Matthias Hertel
Building and room: 051-02-029
|
Bio
After my Bachelor's and Master's degree in Computer Science at the University of Freiburg, I started as a research assistant at the Algorithms and Data Structures chair in 2020. My research interests are Deep Learning, Natural Language Processing, Efficient Algorithms and Energy Informatics and Sustainability.
Projects
- Gruppenrechner für den Wechselunterricht während der Corona-Pandemie / Class splitting for schools during the COVID-19 pandemic (2020): wechselunterricht.org
- Tokenization Repair: Correction of Whitespace Errors in Digitized Documents (2020-21): Paper, Website
- Spelling Correction with Neural Networks (ongoing): Master's thesis
- Entity Linking for Wikipedia (ongoing)
- Fast and Accurate Sentence Segmentation (ongoing)
- Public Transit Route Planning with Transfer Patterns (2015-16): Paper, Google AI blog post
Publications
- Tokenization Repair in the Presence of Spelling Errors
Hannah Bast, Matthias Hertel, Mostafa M. Mohamed
CoNLL, 2021 - Tokenization Repair in the Presence of Spelling Errors
Hannah Bast, Matthias Hertel, Mostafa M. Mohamed
arXiv, 2020 - Scalable Transfer Patterns
Hannah Bast, Matthias Hertel, Sabine Storandt
ALENEX, 2016
Teaching
- WS 2020/21: Deep Natural Language Processing (Seminar)
- SoSe 2021: Algorithmen und Datenstrukturen (Tutor)
- WS 2021/22: Information Retrieval (Tutor)
Supervision of student projects and theses
Please check our Teaching Wiki for a list of available topics, or send me an e-mail if you want to propose your own topic. Past projects and theses:
- Spelling Correction and Autocompletion for Mobile Devices (project by Ziang Lu, 2021)
- Tokenization repair using Transformers (project by Sebastian Walter, 2021)
- Sentence Segmentation (project by Krisztina Agoston, 2021)
- Expansion Planning of Low-Voltage Grids Using Ant Colony Optimization (thesis by Lukas Gebhard, 2021, together with Fraunhofer ISE)
- Planung von Mittelspannungsnetzen mithilfe eines Ameisenalgorithmus (thesis by Jonathan Zeller, 2021, together with Fraunhofer ISE)