Skip to main content
Menu

Language modelling researchers to present at North American Chapter of the Association for Computational Linguistics

 Visualisation on shifts in the ideology of subReddits over time. Orange: Sino, a subreddit originally devoted to geopolitics that moved to a left-wing position; green and red: FreeSpeech and POLITIC, two originally moderate subreddits that moved to a right-wing position.

Visualision from one of the conference presentations. The methods developed in Hofmann et. al. can be used to track shifts in the ideology of subReddits over time.

Oxford e-Research Centre and Faculty of Linguistics language modelling researchers and students have had several papers accepted for the 2022 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2022).

The 2022 conference is being held from 10th to 15th July, and is being run as a hybrid meeting, held in Seattle, Washington, with other participants joining online from around the world.

Seattle Skyline. By Chait Goli Seattle Skyline. By Chait Goli

 

The conference invited the submission of long and short papers featuring substantial, original, and unpublished research in all aspects of Computational Linguistics and Natural Language Processing.

The following papers from our Language modelling researchers and students have been accepted for the conference;

  • Drinkall, Zohren & Pierrehumbert. “Forecasting COVID-19 Caseloads Using Unsupervised Embedding Clusters of Social Media Posts.” (main session).
  • Röttger, Vidgen, Hovy, Pierrehumbert “Two Contrasting Data Annotation Paradigms for Subjective NLP Tasks” (main session).
  • Kirk, Vidgen, Röttger, Thrush & Hale “Hatemoji: A Test Suite and Adversarially-Generated Dataset for Benchmarking and Detecting Emoji-based Hate” (main session).
  • Hofmann, Pierrehumbert, Schütze, “Modeling Ideological Agenda Setting and Framing in Polarized Online Groups with Graph Neural Networks and Structured Sparsity” (Findings).
  • Röttger, Seelawi, Nozza, Talat, Vidgen “Multilingual HateCheck: Functional Tests for Multilingual Hate Speech Detection Models”.

Valentin Hofmann, Felix Drinkall, and Paul Röttger will attend in the event in person. Paul Röttger will present the annotation paradigms article at the main conference and will also present work on Multilingual HateCheck at the Workshop for Online Abuse and Harms at the conference.

University of Oxford Language Modelling group

The Language Modelling group, led by Professor Janet Pierrehumbert, develops and applies methods in computational linguistics and natural language processing (NLP) to address a variety of questions about linguistic patterns in human communities. The group interest is in creating robust and beneficial language technology and formalizing and testing theories in the language sciences. The group uses classical algorithms, deep learning models, and experimental data to address them.

Find out more