Sen CutlerData Engineer

Hybrid Embedding and BM25 Semantic Matching Pipeline

The project goal is to match text chunks against a document set comprising 13 cleaned ebook texts. The process combined embedding similarity with relevance scoring functions built using a curated term cloud, with the goal of surfacing emphasizes distinctive passages. The rating step used BM25/centroid scores as weights together with embeddings as scores, deriving the rating of each chunk strongest-match, top-k, and 95th percentile evidence. It produces clearer results than plain cosine similarity when many documents share similar language.

  • Machine Learning
  • Natural Language Processing
  • Latent Semantic Analysis
  • Python
  • Word Embedding

AI · Statistics · Machine Learning