Search Engine

An information-retrieval system with dense and sparse retrieval

An information-retrieval system built from the ground up in Python, combining classical and neural retrieval over a Persian-language corpus.

Highlights

  • TF-IDF ranking for sparse lexical retrieval
  • ParsBERT embeddings for dense semantic retrieval
  • K-means clustering to organize the document space
  • FAISS for efficient approximate nearest-neighbor search at query time