Tricks, tips and topics in Text Analysis
summary
Pipeline shows concrete choices: spaCy for POS, NER and lemmatization, gensim LDA on a 300‑article news corpus to surface ~5–10 rough topics, scikit‑learn + TF‑IDF + SVD for classification, and Keras LSTM for character/word generation. The tutorial documents exact preprocessing (stop lists, lowercasing, lemmatization, bigrams), notes why each step matters, reports quick LDA outputs (some fire/politics topics), and delivers a working text generator trained on the same news data.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 6:05Text analysis and Python
- 12:05Gathering text data
- 25:39Computational linguistics with SpaCy
- 39:24Topic modeling with Jensen
- 1:08:17Text classification with scikit-learn
- 1:19:36Deep learning for text generation
- 1:27:41Summary
related talks
37mData Science for Digital Humanities: Extracting meaning from Images and Text
24mBertTopic, From Free-Text feedbacks to Calls for Action
39mTemporal Analysis on Topics Utilizing Word2Vec
38mAutomatically categorizing news articles on nrk.no
35mUsing Scattertext and the Python NLP Ecosystem for Text Visualization
01h 48mApplied Data Science
37mBuilding a Scientific Taxonomy at Scale
49mThe CV: A Data Scientist's View