What's this?

This is just me learning a bit about NLP (Natural Lenguage Proccesing). I got interested in this when I heard about sentiment analysis. I'll put in this readme the main resources that I find interesting.

Learning material

I'll try to update here with all the resourses I find helpfull for learning and doing things related with NLP

1 - NLP || Dan Jurafsky || Stanford University

I got introduced to NLP in this YouTuve playlist. Is a series of short introductory videos explaining the base concepts.

2 - NLTK Book (I'm here now)

nltk is a Python framework to work with human language language. They have a book explaining the software and how it works

A note on Jupyter Notebooks

From jupyter.org website:

The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text.

So this is great. I started my Dockerfile from on Dockerfile they maintain with python package for science and added the packages I need like nltk.

Follow the book with Jupyter Notebooks and NLTK (The software)

To interact with this Python package I'm going to use a Docker image with latest Python and nltk + dependencies. This is how to build and open the playground To build the Docker images go to the root of this repo and do: nltk is the name of the image we built

 docker build --rm -t jupyter/nltk-book .

This command will start an server on port 8888 on your browser. Follow the instructions

 docker run -p 8888:8888 --volume={PATH_IN_YOUR_COMPUTER}:/home/jovyan jupyter/nltk-book:latest

Note: PATH_IN_YOUR_COMPUTER has to be a valid path in your computer. Mine for example is: /Users/andres/code/andresgutgon/nlp/book. This way when Jupyter writes a book page it is storage in my machine because the volume is share between Docker and the host (my machine).

Download the book with nltk command line

Create a note book in Jupyter UI and run these commands. This will save in PATH_IN_YOUR_COMPUTER the book chapters and all NLP code necesary

python
>> import nltk
>> nltk.download()

After that follow the book. To import

Ideas

1. Do a Node application with Svelte as UI framework and a backend that analyse sentiment analisis with retextJS. This can be expanded to input a Tweet and analyse the overall sentiment of all the replys to that tweet.