avatar

Patrick Parschan

Ph.D. Candidate
Department of Media and Communication at LMU Munich
p.parschan(at)lmu.de


Welcome!

My name is Patrick, I am currently a Ph.D. candidate in Computational Communication Research at the Department of Media and Communication at Ludwig-Maximilians University in Germany. My Ph.D. thesis is supervised my Prof. Mario Haim and Prof. Barbara Plank.
In my thesis I use deep learning methods to find valid numerical measurments of political positions from textual data.

Research Interests

Publications

Journal papers

  1. Computational measurement of political positions: a review of text-based ideal point estimation algorithms
    Patrick Parschan, Charlott Jakob
    Quality & Quantity, 60, 5663–5690 (2026)
  2. Aligning agent-based testing (ABT) with the experimental research paradigm: a literature review and best practices
    Patrick Parschan, Mario Haim, Julian Unkel
    Journal of Computational Social Science, 7, 1625–1644 (2024)
  3. "You are doomed!" Crisis-specific and Dynamic Use of Fear Speech in Protest and Extremist Radical Social Movements
    Simon Greipl, Julian Hohner, Heidi Schulze, Patrick Parschan, Diana Rieger
    Journal of Quantitative Description: Digital Media, 4 (2024)

Conference proceedings

  1. PolBiX: Detecting LLMs' Political Bias in Fact-Checking through X-phemisms
    Charlott Jakob, David Harbecke, Patrick Parschan, Pia Wenzel Neves, Vera Schmitt
    Findings of the Association for Computational Linguistics: EMNLP 2025, pages 17192–17210, Suzhou, China (2025)

Book chapters

  1. Python for Text-as-Data: Using Word Embeddings to Assess the Diversity of Election-Related Search Queries
    Patrick Schwabl, Valerie Hase
    In Bernauer, J. & Wohlmann, A. (eds.), Doing Quantitative Text Analysis with R: Scraping, Preparing, Visualising and Modelling Data. London: Sage, in press
  2. Vielfalt bei Google? Vielzahl, Ausgewogenheit und Verschiedenheit wahlbezogener Suchergebnisse [Diversity on Google? Multiplicity, balance and diversity of election-related search results]
    Patrick Schwabl, Julian Unkel, Mario Haim
    In C. Holtz-Bacha (Ed.), Die (Massen-)Medien im Wahlkampf: Die Bundestagswahl 2021. Wiesbaden: Springer V, 2023
    Article Code & Data

