How to beginner · 3 min read

Search returning irrelevant results fix

Q: Search returning irrelevant results fix

To fix search returning irrelevant results, use embedding-based semantic search with a high-quality vector store and ensure your query and documents are properly preprocessed. Use OpenAI embeddings like text-embedding-3-small and tune retrieval parameters such as top_k to improve relevance.

Quick answer

To fix search returning irrelevant results, use embedding-based semantic search with a high-quality vector store and ensure your query and documents are properly preprocessed. Use OpenAI embeddings like text-embedding-3-small and tune retrieval parameters such as top_k to improve relevance.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install openai>=1.0
pip install faiss-cpu

Setup

Install required packages and set your OpenAI API key as an environment variable.

bash

pip install openai faiss-cpu

output

Collecting openai
Collecting faiss-cpu
Successfully installed openai-1.x.x faiss-cpu-1.x.x

Step by step

This example shows how to embed documents and queries using OpenAI embeddings, store them in FAISS, and perform a semantic search that returns relevant results.

python

import os
from openai import OpenAI
import faiss
import numpy as np

# Initialize OpenAI client
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

# Sample documents
documents = [
    "Python is a popular programming language.",
    "OpenAI provides powerful AI models.",
    "Semantic search improves search relevance.",
    "Cats are common household pets.",
    "Machine learning enables AI applications."
]

# Embed documents
response = client.embeddings.create(
    model="text-embedding-3-small",
    input=documents
)
embeddings = np.array([data.embedding for data in response.data]).astype('float32')

# Create FAISS index
index = faiss.IndexFlatL2(embeddings.shape[1])
index.add(embeddings)

# Query to search
query = "How to use AI for better search?"

# Embed query
query_response = client.embeddings.create(
    model="text-embedding-3-small",
    input=[query]
)
query_embedding = np.array(query_response.data[0].embedding).astype('float32')

# Search top 3 relevant documents
k = 3
D, I = index.search(np.array([query_embedding]), k)

print("Top relevant documents:")
for idx in I[0]:
    print(f"- {documents[idx]}")

output

Top relevant documents:
- Semantic search improves search relevance.
- OpenAI provides powerful AI models.
- Machine learning enables AI applications.

Common variations

Use async calls with asyncio and await for embedding requests.
Try different embedding models like text-embedding-3-large for better quality.
Adjust top_k in FAISS search to balance recall and precision.
Use other vector stores like Chroma or Pinecone for scalable search.

Troubleshooting

If search returns irrelevant results, verify your documents and queries are clean and well-formatted.
Ensure embeddings are generated with the same model for both documents and queries.
Increase top_k to retrieve more candidates and rerank if needed.
Check your API key and network connectivity if embedding calls fail.

✅

Key Takeaways

Use semantic embeddings and vector search to improve search relevance.
Keep document and query embeddings consistent with the same model.
Tune retrieval parameters like top_k to balance relevance and recall.

Verified 2026-04 · text-embedding-3-small, gpt-4o-mini

Verify ↗