From Student to Modern Research Assistant — a Workout for Social Scientists
So you have a research project where you use standard machine learning approaches to analyse text. A grant brought you money for a research assistant. But: students in the department have to overcome a bit of a knowledge gap if you want to employ them as RAs on your project.
While nowadays students in social sciences often do have a basic training in quantitative research methods, they are often not at a level where they can help doing data science with text.
I put together a couple of sources that I found useful to bring students up to speed. The workout is a mixed bag of a) learning how to handle text b) how to use Python and c) how to analyse text data using machine learning.
There is no particular order. The best way to approach it is to work on a problem from the project — and then figure out the knowledge and skills that are still lacking.
Learning Machine Learning for Text in Python
https://scikit-learn.org/stable/tutorial/index.html
Handling Text
A Nice Intro Class from PolSci: Pablo Barberá & Ken Benoit at LSE
https://lse-my459.github.io/#week-5-machine-learning-for-texts
Learning Machine Learning
Book
Kuhn, Max and Johnson, Kjell: Applied Predictive Modeling (R)
https://www.springer.com/gb/book/9781461468486
Online Course
https://www.coursera.org/learn/machine-learning
Learning Python
Running Python Notebooks Online
https://colab.research.google.com/notebooks/welcome.ipynb
Online Class With Fundamentals
https://www.codecademy.com/learn/learn-python-3
