Lukas Schmelzeisen

Research: Ongoing work on my PhD dissertation. Cocreated DFG-grant application "Open Argument Mining" (funded for three years). Reviewing duties. • Teaching: Independently organized the tutorial on "Web Information Retrieval". Main instructor for the research practical "Artificial Intelligence Approaches for RTS Games". Supervised three bachelor theses on the topics of "Word Embedding Modeling Choices", "Graph Convolutional Networks for Word Sense Disambiguation", and "Image Recognition for RTS Games".

Personally designed and implemented the Java-based "Generalized Language Modeling Toolkit", a toolkit for evaluating experimental language modeling approaches in the area of natural language processing. • Core development role in the relaunch of the social media website "Metalcon" in a team of five students. My focus area was building an asynchronous, distributed, and extensible system

Working title of dissertation: "Ontology Learning via Word Embeddings" Research areas: Machine Learning, Natural Language Processing, Semantic Web

Permission to start a PhD without a Master's degree (very rare in Germany) Exam Subjects: Artificial Intelligence, Machine Learning, Natural Language Processing, Semantic Web

Overgrall grade: 1.5 (German scale, equivalent to A). Bachelor thesis on “Fast and Non-Approximative Language Model Prefixquries for Word Prediction using Top-k Joining Techniques”, grade: 1.0 (German scale, equivalent to A+).