Vector Embedding

Vector Math and Proximity:

In a 300-dimensional embedding space, the vector coordinates for related words align based on semantic attributes:

• embedding("King") - embedding("Man") + embedding("Woman") approx embedding("Queen")
• If you compute the cosine similarity between vector("Deep Learning") and vector("Neural Networks"), the score will be high (e.g. 0.89), whereas the score between vector("Deep Learning") and vector("Baking Cakes") will be near-zero (e.g. 0.12).

Typical Usage in Python (using SentenceTransformers):

from sentence_transformers import SentenceTransformer, util
model = SentenceTransformer('all-MiniLM-L6-v2')

# Generate embeddings
emb1 = model.encode("Machine Learning courses in Chennai")
emb2 = model.encode("Python AI training in Tamil Nadu")

# Compute cosine similarity
similarity = util.cos_sim(emb1, emb2)
print(f"Semantic Relevance Score: {similarity.item():.4f}") # Outputs near 0.78