Bernhard Kratzwald

Researcher at ETH Zurich

Research Interests:

My goal is to reshape and re-think how people interact and work with AI. Currently: Passionately working on new AI solutions for manufacturing beyond academic leaderboard challenges. Focusing on human-AI interactions, continuous learning, and human-in-the-loop settings.

In my free time, I am passionate about good food, running, and traveling.

About Me

I did my Ph.D. at the Chair of Management Information System, ETH Zurich. I started in October 2017 working with Prof. Stefan Feuerriegel. In 2019/20 I vistied Prof. Huan Sun at The Ohio State University and in 2017 I visited the National Institute of Informatics in Tokyo, Japan.

Before starting my Ph.D. I received an MSc degree in Computer Science from ETH Zurich, where I wrote my master's thesis on unsupervised video generation. I received my BSc degree with distinction in Computer Science from the Technical University of Vienna.

News

  • 03/2021 | I successfully passed my Ph.D. examination!
  • 02/2021 | I got accepted into the Talent Kick program of the ETH AI Center
  • 01/2021 | I got interviewed for a podcast (language GER) about AI, NLP, and many more topics, listen here: (WWW/Spotify)
  • 10/2020 | One paper(long) was accepted at COLING 2020, code and preprint available below.
  • 09/2020 | One paper(long) was accepted at EMNLP 2020, code and preprint available below.
  • 02/2020 | Our AAAI abstract won the best student presentation award.
  • 01/2020 | As Co-PI I received funding of $25,000 via the Oracle Cloud for Research program.
  • 10/2019 | One paper with Nil-Jana Akpinar (CMU) was accepted as AAAI-20 student abstract
  • 09/2019 | From October until end of January I will visit Prof. Huan Sun at Ohio State University, USA
  • 08/2019 | I am currently attending the Machine Learning Summer School (MLSS 19) in Moscow, Russia

Publications

Learning a Cost-Effective Annotation Policy for Question Answering

Bernhard Kratzwald, Stefan Feuerriegel, and Huan Sun

Empirical Methods in Natural Language Processing (EMNLP), 2020

Paper, BibTeX | Supplements: Source Code, Reviews

IntKB: A Verifiable Interactive Framework for Knowledge Base Completion

Bernhard Kratzwald, Guo Kunpeng, Stefan Feuerriegel, and Dennis Diefenbach

International Conference on Computational Linguistics (COLING), 2020

Paper, Live Demo, BibTeX | Supplements: Source Code, Reviews

Sample Complexity Bounds for Recurrent Neural Networks with Application to Combinatorial Graph Problems

Nil-Jana Akpinar, Bernhard Kratzwald, and Stefan Feuerriegel

Thirty-Fourth Conference on Artificial Intelligence AAAI (Student Abstract), 2020

* Winner best student presentation award *

Abstract , BibTeX | Supplements: Long Version, Source Code, Poster, Reviews

RankQA: Neural Question Answering with Answer Re-Ranking

Bernhard Kratzwald, Anna Eigenmann, and Stefan Feuerriegel

Annual Meeting of the Association for Computational Linguistics (ACL), 2019

Paper, BibTeX | Supplements: Source Code, Reviews

Learning from On-Line User Feedback in Neural Question Answering on the Web

Bernhard Kratzwald and Stefan Feuerriegel

The Web Conference (WWW), 2019

Paper, BibTeX | Supplements: Poster, Message me for source code.

