TUW NLP

A pre-doc position is available at the Data Science Research Unit of the @tu_wien in Vienna, Austria. Areas of Information Retrieval, Natural Language Processing, or Recommender Systems, with an emphasis on transparency, fairness, or reproducibility. jobs.tuwien.ac.at/Job/200421

Our paper on Offensive text detection across languages and datasets using rule-based and hybrid methods, presented by Kinga Gémes at AIMLAI at @cikm2022 , has just been published in the workshop's proceedings, read it here: ceur-ws.org/Vol-3318/short…

@adaamko presented the POTATO framework at the demo session of #cikm2022 in Atlanta. Check out the paper at dl.acm.org/doi/10.1145/35…

Our article "Explainable lexical entailment with semantic graphs" has been published in Natural Language Engineering, under Open Access: cambridge.org/core/journals/… Congrats to @adaamko and all co-authors!

Read about our tool POTATO in this @TDataScience article by @adaamko

@adaamko presented the POTATO tool at our first seminar of 2022. Video: youtu.be/6R_V1WfIjsU Slides: nlp.ec.tuwien.ac.at/seminar/sessio… Tool: github.com/adaamko/POTATO Paper: arxiv.org/abs/2201.13230

Attending EACL. Some good stuff, but very depressing at times. Models with dubious eval, datasets, and/or pointless tasks; new eval metrics with no validation again human eval. Even worse, the authors (and presumably reviewers) dont understand why above is a problem...