CORE Tutorial 2020

This repository contains the materials required for the tutorial presented at CORE Congress 2020, concerning "Exploratory text analysis for computational social science." The goal of the tutorial is to showcase some simple text analysis techniques that can reveal insight into computational social science questions, such as the content of fake news.

The code for the tutorial is contained in the notebooks word_frequency.ipynb, topic_modeling.ipynb, and word_embeddings.ipynb, which are hosted on Colab at the following links:

1. word_frequency.ipynb: https://colab.research.google.com/drive/1GeB_l0nc_30J3yV9f88NjRrwxM04-KV5
2. topic_modeling.ipynb: https://colab.research.google.com/drive/1Cz519-LH0tR2mV0xyy0RlrJiLRfFoB82
3. word_embeddings.ipynb: https://colab.research.google.com/drive/1HYZ_6ysD_3Yj_JsDV6WtPM-wad-3bQPz

To use one of the notebooks, open the link and choose "Copy to Drive" to make your own version. You will need to log into your Google account to do this.

Instructions to run code locally

NOTE: the following steps are optional for the tutorial, I encourage everyone to use the Colab links above if possible.

You can run the notebooks locally on your own machine, but be warned that installing all the packages may be difficult without conda. The code is compatible with Python 3. We assume that you have installed Python on your machine and are familiar with installing packages using pip or a self-contained environment like anaconda.

Before running the Jupyter notebooks, please install all necessary packages and data (~ 220 MB) in your environment by opening the command line and issuing the following commands. Note that the download may take from 5-30 minutes depending on your download speed.

Linux:

conda create -n CORE_tutorial --file=spec-file.txt
wget https://bitbucket.org/istewart6/core_tutorial_2020/raw/36e69f9d777319ae2cc94354cf57bd01f3e080b3/data.zip . 
unzip data.zip

Mac:

conda create -n CORE_tutorial --file=spec-file.txt
curl https://bitbucket.org/istewart6/core_tutorial_2020/raw/36e69f9d777319ae2cc94354cf57bd01f3e080b3/data.zip -o data.zip
unzip data.zip

Next, launch Jupyter locally to start using the notebooks:

jupyter-notebook .

Acknowledgments

Thanks to Rada Mihalcea and Laura Biester for feedback on early versions of the tutorial.