from sklearn.feature_extraction.text import TfidfVectorizer  
from sklearn.metrics.pairwise import cosine_similarity  
import numpy as np  

# A small collection of descriptions  
documents = [  
    "Machine learning algorithms learn patterns from data",  
    "Deep learning uses neural networks with many layers",  
    "Natural language processing handles text and speech",  
    "Computer vision processes images and video",  
    "Neural networks are inspired by the human brain"  
]  

# Fit and transform  
vectorizer = TfidfVectorizer()  
tfidf_matrix = vectorizer.fit_transform(documents)  

# A search query  
query = ["neural networks and deep learning"]  
query_vector = vectorizer.transform(query)  

# Calculate similarity between query and all documents  
similarities = cosine_similarity(query_vector, tfidf_matrix).flatten()  

# Rank documents by similarity  
ranked = np.argsort(similarities)[::-1]  
print("Most similar documents:")  
for i in ranked:  
    print(f"  {similarities[i]:.3f} — {documents[i]}")