Putting Question-Answering Systems into Practice: Transfer Learning for Efficient Domain Customization

Bernhard Kratzwald and Stefan Feuerriegel

ACM Transactions on Management Information Systems (TMIS), 2019; Volume 9 Issue 4 Paper No. 15

Paper, BibTeX | Supplements: ACM DL

Adaptive Document Retrieval for Deep Question Answering

Bernhard Kratzwald and Stefan Feuerriegel

Short paper in Empirical Methods in Natural Language Processing (EMNLP), 2018

Paper, BibTeX | Supplements: Source Code, Oral Presentation

Deep learning for affective computing: text-based emotion recognition in decision support

Bernhard Kratzwald, Suzana Ilic, Mathias Kraus, Stefan Feuerriegel, and Helmut Prendinger

in Decision Support Systems (DSS), 2018; Volume 115, Pages 24-35

Paper, BibTeX

Preprints

Practical Annotation Strategies for Question Answering Datasets

Bernhard Kratzwald, Xiang Yue, Huan Sun, and Stefan Feuerriegel

Last Update: March 2020

Paper

Estimating the impact of non-pharmaceutical interventions on documented infections with COVID-19: A cross-country analysis

Nicolas Banholzer, Eva van Weenen,Bernhard Kratzwald, Arne Seeliger, Daniel Tschernutter, Pierluigi Bottrighi, Alberto Cenedese, Joan Puig Salles, Stefan Feuerriegel, and Werner Vach

Last Update: May 2020

Paper

Improving Video Generation for Multi-functional Applications

Bernhard Kratzwald, Zhiwu Huang, Danda Pani Paudel, Acharya Dinesh, and Luc Van Gool

Last Update: November 2017

Paper | Supplements: Source Code, Project Page

Education

  • 10/2017 - 02/2020 | Ph.D., ETH Zurich
  • 10/2019 - 01/2020 | Visiting Researcher (Ph.D.) at the Ohio State University, USA
  • 11/2017 - 12/2017 | Visiting Researcher (Ph.D.) at National Institute of Informatics, Tokyo, Japan
  • 09/2015 - 09/2017 | MSc in Computer Science, ETH Zurich
  • 10/2011 - 06/2015 | BSc in Computer Science with distinction, Technical University of Vienna
  • 09/2005 - 06/2010 | HTL Villach (high school - 5 years), graduated top of the class

Work Experience

05/2014 - 05/2015

Project Assistant (JAVA)

ETH Zurich

I developed a framework for academic information retrieval courses ( GitHub Link). A RESTful web service is wrapped around an inverted lucene index. This allows students to query useful statistics and build their own retrieval models, without having unlimited access to the index itself.

05/2014 - 05/2015

Scientific Programmer (JAVA)

Technical University of Vienna

I designed an implemented the prototype for a model-driven multitenant ERP system operating in the cloud. I was responsible for the technical realization of the project and co-authored two peer-reviewed papers.

03/2013 - 04/2014

Junior Researcher

Research Studios Austria

I was a junior researcher in the TraCo project, where I implemented a scientific prototype in C#. We published a paper at IEEE SOCA which I presented in Matsue, Japan.

08/2010 - 08/2011

Paramedic

Austrian Red Cross

I was voluntarily working for the Austrian Red Cross since 2007. From 08/2010 till 05/2011 I conducted my obligatory civil service there. From 05/2011 till 08/2011 I worked full-time and served voluntarily in the catastrophe unit. I successfully finished the training for first-aid teachers.

07/2008 - 01/2010

Software Engineer (Java)

trinitec IT Solutions & Consulting GmbH

Full time in summer, 5h/week next to high school. We implemented a plug-in for andro-mda to generate code artifacts from complex UML models.

Old:

During the school year of 2007/2008 I was working self-employed for the Baustoffprüfstelle Villach automating office processes. In the summer of 2007 I interned at the technical customer support of Telekom Austria.

Service

  • Program Committee/Reviewer: NAACL 2021, ACL 2021/2019, ACL NLI 2020
  • Reviewer (Journals): Expert Systems with Applications, Decision Support Systems, Information Sciences

Extracurricular Activity

  • 2018 - now | Board member of the Scientific Staff @ MTEC, ETH Zurich
  • 2016 - 2017 | Member of the Committee for ETH Students without ETH bachelor, ETH Zurich
  • 2013 - 2015 | Elected student representative (ÖH), Technical University of Vienna
  • 2007 - 2011 | Voluntary paramedic and first-aid teacher, Austrian Red Cross