This page consists of the links to lecture slides and assignments. Lecture videos are shared in a private folder to registered participants. For details on the deadlines for assignments/termpaper check your emails.
Session 1: Introduction
5th October 2022
An overview of the course, NLP, and NLP in Economics. slides
Session 2: Python overview, and an introduction to text analysis in python
6th October 2022
Python overview Jupyter notebook. 
Python and text overview Jupyter notebook.
Regular expressions - slides
Older files, which may still be useful for Python basics:
- Python basics - overview - covers variables, conditionals, loops, functions, error handling, files etc
- Python data structures - overview - covers strings, lists, dictionaries, tuples etc
- An example code covering all python basics
Session 3: NLP and Machine Learning Methods: An Introduction
7th October 2022
NLP methods - an overview (slides)
NLP with Spacy/Textacy Jupyter notebook.
Session 4: Text Classification and Topic Modeling
11th October 2022
Text classification and topic modeling (slides)
Text classification - Notebook.
Data file for the above notebook
Text classification with finetuning - Notebook.
Session 5: NLP with little or no training data
12th October 2022
Session 6: NLP and Economics, Group Discussions
13th October 2022
Instructions:
-
Students are grouped into teams of 2-3 people and can choose a paper related to economics research that uses NLP methods. You can choose one of the papers from the readings list in course materials or pick one on your own.
-
Teams should present for 10-15 min, we can have a 5 min additional discussion per team. Slides are optional.
-
What to cover:
- What problem relevant to economics is covered in the paper
- What NLP methods are used?
- How do they fare? What is your opinion about the paper?
- What are some good ideas for future if you get a chance to work on this topic?
(Optional) Session 7: Review
17th October 2022
Term papers
Deadline: 27th October 2022
Write a problem statement for a research question in your discipline that you think can benefit from using NLP methods. Give some background about the problem, existing ways of addressing it, and how NLP can help. Write about what NLP approaches you learnt in this class can be useful for this problem. You are free to read any related literature, but I want you to come up with a problem statement of your own. Your writeups can be 2-4 pages long, and contain all the above mentioned information.
There is no meeting afterwards. I will send you feedback by email.