Multi-query retrieval for better recall in LangChain

Introduction

Multi-query retrieval helps find better answers by asking several questions instead of just one. It improves how well you get the right information.

When you want to get more complete answers from a large set of documents.

When a single question might miss important details in the data.

When you want to combine results from different angles or keywords.

When you need to improve recall in search systems or chatbots.

When working with complex topics that need multiple queries to cover fully.

Syntax

LangChain

from langchain.vectorstores import FAISS
from langchain.chains import MultiQueryRetriever

# Create a vector store (example with FAISS)
vector_store = FAISS.load_local("path_to_index", embedding_function)

# Create a MultiQueryRetriever
multi_query_retriever = MultiQueryRetriever(
    retriever=vector_store.as_retriever(),
    queries=["query 1", "query 2", "query 3"]
)

# Retrieve documents
results = multi_query_retriever.get_relevant_documents("main question")

The MultiQueryRetriever takes multiple queries to improve recall.

You need a vector store retriever like FAISS or others to use it.

Examples

Using two queries to find documents about climate change and global warming for the main question.

LangChain

multi_query_retriever = MultiQueryRetriever(
    retriever=vector_store.as_retriever(),
    queries=["climate change", "global warming"]
)
results = multi_query_retriever.get_relevant_documents("effects on environment")

If no extra queries are given, it works like a normal single query retriever.

LangChain

multi_query_retriever = MultiQueryRetriever(
    retriever=vector_store.as_retriever(),
    queries=[]
)
results = multi_query_retriever.get_relevant_documents("renewable energy")

Multiple related queries help find more documents about ethical AI topics.

LangChain

multi_query_retriever = MultiQueryRetriever(
    retriever=vector_store.as_retriever(),
    queries=["AI ethics", "machine learning fairness", "bias in AI"]
)
results = multi_query_retriever.get_relevant_documents("ethical AI development")

Sample Program

This program loads a saved FAISS vector store, sets up a multi-query retriever with three related queries about AI history, and retrieves documents for a main question. It then prints the titles of the found documents.

LangChain

from langchain.vectorstores import FAISS
from langchain.embeddings import OpenAIEmbeddings
from langchain.chains import MultiQueryRetriever

# Setup embeddings and vector store
embedding_function = OpenAIEmbeddings()

# Assume we have a FAISS index saved locally
vector_store = FAISS.load_local("./faiss_index", embedding_function)

# Create MultiQueryRetriever with multiple queries
multi_query_retriever = MultiQueryRetriever(
    retriever=vector_store.as_retriever(),
    queries=["history of AI", "AI milestones", "AI breakthroughs"]
)

# Retrieve documents for a main question
documents = multi_query_retriever.get_relevant_documents("important events in AI development")

# Print titles of retrieved documents
print("Retrieved document titles:")
for doc in documents:
    print(doc.metadata.get('title', 'No Title'))

Important Notes

Multi-query retrieval improves recall by combining results from several queries.

Time complexity depends on the number of queries and size of the vector store.

Common mistake: Using unrelated queries can reduce result quality.

Use multi-query retrieval when you want broader coverage than a single query.

Summary

Multi-query retrieval asks several related questions to get better search results.

It works by combining results from multiple queries in a vector store retriever.

This method helps find more complete and relevant documents for complex topics.

Practice

1. What is the main purpose of multi-query retrieval in langchain?

easy

A. To store data in multiple vector stores

B. To reduce the number of queries sent to the database

C. To speed up the response time by using a single query

D. To ask several related questions to get better search results

Solution

1. Step 1: Understand the concept of multi-query retrieval

   Multi-query retrieval involves asking multiple related questions instead of just one to improve the quality of search results.

2. Step 2: Identify the main goal of this approach

   The goal is to get better and more complete search results by combining answers from several queries.

3. Final Answer:

   To ask several related questions to get better search results -> Option D

4. Quick Check:

   Multi-query retrieval = multiple related questions for better results [OK]

Hint: Think: multiple questions improve search quality [OK]

Common Mistakes:

• Confusing multi-query with reducing query count
• Thinking it stores data in multiple places
• Assuming it speeds up by using fewer queries

2. Which of the following is the correct way to create a multi-query retriever in langchain?

easy

A. MultiQueryRetriever(vectorstore, llm=llm)

B. MultiQueryRetriever(queries=["query1", "query2"])

C. MultiQueryRetriever.from_llm(vectorstore.as_retriever(), llm)

D. MultiQueryRetriever.from_llm(vectorstore, llm)

Solution

1. Step 1: Check the required parameters for MultiQueryRetriever

   The MultiQueryRetriever is created using from_llm which requires a BaseRetriever (from vectorstore.as_retriever()) and an llm.

