Graduation Seminar

General Information

Lecturer Dr. Christian Kahmann
Workload 2 SWS Lecture
Seminar Thursday, 09:15 - 10:45
Contact Email or Moodle
Intended subjects Bachelor's, Master's or doctoral thesis in the departments TEMIR and ASV or from the fields of NLP and machine learning in general.


  • The seminar will be carried out online via BigBlueButton.
  • The first seminar is expected to take place on 14th October 2021. Here, the presentation slots will be distributed across the following weeks. In general, it is intended that you give 2 presentations. One at the beginning of your work and the second talk at about 2/3 of your thesis. In the first presentation, it is important to clarify the motivation and the research question, as well as the planned operationalisation. The expected time for this is about 10-15 minutes. In the second presentation, the algorithms and procedures used should be explained and clarified to the audience. If possible, it is of course also welcome to present and discuss results that have already been achieved. The length of this presentation should be about 20-25 minutes. The final grade of the seminar is based on the two individual grades of the presentations.
    If explicitly requested, it is also possible to give a single detailed presentation. This should then last approx. 35 minutes. This solution is not recommended though, as less feedback can be given early in this way.
  • After each talk, there will be a short scientific discussion in which ALL listeners are invited to ask questions and make comments. With this in mind, it is appreciated that you are present for as many of the seminar dates as possible.
  • The schedule of lectures will be displayed and updated on this website. Please check back regularly.
  • Please enroll in the associated Moodle course. If necessary, I will contact you this way in the future.


  • Seminars are live. The first session is for organizing the dates for your presentations. Please be there! [Big Blue Button]
  • Materials will be organized in [Moodle]
  • Communication
    • Email - important announcements will be sent out via mail (Using Moodle).


  • 14.10.2021
    Scheduling appointments via [BigBlueButton]

  • 21.10.2021

  • 28.10.2021

  • 04.11.2021
    • Jonas Richter: Netzwerk des Widerstands (2)
    • Lucy Betke: Analysing Entity Coverage in News Summarization (2)

  • 11.11.2021
    • David Hanslischeck: Quantification of the Overton Window using Text Mining (2)

  • 18.11.2021
    • Christopher Schröder: Active Learning
    • Hannes Hansen: From contextual to static word embeddings (1)

  • 25.11.2021
    • Henrik Bininda: Crawling and Analyzing the Novelupdates Corpus (1)
    • Yannick Dannies: Investigating Stopping Criteria for Active Learning (1)
    • Maximus Gerner: Automatische Generierung von Schachberichten (2)

  • 02.12.2021
    Dies academicus

  • 09.12.2021

  • 16.12.2021
    • Wolfgang Kircheis: Analyzing the History Section of Wikipedia Articles (1)
    • Kim Bürgl: tba

  • 23.12.2021

  • 30.12.2021

  • 06.01.2022
    • Robby Wagner: Impact of Main Content Extraction on large scale near duplicate detection (L)
    • Mathias Halbauer: Analyse der Reaktionen von Coronamaßnahmen auf Basis von Twitterdaten (L)

  • 13.01.2022
    • Niklas Deckers: The Importance of Suppressing Domain Style in Authorship Analysis
    • Tariq Youssef: tba

  • 20.01.2022
    • Lukas Göhlich: Detecting Bias in Media (2)
    • Christopher Akiki: Query Expansion in the Age of Large Language Models (1)

  • 27.01.2022
    fällt aus

  • 03.02.2022
    • Yannick Dannies: Investigating Stopping Criteria for Active Learning (2)
    • Max Martius: Synonymanalyse (L)
    • Erik Schmidt: A New Controlled Dataset for Investigating Deliberation on Wikipedia (2)