Natural Language Processing in Python
summary
Scrapes and compares transcripts from 12 comedy specials with requests and BeautifulSoup, builds a document-term matrix (~7,000 columns) via scikit‑learn CountVectorizer, and pickles intermediate objects. Cleans/tokenizes text with regex, runs EDA (word clouds, vocabulary sizes, profanity rates — e.g., ~2 F‑words/min for one comic vs 0 for another), scores sentiment with TextBlob over ten slices, extracts LDA topics (nouns+adjectives, 4 topics after many passes) with gensim, and generates sample routines using Markov chains — the end‑to‑end pipeline executes and yields interpretable results.
AI-written summary · about
chapters (26)
- 0:00Introduction
- 2:30What is NLP
- 3:22Examples of NLP
- 7:31What is Data Science
- 12:11Data Science Workflow
- 19:21Data Gathering
- 20:57Scope
- 22:31Data Cleaning
- 24:10Corpus
- 24:55Term Matrix
- 25:59Regular Expressions
- 26:56Tokenization
- 27:26Remove Stop Words
- 28:09Bag of Words
- 28:53Matrix
- 29:17Document Term Matrix
- 30:09Create Document Term Matrix
- 30:36Data Format
- 30:58Output
- 31:26Jupiter Notebook
- 33:41Scraping Data
- 35:39Cleaning Data
- 37:07DataFrame
- 44:48TermMatrix
- 49:03ETA
- 56:58Pickle Files
related talks
01h 39mNatural Language Processing (NLP) in Python - From Zero to Hero
44mTeaching Computers ABCs: A Quick Intro to Natural Language Processing
35mHow does text become data?
33mWhen Stand-up Comedy Meets Machine Learning
23mMultilingual Natural Language Processing
01h 27mTips, Tricks and Topics in Text Analysis
28mLifecycle of Web Text Mining: Scrape to Sense
30mAnalysing a TwitterBot using TextBlob, NLTK and Python