2. Step 2: Identify the correct syntax

   MultiQueryRetriever.from_llm(vectorstore.as_retriever(), llm) correctly uses the classmethod with proper arguments.

3. Final Answer:

   MultiQueryRetriever.from_llm(vectorstore.as_retriever(), llm) -> Option C

4. Quick Check:

   Correct syntax includes from_llm, .as_retriever(), and llm [OK]

Hint: Look for from_llm(vectorstore.as_retriever(), llm) syntax [OK]

Common Mistakes:

• Passing queries as a list to constructor
• Using direct constructor without llm_chain
• Forgetting .as_retriever() on vectorstore

3. Given the following code snippet, what will be the output type of results?

retriever = MultiQueryRetriever.from_llm(vectorstore.as_retriever(), llm)
results = retriever.get_relevant_documents("AI")

medium

A. A single document matching the first query only

B. A list of documents combining results from both queries

C. A dictionary with queries as keys and documents as values

D. An error because get_relevant_documents requires no arguments

Solution

1. Step 1: Understand MultiQueryRetriever behavior

   It [uses the LLM to] runs multiple queries and combines their results into a single list of documents.

2. Step 2: Analyze the method call

   Calling get_relevant_documents("AI") triggers retrieval for all [generated] queries, returning combined documents.

3. Final Answer:

   A list of documents combining results from both queries -> Option B

4. Quick Check:

   Multi-query retrieval returns combined document list [OK]

Hint: MultiQueryRetriever returns combined list of documents [OK]

Common Mistakes:

• Expecting a dictionary instead of a list
• Thinking it returns only first query results
• Assuming get_relevant_documents takes no arguments

4. Identify the error in this code snippet using multi-query retrieval:

retriever = MultiQueryRetriever.from_llm(vectorstore, llm)
results = retriever.get_relevant_documents("AI")

medium

A. vectorstore.as_retriever() is missing

B. get_relevant_documents requires a query argument

C. vectorstore is not defined

D. MultiQueryRetriever does not exist in langchain

Solution

1. Step 1: Check the constructor parameters

   from_llm expects a BaseRetriever as the first argument, but vectorstore is a VectorStore; use vectorstore.as_retriever().

2. Step 2: Verify other parts of the code

   get_relevant_documents has argument; llm assumed defined; class exists.

3. Final Answer:

   vectorstore.as_retriever() is missing -> Option A

4. Quick Check:

   from_llm requires BaseRetriever [OK]

Hint: from_llm needs vectorstore.as_retriever() [OK]

Common Mistakes:

• Passing vectorstore directly instead of retriever
• Thinking get_relevant_documents needs no arg when it does
• Assuming vectorstore undefined without context

5. You want to improve recall for a complex topic by using multi-query retrieval with three related queries. Which approach best ensures you get the most relevant documents without duplicates?

hard

A. Use MultiQueryRetriever.from_llm(retriever, llm) to generate multiple queries and deduplicate results

B. Run each query separately and pick the first result from each

C. Combine all queries into one long string and run a single query

D. Use only the query with the highest expected relevance

Solution

1. Step 1: Understand multi-query retrieval goal

   It aims to combine results from multiple related queries to improve recall.

2. Step 2: Identify best practice to avoid duplicates

   Using MultiQueryRetriever.from_llm(retriever, llm) generates multiple queries via LLM and deduplicates results for completeness without repeats.

3. Final Answer:

   Use MultiQueryRetriever.from_llm(retriever, llm) to generate multiple queries and deduplicate results -> Option A

4. Quick Check:

   MultiQueryRetriever generates + deduplicates = best recall [OK]

Hint: MultiQueryRetriever generates queries then deduplicates for best recall [OK]

Common Mistakes:

• Running queries separately without combining results
• Merging queries into one string loses query specificity
• Using only one query misses relevant documents