Back to Projects

Machine Learning

Recipe Text Clustering

An unsupervised text clustering project comparing bag-of-words and sentence embeddings for recipe-name clustering, using K-means, DBSCAN, and UMAP visualization.

Period

May 2026

Tools / Tech

PythonScikit-learnK-meansDBSCANSentenceTransformerUMAP

Why I built it

This project helped me test how different text representations change clustering behavior on short, noisy recipe names.

Links

What it includes

  • Compares bag-of-words features with sentence embeddings.
  • Runs K-means and DBSCAN clustering.
  • Uses UMAP to visualize clusters and inspect whether groups are meaningful.

What I worked on

  • Prepared text representations and clustering experiments.
  • Compared algorithm behavior across sparse text features and dense embeddings.
  • Completed the clustering notebook myself and reviewed outputs qualitatively instead of assuming labels were correct.

What I Learned

  • Learned why short-text clustering is sensitive to representation choices.
  • Practiced evaluating unsupervised learning without ground-truth labels.
  • Built intuition for when visualization helps and when it can mislead.