September 2015. Working through the NLTK book. Scraping academic papers, analyzing them. Hypothesisless.

• Natural Language Processing with Python

1. git.
2. Make def for scraping text, converting PDF to TXT.
3. Basic text data cleaning: eliminate \\n, etc.
4. FreqDist common words.
5. Language diversity?
6. Data visualizations of everything.
7. Add more sources/corpi. (Compare to non-academic? Compare over time?)
8. Refactor code into: scrape.py, clean.py, analysis.py, visualize.py, or something.
9. Most common word endings.
10. Sentence length.
11. FreqDist: instead of abs values, convert to % of total wordcount.
12. Return a random sentence.