for your data, learn about Vector databases, Vectorization of your data, similarity metrics, and more.