Analysing a TwitterBot using TextBlob, NLTK and Python
summary
Downloads about 2600 tweets from the Magical Realism Bot, uses Tweepy, TextBlob and NLTK to POS-tag each tweet and build parts-of-speech sequences and per-tag word lists. Encodes those tag sequences into a directed acyclic word graph, samples tags to generate templates and fills them with observed words. Simple templates (e.g. determiner + adjectives + noun, or “made of” patterns) produce grammatical outputs, but random fills often lack semantic coherence, so full reverse engineering needs richer grammar, context or template metadata.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:30Natural Language Processing
- 2:08conversational interfaces
- 3:50what is a TwitterBot
- 6:25how TwitterBot is generated
- 8:43collecting tweets
- 9:30using TextBlob
- 12:10TextBlob capabilities
- 17:49Part of speech tagging
- 19:19Directed acyclic word graph
- 22:14Adding structure
- 28:11Questions
related talks
01h 51mNatural Language Processing in Python
01h 09mAutomatic Tagging of Short Texts with scikit-learn and NLTK
19mDeep Into the Tweet
01h 39mNatural Language Processing (NLP) in Python - From Zero to Hero
55mDemystifying Natural Language Processing across several industry verticals
33mSyntax Isn't Everything: NLP for Rubyists
17mExploring Natural Language Processing
44mTeaching Computers ABCs: A Quick Intro to Natural Language Processing