Daniil Sorokin

I am an Applied Scientist for Alexa AI at Amazon. Before that, I was a PhD student at the UKP Lab, Technische Universität Darmstadt. For my PhD project, I worked on linking texts to knowledge bases and on automatic question answering. I have recently published on the topics of relation extraction, semantic parsing for question answering and weak supervision.

I have more than 8 years of research experience in natural language processing and machine learning. Before I started the PhD program, I have completed a master degree in NLP at the University of Tübingen and had an experience working at a machine translation company. From November 2018 to January 2019, I was interning at Factmata and working on automatic evidence-based claim verification.

I have published in EMNLP, COLING, ESWC, *SEM and other venues. I was a PC member of/reviewer for ACL 2018-2021, CONLL 2018, EMNLP 2018-2020, NAACL 2020-2021, RepL4NLP 2019, GSCL-2017, Journal of Natural Language Engineering. At Technische Universität Darmstadt I have supervised 7 student theses and have taught a course on Natural Language Processing with Python for 3 years.

Daniil Sorokin portrait



  • Invited talk at the Seminar on Advanced Topics in Big Data of the Department of Computer Science on "Leveraging User Paraphrasing Behavior In Dialog Systems To Automatically Collect Annotations For Long-Tail Utterances", January 2021, Royal Holloway University of London (paper)
  • Invited talk at the Institute for Computational Linguistics on "Graph Neural Networks for Knowledge-Informed Natural Language Understanding", June 2019, University of Zurich
  • Leading the meeting on Building a Knowledge Base Question Answering Pipeline at the London Data Science Journal Club, January 2019 (slides)
  • Invited talk at the NLP Seminar of the Department of Computing on "Encoding Knowledge with Graph Neural Networks", January 2019, Imperial College London (slides)
  • Invited talk at the South England NLP Meetup organized by the UCL Machine Reading group on challenges of applying Graph Neural Networks to Knowledge Base Question Answering, January 2019, (slides)
  • Invited talk at the NLP Meeting, December 2018, the Alan Turing Institute (slides)
  • Invited talk at NLIP Seminar Series on Graph Neural Networks for Knowledge Base Question Answering, November 2018, Computer Laboratory, University of Cambridge (slides)
  • Modeling Semantics with Gated Graph Neural Networks for Knowledge Base Question Answering, August 2018, COLING 2018 (slides, paper)
  • Invited talk at the "Natural Language Processing and the Web" seminar on Knowledge Base Question Answering, December 2017, TU Darmstadt
  • System presentation at the 7th Open Challenge on Question Answering over Linked Data (QALD-7), May 2017, ESWC 2017
  • Web-based Visualisation of Toponyms and Their Derivational Patterns, September 2014, Advances in Visual Methods for Linguistics 2014 (AVML)
  • Visualizing toponym clusters on an interactive map, March 2013, 35th Annual Conference of the German Linguistic Society, Workshop on the Visualization of Linguistic Patterns (slides)


  • Outstanding Reviewer, the 56th Annual Meeting of the Association for Computational Linguistics (ACL), 2018
  • Teaching Award in Computer Science, Technische Universität Darmstadt, 2017
  • Best Challenge Paper on Question Answering, Extended Semantic Web Conference (ESWC), 2017
  • Fully-funded Master’s scholarship, German Academic Exchange Service (DAAD), 2011-2013