Ruprecht-Karls-Universität Heidelberg

Dr. Yannick Versley

I was a visiting professor at ICL from October 2013 to March 2016, and a junior research group leader in the Leibniz ScienceCampus Empirical Linguistics and Computational Language Modeling from April 2015 to March 2016.
From April 1st on, I will work at LinkedIn EMEA in Dublin.
Contact details
Emailversley <at> cl.uni-heidelberg.de


Topics of Interest

I have worked on a number of topics, including statistical parsing (Ver05, VR09, Ver14), coreference resolution (Ver06, Ver07, VPP+08, VMPY08), discourse relations (Ver10, Ver11, Ver13a, Ver13b) as well as (distributional) semantics of nouns, verbs, and word relations (Ver07, Ver08, Ver12a, Ver12b).

My current interests are centered around efficient structure-building mechanisms for natural language processing (parsing, compositional sentiment analysis), and how to scale them to a multitude of languages and domains.

Teaching

Winter 2015/2016

  • Introduction to Computational Linguistics
  • Research Module for MA students

Summer 2015

Winter 2014/2015

Summer 2014

Winter 2013/2014

Completed Theses

  • Lena Maldacker (2015): Information Extraction aus historischen Daten mittels Integer Linear Programming Inference
  • Julian Gerhard (2015): Genre Adaptive Optimization of Normalization Algorithms
  • Sebastian Spaar (2015): Ein Greedy-Ansatz zur dokumentenspezifischen Zeichenerkennung in frühneuhochdeutschen Texten
  • Hanna Hees (2015): Prognose potentieller psychischer Erkrankungen aus Forenbeiträgen
  • Raphael Schumann (2015): Literaturbasierte Konferenzempfehlungen für Autoren und Verleger
  • Xiaolin Bao (2015): Diskontinuierliches Konstituentenparsing durch Konversion aus Dependenzen
  • Angelika Kirilin (2015): Crosslinguale Analyse von Parserverhalten mittels Graphmotiven
  • Alexandra Hagelstein (2015):
    Data-driven Comparison of discourse connectives in English, German and French
  • David Grimm (2015): Verbesserte Lemmatisierung in domänenfremden Texten
  • Michael Haas (2015): Weakly Supervised Learning for Compositional Sentiment Recognition
  • Elisa Starke (2014): Morphologiesensitives Wortclustering für die Domänenadaptation im Parsing
  • Sabrina Mänz (2014): Experimente mit einem multimodalen E-Mail-Korpus für die Sprachsynthese
For co-supervised theses, see here

zum Seitenanfang