This topic describes how to build vector search applications based on OceanBase Database.
Vector search practice topics
Build a Q&A robot with OceanBase Database
The Q&A robot is an application based on the OceanBase database that stores documents in bulk as vectors. Users can ask questions through the UI interface, and the Q&A robot will embed the question into a vector using the BGE-M3 model, then retrieve similar vectors from the database. Once the corresponding document content for the similar vectors is found, the application sends them along with the user's question to the LLM. The LLM then generates more accurate answers based on the provided documents.
Build an image search application with OceanBase Database
The image Search application is another solution built on the OceanBase database, designed to store image libraries as vectors. Users can upload images through the UI interface to perform a search. The application converts the uploaded image into a vector and retrieves similar vectors from the database. Finally, it displays the matching images on the UI in a visually intuitive way.
