Twitter Bot

twitter-bot-python-darioncassel

1. Let's Build a Twitter Bot

• Rearrange words
  --Stretch--
• String, sentence reversals
• Interactive mad libs

2. Random Dictionary Words

• Random "sentences"
  --Stretch--
• Autocomplete
• Anagrams (generates sentences)

3. Analyze Word Frequency in Text

• Histogram, unique_words, frequency
  --Stretch--
• Histogram with tuples (histogram_tuple())

4. Stochastic Sampling

• Random select, weighted select, test_probability
  --Stretch--
• Speed, memory optimized histograms

5. Data Structure Comparisons and Performance Analysis

• Flat Associative Array
• Sorted Singly Linked List
• Doubly Linked List
• Binary Search Tree
• Analysis & benchmarks

6. Hash Table from Scratch

• HashTable

7. Picking a Corpus

• Sources
• Diffbot Script
• Corpus

8. Parsing Text and Cleanup

• Cleanup Parser
• Clean Corpus
  --
• Clean Corpus 2 (with periods)
• Clean Corpus 3 (with more punctuation)
• Clean Corpus 4 (with more punctuation and parenthesis)
• Clean Corpus 5 (without parenthesis)

9. Tokenization

• Tokenizer

10. Counting Tokens

• Random Weighted Select of tokens
• BinaryHeap

11. A Word about Application Architecture

--N/A--

12. Generating Sentences with Markov Chains

• Markov Model
• Markov Model Tester
• Model retain script
  --Models--
• (Pickled) Models with Corpus 1 (n is lookback):
  • Trained Model n=1
  • Trained Model n=2
  • Trained Model n=3
  • Trained Model n=4
  • Trained Model n=5
  • Trained Model n=6
  • Trained Model n=7
• (Pickled) Models with Corpus 2:
  • Trained Model n=2
  • Trained Model n=3
  • Trained Model n=4
  • Trained Model n=5
• (Pickled) Models with Corpus 3:
  • Trained Model n=2
  • Trained Model n=3
• (Pickled) Models with Corpus 4:
  • Trained Model n=1
  • Trained Model n=2
  • Trained Model n=3
  • Trained Model n=4
  • Trained Model n=5
• (Pickled) Models with Corpus 5:
  • Trained Model n=1
  • Trained Model n=2
  • Trained Model n=3
  • Trained Model n=4

13. Flask Web App

• Markov Model Runner
• Web app source
• Philisobot site

14. Time to Tweet

• Philisobot site
• @philisobot