I am a computational linguist and work as a Language Engineer at Amazon.
Before that, I was as a postdoctoral researcher in Natural Language Processing at the Applied Computational Linguistics Lab / Goethe University Frankfurt, Germany.
Research areas:
• Machine/Deep Learning in NLP
• Shallow Discourse Parsing (PDTB-/CDTB-style)
• Implicit Semantic Role Labeling (PropBank/NomBank/FrameNet)
• Text Summarization and Text Generation
• Sentiment/Opinion Mining and Subjectivity Analysis
October 2020 - Present
April 2013 - Present
• Set up a server infrastructure for corpus linguistic research.
• Involved in academic organization, grant applications, teaching (Corpus Linguistics & NLP), advising of students (B.A./M.A. Theses).
July 2019 - September 2019
• Developed a topic tracking framework for dialog systems.
October 2010 - March 2011
• Development of a web crawler for online reviews and implementation of a multi-language sentiment analysis program.
• Prototype development of a GUI and annotation guidelines for sentiment annotation.
• Successful patent grant
August 2009 - January 2010
• Development of a pattern-based linguistic component for sentiment analysis + integration into IBM product: primarily, based on user-generated content from social media, forums, message boards, blogs.
January 2015 - September 2019
Final grade: summa cum laude (excellent)
• Thesis on Implicit Discourse Parsing & Implicit Semantic Role Labeling:
Retrieving Implicit Relations from Text. Hidden Semantics and Natural Language Processing.
State-of-the-art Chinese Implicit Discourse Parser (until 2019)
April 2011 - March 2013
Final grade: 1.0 (A)
Thesis: A Suffix Array-based Algorithm for the Extraction of Enhanced N-Grams
October 2007 - October 2010
Final grade: 1.3 (A-)
Thesis: A Finite-State Approach to Sentiment Analysis
I am familiar with a wide range of programming and scripting languages including Java, Python, and R. I have knowledge in employing deep learning (Transformer-based architectures, CNNs, LSTM models) to complex natural language processing tasks such as discourse parsing or natural language inference. In particular, I have developed open source software using theano, keras, deeplearing4j, and huggingface transformers. Beyond that, I have set up web applications with Google Web Toolkit, ZK, and RESTful Web Services using JAX-RS.
Collaboration together with Markus Kaindl and Stephanie Preuss, colleagues at Springer Nature
Computer-generated content, Springer NatureMentor for Bakhtiyar Syed who successfully implemented the first SRL parser for Sumerian!
Open Source Code, Google Summer of CodeLink to Github repository and implementaiton by Samuel Rönnqvist.
Implicit Discourse Parsing, CDTB, CoNLL 2016 Shared Task
n.schenkem.uni-frankfurt.de
Affiliations:
Fachbereich Neuere Philologien
Institut für England- und Amerikastudien, Abteilung Linguistik
Goethe-Universität Frankfurt am Main / Campus Westend
Norbert-Wollheim-Platz 1, IG-Farben Haus, Raum: 3.216
60629 Frankfurt am Main
Fachbereich Informatik und Mathematik
Institut für Informatik, Angewandte Computerlinguistik
Goethe-Universität Frankfurt am Main / Campus Bockenheim
Robert-Mayer-Str. 11-15, Raum: 109
60325 Frankfurt am Main