Prof. Dr. Melanie Siegel

Language - our most important means of communication - is the topic that drives my research, my practical work and my university teaching. I have always been fascinated by analyzing languages and seeing how the differences in the way cultures think and act are reflected in language and how this makes intercultural exchange exciting. As a student, I went to Japan for a year on a language course (and interrupted my studies to do so). This year had a significant impact on my personal development, but also on my research interests. When I returned to Bielefeld University, Prof. Dr. Metzing gave me the opportunity to use my newly acquired language skills directly in a research project on the automatic generation of directions (there was no Google Maps back then).

I was fascinated by computational linguistics. By trying to reproduce linguistic phenomena, we develop a deep understanding of language and ultimately the thinking behind it. For my dissertation, I worked on so-called "translation mismatches" in machine translation and discovered, for example, that the extra-linguistic context is much more important for comprehension in Japanese than in German. For example, the social relationship between the speaker, the addressee and the third party being spoken about is directly reflected in the Japanese word forms. Reproducing this in machine translation requires much more than simply analyzing the sentence - and this problem is still far from being solved.

My research into the grammar of the Japanese language has always been characterized by the idea of how the results can be used to support people in their communication and work. The book that summarizes this work was published by CSLI Stanford in 2016 under the title "Jacy: An Implemented Grammar of Japanese". Right after the introduction, we describe how the grammar has been applied in different projects before going into the details of the grammar.

In 2006, I first went fully into application. As a language technologist at the Berlin-based company Acrolinx, I developed technology for grammar, spelling, style and terminology checking to help authors write good technical documentation. But here, too, the "old" issues soon resurfaced: I helped the company develop language resources for the Japanese language, establish contacts in Japan and set up another office in Tokyo. I then got involved with automatic preediting for machine translation and worked on a research project. Finally, I came across the topic of "easy language" - texts for people with limited language comprehension - and developed technologies to effectively support the production of these texts.

I have been a professor at Darmstadt University of Applied Sciences since 2012. Until 2024, I taught in the Bachelor's and Master's programs in Information Science. Since the winter semester 2024/2025, I have been working in computer science and data science. I really enjoy telling students about my experiences and getting them excited about language and technology. I've been involved with teaching methods in face-to-face and e-learning, in lectures, seminars and projects right from the start and try to develop at least one new seminar a year. In research, the topic of sentiment analysis came back on the table, which I had already dealt with in 2005 but had lost sight of again. This was also because the students showed great interest in this topic. My experience in technical documentation turned out to be a good teaching topic. Collaborations with other degree courses (e.g. business studies, computer science and journalism) have led to research ideas and joint courses. For several years, I have been working on the topics of recognizing and combating hate comments and disinformation.

News:
In November 2017, we founded the Promotionszentrum Angewandte Informatik of the Hessian universities. I am a member of the doctoral committee. Together with colleagues, I am currently supervising these doctoral theses:

• Mina Schütz: Disinformation Detection: A Visual and Explainable Semi-Supervised Transfer Learning Approach
• Jian Xi: Detektion und Bewertung semantischer Inhalte in Bildern für eine erweiterte Kommunikationsanalyse
• Midhad Blazevic: Visual Collaborative Research Platform Based on Data Analytics and Natural Language Processing
• Fabian Sturm: Human Action Recognition in Assembly Lines
• Lennart Sina:Visual Analytics for Corporate Foresight through Automatic Detection and Forecasting of Emerging Trends
• Margot Madina: Easy-to-Read Language: Development of Linguistic Resources and Automatic Text Adaptation Tools
• Christina Barz: From Ingroup to Intergroup Cooperation: A Mixed Methods and Interdisciplinary Investigation of what Motivates Cooperation between Environmental Movement Organizations
• Anne Kathrin Berg: Bedeutung und Akzeptanz öffentlich-rechtlicher Nachrichten in "Leichter Sprache“
• Sabine Richter: Automatisierte Erstellung eines personenspezifischen Rigs zum Vergleich einer Person im Videomaterial mit Personen einer Vergleichsgruppe
• Florian Meyer: Ausbreitung von Hass in sozialen Netzen und die Bestimmung der dynamischen Toxizität
• Uliana Eliseeva: A Modular Approach to Orchestrate Data Pipelines for Visual Analytics

The research project DeTox investigates the automatic classification of toxic comments in social networks. It is a collaboration between Darmstadt University of Applied Sciences, Fraunhofer SIT and the Hessen Cyber Competence Center, funded by the Hessian Ministry of the Interior and Sport. A video of a presentation on this is here.
The research project BoTox is dedicated to researching automated methods for bot and context recognition of hate comments online. The aim is to protect social discourse and identify criminally relevant content, because in an increasingly networked digital space, the impact of hate speech and manipulative bots on our society poses a major challenge. This project is a collaboration between Darmstadt University of Applied Sciences, Fresenius University of Applied Sciences and the Hessen Cyber Competence Center. BoTox is funded by the cyber security research funding of the Hessian Ministry of the Interior, Security and Homeland Security.

