FuzzyTM: a Python package for fuzzy topic models
summary
FuzzyTM is a Python library that trains fuzzy topic models (FSA, FSAW, FSAV, FSAE) and produces topic embeddings for downstream tasks; it implements LSA/SVD pipelines, fuzzy clustering, sparse document–term matrices and save/load with pickle because runs are non‑deterministic. Experiments run each algorithm 10 times, report a coherence of ~0.53 on M10, and handle corpora with ~90,000 unique words and ~4,500 documents (≈300 million dense cells) using sparse arrays for speed. FSAW is the highlighted model.
AI-written summary · about
chapters (5)
- 0:00Start
- 2:10Topic Models are a group of unsupervised natural language processing algorithms that extract latent topics in texts.
- 6:13Probabilities vs. Fuzziness
- 10:21FLSA-W, one of the fuzzy topic models, outperforms other state-of-the-art topic models on open datasets in most settings
- 15:27We use pyFUME for fuzzy clustering
related talks
30mFuzzy Matching at Scale
01h 01mWorkshop: Intro to topic modeling using Python
01h 19mGenerating Customers Insights with Topic Modelling and HuggingFace SetFit Method
45mDecoding Topics: A Comparative Analysis of Python’s Leading Topic Modeling Libraries Using Climate C
48mTopic Modelling (and more) with NLP framework Gensim
13mThinking of Topic Modeling as Search
45mEvaluating Topic Models
01h 26mBagging to BERT- A tour of applied NLP