Prof. Dr. Felix Naumann

Solving the Climate Crisis with Text Mining

Obviously, our seminar title is probably an overstatement. Nonetheless, we are looking forward to an interesting seminar with devoted students to work on problems that directly result from the work of researchers at the Mercator Research Institute on Global Commons and Climate Change (MCC). They are the ones, who, among other things, compile the climate report by sighting thousands of research articles published within a certain time frame.

Text Mining can significantly support their work. That's why we are excited to offer this collaborative project seminar between MCC and HPI! We already have a few datasets, for example:

  • all Bundestag speeches since 1949
  • 25M tweets by German politicians and other users about the coal discussion
  • 220k seemingly relevant research articles on sustainability
  • 400k research articles on climate change

... and created a list of interesting (research) problem statements to inspire students to define their own goal for the semester. The only limits are time and that it should stay on the topic of text mining and has at least some relevance to MCC.

This project seminar will be a great opportunity in preparation for a master's thesis... (and obviously to make our world a better place).


Time: Mondays, 11:00am
Location: The internet

20.04.Introduction and Organization (on this website)
Videos [1] [2] [3] [4]
27.04.Form groups (via Moodle)
 Weekly (virtual) meetings
11.05.Project pitches
(5-10 min video submissions to Moodle)
 Weekly (virtual) meetings; weekly blog posts of progress
15.06.Student presentations on Text Mining methods
 Weekly (virtual) meetings; weekly blog posts of progress
13.07.Final presentations

(still updated, subject to change)


To join the seminar:

  1. Send email to Tim Repke
    Subject: "[CTM] Request to participate"
    Deadline Wednesday 22.04., 11am
  2. We randomly draw 12 students and contact you by the end of Wednesday
  3. If you were selected, contact the "Studienbüro" to officially sign up

Learning Objectives

Students will learn to...

  • perform common text analytics tasks to explore and understand a given text dataset.
  • understand research papers and report on their own projects' progress.
  • transfer and apply existing methods and research to new use cases.


  • 50% Project
  • 20% Presentation
  • 30% Written report

(subject to change)


There are no specific prerequisites for this seminar. However, basic knowledge of text mining / NLP is an advantage. Having some experience with Pyhton is recommended, as many existing libraries are written in Python. We certain students that are open and motivated to learn about text analysis are able to successfully pass the seminar with the help of materials provided and self-study.


This course is organised by Tim Repke and Dr. Ralf Krestel and will be jointly supervised by Finn Müller-Hansen, Max Callaghan, and Prof. Dr. Jan Minx from MCC.


  • Related web links, articles, books soon