Teaching:

SoSe 2024, SoSe 2022, SoSe 2021, SoSe 2020:Sentiment-Analyse (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2019/20:Informationsqualität (Hochschule Darmstadt, Studiengang Informationswissenschaft)

September 2019: Introduction to Natural Language Processing (Ulyanovsk State Technical University)

SoSe 2019: Chancen und Risiken der Digitalisierung (Hochschule Darmstadt, Studiengang Informationswissenschaft, zusammen mit Prof. Dr. Stefan Schmunk)

WiSe 2022/23, WiSe 2020/21, WiSe 2018/19, SoSe 2018: Angewandte Methoden der Sprachverarbeitung (Hochschule Darmstadt, Studiengang Informationswissenschaft)

SoSe 2023, SoSe 2022, SoSe 2020, SoSe 2017: Grundlagen der Sprachverarbeitung (Hochschule Darmstadt, Studiengang Informationswissenschaft)

SoSe 2017: Roboterjournalismus und Social Bots (Hochschule Darmstadt, Studiengang Informationswissenschaft, zusammen mit Prof. Dr. Lorenz Lorenz-Meier)

SoSe 2016: Maschinelle Übersetzung (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2020/21, SoSe 2019, WiSe 2015/16: Text Mining (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2014/15, SoSe 2016, SoSe 2018: Texttechnologie (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2018/19, SoSe 2017, WiSe 2015/16, SoSe 2014, SoSe 2013: Opinion Mining (Hochschule Darmstadt, Studiengang Informationswissenschaft)

SoSe 2024, SoSe 2023, SoSe 2022, SoSe 2021, SoSe 2020, SoSe 2019, SoSe 2018, SoSe 2017,SoSe 2015, SoSe 2014, SoSe 2013: Semantik II (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2023/24, WiSe 2022/23, WiSe 2021/22, WiSe 2020/21, WiSe 2019/20, WiSe 2018/19, WiSe 2016/17, WiSe 2015/16, WiSe 2014/15, WiSe 2013/2014, WiSe 2012/2013: Semantik I (Hochschule Darmstadt, Studiengang Informationswissenschaft)

WiSe 2019/20, WiSe 2016/17, WiSe 2015/16, WiSe 2013/2014: Methoden der Technischen Dokumentation (Hochschule Darmstadt, Studiengang Informationswissenschaft)

SoSe 2011, SoSe 2010: Technische Dokumentation für den Internationalen Markt (Hochschule Karlsruhe, Studiengang Technische Redaktion)

SoSe 2011, SoSe 2010, SoSe 2009, SoSe 2008: Maschinelle Übersetzung (Universität Bielefeld, Linguistik)

WiSe 2009/2010: Wissensbasierte Systeme (Fachhochschule Hannover, Masterstudiengang Informations- und Wissensmanagement)

WiSe 2009/2010: Language Checking (Universität Saarbrücken, Fakultät für Computerlinguistik)

WiSe 2008/2009: Multimodale Integration von Wissen (Fachhochschule Hannover, Masterstudiengang Informations- und Wissensmanagement)

WiSe 2005/2006: Python for Computational Linguistics (Universität Saarbrücken, Fakultät für Computerlinguistik)

SoSe 2004, SoSe 2002: Practical Grammar Engineering Using HPSG (Universität Saarbrücken, Fakultät für Computerlinguistik)

SoSe 2001: Probleme der Verarbeitung des Japanischen (Universität Saarbrücken, Fakultät für Computerlinguistik)

WiSe 1998/1999: HPSG für das Japanische (Universität Saarbrücken, Fakultät für Computerlinguistik)

WiSe 1997/1998: Phänomene der Topikalisierung (Universität Saarbrücken, Fakultät für Computerlinguistik)
br>