OpenSearch Vector Search Edition supports a variety of data query methods to accommodate different query requirements.
Hybrid query: A hybrid query is performed based on text and vectors, leveraging both structured and unstructured data to enhance the accuracy of query results.
Vectors include dense vectors and sparse vectors. The overall score of vectors is calculated based on the sum of the distance of dense vectors and the distance of sparse vectors. In the case of the Euclidean distance (SquareEuclidean), the closer the distance between vectors, the higher the similarity of the vectors.
The overall score of text is determined by the keyword matching degree. The higher the matching degree, the higher the score of the text. The final comprehensive score is the sum of the vector score and text score. A small vector score and a large text score are preferred. The larger the comprehensive score, the higher the document relevance. The OpenSearch Vector Search Edition allows you to configure weights to balance the impact of the vector score and text score. For example, you can decrease the weight of the vector score and increase the weight of text score to achieve better search effects.
Vector-based query: You can import generated vector data into a Vector Search Edition instance and perform vector-based queries.
Multi-vector-based query for a single document: You can store multiple vectors in a vector field of a document for data query.
Multi-query: You can send multiple query requests at a time. This enhances the query efficiency and response speed.
Prediction-based query: After you convert text and images into vectors by using the built-in vectorization model of Vector Search Edition, you can use text or images to perform prediction-based queries.
Primary key-based query: A document can be identified based on a unique identifier,. This ensures efficient data access.
Filter expression: You can specify filter conditions to filter documents. This improves data filtering.
Inverted index: Inverted indexes can help accelerate full-text queries, quickly identify documents, and improve query performance.