Definition
The process of using an LLM to transform a raw user input into one or more optimized queries—through expansion, decomposition, or rephrasing—to bridge the semantic gap between a user's natural phrasing and the underlying indexed knowledge in a RAG pipeline.
Not traditional SEO keyword optimization; it is a pre-retrieval step used to improve semantic alignment for vector or hybrid search.
"A research assistant who takes a vague question and re-writes it into five specific library search terms to ensure all relevant books are found."
Conceptual Overview
The process of using an LLM to transform a raw user input into one or more optimized queries—through expansion, decomposition, or rephrasing—to bridge the semantic gap between a user's natural phrasing and the underlying indexed knowledge in a RAG pipeline.
Disambiguation
Not traditional SEO keyword optimization; it is a pre-retrieval step used to improve semantic alignment for vector or hybrid search.
Visual Analog
A research assistant who takes a vague question and re-writes it into five specific library search terms to ensure all relevant books are found.