Scientific Paper Clustering for Similarity Discovery
What it is
A Python notebook-based pipeline that clusters 10,000+ arXiv abstracts by semantic similarity using TF-IDF vectorization and K-Means. It includes dimensionality reduction via PCA, t-SNE, and UMAP for visualization, along with topic modeling to label clusters and benchmarks for NLP preprocessing strategies.
The problem it solves
Researchers face difficulty finding papers with similar research backgrounds due to the vast number of loosely tagged publications, requiring manual reading of dozens of abstracts to identify relevant clusters.
What you could use it for
- Automating the discovery of related literature in large academic repositories
- Visualizing the thematic landscape of a specific research field
- Benchmarking the impact of NLP preprocessing steps on clustering quality
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.