Research interests

• Event and Temporal Relation Extraction
  Millions of news articles are published every day. Finding, extracting and structuring the important information in an efficient way is a crucial task in many domains, for example in algorithmic trading. I do research on the automatic event and relation extraction on news articles. Each event is characterized by the involved entities, by a connecting property like a certain action, and a temporal information.Subsequently, temporal and causal relations are extracted for those events.
• Deep Learning
  For my research, I apply various super- and semi-supervised machine learning techniques for the event and relation detection. A special research focus lies on deep learning, which allows to model high-level abstraction of data. I use and extend deep learning techniques in order to extract events and relations within documents and across documents.

Projects

Currently I work in two projects:

• DARIAH DE-II: I'm involved in the development of quantitative methods to detect the evolution of narrative techniques in (German) novels over the past centuries.
• Structuring Story-Chains: As thousands of news articles are published daily, it is challenging to stay up-to-date on every topic. The goal of the project is to help readers to tackle the information-overload. This is done by extracting and analyzing the causal connections between articles. The results are useful for various tasks, e.g. to get a quicker overview on a certain topic.

Awards & grants

• Grant for a Software Campus project, 2014

Challenges

• I participated in the Event Nugget Detection and Classification task at TAC 2015. The created deep neural network scored a F1-measure of 65.31% and is placed first among 14 systems.
• I participated in the German Named Entity Recognition task at KONVENS 2014. The created state-of-the-art deep neural network scored a F1-measure of 75.1% and is placed 2nd among 11 systems.

Teaching

• 11/2017: Seminar Deep Learning for NLP at the University of Duisburg-Essen
• WS 2015/2016: Deep Learning for NLP
• WS 2014/2015: Foundations of Language Technology / Grundlagen Intelligenter Systeme

Student guidance

I'm supervising Bachelor- and Master-theses in the field of Natural Language Processing in combination with Machine Learning. In case you like to write a thesis in the following fields, feel free to approach me:

• Structuring Story-Chains: You are interested to analyze thousands of news articles and creating an intelligent system, supporting either readers or editors of news websites? This project has the opportunity for various types of theses: Reading recommendation systems, visualization of large datasets & graphs, information extraction from news articles, link discovery, as well as human-machine-interaction.
• The Deep Learning Revolution: Deep Learning uses deep neural networks to achieve state-of-the-art results in all domains of machine learning. In my research I investigate different deep learning approach and apply them to text data. In case you are interested to write a thesis in this field, feel free to approach me.

I have supervised the following theses:

• Philip Beyer. Proposal for a STS Evaluation Framework for STS based Applications. Student Research Paper (Studienarbeit). Computer Science Department. Technical University of Darmstadt. Published at Coling 2016.
• Ziyang Li. Related Articles Discovery in Large Corpora (Masterthesis). Computer Science Department. Technical University of Darmstadt.
• Michael Bräunlein. Multi-Document High Precision Event Extraction (Masterthesis). Computer Science Department. Technical University of Darmstadt.

Biographical information

I hold a M.Sc. (Master of Science) in IT-Security from the Technical University of Darmstadt, a B.Sc. (Bachelor of Science) in Computer Science, and a B.Sc. in Mathematics from the University of Oldenburg.