Conference posters and papers

  1. COMPTEXT25
    Computational ideal point estimation using textual data: a systematic review of existing algorithms
    Patrick Parschan, Charlott Jakob
    7th International Interdisciplinary Conference on the Quantitative and Computational Analysis of Text-, Image- and Video-as-Data (COMPTEXT), Vienna, 24–26 April 2025
  2. ECREA24
    A longitudinal analysis of crisis-specific fear speech use in radical and extremist social movements in 2020–2023
    Heidi Schulze, Simon Greipl, Julian Hohner, Patrick Parschan, Diana Rieger
    10th European Communication Conference (ECREA), Ljubljana, 24–27 September 2024
  3. ICA24
    From Politics to Products: Identifying and Analyzing Advertisements in Far-Right Audio Media
    Joshua Greenfield, Robert Dahlke, Josephine Lukito, Patrick Parschan, Maximilian Stecker, Daphne Walter
    74th Annual Conference of the International Communication Association (ICA), Gold Coast, 20–24 June 2024
  4. ICWSM24
    "You are doomed!" — Crisis-specific and dynamic use of fear speech in protest and extremist radical social movements
    Simon Greipl, Julian Hohner, Heidi Schulze, Patrick Parschan, Diana Rieger
    18th International AAAI Conference on Web and Social Media (ICWSM), Buffalo, 3–6 June 2024
  5. COMPTEXT24
    Echoes of the Reichstag? Decoding Populist Rhetoric Across Eras
    Patrick Parschan, Sean Palicki, Benjamin Bevis, Carl Claessen
    6th International Interdisciplinary Conference on the Quantitative and Computational Analysis of Text-, Image- and Video-as-Data (COMPTEXT), Amsterdam, 2–4 May 2024
  6. COMPTEXT24
    "We are doomed!" Detection of Fear Speech as Radicalization Narrative in Far-Right Communication
    Julian Hohner, Simon Greipl, Heidi Schulze, Patrick Parschan, Diana Rieger
    6th International Interdisciplinary Conference on the Quantitative and Computational Analysis of Text-, Image- and Video-as-Data (COMPTEXT), Amsterdam, 2–4 May 2024
  7. DGPuK24
    More Crises, More Fear? A (Semi-)Automated Analysis of Fear Speech in the Online Discussions of Radical and Extremist Actors 2020–2022
    Heidi Schulze, Simon Greipl, Julian Hohner, Patrick Parschan
    69. Jahrestagung der Deutschen Gesellschaft für Publizistik- und Kommunikationswissenschaft (DGPuK), Erfurt, 13–15 March 2024
  8. CHOC23
    A Little Less Hate, But a Lot More Harm — Fear Speech as Strategic Borderline Communication
    Heidi Schulze, Simon Greipl, Julian Hohner, Patrick Parschan, Diana Rieger
    Conference on Harmful Online Communication (CHOC 2023), Cologne, 16–17 November 2023
  9. ISD23
    Zwischen Furcht und Feindseligkeit: Narrative Radikalisierungsangebote in Online-Gruppen
    Simon Greipl, Julian Hohner, Heidi Schulze, Patrick Parschan, Diana Rieger
    Im toten Winkel — Rechtsextreme Radikalisierung im Netz, Berlin, September 2023, Institute for Strategic Dialogue (ISD)
  10. ICA23
    From auditing to testing: Adequate experimental designs to study algorithmic content curation
    Patrick Schwabl, Mario Haim, Julian Unkel
    73rd Annual International Communication Association Conference (73nd ICA Conference), Toronto, 25-29 May 2023 (Session 2311, Part 5)
    PDF Presentation
  11. COMPTEXT23
    Computational scaling of political positions from textual data using word embeddings
    Patrick Schwabl
    5th ANNUAL COMPTEXT Conference 2023, Glasgow, 12-13 May 2023 (Panel 17)
    PDF
  12. DGPuK Methods Conference 22
    Do we need Multimodality? Experiments with Tweets from European Union Executives
    Sina Özdemir, Patrick Schwabl
    23. Tagung der Fachgruppe Methoden der Publizistik- und Kommunikationswissenschaft der DGPuK, Munich, Germany, 5-7 October 2022 (Session Panel: Visualisierung, Part 2)
    PDF
  13. TADA22
    Scaling latent political positions from textual data using word embeddings
    Patrick Schwabl
    4th Text as Data Conference 2022, Cornell Tech, New York City, 6-7 October 2022 (Session A, Part 8)
    PDF
  14. ICA22
    Searching for biased information? Informational strategies and algorithmic curation
    Patrick Schwabl, Mario Haim, Julian Unkel
    72nd Annual International Communication Association Conference (72nd ICA Conference), Paris, 26-30 May 2022 (Session 3226, Part 6)
    PDF

Other

  1. Talk
    How (not) to measure political positions using contextual word embeddings in multi-party systems
    Patrick Schwabl
    Presented to NLP Munich NLP Munich, 20 April 2023
  2. Talk
    How (not) to measure political positions using contextual word embeddings in multi-party systems
    Patrick Schwabl
    Presented at Lunchbreak event of the Junior Scholar Network Political Communication (NaPoKo), 18 April 2023
  3. MA
  4. BA
  5. Talk
    Political position scaling with word representations (from contextualized word embeddings)
    Patrick Schwabl
    Presentation at the workshop on AI-Humanity-Society hosted by LMU Munich and TAU University, 05-09 December 2022
  6. Talk
    Algorithmische Biases in Suchmaschinen im Bundestagswahlkampf 2021 [Algorithmic biases in search engines during the German federal election 2021]
    Patrick Schwabl, Julian Unkel, Mario Haim
    Presented at the sprint review of the Bavarian Research Institute for Digital Transformation, 21 September 2021

University teaching & other courses

Services

Honorary offices

Conference Reviewer

Other

I co-host and organize the in-person Computational Exchange on Wednesday coding club at LMU. See the dedicated page for details, past sessions, and how to join.

Funding and scholarships



This website is based on the minimal light theme by Yaoyao Liu.