This is a short, incomplete and informal description of my professional journey, heavily focusing on my experience in the field of Natural Language Processing (NLP). For a complete, more structured, up-to-date summary of my academic and professional career achievements, please check out my resume here.

I am currently a fourth year Ph.D. candidate in Computer Science at the NLP Lab of the University of British Columbia (UBC). My research focuses on designing and implementing new computational models that enable automated systems to better understand and generate natural language. In addition to my primary research on data-driven discourse parsing — a crucial upstream task to better understand natural language — I am generally interested in creative applications of modern machine learning methods and algorithms.

My interest in NLP already sparked during my time as a Master’s student at the Karlsruhe Institute of Technology (KIT), where I got exposed to a variety of specialized courses on different topics of Machine Learning and Artificial Intelligence (AI). Out of these many course topics I especially liked the topic of NLP for maily three reasons:

1. Natural language is everywhere, and computationally understanding, extending and generating language holds great potential.

2. In comparison to Computer Vision, which already achieved great improvements through the use of deep-learning, NLP is still full of major unanswered questions — down to very essential problems.

3. Natural language itself is ambiguous, unstructured, yet incredibly powerful.

Out of this initial interest grew the amazing opportunity to work with Professor Dr. Alex Waibel, joint professor at the KIT and the Carnegie Mellon University (CMU). Under his supervision, I started working on contextual systems, mostly focusing on language modelling and translation. To assess the ability of a deep-learning system to obtain context-knowledge, we provided the community a new dataset, published at LREC 2018. I used the dataset myself to train and evaluate a context-aware language model, inspired by state-of-the-art translation systems at the time for my Master’s thesis.

With my initial insight into the workings of NLP related research during my Master’s thesis and my previous industry experience as a working student and intern in Germany and Canada, I had a tough decision to make — either continue to pursue an academic career or transfer into a full-time job.

I decided to continue my career in academia by applying at the University of British Columbia (UBC) in Vancouver. My decision to pursue a Ph.D. degree was mostly influenced by the wish to dedicate an extended time-span to an open problem that I am genuinely curious about, with the chance to potentially have a positive impact on the world.

Since I started my Ph.D in September 2018, I mainly focus on the area of data-driven document-level semantic and pragmatic analysis of textual data – In my opinion a crucial building block for the next generation of NLP models and applications, aiming at truely understand the authors intention and communicative goal.