Query rewriting

As explained in Essential GraphRAG

Query rewriting is the transformation of a user query into a formulation that better matches the language and context of target documents.

Query rewriting transforms a user query into alternative phrasings that better match the terminology and context of target documents. It addresses a common vector-retrieval failure: semantically important documents may be missed when query and corpus language differ. section 3.0

Use query rewriting to improve retrieval when:

• The query embedding does not align closely with documents containing the needed information. section 3.0
• The user’s intent is expressed differently from the language used in information-rich documents. section 3.0
• The retriever needs contextual summaries, paraphrases, or hypothetical questions alongside exact text. section 3.4

Two rewriting strategies

Hypothetical document retriever and step-back prompting reformulate the retrieval input so it better bridges user intent and document content. section 3.0 section 3.4

Figure 3.1 Query rewriting by using the step-back technique to increase the vector retrieval accuracy

Shows how step-back rewriting can increase vector-retrieval accuracy. section 3.0

How Essential GraphRAG covers it

How deep it goes

Explains the terminology gap, names major strategies, and extends matching to summaries, paraphrases, and hypothetical questions.

Examples

Uses hypothetical document retrieval, step-back prompting, contextual summaries, paraphrases, and hypothetical question embeddings.

What it assumes

Assumes text embeddings and vector similarity search, while motivating rewriting through their incomplete retrieval accuracy and recall.

What you can do after

Apply it via hypothetical document retrieval or step-back prompting to improve document matching and retrieved-answer relevance.

How it connects

Connects query rewriting with text embeddings, vector similarity search, RAG, hypothetical document retrieval, step-back prompting, and parent document retrieval.

liveBookmanning.com

Read Essential GraphRAG on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  AI Agents and ApplicationsQuery rewriting is a technique that transforms an original search request into one or more clearer or more targeted queries to improve retrieval results.
• 

  liveBookmanning.com
  AI-Powered SearchQuery rewriting is the transformation of an original search query into an alternative query that preserves or enhances its intended retrieval behavior.
• 

  liveBookmanning.com
  The Art of AI Product DevelopmentA retrieval technique that rewrites an input query into a form better suited to searching a target information collection.
• 

  liveBookmanning.com
  Grokking AI ApplicationsA technique that rephrases a user query to improve its semantic match with retrievable document content.

Compare how 5 books explain query rewriting →

Related concepts

• language modeluses · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
• step-back promptingincludes · Step-back prompting is a prompting technique that reframes a specific question as a broader, more abstract question before addressing the original question.
• HyDEincludes · HyDE is a retrieval technique that generates hypothetical document or answer content from a query and uses that content or its embedding to retrieve relevant documents.
• ruleimplemented by · A PostgreSQL rule is a query-rewrite definition attached to a database relation that modifies or adds statements when specified events occur.
• retrieval-augmented generationpart of · Retrieval-augmented generation is a technique that retrieves relevant information from external sources and supplies it as context to a generative model when producing an output.

This explainer was generated by Manning's knowledge graph from the text of Essential GraphRAG. The section links point to the parts of the book it draws on.