Deborah Jakobi

Deborah Jakobi, M.A.

PhD candidate
Digital Linguistics

Room number: AND 2.18

E-Mail

I am a second-year PhD student in Computational Linguistics, supervised by Prof. Dr. Lena A. Jäger and co-supervised by Prof. Dr. Rico Sennrich. I have a Master's degree in Computational Linguistics & Language Technology and Informatics from UZH.

My interests lie at the intersection of computational psycholinguistics and machine learning reserach. On the one hand this includes the development of new methods and tools to process, evaluate and analyse cognitive data and on the other hand implementing machine learning methods to predict reader or language properties from cognitive data, and how to use cognitive data to improve, evaluate and analyse machine learning methods.

Publications
Talks and Posters
Teaching
Teaching Assistance / Student Tutor

Publications

Neural Additive Models Uncover Predictive Gaze Features in Reading, ETRA 2025
Deborah N. Jakobi, David R. Reich, Paul Prasse, Lena A. Jäger

Data Management Plan (DMP) for COST Action "MultiplEYE" (CA21131), PsychArchives
Marie-Luise Müller, Deborah N. Jakobi, Ramunė Kasperė, Nora Hollenstein, Lena A. Jäger
[http | bib]

MultiplEYE Data Sharing Policy, PsychArchives
Marie-Luise Müller, Ramunė Kasperė, Deborah N. Jakobi, Leah Bradshaw, Nora Hollenstein, Lena A. Jäger
[http | bib]

🏆 Best student late breaking work award
Reporting Eye-Tracking Data Quality: Towards a New Standard, ETRA 2024
Deborah N. Jakobi, Daniel G. Krakowczyk, and Lena A. Jäger
[http |arxiv preprint | bib | poster (PNG, 997 KB)]

PoTeC: A German Naturalistic Eye-tracking-while-reading Corpus, Behavior Research Methods
Deborah N. Jakobi, Thomas Kern, David R. Reich, Patrick Haller & Lena A. Jäger
[arxiv preprint |data | bib]

ScanDL: A Diffusion Model for Generating Synthetic Scanpaths on Texts, EMNLP 2023
Lena S. Bolliger, David R. Reich, Patrick Haller, Deborah N. Jakobi, Paul Prasse, Lena A. Jäger
[ http |Video|Poster|bib]

pymovements: A Python package for eye movement data processing, ETRA 2023
Daniel G. Krakowczyk, David R. Reich, Jakob Chwastek, Assunta Süss, Paul Prasse, Deborah N. Jakobi, Oleksii Turuta, Paweł Kasprowski and Lena A. Jäger
[https | PyPI |bib]

Talks and Posters

MultiplEYE: A multi-lab effort to establish new standards for the collection and pre-processing of multi-lingual eyetracking-while-reading data, X-PPL 2023 (talk)
Deborah N. Jakobi, Maroš Filip, Ramunė Kasperė, Nora Hollenstein, Lena A. Jäger & The MultiplEYE Team

The MultiplEYE Reading Corpus, X-PPL 2023 (poster)
Deborah N. Jakobi, Maroš Filip, Ramunė Kasperė, Nora Hollenstein, Lena A. Jäger & The MultiplEYE Team
[Poster Gallery ]

Teaching

Advanced Machine Learning, University of Zurich, Spring Semester 2025

First, this course covers an in-depth discussion of state-of-the-art methods in supervised and unsupervised machine learning including (Retrieval) Transformers, Graphical Neural Networks and Diffusion Models as well approaches to combine elements from these architectures. We will further discuss how transfer learning (including zero- and N-shot learning) can be applied in different types of problem settings. Second, this course provides an introduction to Reinforcement Learning that introduces the reinforcement problem setting as Markov Decision Process and covers Dynamic Programming approaches, Monte Carlo methods, Temporal Difference Learning and approximate solution methods.

Course Overview

Quantitative Methods, University of Zurich, Fall Semester 2024

This course introduces the basic concepts of statistical analysis as used in modern linguistics, covering data description and visualization as well as basic techniques of machine learning and modelling. The course also introduces the basic concepts of frequentist vs Bayesian approaches and the use of simulations and baseline models.

Course Overview

Summer School, University of Potsdam, August 2023

Teaching at the SFB 1287 Summer School at the University of Potsdamtogether with Prof. Dr. Lena A. Jäger and Shuwen Deng. We have been teaching on "Generating human-like Scanpaths in Reading using Deep Neural Sequence Models" specifically introducing Eyettention.

COST action training school, April 2023

Trainer at MultiplEYE COST action training school, April 26 - April 28, 2023, in Kaunas, Lithuania. My sessions were about running and implementing an eye-tracking experiment with Python specifically for researchers without programming experience.