Originally created by @Orangerccc on GitHub (Jan 15, 2025).
Can a multi-retrieval strategy be incorporated into RAG? For different types of texts, data retrieval and recall should be conducted on multiple databases because different formats of data are suitable for different database scenarios. For instance, structured data is suitable for relational databases, while unstructured data is suitable for vector databases. For the same question, multi-retrieval should be carried out on relational databases, full-text search engines, vector databases, knowledge graphs, etc., to make the RAG function more powerful.
Originally created by @Orangerccc on GitHub (Jan 15, 2025).
Can a multi-retrieval strategy be incorporated into RAG? For different types of texts, data retrieval and recall should be conducted on multiple databases because different formats of data are suitable for different database scenarios. For instance, structured data is suitable for relational databases, while unstructured data is suitable for vector databases. For the same question, multi-retrieval should be carried out on relational databases, full-text search engines, vector databases, knowledge graphs, etc., to make the RAG function more powerful.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @Orangerccc on GitHub (Jan 15, 2025).
Can a multi-retrieval strategy be incorporated into RAG? For different types of texts, data retrieval and recall should be conducted on multiple databases because different formats of data are suitable for different database scenarios. For instance, structured data is suitable for relational databases, while unstructured data is suitable for vector databases. For the same question, multi-retrieval should be carried out on relational databases, full-text search engines, vector databases, knowledge graphs, etc., to make the RAG function more powerful.