lanzhang76/dorkshop_nlp

Teaching notes for nlp dorkshop at Parsons

Natural Language Processing with Python

Summary

An beginner introduction to Natural Language Processing (NLP) topics using Python, NLTK, and Jupyter notebook. Presented via a two-hour Zoom video as part of Parsons Design & Technology peer-to-peer 'dorkshop' series.

Installations and setup

This workshop uses the following tools:

• Python3
• nltk library
• Jupyter Notebook web application
• GitHub

For help with installations and downloads, see installations.md.

Schedule

1. Intro: What is NLP? what might we do with it in DT?
2. 00_setup.ipynb: What are some Python string basics, and how do we run them in a Jupyter notebook?
3. 01_nltk_mini_project.ipynb: What is NLTK? What is a tokenizer? What is a part-of-speech tagger? How might we make something fun with these?
4. 02_nltk_sticker_sheet.ipynb: What else can we do with the NLTK library?

Learning outcomes

• Basic familiarity with common topics in NLP
• Pointers to tools, methods, and projects related to NLP
• Create a fun, small text-based program using Python in Jupyter Notebook

References for more

Lessons

• Intro to Python for Text and Data (Alison Parrish)
• Computational Form: Generating Text (Justin Bakse)

Tools and libraries

• Natural Language Toolkit - General NLP library for Python
• Markovify - Markov text generation library for Python
• Natural - General NLP library for Javascript
• Tracery.js - Context-free-grammar library for Javascript
• P5.speech - P5.js library for speech recognition and speech synthesis
• Talk to Transformer - Web application for running GPT2 text generator

Projects

• NaNoGenMo (based on an idea by Darius Kazemi)
• 1 the Road (Ross Goodwin)
• Predictive Writer by Botnik Studio
• Toast - a Word Guessing Game (by Anna & Lan)